# Bernard Williams on Philosophy and History Author: Matthieu Queloz Published in: Edited with Marcel van Ackeren. Oxford: Oxford University Press. 2025. [doi:10.1093/9780191966361.001.0001](https://doi.org/10.1093/9780191966361.001.0001) DOI: [10.1093/9780191966361.001.0001](https://doi.org/10.1093/9780191966361.001.0001) Canonical entry: https://www.matthieuqueloz.com/books/bernard-williams-on-philosophy-and-history/ Published PDF: https://philpapers.org/archive/VANMSO-5.pdf Machine-readable text companion generated from the PDF. Page markers follow the journal pagination. Pages 2–15 contain title pages, copyright, contents, contributor biographies, and an epigraph. They are omitted from this companion so the machine-readable text begins with the book’s main prose. [p. 16] Foreword: Williams on Philosophy, History, and History of Philosophy A. W. Moore There is a trajectory discernible in the course of Bernard Williams’s philosophical work that is both distinctive and revealing. In his earliest work he tries to make sense of some of the most basic aspects of our thought and experience: for instance, by confronting such metaphysical issues as what gives each of us a separate identity. In subsequent work he tries to make sense, more specifically, of our ethical thought and experience. And towards the end of his career he becomes increasingly self-​conscious about the very nature of the discipline to which his attempts to make sense of these things have been a contribution: that is to say, about the scope and limits of philosophy itself.1 It is hard not to see the transition from the first to the second of these as prompting the transition from the second to the third. For the question of what claim this abstract, reflective, dialectical enterprise can have on our attention becomes all the more urgent when the problems with which it is concerned are of the most practical and most important kind of all, problems about how one should live. The very title of what is often reckoned to be Williams’s greatest work, Ethics and the Limits of Philosophy (Williams 2006a)—​a middle period work—​indicates how far such self-​consciousness already informs what he is doing even before it comes to define it. Williams’s exploration of the nature of philosophy has as one of its primary concerns what is also the primary concern of the present volume: the relation between philosophy and history. Before he died Williams had hoped to publish a volume of essays in the history of philosophy. This volume would have included a substantial introduction on what the history of philosophy is, how it should be executed, and why it matters. This introduction, by its very nature, would have further exemplified the trajectory that I sketched above. Unfortunately, Williams did not live to see the project through. A volume of his essays in the history of philosophy, rather more extensive than what he had envisaged, did appear posthumously (Williams 1 Not (of course) that the trajectory is without deviation. Williams (1969) is a comparatively early essay on the nature of philosophy. But he produces much more in this vein later. Examples include Williams (2006f, 2006g, 2014). A. W. Moore, Foreword: Williams on Philosophy, History, and History of Philosophy In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. DOI: 10.1093/​9780191966361.003.0001 [p. 17] A . W. Moore c). But all that we have of the proposed introduction are some brief notes that he left behind. Still, there is enough in these notes to highlight one extremely significant aspect of how Williams conceives the relation between philosophy and history: namely, that the importance of the latter to the former is not merely the importance to philosophy of its own history. Williams is just as interested in the importance to philosophy of—​simply—​history. Indeed, he writes in his notes that, while it is clear that philosophy needs history, it is less obvious why it needs the history of itself. There is no implication that it does not need the history of itself. His own work bears ample witness to why it does have this need, and to how the need both can and should be met. But it is noteworthy that Williams takes the other need, that of philosophy for history tout court, to be the more evident of the two. In what follows I shall divide my comments between these two needs, or rather between the two more general relations of which they are instances: that of philosophy to its own history and that of philosophy to history other than its own. I shall begin with the former. ## 1. Philosophy and Its Own History The notes for Williams’s proposed introduction indicate that he intended to return to a distinction that first appeared in the preface to his book on Descartes (Williams 1978: 9–​10) and that he amplified in one of his later essays, likewise on Descartes (Williams 2006d: 257–​8). This is the distinction between the history of philosophy and the history of ideas. The history of philosophy, to which Williams’s own historical work belongs, is philosophy before it is history; with the history of ideas it is the other way round. Williams never pretends that this distinction is a sharp one. But he does take it to be an important one. He is surely right to do so. Moreover, part of its importance is the particularly stark way in which it subserves the attempt to achieve a self-​conscious understanding of the nature of philosophy itself. For of no other value of X is it true to say, in quite the same way as it is in the case of philosophy, that the history of X is, or even can be, part of X.2 In particular, this is not true of either science or the arts. The history of each of these is always in the first instance history. That is, a contribution to the history of science is never going to be a significant contribution to science, nor a contribution to art history a significant contribution to the arts. In 2 Marcel van Ackeren has raised the issue of historiography with me. This is an interesting case, and it perhaps shows how much work the qualification—​‘in quite the same way as it is in the case of philosophy’—​needs to do, maybe enough to warrant an argument to the effect that the history of historiography could only be part of historiography ‘in quite that way’ if, or to the extent that, historiography was essentially the philosophy of history. I hope that some of what I go on to say will prevent this from reducing to vacuity. [p. 18] Williams on Philosophy, History, and History of Philosophy many respects philosophy is poised between science and the arts, sharing certain features with the one but not the other and sharing certain other features with the other but not the one. In this critical respect, however, it is philosophy that is the outlier. To understand why is to appreciate something pivotal about what makes philosophy unique. Let us first consider the contrast between philosophy and science. The reason why the history of science is in the first instance history is that it is a history of discovery. Even when it serves as a partial vindication of what scientists think and say and do today, as opposed to what they thought and said and did before, this is only because the scientists of today have themselves already done the work of discrediting the theories of their predecessors. Whatever interest the history of science may hold for scientists, it has little claim on their attention qua scientists. And the little claim it does have is to indicate pitfalls into which scientists have fallen in the past and which they must now try to avoid (cf. Williams 1969: 148–​9; The history of philosophy is markedly different.3 True, there has been progress in philosophy, or at least there has been evolution. Some ideas have developed in ways that have made them better equipped to survive the ravages of hostile cross-​ examination; others have died out because they were not equipped to survive these ravages. (In the first case, think of Frege’s sharpening of Kant’s idea of an analytic truth—​an example that is apposite even if Quine’s subsequent repudiation of the idea is justified, for Quine would have had a much softer target if he had had only Kant’s idea to reckon with. And in the second case, think of Descartes’ idea of an uncreated substance having dominion over one created extended substance and a multitude of created thinking substances.) But it would be wrong to think that there has been progress of the same sort as there has been in science or, in so far as there has not, that this is to the detriment of philosophy. We still have a great deal to learn from Aristotle. And while we also have a great deal to learn from Quine, we certainly have no more to learn from him just because he was working over two thousand years later. Historians of philosophy must engage philosophically with past philosophical ideas in a way in which only crackpot historians of science could think it appropriate to engage scientifically with past scientific ideas. Historians of philosophy must try to derive positive lessons from past ideas which will assist them in making their own philosophical sense of things. What then of the contrast between philosophy and the arts? Well, even if great philosophical ideas of the past have an enduring value that great works of art of the past have and that great scientific theories of the past, for all their greatness, lack, that is not what makes contemporary philosophers’ engagement with them a 3 But it is not completely different. Thus the history of philosophy can likewise indicate pitfalls into which philosophers have fallen in the past and which they must now try to avoid: see Williams [p. 19] A . W. Moore historical exercise. Taking inspiration from Beethoven’s ‘Hammerklavier’ in composing a piano sonata, or for that matter playing Beethoven’s ‘Hammerklavier’, is not a historical exercise, at least not in anything like the way in which writing a commentary on Spinoza’s Ethics is. As Williams himself emphasizes, the fact that the history of philosophy is philosophy before it is history should not obscure the fact that it is history too (Williams 2006d: 258–​9). And part of what makes it history is that it must include an attempt to re-​enact former ways of making sense of things, by attending to the questions that were being asked, the presuppositions that were being made, the authorities that were being invoked, and suchlike. It has a crucial transchronological aspect.4 This in turn gives the lie to yet another enticing and prevalent misconception about the history of philosophy that Williams works hard to scotch: that its principal value lies in its indicating voices of yore that can be heard as participating in contemporary philosophical discussions. On the contrary, Williams insists, its principal value lies in its indicating voices of yore that cannot be heard as participating in contemporary philosophical discussions. It indicates voices that challenge whatever presuppositions make contemporary philosophical discussions possible, and it furthers philosophical reflection by signalling alternatives to what, if we remain rooted in the present, we will unthinkingly take for granted. Its business is to make the familiar look strange and the strange familiar. And our hope, as Williams says in the essay on Descartes to which I have already referred, is the same as the hope that Nietzsche expressed concerning classical philology: that in acting against the present age it benefits an age to come (Williams 2006d: 259). If it succeeds in doing this, then it will do so by spurring philosophers into participating in alternatives to their contemporary philosophical discussions. But those alternatives will not simply be bygone philosophical discussions. There is no reliving the past. And, even if there were, there would be no clear motive for doing so. The right response to the voices of yore is not to try to join in with them, any more than it is to turn a deaf ear to them. The right response is to connect what they are saying with what is being said now, and to search for ways of saying something that makes creative use of both. And this may mean continuing to say what is being said now, but with a more acute sense of what else might be said and of what is therefore being taken for granted. Come to that, it may mean continuing to say what is being said now, but with a renewed confidence in what is being taken for granted. The point is this. In the best history of philosophy there is an active sense 4 The two examples of how a musician may engage with Beethoven’s ‘Hammerklavier’ are intended to downplay the assimilation of philosophy to one of the arts. But I do not mean to suggest that an artist’s engagement with a great work of art of the past is always like that. Sometimes it is far more like the engagement of a historian of philosophy with a great philosophical idea of the past. Consider Lichtenstein vis-​à-​vis Monet. None of the contrasts that are being drawn here are sharp. But that just adds to their interest: it does not detract from their importance. [p. 20] Williams on Philosophy, History, and History of Philosophy of the possibilities for the future that the tendencies of the past have opened up, be the disruptions and continuities in those possibilities as they may. In sum, then, there are several inappropriate models for the relation between philosophy and its own history. It is unlike the relation between science and its own history. It is unlike the relation between the arts and their own history. It is unlike the relation between trying to solve some philosophical problem and consulting the latest issues of the journals to see how others try to solve it. Nor, unless we are committed Hegelians, shall we think of the history of philosophy as the telling of some grand dialectically necessary story that is destined to end with the narrator’s arriving at a kind of self-​knowledge. Williams himself invokes a musical analogy: he cites Stravinsky’s Pulcinella ‘in which the melodic line is Pergolesi’s, the harmony and orchestration Stravinsky’s’ (Williams 1978: 10—​in fact, the attribution of the melodic line to Pergolesi has since been shown to be spurious, but that is not the point). The melodic line here is intended to correspond to the philosophical concerns under investigation, the harmony and orchestration to the principles governing their rational reconstruction by the investigator. Another musical analogy that I think works well is that of theme and variation. Historians of philosophy work with what philosophers of the past have bequeathed, learning from it, questioning it, probing it, reworking it, appropriating it, and developing it. This is what makes them philosophers before they are historians. The best philosophy of this kind is a creative and edifying exercise that is fundamentally of a piece with the best philosophy of any other kind. ## 2. Philosophy and History Other than Its Own Very well, what are the benefits to philosophy of history other than its own? In part—​in large part—​they are the same. In particular, they include the salutary effects of listening to voices of yore that cannot be heard as participating in contemporary discussions. But it is no longer just a question of voices that cannot be heard as participating in contemporary discussions between philosophers; rather, it is a question of voices that cannot (even) be heard as participating in contemporary discussions of philosophical interest, say discussions that involve such traditional foci of philosophical attention as beauty, justice, or truth. These are not the preserve of philosophers. A judgement can be of aesthetic interest, for instance, indeed it can be an aesthetic judgement, without being a contribution to the philosophy of art. And this helps to explain why a piece of history can have the salutary effects in question without being a contribution to the history of philosophy. The fact remains that the salutary effects are salutary philosophical effects. The philosophy of art is bound to benefit from a due appreciation of the aesthetic judgements that people make and have made in the past. Likewise where other branches of philosophy are concerned. In particular, ethics, whose problems, as I emphasized [p. 21] A . W. Moore earlier, are nothing less than problems about how one should live, really needs to be informed by a due appreciation of the ethical judgements that people make and have made in the past, just as it needs to be informed, more generally, by a due appreciation of the ways that people live and have lived in the past. Only then is it liable to avoid the danger of mistaking mere contingencies about how we do these things for necessities. Only then, for that matter, is it liable to avoid the converse danger, to which it is perhaps only slightly less prone, of mistaking genuine necessities about how we do these things for contingencies. The first danger is fuelled by ignorance of alternatives. The second danger is fuelled by overactive imagination. Knowledge of what has actually worked in the past mitigates both (cf. Williams b: 36). Here it is helpful to invoke a distinction that is often drawn by philosophers, most famously perhaps by Rawls with respect to justice, between a conception and a concept. Various relatively determinate ‘conceptions’ of something can all be said to correspond to the same relatively indeterminate ‘concept’ of that thing. Granted this distinction, one way in which we may need to confront the contrast between the necessary and the contingent to which I have alluded is as follows: we may need to acknowledge a necessity about the fact that we possess some given concept alongside a contingency about which corresponding conception we possess. To be sure, there are no sharp lines here. To press too hard on the question of how much two conceptions can differ while still corresponding to the same concept would be like pressing too hard on the question of how much the rules of a game can change while still counting as rules of the same game. (Some time in the fifteenth century the pawn was first allowed to move forward two squares in chess. Was that the point at which people first began to play chess? Or had the game already existed for a long time and merely underwent a change then? It really does not matter. As Williams marvellously remarks in a related connection, ‘it would be arrant scholasticism to go on about it’ (Williams 2014: 406).) There nevertheless are cases—​one of which is provided by Rawls’s concern, justice, and another of which is provided by one of Williams’s chief concerns, truthfulness—​where it is important to recognize that there is a single corresponding concept, and where part of what makes it important is the very contrast that is at issue here, between the necessity of the concept and the contingency of the conceptions. Thus it is important for us to recognize that, even if we cannot dispense with the concept of truthfulness, there are alternatives to our conception of truthfulness, and there may even be alternatives that we would be better off adopting (cf. Williams 2014: 406–​12). As for the comparison that I have been drawing between the benefits to philosophy of its own history and the benefits to philosophy of history other than its own, what it comes to, very crudely, is this.5 Both kinds of history assist philosophy 5 See Williams (2006g: §5) for something related but altogether less crude. [p. 22] Williams on Philosophy, History, and History of Philosophy in making sense of how we make sense of things; and both do this by signalling bygone alternatives; but the former does it by signalling bygone alternatives at the upper level—​that is, bygone alternatives to how we make sense of how we make sense of things—​whereas the latter does it by signalling bygone alternatives at the ground level—​that is, bygone alternatives to how we make sense, simply, of things. And the latter thereby fosters a philosophical appreciation of respects in which we could have made different sense of things; of respects in which we could not have done so; of respects in which we still could do so; of how the forces that have brought us to where we are now may themselves provide some of the impetus for us to do so; and of what gains and losses this would involve. And it is worth adding that these benefits to philosophy will be all the greater if, as I believe, and as I think Williams believes,6 one of the tasks for philosophy is to create new ways of making sense of things. But it is not only ‘real’ history that can help philosophy in these ways. Provided that we have assimilated enough real history to discipline our imaginations, then imaginary history can help too. And here we come upon that important method for making sense of how we make sense of things which Nietzsche called ‘genealogy’. This is a method that Williams himself both practises and preaches, most notably in his last book Truth and Truthfulness, to which he gives the subtitle An Essay in Genealogy (Williams 2002). A genealogy is an account of the place that some way of making sense of things has in our lives and of how it came about or could have come about. It can include elements of real history. But it can also include elements of imaginary history, describing how certain processes could have led us to make sense of things in the way in question, in particular how they could have done so by virtue of serving various human interests. Such imaginary history can be useful in helping us to see how something that does not immediately present itself as performing any such function in our lives nevertheless does so. And it can play a similar role to real history: it can expose some aspects of how we make sense of things as more or less necessary (in as much as something had to fulfil this function), others as more or less contingent (in as much as this did not have to do so). Note that a genealogy can never be entirely value-​neutral. That is, it can never just be a story about the place that some way of making sense of things has in our lives, with no implications concerning either the value of anything’s occupying that place or the comparative value of other things’ occupying it instead. The very fact that the genealogy reveals this way of making sense of things as fulfilling some function in our lives—​as serving interests of ours—​precludes that. Indifference is not an option here. There are, however, many forms that our evaluation can take, depending on how we assess the function itself and on how we assess all the relevant side effects. In particular, and most basically, it can take the form of overall 6 Cf. Williams (2006e: 31–​3). Cf. also his insistence that we have too few ethical ideas (Williams [p. 23] A . W. Moore approval or it can take the form of overall disapproval.7 Hume provided a genealogical account of our idea of justice that served as a vindication of it; Nietzsche provided a genealogical account of our idea of God that served as a subversion of it. Even here, however, there is a vital distinction to be drawn. Let s be one of our ways of making sense of things. A genealogy may expose s as involving false beliefs, conflicting values, or defective concepts, say. So far, so subversive. Or, conversely, it may herald s as involving true beliefs, coherent values, or effective concepts. So far, so vindicatory. These are a matter of how s measures up as a way of making sense of things. We can call subversion and vindication in this respect ‘primary’ subversion and ‘primary’ vindication. But there is another dimension of assessment, having to do with the capacity for a way of making sense of things to perform some function in our lives that goes beyond equipping us to make sense of anything. The sheer fact that s has been primarily subverted or primarily vindicated does not settle whether it is to be deplored or celebrated for performing a function of this kind. If s is exposed as involving false beliefs, for example, it can still be heralded as providing us, or as having provided us, with much-​needed comfort, and it can still be commended as a result. Again, if s is heralded as involving true beliefs, it can still be exposed as causing us, or as having caused us, a great deal of pain, and it can still be condemned as a result. Or suppose that s is exposed as involving a conflict of values, and suppose that we would rather live with the conflict—​even if we cannot thereby make unified sense of anything—​than sacrifice any of the values (cf. Williams 1981). Then s can still be heralded as enabling us, or as having enabled us, to resist the pressures of systematization that would force us to make such a sacrifice. There thus arise what we can call ‘secondary’ vindication and ‘secondary’ subversion, where these are compatible, respectively, with primary subversion and primary vindication. Very well, suppose that we have acceded to a primary subversion of s. Does it follow that there might still be some rationale for continuing to take s seriously, to the extent of continuing to adopt it? Well, here of course we run into a variant of Moore’s paradox.8 We cannot both accede to the primary subversion of s and continue to adopt it. Or at least, we cannot do so in the full light of reflection. But could we perhaps accede to the primary subversion of s, from a position of reflection, and then (re-​)adopt it by (re-​)immersing ourselves in some relevant unreflective practice? Williams himself adduces reasons to deny that we could do even this (Williams 2006a: 163–​4)—​never mind that it would involve paying a price of self-​deception that we should surely be wary of paying. 7 I do not mean to suggest that these are exhaustive alternatives. It can also take the form of neither: the value involved may ultimately balance out. That is still different from there not being any value involved at all. 8 This is the paradox that, even though it is possible for me to have false beliefs, there is nevertheless an incoherence in my claiming, with respect to any given belief of mine, that it is a case in point. [p. 24] Williams on Philosophy, History, and History of Philosophy Let us grant, for the sake of argument, that Williams is right about this. There might nevertheless be a rationale for continuing to take s seriously in some less committed way. For one thing, we still have the history to look back on: we may quite reasonably applaud s for how it served us in the past. But there might also be a rationale for continuing to take s seriously that bears on the future. For not only might s be subject both to primary subversion and to secondary vindication; s might also be subject, given the prospect of new sets of circumstances affording new perspectives from which s can be put to new work, to potential future primary vindication. This might even involve a kind of creative transformation of the false into the true. A very simple model of such a transformation would be the case of a man who thinks that he is facing north when in fact he is facing south—​hence a man who thinks something false—​and who later, having been reoriented so that he is indeed facing north, still thinks that he is—​hence a man who later thinks something true. I shall shortly consider what Williams might make of such a suggestion. But first I note that it would be by no means completely foreign to Nietzsche. Nietzsche continually emphasizes the importance of perspective to our sense-​making. Not only that. He insists that only a concept that does not have a history can be defined. A variant of this is that only a way of making sense of things that does not have a history can have its application circumscribed. If we combine this with the idea that, among those ways of making sense of things that are currently on our radar, only one that does have a future, or at least that has a possible future, cannot have its application circumscribed—​the thought being that nothing precludes such circumscription except the potential for disruption in what we cannot now foresee—​ then we arrive at the conclusion that, among those ways of making sense of things that are currently on our radar, only one that has a future, or at least a possible future, has a past. This is less paradoxical than it may sound. What it comes to is this: once sense has been made of things in some way, then that way of making sense of things can at most be lost, never destroyed, and, as long as we are aware of it, it is always available to be adapted and extended, possibly through the adoption of new conceptions corresponding to old concepts, so as to serve afresh as a good way to make sense of things. Such ideas, I say, would not be uncongenial to Nietzsche. They nevertheless sound uncongenial to Williams. For Williams thinks that ways of making sense of things can be destroyed—​indeed that they can be destroyed by reflection, including the kind of reflection that is characteristic of both primary and secondary subversion (Williams 2006a: 167–​9).9 This is not the place to dwell on all the issues that this raises—​though it is worth noting in passing that they include some fascinating issues concerning which of destruction and preservation, if both are 9 Nietzsche, for that matter, proclaimed the death of God. [p. 25] A . W. Moore possible, better serves the aim that Williams advocates of not imposing too much on our descendants’ sense-​making (see Williams 2006a: 172–​3). Suffice to observe that, even if Williams is wrong about how much damage reflection can wreak, reflection is not what is required to revive a way of making sense of things that has been primarily subverted; what is required is a new set of circumstances affording a new perspective on things. Reflection, for reasons of the very sort that Williams adduces in his discussion of these issues, is liable to be a positive hindrance here. So, given that philosophy is reflective if it is anything, there is little that philosophy can do to foster the kind of creative transformation of bygone ways of making sense of things that I have just been portraying. What it can do is to bring the very possibility of such a thing to our attention, along with its potential benefits and dangers. This is reminiscent of the way in which, on Williams’s view, philosophy can reveal the importance of our having confidence in our various ways of making sense of things, even though it does not itself contribute much to such confidence, indeed even though it is one of the main causes of our need for it (Williams 2006a: ch. 9). More generally, it is reminiscent of the way in which, on Williams’s view, the insights that philosophy affords include insights into its own scope and limits. And here I think we have come full circle. We are now back with Williams’s self-​ consciousness about the nature of his discipline and his corresponding sense of the importance to that discipline of history, including its own history. Everything about Williams’s understanding of the relation between philosophy and history is born of his awareness of what philosophy importantly can, and importantly cannot, do—​where the former includes what it can do with the help of history while the latter includes what it cannot do without such help—​together with his lifelong crusade to assist it in trying to do the former while also assisting it in trying not to do the latter.10 ## References Williams, Bernard. 1969. ‘Philosophy’. In General Education: A Symposium on the Teaching of Non-​specialists. Edited by Michael Yudkin, 138–​64. Harmondsworth: Penguin. Williams, Bernard. 1978. Descartes: The Project of Pure Enquiry. Harmondsworth: Penguin. Williams, Bernard. 1981. ‘Conflicts of Values’. Reprinted in Moral Luck: Philosophical Papers Cambridge: Cambridge University Press. Williams, Bernard. 2002. Truth and Truthfulness: An Essay in Genealogy. Princeton: Princeton University Press. Williams, Bernard. 2006a. Ethics and the Limits of Philosophy. London: Routledge. 10 I am very grateful to Marcel van Ackeren and Matthieu Queloz for comments that helped me to improve this Foreword. [p. 26] Williams on Philosophy, History, and History of Philosophy Williams, Bernard. 2006b. ‘Pluralism, Community and Left Wittgensteinianism’. Reprinted in In the Beginning Was the Deed. Edited by Geoffrey Hawthorn, 29–​39. Princeton: Princeton University Press. Williams, Bernard. 2006c. The Sense of the Past: Essays in the History of Philosophy. Edited by Myles Burnyeat. Princeton: Princeton University Press. Williams, Bernard. 2006d. ‘Descartes and the Historiography of Philosophy’. Reprinted in The Sense of the Past: Essays in the History of Philosophy. Edited by Myles Burnyeat, 257–​64. Princeton: Princeton University Press. Williams, Bernard. 2006e. ‘Metaphysical Arguments’. Reprinted in Philosophy as a Humanistic Discipline. Edited by A.W. Moore, 22–​33. Princeton: Princeton University Press. Williams, Bernard. 2006f. ‘Philosophy as a Humanistic Discipline’. Reprinted in Philosophy as a Humanistic Discipline. Edited by A. W. Moore, 180–​99. Princeton: Princeton University Press. Williams, Bernard. 2006g. ‘What Might Philosophy Become?’. In Philosophy as a Humanistic Discipline. Edited by A. W. Moore, 200–​13. Princeton: Princeton University Press. Williams, Bernard. 2014. ‘Why Philosophy Needs History’. Reprinted in Essays and Reviews, Princeton: Princeton University Press. [p. 27] Doing History Philosophically and Philosophy Historically Marcel van Ackeren and Matthieu Queloz Bernard Williams’s historical work has tended to be overshadowed by his work as a systematic philosopher. If the influence of that historical work is harder to discern, however, it is in good part due to its remarkable range: it covers not only Homer, the Greek Tragedians, Plato, and Aristotle, but also Enlightenment figures such as Descartes, Hume, and Kant as well as modern thinkers such as Nietzsche, Collingwood, and Wittgenstein. Moreover, Williams’s historical work demonstrates remarkable range also in relating philosophy and history in a variety of different ways. We can distinguish at least four—​three having to do with philosophy’s relations to its own history and the fourth with its relation to history more generally. In his characteristically compressed style, Williams sometimes moved swiftly between these, or combined several of them at once; but they are worth distinguishing analytically even when Williams combines them in practice, because this enables us to understand them as four independently illuminating connections between philosophy and history.1 First, Williams believes that if one is going to study philosophy, one must come to know some of its history: ‘someone learning philosophy itself will need to learn something of its history’, he declares already in a little-​known text on the discipline of philosophy from 1969, ‘and this will involve, if it is to be useful, study of the actual writings of earlier philosophers’ (1969: 148). True to his convictions, Williams left behind a large body of work in which he engages in detail with the work of past philosophers. He published a book on Plato and one on Descartes, and Shame and Necessity dealt at length with Homer and the Greek Tragedians. Ethics and the Limits of Philosophy engages with Socrates, Plato, Aristotle, and Kant, while Truth and Truthfulness offers some exegesis of Hume, Rousseau, and Nietzsche. His writings in social and political philosophy are also explicitly informed by the work of Hobbes, Machiavelli, and Weber. And the posthumously published collection The Sense of the Past includes no fewer than twenty-​five essays, written over 1 For an account of the current debate, see van Ackeren (2018). Marcel van Ackeren and Matthieu Queloz, Doing History Philosophically and Philosophy Historically In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. DOI: 10.1093/​9780191966361.003.0002 [p. 28] Doing History Philosophically the course of forty-​one years, running from the sixth century bc to the twentieth century ad. As he clarified towards the end of his career, Williams regards ‘those who think that philosophy can ignore its own history’ as making ‘enormous and implausible assumptions’ (2006c: 192). Philosophy cannot ignore its own history in the way that science tends to ignore its history, because for philosophy to ignore its history would be for it to implausibly assume that its history was entirely vindicatory—​a history of discovery, which represents us as having got some mind-​independent fact right, or at least a history of progress, in which later philosophers can be seen as having got something right by the lights of earlier philosophers, so that later outlooks can definitively supersede earlier outlooks. As Williams sees it, however, much of philosophy has a history that is nothing like this. It has, in good part, what might be called a history of change as opposed to a history of discovery or progress.2 The history of moral and political philosophy, and even the history of metaphysics, involves changes too radical to plausibly sustain the systematic depiction of later philosophers as having won an argument with earlier philosophers. That is to say, the history of philosophy is partly the history of the very argumentative frameworks within which victory over rival arguments first becomes intelligible. This is not to deny that there can be progress in philosophy. The peculiarity of the relation of philosophy to its history, for Williams, ‘lies in the fact that great philosophical works of the past are still philosophically (and not merely historically) interesting, while at the same time there is such a thing as progress in philosophy’ (1969: 148). This combination of features is peculiar in that it sets philosophy apart from both the sciences and the creative arts: Great works of the past have contributed to advances in philosophy, and can themselves be criticized by reference to modern developments; at the same time they are still to be read and studied in their own right, and illuminated (as with works of art) by historical understanding of the situation in which they were written. The pull between these two approaches is constant, and also valuable: one has to resist the over-​simplifications which would try to dump the one approach or the other. (1969: 150) The contrary pull between evaluating past philosophers’ arguments by present-​ day lights and historicizing them by setting them in their historical context is a productive one, Williams thinks, and there are several ways in which philosophical and historical inquiry can be combined to exploit it. 2 For an elaboration of the distinction between the history of change, the history of progress, and the history of discovery, which is itself meant to explicate distinctions implicit in Williams’s own remarks, see Queloz (2017). [p. 29] Marcel van Ackeren and Matthieu Queloz One of these productive combinations—​and this is the second interconnection between philosophy and history—​is that someone who is going to engage with the history of philosophy will in turn need philosophy to make proper sense of that history. The history of philosophy without philosophy is ‘self-​defeating’ Williams affirms. It cannot succeed even as history. This is because ‘it is impossible to have any living interest in the philosophers being expounded, indeed impossible even to understand what they are at, without a genuine feeling for the problems they were concerned with’ (1969: 150). The present-​day interest of past philosophers lies not merely in the conclusions they reached, but in the arguments and considerations that moved them. Merely reporting their conclusions without doing the philosophical work of trying to understand their problems and arguments is bound to prove unilluminating even historically, let alone philosophically: The historical understanding of a philosopher is, after all, supposed to be understanding: it is supposed to make comprehensible what he was at, what his problems and arguments were. That is not going to be done by merely repeating what he said. It involves both a philosophical sense of what a philosophical problem is, and some use of philosophical concepts and distinctions to explain (in some cases, even to translate) the philosopher’s words. This already puts one in a position where one is involved in a critique and an evaluation. Even when work in the history of philosophy primarily aims to produce history rather than philosophy—​in other words, when it aims to produce what Williams referred to as ‘the history of ideas’—​it will require some philosophy, and will have to be informed to some degree by our sense of what makes sense as a philosophical problem or argument, if it is to understand what it is the history of. As Williams wrote in 1965, there cannot ‘be a history of ideas without identification of ideas; and to identify what ideas are embodied in a text, particularly a philosopher’s text, is no less a matter of philosophical comprehension than of anything else’ a: 54). At the same time, Williams is known for emphasizing that the history of philosophy can also be done philosophically, in a way that yields philosophy before it yields history (2005a: xiii). This third connection between philosophy and history involves practising the history of philosophy as opposed to the history of ideas. The key question then becomes what the historical texts mean to us rather than what they meant back then. To this end, one must reconstruct the problems and arguments of past philosophers as clearly and forcefully as one can. This requires one to be mindful of the present state of the discipline—​not just because a forceful rational reconstruction of past philosophers’ arguments is apt to profit from drawing on the most illuminating concepts and distinctions philosophy now has to offer, [p. 30] Doing History Philosophically but also because what makes sense to us now as a pressing problem or a strong argument is, to a considerable degree, a function of current debates and concerns.3 At first glance, Williams’s aspiration to produce history of philosophy as opposed to history of ideas seems simply to echo Grice’s maxim that ‘we should treat those who are great but dead as if they were great and living, as persons who have something to say to us now’ (Grice 1989: 66). This maxim, Williams recalls, was approvingly cited by Peter Strawson as something that ‘all Oxford philosophers would agree with’ (2006b: 344). And indeed, Williams himself does not disagree with it. But he has a rather more ambiguous relationship to it than some of his Oxford mentors and colleagues. Responding to Bryan Magee’s characterization of Oxford philosophers in the s as arguing with the ideas of Locke, Descartes, or any other dead philosopher as if they were a colleague in the Common Room, Williams grants that ‘the approach to a lot of the philosophy of the past had what might be called a sturdily anachronistic character’ (1982: 121). He describes his principal mentor, Gilbert Ryle, as urging his students to treat a text written by Plato ‘as though it had come out in Mind last month’ (Williams 2006c: 181). But Williams observes that this ‘rather odd way of doing’ the history of philosophy ‘is rather productive and stimulating, and has in fact had a more robust legacy than some kinds of the history of philosophy which are just passively guided by an excessive concern for not being anachronistic’ (1982: 121). As comes out in this remark, Williams, self-​consciously following Nietzsche, is critical of antiquarian approaches to the history of philosophy that entirely neglect to relate the past to present concerns. Gathering ‘any old facts, merely for their own sake’, he writes in Truth and Truthfulness, ‘can sustain an individual life, but in a larger scheme of things historical research will not make sense unless it is driven by some question’ (2002: 146). It is a ‘Platonic misunderstanding’ (2002: 142) of the ideal of disinterested inquiry to demand that historical inquiry should remain pure of any connection to present concerns. Just as ‘a desire for fame does not corrupt or undermine the search for truth, if what one will be famous for (if all goes well) is having found the truth’ (2002: 142), so a connection to some present concern does not corrupt or undermine the disinterestedness of historical inquiry, if what will satisfy the concern (if all goes well) is knowledge of what actually happened, irrespective of what we would like to have happened. Which aspects of the past ‘strike us and strike our historical curiosity’ (Williams 2005a: xiv) can be a function our present situation even when what we are curious about is, in Ranke’s famous phrase, wie es eigentlich gewesen ist—​how things actually were. At the same 3 Arguably, Williams’s problem-​oriented elaboration of the genealogical method can be deployed to this end in the history of philosophy: it can make past thinkers speak to us by revealing how their concepts relate to our concerns in the sense of forming specific solutions to more generic problems we still face in some form; see Queloz (2023). [p. 31] Marcel van Ackeren and Matthieu Queloz time, the voices of yore will only mean much to us if they are imbued with meaning by our present concerns. Williams approvingly quotes Nietzsche’s remark that it is ‘only if we give them our soul, that [the works of earlier times] can go on living: it is our blood that makes them speak to us. A really “historical” presentation [Vortrag] would speak as a ghost to ghosts’ (Williams 1993: 174n36). However, Williams also acknowledges that if one seeks to tie past philosophical ideas as closely as possible to present concerns, one ends up doing so at the expense of one’s ability to recover the distinctive significance they carried in their original context. There is a trade-​off between the pursuit of present-​day relevance and the pursuit of authenticity. The two cannot be pressed all the way in concert without eventually coming into conflict—​rather as ‘Impressionism, by exploring as intensely as possible the surface effects of light, was thereby debarred from giving as much information about structure as was accessible to some other styles of painting’ (2006a: 257). There comes a point, for instance, at which our willingness to count the sheer boringness of a reconstructed philosophical view as a criticism shows that we are more interested in producing philosophy than history d: 166). Similarly, we must realize that pursuing the maximally consistent interpretation of a philosophical text is not necessarily going to lead to the most authentic representation of it. As Williams himself writes of his attempts to make sense of Plato, for example: people spend enormous time (I have spent some myself ) on trying to find interpretations of Plato’s Sophist which make Plato’s theories consistent. But if Plato’s Sophist is about what we think it is about (and granted his theories about these very difficult subjects came when they did) it is wildly improbable that his theories on those subjects would succeed in being consistent. (2006d: 166) This is why, alongside the history of philosophy that yields philosophy before it yields history, an important role remains for the history of ideas, according to Williams—​as long as it is sufficiently informed by philosophy to achieve a proper grip on the ideas it traces the history of. Moreover, Williams has strong reservations about ways of doing history philosophically that presuppose the transhistorical identity of the questions being asked. If we treat the great and dead as having something to say to us now because they were fundamentally asking the same questions that we are now asking, we ignore everything that is historical about the history of philosophy. But by flattening the differences between our situation and the past, we deprive ourselves of the main rationale for turning to the past to begin with. This is why Williams came to dismiss Ryle’s injunction to treat a text written by Plato as though it had come out in Mind last month as ‘an idea which, if it means anything at all, means something that destroys the main philosophical point of reading Plato at all’ (2006c: 181). Why bother learning Greek and reading Plato if all we find is what we were familiar with [p. 32] Doing History Philosophically already, because we remain resolutely insensitive to anything that could set ancient philosophical thought apart from the latest journal articles? ‘To justify its existence’, Williams believes, the history of philosophy must not only be done philosophically, to ‘sustain its identity as philosophy’, but also ‘maintain a historical distance from the present’ (2006a: 259). Just in so far as what the great and dead have to say to us now differs from what the living have to say to us can the past help us understand our ideas better by revealing unquestioned presuppositions and unasked questions in our current ways of going on. Again following Nietzsche, Williams sees the value of the history of philosophy as lying notably in its capacity to be untimely, to act against the age for the benefit of future ages (1993: 4; 2006a: 259). For it is in virtue of ‘the possibility of the past philosophy’s being untimely’ that it can help ‘to make strange what is familiar in our own assumptions’ (2006a: 263). This ‘alienation effect’ allows us to use the respects in which the past is a foreign country to our advantage.4 If we are to succeed in this, however, we must not only draw on as much of our own philosophical understanding as necessary to recover from the past something that counts as philosophy; we must also draw on as much historical understanding as possible to ensure that what comes out is not simply our philosophy: What we must do is to use the philosophical materials that we now have to hand, together with historical understanding, in order to find in, or make from, the philosophy of the past a philosophical structure that will be strange enough to help us to question our present situation and the received picture of the tradition, including those materials themselves. (2006a: 264) By becoming familiar with the strangeness of past philosophy, and recovering the strangeness of what is most familiar in our current ways of philosophizing, we put ourselves in a position to challenge the presuppositions, questions, and answers characteristic of contemporary debates instead of blindly perpetuating them. A fourth way in which Williams connects philosophical and historical inquiry, finally, is by insisting on the need for philosophy to involve itself not just in its own history, but in history more generally. In an interview conducted in December Williams speaks of a ‘historicist turn’ (2009: 198) having become increasingly prominent in his work over the previous ten to fifteen years. In the same year, he published a piece in the London Review of Books entitled ‘Why Philosophy Needs History’, which was much discussed at the time.5 And in Truth and Truthfulness, he insists that ‘philosophy cannot be too pure if it really wants to do what it sets out to do’—​it ‘must involve more than abstract argument, and . . . must engage itself in history’ (2002: 39). When dealing with concepts that have a significant history, such 4 On the alienation effect and the uses Williams himself makes of it, see van Ackeren (2019). 5 Reprinted in Williams (2014b). [p. 33] Marcel van Ackeren and Matthieu Queloz as the values of liberty, truthfulness, justice, or equality, that history ‘stands in the way of their simply having a definition’ (2001: 91), Williams argues, paraphrasing Nietzsche. That is why Williams felt the need to turn—​as his close friend Isaiah Berlin had done in his own way—​‘from a form of philosophy which ignored history to a form of philosophy which did not ignore history’ (2001: 92). It is in terms of this fourth connection between philosophy and history that we can make sense of Williams’s embrace, in his late work, of a form of the genealogical method. This involves not so much doing history philosophically as doing philosophy historically. Distinguishing these two aspirations allows us to reconcile Williams’s emphasis, in his remarks on the history of philosophy, on how different past philosophical thought is with his seemingly contrary emphasis, in his genealogical philosophizing, on unsuspected commonalities across history. For what the state-​of-​ nature story opening his genealogical narrative in Truth and Truthfulness reveals, after all, is that truthfulness possesses ‘a common core . . . developed or expressed . . . in different ways’ across history; in other words, Williams proposes to understand truthfulness in terms of ‘historical variation’ around a ‘central core’ (2014b: 407). Similarly, his historical reflections on the notions of responsibility and voluntary action in the ancient Greek world lead him to conclude that these notions are practically indispensable to any human society anywhere, because their indispensability follows already ‘from some universal banalities’ (1993: 55). And his genealogical sketch of the development of varying conceptions of liberty over the course of history turns on the idea that ‘there must be a core, or a primitive conception, perhaps some universal or widely spread human experience, to which these various conceptions relate’ (2005b: 76). To some of his critics, this emphasis on transhistorical constants has seemed to be deeply at odds with his championing of historical difference. Jonathan Barnes, for example, concludes that ‘Williams doesn’t practice the HP [history of philosophy] he preaches’ (2011: 21). If ‘the point of reading philosophers of the past is to find in them something different from the present’ (Williams 2006b: 344), why does Williams spend so much time arguing that people in very different historical circumstances fundamentally conceive of truthfulness, voluntariness, responsibility, and liberty in much the same way as we do? The answer is that doing philosophy historically is a very different enterprise from doing history philosophically. While the latter aims to reconstruct the philosophy of the past without turning it into an echo chamber of contemporary debates, the former aims to use the history behind our ideas—​as opposed to the history of their discussion in philosophical texts—​in order better to understand how and why we came to live by these ideas, and why they have the contours they do. How we peculiarly conceive of truthfulness, voluntariness, responsibility, and liberty now and around here is one thing. But in order to place other conceptions [p. 34] Doing History Philosophically of these things ‘in a philosophical and historical space’, as Williams puts it, we also need ‘a more generic construction or plan’ (2005b: 76). That is what we can gain from reflection on ‘universal banalities’ and from their representation using the narrative device of the ‘State of Nature’. Grasping in more abstract terms what the central concerns are that most basically animate preoccupation with anything like these ideas allows us to reidentify them across different historical contexts (how would we know what counted as different expressions of the same underlying ideas otherwise?). Moreover, this can help us to understand what kind of facts—​whether general or parochial—​an idea derives its importance from, and what might have led other societies to conceive of an idea differently from the way we conceive of it now. The concerns of the philosophical genealogist are therefore not at all the same as those of the historian of philosophy. Williams thus connects philosophy and history in numerous ways between his work in the history of philosophy, on the history of philosophy, and on history’s importance to philosophy. Better understanding these aspects of his oeuvre and how they relate is not only central to understanding his own thought, but also has valuable insights to offer to several currently raging debates—​on the methodology of philosophy, on how and why to do history of philosophy, and on the relevance of the historical perspective to systematic philosophy. Given that Williams himself insisted that historical and philosophical inquiry were significantly intertwined, there is also reason to think that his historical work informed and might elucidate his more systematic work. The present volume assembles interpreters of Williams’s work and well-​known experts in the various fields of scholarship it touches on to address the following four sets of questions: 1. Critical appraisal of his historical work: How does Williams interpret past philosophers? In what way are these interpretations influenced by his own systematic views? What is the influence of his historical work on current exegetical debates? 2. Systematic significance of his historical work: How does Williams’s engagement with historical texts shape his own systematic views? What are his intellectual debts to thinkers of the past, and which insights, including negative insights, did he carry over into his systematic work? 3. Methodology of the history of philosophy: How can the history of philosophy be done philosophically (as judged by Williams’s own example)? What should work in the history of philosophy aim to do for us, and how can these aims guide the way we do it? What mistakes should we avoid? 4. Metaphilosophy: How should we do systematic philosophy in a way that is informed by history? What methods did Williams develop to this end? Why does philosophy need history? [p. 35] Marcel van Ackeren and Matthieu Queloz The volume’s thorough investigation of these various ways in which philosophy and history are intertwined in Williams’s thought promises to be of interest not just to Williams scholars, but also to the growing number of people interested in how history can be done philosophically and how philosophy can learn from history. The bulk of the volume is devoted to Williams’s work in the history of philosophy and how it informs his systematic work. Three groups of four to five chapters each examine Williams’s engagement with figures from antiquity, the Enlightenment, and modernity. In ‘Psychology, Ethics, and “Ethicized Psychology”: Bernard Williams on Greek Thought and Greek Philosophy’, Terence Irwin addresses Williams’s contention that pre-​Platonic Greek thought beyond philosophy—​especially Homer and the tragedians, but also Thucydides—​expresses ethical views superior to those of Plato and Aristotle, notably because the pre-​Platonic Greeks lack a problematically ‘ethicized’ psychology revolving around a certain kind of belief in a will. Irwin challenges the contrast that Williams attempts to draw, arguing that the pre-​Platonic Greeks in fact did believe in a will in just this sense, and would have been worse off if they had not. Nevertheless, Irwin sees philosophical value in Williams’s approach to these ancient sources, and endorses the view that literary and historical argument can be appropriately combined with philosophical argument. Sophie Grace Chappell continues the critical examination of Williams’s use of Greek tragedy in ‘Agamemnon at Aulis: A Misfiring Example in Williams’, focusing on Williams’s discussion, in his classic paper on ethical consistency, of the tragic dilemma that Aeschylus represents Agamemnon as facing at the Greek port town of Aulis. After situating the Agamemnon example within Williams’s use of examples more generally, Chappell argues that this turns out to be an unfortunate choice of example in light of its wider textual context. She proceeds to prise apart what Williams wants to say from what actually goes on in Aeschylus’s play. In ‘Bernard Williams on Truth and Plato’s Republic on Justice: What Are Genealogical Arguments Good For?’, Catherine Rowett compares Williams’s genealogical method in Truth and Truthfulness with Plato’s use of a similar method in the Republic: Plato, she argues, uses a naturalistic genealogy to explain and defend the value of justice in both political and individual ethics. Plato uses genealogy to ask not only whether justice pays, but also what justice is. This can help to avoid a certain form of reductionism that Rowett perceives in Williams’s approach. It also secures a good fit between lived practice and theoretical analysis. She ends by speculating why Williams did not include Plato’s Republic among his exemplars of naturalistic genealogies. In ‘The Invention of the Humanistic Discipline: Williams on Plato on Philosophy’, Marcel van Ackeren then examines Williams’s relation to Plato more broadly. Given how critical of Plato’s ethical views Williams was in Ethics and the Limit of Philosophy and Shame and Necessity, one might be forgiven for thinking that Williams took a dim view of Plato. But as van Ackeren shows, Williams [p. 36] Doing History Philosophically expressed great admiration for Plato in other places, and he offered an interpretation of Plato’s conception of philosophy that he seems to have endorsed to a remarkable extent. Van Ackeren highlights three aspects of this interpretation: those concerning the role of the dialogue form, philosophy’s aim to improve our lives, and the limits of philosophy. Through these reflections, Williams not only makes a lasting contribution to our understanding of Plato, but also helps us understand his own position. John Cottingham turns to Williams’s relationship to Descartes in ‘Pure Enquiry, the Absolute Conception, and Convergence: Bernard Williams in Dialogue with Descartes’. After exploring why Williams ended up writing an entire book about Descartes, of all people—​a committed theist who did not do much moral philosophy, thought of reason as a God-​given and unerring endowment, and was a system-​builder par excellence—​ Cottingham considers Williams’s account of Descartes’s project of ‘pure inquiry’. On this account, pure inquiry aims at a special kind of truth about how things are independently of our local ways of conceiving them; it aims at an ‘absolute’ conception of reality’—​a conception of the world ‘as it seems to God, and therefore as it really is’. After considering Williams’s argument that we need something like this conception if there is to be any knowledge, Cottingham evaluates Williams’s notably unCartesian attempt to defend the idea of an absolute conception without invoking God. Finally, Cottingham turns to the notion of convergence and Williams’s equally unCartesian suggestion that there is a radical asymmetry between the scientific and the ethical. In ‘Getting Round the Cartesian Circle’, Gerald Lang takes a critical look at Williams’s insightful solution to the notorious problem of the ‘Cartesian Circle’ in Descartes’s Meditations. Williams’s solution is to allow the meditator’s intuitions that the proofs of God’s existence are true to sustain a rule of ongoing acceptance of clear and distinct perceptions, even when these are not being experienced. This allows the meditator to go from momentary certainties to more durable items of knowledge and defeat scepticism. Lang quarrels with Williams’s solution and offers an alternative interpretation, however. He argues that Descartes’s principal worry lay in distinguishing between perceptions which are clear and distinct, and perceptions which merely seem clear and distinct. Descartes never questioned the epistemic status of the former class of perceptions, and the clear and distinct perceptions which matter to the cosmological argument for God can earn an exemption from sceptical doubt. Consequently, the Circle can be avoided. With Lorenzo Greco’s ‘A Humean Williams and a Williamsian Hume’, we come to Williams’s relationship to a figure who does not loom as large in Williams’s oeuvre as Plato or Descartes. Yet Greco shows that there are many elements in Williams’s approach—​from his early papers to his late reflection on genealogy—​that can be interpreted in a Humean light, revealing more than just a superficial resemblance. And if Hume is read in light of Williams’s philosophical concerns, he also appears to have been less of an optimist than Williams believed. Thus, Williams, [p. 37] Marcel van Ackeren and Matthieu Queloz although not a card-​carrying Humean, turns out to have more in common with the Scotsman than first appears. Paul Russell elaborates on this point in ‘Recasting Responsibility: Hume and Williams’. Williams found in Hume a powerful philosophical ally who likewise resisted the impositions of what Williams called ‘the morality system’. As Russell demonstrates, Hume and Williams share a commitment to providing a more ‘truthful’ and ‘realistic’ understanding of moral responsibility and our human ethical predicament, an understanding which echoes that of the ancient Greeks. Significant differences remain, however. They include Williams’s view that a naturalistic project of the kind that Hume pursues is of limited value when it comes to making sense of the morality system’s illusions about responsibility and blame. It is, Williams maintains, only when we consider moral responsibility in genealogical terms, which give attention to the importance of culture and history, that we can find a way of exposing the various prejudices and illusions of the morality system. In ‘The Predicament of Temporality: Williams’s Challenges to Kant’s Practical Reason’, Carla Bagnoli argues that Williams’s criticisms of Kant’s account of morality must be understood in light of their disagreement over the function of reason. The fundamental challenge they both respond to, she argues, is the tension between the temporal features of human agency and the allegedly categorical authority of some normative claims. This predicament is central to any theory of practical reason. For Kant, its roots lie in human embodiment, finitude, and fragility, and the remedy is the normative standard of reason, which plays a constitutive role in unifying the agent across time. For Williams, by contrast, mortality is a condition of the possibility of valuing life, and agential unity is both unfeasible and undesirable. Bagnoli assesses the respective merits of these contrasting views. Peter Kail, in ‘Genealogy: Williams, Hume, and Nietzsche’, critically examines the claims that Williams makes about the historical roots of his genealogical method in Truth and Truthfulness. In particular, Williams illustrates the principal features of his method by harking back to the work of Friedrich Nietzsche and David Hume. But Williams does not discuss these two thinkers in any great detail in this connection. Kail remedies this with a detailed discussion of how Williams’s conception of genealogy relates to the genealogical approaches to be found in Nietzsche and Hume. This leads Kail to contest some of the key claims that Williams makes about these two thinkers. In ‘Ethics, Untimeliness and Redlichkeit: On the Character of Williams’s Relationship to Nietzsche’, David Owen offers the most detailed study to date of Williams’s relationship to Nietzsche. He looks at the character and development of this relationship, and asks what difference Williams’s encounter with Nietzsche made to Williams’s philosophy. After sketching the ways in which Nietzsche appears in Williams’s thinking prior to Ethics and the Limits of Philosophy, Owen argues that this work marks Williams’s recognition of his affinities with Nietzsche, which come to fruition in the transition of Williams’s philosophy from a broadly [p. 38] Doing History Philosophically Humean to a distinctively Nietzschean sensibility. The key question, however, is how we should understand the character of Williams’s relationship to Nietzsche. Owen offers a nuanced answer by addressing Williams’s engagement with classical Greek literature and philosophy, with naturalism and genealogy, and with truth and truthfulness. In ‘The Sense of the Past: Williams and Collingwood on Humanistic and Scientific Knowledge’, Giuseppina D’Oro and James Connelly explore significant commonalities between Williams and R. G. Collingwood: in particular, their conception of what it takes to understand the past from a humanistic perspective and their commitment to the irreducibility of humanistic to scientific knowledge. D’Oro and Connelly argue that this defence of humanistic understanding against the threat of scientism aims to overcome the hegemony of scientific knowledge, but without endorsing a historicism that would be guilty of the reverse error, reducing all knowledge to humanistic knowledge. Williams and Collingwood attempt to tread a middle path between the view that all knowledge is at bottom scientific knowledge and the converse view that all knowledge is at bottom historical knowledge. As D’Oro and Connelly point out, however, this attempt was widely misunderstood: Williams was accused of identifying scientific knowledge with an ‘absolute’ conception of reality, while Collingwood was charged with relativizing all knowledge, including scientific knowledge. Yet their philosophical contribution, D’Oro and Connelly argue, lies precisely in outlining a path between scientism and historicism. In ‘Williams’s Debt to Wittgenstein’, Matthieu Queloz and Nikhil Krishnan argue that several aspects of Williams’s style, methodology, and metaphilosophy can be read as evolving dialectically out of Wittgenstein’s own. After considering Wittgenstein’s stylistic influence on Williams, especially regarding ideals of clarity, precision, and depth, they examine Williams’s methodological debt to Wittgenstein—​including his anthropological interest in thick concepts and their point. Williams’s explicit association, in the 1999s, with ‘Left Wittgensteinianism’ is then shown to be not a sudden conversion, but the product of a longstanding critical engagement with Wittgenstein’s methodology and metaphilosophy: Williams reaches this position by envisaging a Wittgensteinianism that thinks in concrete sociohistorical terms, embraces genuine explanation, and relinquishes its insistence on the purity of philosophy. Moreover, this critique turns out to be continuous with Williams’s advocacy of philosophy as a humanistic discipline. Finally, Queloz and Krishnan show that Williams inherits from Wittgenstein a certain understanding of how philosophy can help us to live. The volume then continues with three chapters foregrounding Williams’s methodological reflections on the history of philosophy. In ‘Why Bernard Williams Is a Bad Example for Historians of Philosophy’, John Marenbon argues that Williams shifted, in the course of his career, from recommending the ‘rational reconstruction’ of historical texts to advocating a subtler [p. 39] Marcel van Ackeren and Matthieu Queloz method, whereby philosophy from the past makes the familiar strange and the strange familiar. Yet Williams never abandoned his distinction between historians of philosophy and historians of ideas, Marenbon observes, and so never accepted that understanding philosophy from earlier times on its own terms is a genuinely philosophical activity. Nor did Williams question the fundamental assumption that history of philosophy is valuable only because of its value to real, first-​order philosophers. Marenbon concludes that the resulting method should not be imitated, and that it commits one to an excessively restricted view of what counts as philosophy, which leaves out most of the broad tradition of western philosophy. Ralph Wedgwood, in ‘The Iniquity of Oblivion’, starts out from Williams’s distinction between two ways of studying the philosophy of the past—​the history of ideas and the history of philosophy. Wedgwood then examines the point of the history of philosophy thus conceived. He considers two possible answers to this question. The first focuses on the project of rescuing philosophical ideas from being undeservedly forgotten, while the second focuses on Williams’s idea of a genealogy of our concepts. Wedgwood finds that the first answer is importantly true, and the kind of history of philosophy that it recommends is a crucial part of our discipline; by contrast, he concludes that the second answer fails to provide a sufficiently secure basis for the study of the history of philosophy. In ‘Williams, Berlin, and the Vindication Problem’, Garrett Cullity invites us to read Williams as responding to the fact that the ethical questions we face, and the resources we have for answering them, are historically contingent and local. There is no ‘Archimedean point’ from which to seek universal answers to ethical questions about the living of a human life, on Williams’s view. This presents both a philosophical and practical challenge: How can a reflective awareness of the cultural contingency of ethical thought be reconciled with a commitment to the seriousness of ethical questions, and the objectivity that this seriousness requires us to attribute to them? Cullity probes Williams’s own response to that challenge, and makes the case for a different but related response, which draws on themes from one of Williams’s intellectual mentors, Isaiah Berlin. The volume ends with four chapters examining Williams’s contention that philosophy should itself be done historically. In ‘Serpents in the Genealogical Garden of Eden: Why Williams’s Genealogy Is Excessively Historicist and Insufficiently Historical’, Hans-​Johann Glock seeks to put Williams’s influential contribution to the movement of ‘analytic genealogy’ into perspective. First, Glock argues that Williams’s genealogical method is not immune to the anti-​genetic objection that both the content and the validity of most concepts depends on their function rather than their origin. Secondly, Glock criticizes the way in which Williams combines ‘imaginary’ and ‘historical genealogy’. Thirdly, he urges that genealogy must be supplemented and controlled by non-​ genealogical philosophical analysis and non-​philosophical theories about human nature and the evolutionary genesis of practices. The attractions of genealogy need [p. 40] Doing History Philosophically to be balanced against the abiding merits of conceptual analysis, Glock cautions. Moreover, he sees reason to hope that recent advances in evolutionary theory and biological anthropology might replace state-​of-​nature fictions with testable scientific hypotheses. In ‘Internal Reasons and Historical Thinking’, Geraldine Ng explores in what sense Williams can be read as a historicist. She clarifies how Williams’s ethical philosophy is committed to acknowledging the contingency and complexity of ethical life, and how, in the manner of R. G. Collingwood, Williams takes the task of appraising present and past agents to involve historical understanding. She then argues that Williams’s strategy for defending ethical knowledge in relation to ‘thick’ concepts is of a piece with a historicist account of the normative force of internal reasons. Finally, she suggests that this reconsideration of reasons internalism yields a robust response to the common objection that Williams’s ethical philosophy is merely negative. Williams’s ethical philosophy is not merely negative, because his philosophical method is not merely analytic. Rather, in light of his earlier, implicit ‘historicist turn’, Williams’s ethical philosophy emerges as historicist and positive. Amanda Greene and Ilaria Cozzaglio then turn to the role of the historical perspective in Williams’s political philosophy. In ‘The Art of the Possible: Williams on Political Judgement and the Historical Perspective’, they show that Williams places great emphasis on history when it comes to making judgements about political actors and political orders. At the same time, Williams combines this concern for historical context with universal considerations, such as drawing a distinction between order and tyranny. Greene and Cozzaglio show that this duality of contextualism and universalism is anchored in a respect for the limits of politics. They find in Williams a challenge to the common view that judgement calls for philosophical analysis whereas action calls for political analysis. According to Williams, political judgements and political actions each require both a philosophical and a political sensibility. Only then can political critique have a point and make a difference. Finally, Greene and Cozzaglio propose that ‘answerability’ to a historically situated audience is the distinctive trait of Williams’s approach to political philosophy. Finally, Miranda Fricker brings the volume to a close with ‘A Project of “Impure” Enquiry: Williams’s Historical Self-​ Consciousness’, which considers how Williams’s abiding interest in the borderlands between Philosophy and History shape his philosophy. As Fricker notes, Williams accused moral philosophy, and particularly moral theory, of overstepping the boundary marking the real ‘limits’ of the discipline; and in his later work, he explicitly advanced the idea of doing ‘impure’ philosophy, by which he meant philosophy that mixes itself with history. Through an examination of the complex impression left on Williams’s historical self-​consciousness by his engagements with Descartes and Wittgenstein, Fricker identifies several ways in which philosophy and history are closely intertwined for Williams. This allows her to draw out his positive vision of ‘impure’ philosophy—​a [p. 41] Marcel van Ackeren and Matthieu Queloz philosophical style he took to contribute to nurturing philosophy as a humanistic discipline. A project of this nature, aiming to bring new perspectives to a major philosopher’s work, relies heavily on the wisdom of others. In this, we have been exceptionally fortunate to benefit from the help of Patricia Williams and Adrian Moore. Their steadfast encouragement has been a source of great motivation, while their generous counsel has improved the volume at many points. We hope the resulting collection shows how Williams himself can serve as the kind of ‘untimely’ figure he championed—​one whose thought can act against our age for the benefit of the next. References Barnes, Jonathan. 2011. Method and Metaphysics: Essays in Ancient Philosophy I. Edited by Maddalena Bonelli. Oxford: Oxford University Press. Grice, Paul. 1989. Studies in the Way of Words. Cambridge, MA: Harvard University Press. Queloz, Matthieu. 2017. ‘Does Philosophy Have a Vindicatory History? Bernard Williams on the History of Philosophy’. Studia Philosophica 76: 137–​52. Queloz, Matthieu. 2023. ‘Making Past Thinkers Speak to Us through Pragmatic Genealogies’. In Historiography and the Formation of Philosophical Canons. Edited by Sandra Lapointe and Erich Reck, 171–​91. New York: Routledge. van Ackeren, Marcel. 2018. ‘Philosophy and the Historical Debate: A New Debate on an Old Topic’. In Philosophy and the Historical Perspective. Edited by Marcel van Ackeren with Lee Klein, 1–​17. Oxford: Oxford University Press. van Ackeren, Marcel. 2019. ‘Williams (on) Doing History of Philosophy: A Case Study on Ethics and the Limits of Philosophy’. In Ethics beyond the Limits: New Essays on Bernard Williams’ Ethics and the Limits of Philosophy. Edited by Sophie Grace Chappell and Marcel van Ackeren, London: Routledge. Williams, Bernard. 1969. ‘Philosophy’. In General Education: A Symposium on the Teaching of Non-​specialists. Edited by Michael Yudkin, 138–​64. Harmondsworth: Penguin. Williams, Bernard. 1982. ‘The Spell of Linguistic Philosophy: Dialogue with Bernard Williams’. In Men of Ideas: Some Creators of Contemporary Philosophy. Edited by Bryan Magee, 110–​24. Oxford: Oxford University Press. Williams, Bernard. 1993. Shame and Necessity. Berkeley: University of California Press. Williams, Bernard. 2001. ‘Liberalism and Loss’. In The Legacy of Isaiah Berlin. Edited by Mark Lilla, Ronald Dworkin, and Robert Silvers, 91–​103. New York: New York Review of Books. Williams, Bernard. 2002. Truth and Truthfulness: An Essay in Genealogy. Princeton: Princeton University Press. Williams, Bernard. 2005a. Descartes: The Project of Pure Enquiry. London: Routledge. Williams, Bernard. 2005b. ‘From Freedom to Liberty: The Construction of a Political Value’. In In the Beginning Was the Deed: Realism and Moralism in Political Argument. Edited by Geoffrey Hawthorne, 75–​96. Princeton: Princeton University Press. Williams, Bernard. 2006a. ‘Descartes and the Historiography of Philosophy’. In The Sense of the Past: Essays in the History of Philosophy. Edited by Myles Burnyeat, 257–​ 64. Princeton: Princeton University Press. Williams, Bernard. 2006b. ‘An Essay on Collingwood’. In The Sense of the Past: Essays in the History of Philosophy. Edited by Myles Burnyeat, 341–​60. Princeton: Princeton University Press. Williams, Bernard. 2006c. ‘Philosophy as a Humanistic Discipline’. In Philosophy as a Humanistic Discipline. Edited by A. W. Moore, 180–​99. Princeton: Princeton University Press. [p. 42] Doing History Philosophically Williams, Bernard. 2006d. ‘Political Philosophy and the Analytical Tradition’. In Philosophy as a Humanistic Discipline. Edited by A. W. Moore, 155–​68. Princeton: Princeton University Press. Williams, Bernard. 2009. ‘A Mistrustful Animal’. In Conversations on Ethics. Edited by Alex Voorhoeve, 195–​214. Oxford: Oxford University Press. Williams, Bernard. 2014a. ‘The English Moralists, by Basil Willey’. In Essays and Reviews 1959–​ Edited by Michael Wood, 52–​5. Princeton: Princeton University Press. Williams, Bernard. 2014b. ‘Why Philosophy Needs History’. In Essays and Reviews 1959–​2002. Edited by Michael Wood, 405–​12. Princeton: Princeton University Press. [p. 46] Psychology, Ethics, and ‘Ethicized Psychology’ Bernard Williams on Greek Thought and Greek Philosophy Terence Irwin 1. The Errors of Philosophy Bernard Williams’s reflexions on Greek ethical thought extend beyond Greek philosophy. In his view, certain selected literary sources—​especially Homer and the tragedians, but also Thucydides—​express ethical views that are preferable to those we find in Plato and Aristotle. In the Greek philosophers we find views that are in some ways similar to modern views that Williams opposes. While the philosophers share some of the basic errors (as Williams conceives them) that we find in modern views, we can identify these errors partly by recognizing that the earlier Greeks avoided them. These earlier Greeks show that we can develop a complex and, in certain respects, attractive account of ethics without falling into the errors of the philosophers. Once we see that the philosophical framework is dispensable, we may be encouraged to reject it. Williams expresses this view in 1981, in his survey of Greek philosophy. If there are features of the ethical experience of the Greek world which can not only make sense to us now, but make better sense than many things we find nearer to hand, they are not all to be found in its philosophy. Granted the range, the power, the imagination and inventiveness of the Greek foundation of Western philosophy, it is yet more striking that we can take seriously, as we should, Nietzsche’s remark: ‘Among the greatest characteristics of the Hellenes is their inability to turn the best into reflection.’ (1981b: 253) He returns to this theme in 1993, in Shame and Necessity, which argues at length for the unfavourable comparison of the philosophers with earlier Greek writers. His arguments could be amplified by reference to his other works on relevant topics, but I will mostly refrain from pursuing these themes into Williams’s other works. I will concentrate on one argument in Shame and Necessity. Williams argues that Terence Irwin, Psychology, Ethics, and ‘Ethicized Psychology’ In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. [p. 47] 34 Terence Irwin the pre-​Platonic Greeks lack a certain kind of belief in a will, and he explains why he thinks they were better off without Plato’s invention. I will ask whether they really lacked this belief, and whether they should have avoided it.1 My discussion is selective and critical, in so far as it gives reasons for disagreement with one of Williams’s conclusions. But it is also meant to be appreciative, in so far as it follows Williams’s approach to the sources. I agree with his view that literary and historical argument can be appropriately combined with philosophical argument, to throw light on philosophical questions that reasonably concern us no less than they concerned Greek writers.2 2. Will and Freedom To introduce some of the questions that concern Williams, it is useful to consider a contrasting judgement about the contribution of Greek philosophers to the understanding of some of these questions. When Dante is more than half way through Purgatory, he asks Virgil to explain the relation between rational agency, will, and freewill.3 «Le tue parole e ’l mio seguace ingegno», /​rispuos’ io lui, «m’hanno amor discoverto, /​ma ciò m’ha fatto di dubbiar più pregno; /​ché, s’amore è di fuori a noi offerto /​e l’anima non va con altro piede, /​se dritta o torta va, non è suo metro.» ‘Your words and my following wit’, I answered, ‘have made love clear to me, but that has made me more full of questions. For if love is offered to us from outside, and if the soul moves with no other feet, it has no merit whether it goes right or wrong.’ (Purgatorio 18. 40–​6) Virgil’s discussion of love makes Dante wonder how much is in an agent’s control. The primary and necessary object of the rational will is the ultimate good, and our pursuit of this ultimate good is not in our control. But all our rational desires and actions aim at the ultimate good. How, then, can our actions be in our control? Virgil agrees with one of Dante’s assumptions. The primary direction of the will to the ultimate good is natural, and not subject to choice. 1 A perceptive discussion of Shame and Necessity is Gisela Striker’s review, ‘Are We Any Better?’ I discussed some aspects of when it first appeared, in a Critical Notice (Irwin 1994). I have used some parts of that discussion in this essay. 2 I agree with the excellent and appreciative discussions of Williams’s work by Burnyeat (2006) and Long (2007: ch. 6). 3 I use ‘will’ to render ‘voluntas’ (‘voglia’ in Dante), and ‘freewill’ (one word) to render ‘liberum arbitrium’. I will discuss the relations between the two, later in this section. I use ‘free will’ (two words) when I am not speaking specifically of liberum arbitrium. [p. 48] che sono in voi sì come studio in ape /​di far lo mele; e questa prima voglia /​ merto di lode o di biasmo non cape. They are in you just as zeal to make honey is in bees. And this first will allows no meriting of praise or blame. (58–​60, emphasis added) But he argues that this fact about the rational will does not exclude our control over our actions, because the way in which we direct our other volitions to the will for the ultimate good is up to us. Or perché a questa ogn’ altra si raccoglia, /​innata v’è la virtù che consiglia, /​e de l’assenso de’ tener la soglia. /​Quest’ è ’l principio là onde si piglia /​ragion di meritare in voi, secondo /​che buoni e rei amori accoglie e viglia. /​Color che ragionando andaro al fondo, /​s’accorser d’esta innata libertate; /​però moralità lasciaro al mondo. /​Onde, poniam che di necessitate /​surga ogne amor che dentro a voi s’accende, /​di ritenerlo è in voi la podestate. (61–​72, emphasis added) In order that every other will may be conformed to this [first will], there is innate in you the power that counsels and that ought to hold the threshold of assent. This is the principle in which the reason for merit is found, according to how it gathers and winnows good and guilty loves. Those who in their reasoning went to the foundation recognized this innate freedom, and therefore they left morality to the world. Whence, admitting that every love that is kindled in you arises by necessity, the power to control it is in you.4 This, says Virgil, is what is meant by freewill (‘libero arbitrio’, 74). Though the innate will for the ultimate good, which is not subject to praise and blame (‘e questa prima voglia merto di lode o di biasmo non cape’), we also have the power to direct our actions better and worse in relation to this ultimate good. This power operates through deliberation and assent (‘innata v’è la virtù che consiglia, e de l’assenso de’ tener la soglia’). It is this power of deliberation and action that makes us subject to praise and blame (‘Quest’ è ’l principio là onde si piglia ragion di meritare in voi’). The ancient philosophers who got to the root of this question understood the basis of morality (‘Color che ragionando andaro al fondo, s’accorser d’esta innata libertate; però moralità lasciaro al mondo’). Virgil’s account of the connexion between rational will and freedom agrees with Aquinas, who argues that freewill results from our being rational agents who pursue our ultimate good. 4 Translation mostly derived from Singleton 1970–​5. [p. 49] 36 Terence Irwin Some philosophers maintained that man's will is necessarily moved to elect a thing. . . . But that opinion is heretical, for not everything necessary . . . the principle of which is outside: Arist. Nic. Ethics III, 1 (1118a1) and (1110b15) as is said e.g. Q. D. On Truth q. 22, a. 5 in the Response. for it destroys the reason for merit and demerit in human acts: for it does not seem to be meritorious or demeritorious for someone of necessity to do what he cannot avoid. It is even counted among extraneous opinions in philosophy, because not only is it contrary to faith, but it subverts all the principles of moral philosophy. For if nothing is within our power, but we are necessarily moved to will, then deliberation, exhortation, precept, and punishment, and praise and blame, with which moral philosophy is concerned, are nullified. (De Malo, q6 a1). quidam posuerunt, quod voluntas hominis ex necessitate movetur ad aliquid eligendum . . . Haec autem opinio est haeretica: tollit enim rationem meriti et demeriti in humanis actibus. Non enim videtur esse meritorium vel demeritorium quod aliquis sic ex necessitate agit quod vitare non possit. Est etiam annumeranda inter extraneas philosophiae opiniones: quia non solum contrariatur fidei, sed subvertit omnia principia philosophiae moralis. Si enim non sit liberum aliquid in nobis, sed ex necessitate movemur ad volendum, tollitur deliberatio, exhortatio, praeceptum et punitio, et laus et vituperium, cir ca quae moralis philosophia consistit. Though Aquinas does not say here, as Dante says, that the ancient philosophers recognized this connection between freedom and morality, we can see from his discussions of Aristotle that he agrees with Dante. In the present passage he insists that the denial of freewill is not just a theological error, but also a fundamental error in moral philosophy. Both Dante and Aquinas consider two conditions of the will: (1) Our willing the ultimate end, in respect of which we are not free. (2) The direction of the will to actions that result from deliberation, in respect of which we are free. The second condition of the will is freewill. Though the Latin (and Italian) terms are different (‘voluntas’, ‘liberum arbitrium’), the English rendering ‘freewill’ captures the connection that Dante and Aquinas see between will and freedom.5 These passages from Dante and Aquinas give us a brief statement of some views that Williams regards as basic mistakes in moral philosophy and about the history of philosophy. According to Dante, the ancient philosophers explained the possibility of morality by showing that we have freewill. In Williams’s view, this judgement on the ancient philosophers contains two mistakes: (1) In his view, the ancient philosophers, at least if we consider Plato and Aristotle, do not show that 5 On liberum arbitrium, see Irwin (2007: S235n3, §265). [p. 50] we have freewill. They are not even interested in this question. (2) Freewill is not necessary for the possibility of morality; we can, and should, think about ethics without reference to freewill, and Greek thought shows us how to do this. My description of the second error may seem to be unfair to Williams, because I have spoken of morality and of ethics as though they were identical. Williams, however, maintains that they are different. He believes that Greek thought is about ethics, but the ‘peculiar institution’ of morality belongs to a Kantian outlook in which freedom, praise, blame, and morality are connected. Given this distinction, he is not really disagreeing with Aquinas; he is only disagreeing with Dante. This conclusion is mistaken. When these medieval writers speak of ‘morality’ and of ‘moral philosophy’, they are simply putting into Latin what Aristotle says in Greek when he speaks of ‘ethics’ (êthika) and ‘virtue of character’ (êthikê aretê, virtus moralis). In Williams’s terms, they claim that free will is necessary for ethics, and Dante claims that the Greek philosophers showed that this is so. In Williams’s view, the medievals are mistaken about the Greek philosophers, who do not connect freewill with ethics. This is because the Greeks do not introduce freewill into discussions about responsibility at all. The belief in some metaphysical fact that supports true judgements about responsibility has no place in Greek thought, including Greek philosophy (1993: 68). Both compatibilist and incompatibilist accounts of the relation of responsibility to causal determination are irrelevant to Greek views (1993: 215n42). I will eventually come back to Williams’s claims about responsibility and freewill. But first I need to discuss a prior question, about the place of a conception of the will in Greek thought. In his view, Plato and Aristotle have a conception of the will. but no belief in freewill; hence they do not rely on will and freewill to explain responsibility, even though a conception of the will was available for use in explaining free will. Earlier Greek sources, however, have no conception of will, and no conception of freewill; hence a conception of will is not available, as it is in Plato and Aristotle, for the explanation of responsibility. Pre-​Platonic sources explain human action, including action for which we are responsible, without reference to will. This claim needs to be explained and evaluated. Before I go any further, some of the limits of Williams’s discussion and mine need to be noticed. The Greek philosophers he has in mind are Plato and Aristotle, but no one after them. The non-​philosophical sources are earlier than Plato or contemporaries of Plato, and among these sources Homer and the tragedians are discussed most fully. The restrictions on the range of sources sometimes affect the cogency of his arguments, but I will mostly stick to the sources that Williams discusses.6 6 On the temporal limits of Williams’s discussion, see section 8 below. [p. 51] 38 Terence Irwin 3. Will and Ethicized Psychology To form a clearer idea of what is missing in earlier Greek thought, we should explore the conception of the will that Williams attributes to Plato. He distinguishes two conceptions of the will: 1. The will, in contrast to other motives, is essentially motivated by morality, and especially by moral duty (1993: 41). 2. The will is one of those ‘basic operations of the mind’ that need to be ‘classified in ineliminably ethical terms’ (1993: 46). The first of these two conceptions of the will is sometimes attributed to Kant. Williams attributes the second to Plato. There is a less extreme, and historically more important, idea that links notions of action and effort on the one hand to moral or ethical ideas on the other. That is the idea that the basic theory of action itself, the account of what human beings are and how they do anything, is a theory that must be expressed in ethical terms. This is not merely the idea that there must be a psychology of ethics—​that is to say, an account of our psychology inasmuch as we have ethical dispositions, beliefs, and feelings: that is obviously correct, but it is not the idea in question. The idea is rather that the functions of the mind, above all with regard to action, are defined in terms of categories that get their significance from ethics. This is an idea that is certainly lacking in Homer and the tragedians. It was left to later Greek thought to invent it, and it has scarcely gone away from us since. It was invented, it seems, by Plato. (1993: 42) Both conceptions are absent (in Williams’s view) from Homer and from other pre-​Platonic sources. Though Homer has a lot to say about will and action, he lacks both of these conceptions of the will. In Willams’s view, Homer is better off without them. Though Williams’s conception of an ethicized psychology is not explained fully, we can see what he has in mind from his comments on Plato’s partition of the soul. He argues that Plato wants to explain conflicts between ‘rational concerns that aim at the good’ and ‘mere desire’ (1993: 42). It is only in the light of ethical considerations, and certain ethically significant distinctions of character and motive, that Plato’s schema is intelligible. In particular, it is not enough, for the rational part of the soul to be involved, that rational capacities have been exercised in arriving at a course of action, as they may be in finding a way of carrying out some desire—​that would merely involve the superior, rational, part in the conflicts of the desiring part. Reason operates as a [p. 52] distinctive part of the soul only to the extent that it controls, dominates, or rises above the desires. (1993: 43) We can gain more light on Williams’s conception from his remark that ethicized psychology persists in Aristotle and other Peripatetics.7 The Peripatetics, in their usual patronising way, commended Plato for having discovered that there were nonrational elements in the soul, and they remodelled his divisions in a much less dramatic and more realistic direction. But the basic involvement of ethical categories in the theory of action had come to stay. Aristotle’s most famous contribution to these topics is his discussion of akrasia—​a term usually translated either as ‘weakness of the will’ or ‘incontinence’ (neither, for different reasons, very fortunate). His definition of this condition is entirely shaped by ethical interests: ‘The akratês knows that what he is doing is bad, but does it through passion; the enkratês, knowing that the desires in question are bad, does not follow them, because of reason.’ (1993: 44) Williams’s remark on incontinence draws our attention to Aristotle’s treatment of rational desire, which agrees with Plato on the point that Williams rejects. Aristotle recognizes a distinctively rational desire, boulȇsis, which aims at the good. A decision (prohairesis) is a desire resulting from deliberation about how to fulfil a rational wish (boulêsis), not from just any deliberation and desire.8 This connection between boulêsis, reason, and the good helps to explain why Aquinas treats Aristotle’s remarks about boulêsis as remarks about the will (voluntas). Williams does not suppose that an ethicized psychology is the only alternative to a value-​free psychology. He explains: ‘The criticism of Plato and Aristotle is not so much that they took psychological explanation to involve values as that they incorporated into it a particular set of ethically and socially desirable values’ An ethicized conception of the self claims that ‘the functions of the mind, especially with regard to action, are defined at the most basic level in terms of categories that get their significance from ethics’ (1993: 160). 4. Ethicized Psychology in Plato We might take Williams to mean that Platonic and Aristotelian psychology is ethicized because it incorporates specific ethical principles; perhaps this is what he means in speaking of ‘a particular set of ethically and socially desirable values’. Plato would hold this sort of ethicized conception if, for instance, he supposed that 7 Williams cites a passage from the Magna Moralia, which he takes (questionably) to be spurious. 8 This feature of boulêsis is emphasized by Anscombe (1965). [p. 53] 40 Terence Irwin the rational part of the soul essentially aims at (e.g.) the virtue of justice. According to this supposition, it is essential to human beings that they want to be just, in the sense of ‘just’ that includes concern for the good of others. Any who entirely lacked this concern would not have a rational part in their souls, and would not be rational agents. Williams does not say that this is how Plato conceives rational desire. If he did say this, he would be mistaken. For Plato believes that many people have a rational part in their soul, but do not care about being just or about doing just actions. Such people present Plato with his main question in the Republic; in Books I–​II we learn that many people do not care about being just, but only about appearing just. If we reply that these people have souls with no rational part, we have to explain why Books VIII–​I X describe people who subject their rational part to non-​rational desires in different ways; if they subject their rational part, they must have one, even when it has been subjected to other parts. Despite these objections, we might still say that the rational part, according to Plato, is ethicized because, if it were not subjected to a non-​rational part, but could be guided by reason without non-​rational distractions, it would eventually want to be just. In this respect, we might say, justice is its destination. Williams describes Plato’s view in these terms in an earlier essay: The Platonic aim, then, can be seen as this, to give a picture of the self such that if one properly understands what one is, one will see that a life of justice is not external to the self, but an objective which it must be rational for one to pursue. b: 247) If Plato believes this, he takes something other than the concern for justice to be a more fundamental characteristic of the rational part. If justice is the destination, we want to know why the soul sets out on this journey, and why arrival at any other terminus than justice is the result of some mistake. We can explain these features of the rational part if we suppose that it is fundamentally concerned with one’s good as a whole.9 If we adapt Plato’s political analogy, the rational part is impartially concerned with the good of the whole soul, not with the aims of a single part, and with one’s good over time, not with the fulfilment of this or that desire at this or that time. We may describe this as the holistic concern of the rational part. Since the rational part has this holistic concern, and since, in Plato’s view, justice fulfils this holistic concern, justice is the destination, and those who do not choose justice are mistaken. Non-​rational impulses and 9 One might argue that the aim of the rational part is psychic harmony. This is not an alternative to holistic concern. The type of agreement between the parts that the rational part seeks is not just any sort of agreement, but agreement with the rational part in its pursuit of the good of the whole soul and each part. [p. 54] short-​term advantages may persuade us that we have no reason to care about being just, but we are mistaken if we are persuaded by these considerations. For present purposes, we need not ask whether Plato is right to hold that the destination of the rational part, so conceived, is justice. We can confine ourselves to inquiring into the holistic concern that is supposed to guide us to this destination. Having seen that Plato attributes this holistic concern to the rational part, we can reasonably attribute the same conception of the rational part to Aristotle. Rational desire does not aim simply at this or that good, but at the greater good, understood as greater over time (De Anima 434a5–​10) and over the different aspects of the self. Aristotle describes this greater good as the human good (to anthrôpinon agathon) or happiness (eudaimonia). The non-​rational aspects of different virtues of character allow practical reason to do its proper work, and thereby to discover that the life of virtue, including justice, is the one that achieves the human good. 5. What Is Wrong with Ethicized Psychology? After this survey of the Platonic and Aristotelian conception of the rational part, we can return to Williams’s objections to ethicized psychology. If Plato held that the rational part essentially aims at justice, we could see why this psychology is objectionably ethicized, and in what respect Plato holds that ‘the functions of the mind . . . are defined at the most basic level in terms of categories that get their significance from ethics’ (1993: 160). If we attribute an essential pursuit of justice to the rational part, we make it difficult to see what the various unjust people whom Plato describes have in common with just people. They seem to have rational desires, but to disagree with just people about the proper destination of these rational desires. Plato supposes correctly that these souls also have rational parts. To explain where they go wrong, we need to describe what they have in common with just souls, and this is what Plato tries to do. If the ‘categories that get their significance from ethics’ were such categories as justice, Plato’s ethicized psychology would be open to objection for its descriptive inadequacy. What should we say, however, about the attribution of holistic concern to the rational part? Does this count as ethicized psychology of the sort that Williams rejects? We may be inclined to answer No. Perhaps holistic concern is too indeterminate to qualify as a category that gets its significance from ethics. We could not identify holistic concern if we had no conception of an overall good; if ‘good’ is a category that gets its significance from ethics, then it introduces an ethicized psychology. But we may reasonably doubt whether ‘good’ is such a category; we can identify what is good and bad for different organisms without being concerned about any ethical goodness, if that is the sort of goodness that belongs to ethical virtues and to virtuous actions. Since Plato speaks of good and bad conditions of [p. 55] 42 Terence Irwin knives and of animal and human bodies, he does not seem to treat goodness as an essentially ethical property. Williams, however, seems to treat holistic concern as an aspect of ethicized psychology. When he describes a practical function of reason that does not require an ethicized psychology, he says: ‘rational capacities have been exercised in arriving at a course of action, as they may be in finding a way of carrying out some desire’. The ethicized conception of the rational part is involved ‘to the extent that it [sc. reason] controls, dominates, or rises above the desires’ (1993: 45). This form of expression suggests that, according to Plato, no desires belong to the rational part; if some desires belonged to it, the rational part could hardly control, dominate, or rise above, all desires. Plato, however, attributes some desires to the rational part (Resp. 580d7–​9), and Aristotle attributes them to it, in the respect that wish (boulêsis) is a desire (orexis) that belongs to the rational part. If we correct Williams’s form of expression on this point, we can take him to say that Plato and Aristotle accept an ethicized psychology to the extent that they believe that the rational part and its desires control, dominate, or rise above, the desires of the non-​rational parts. Perhaps Williams believes that the holistic conception is to be rejected because it takes holistic concern to be rationally superior, and is ethical in this respect. Plato supposes that the aims of the rational part are superior not because they are always psychologically dominant, but because they claim rational authority.10 Holistic concern ‘rises above’ non-​rational desires in so far as it aims at a good that they do not conceive. It seems, then, to belong to the sort of psychological description that Williams rejects. It does not exactly control or dominate desires of the rational part, but it forms them by providing them with the sort of object to which they respond. A true view about rationality, according to Williams, asserts that ‘rational capacities have been exercised in arriving at a course of action, as they may be in finding a way of carrying out some desire’. We may loosely describe this function of reason as instrumental.11 Hence we may attribute to Williams the claim that the only practical function of reason is instrumental. This may be an unwarranted generalization from the example that Williams offers; perhaps he intends it only as an example of a non-​ethicized conception of practical reason, without implying that it is the only example. But, since it is the only example he offers, and since he does not say how we might recognize other examples, we may reasonably proceed on the assumption that he does not believe there are any other examples. It is not clear, however, that holistic concern involves a claim about superiority that is essentially ethical. Admittedly, Plato believes that justice is the destination of the rational self, ‘an objective that it must be rational for one to pursue’ 10 This use of ‘authority’ is derived from Butler, Sermons ii 14. 11 Loosely, because Williams has more to say. See Williams (1981a). [p. 56] b: 247), as Williams puts it. Let us grant, for the sake of argument, that Plato and Aristotle have no good reason to believe that justice is the destination. One might apparently grant this while still believing they are right to attribute holistic concern to rational desire. To see that one might (rightly or wrongly) attribute holistic concern without agreeing that justice is the destination, we may consider Sidgwick’s ‘dualism of practical reason’. Sidgwick defends the view that justice is the destination, since he believes that a good argument can be offered to show that some axioms of practical reason support morality, in its utilitarian version. But though he believes the argument from practical reason to morality is a good argument, he does not believe it is conclusive. On the contrary, he believes that a form of egoism can plausibly claim to be the destination of practical reason, so that there is insufficient reason to go beyond egoism to morality. In Sidgwick’s view, we have a good, but non-​conclusive case both for egoism and for morality, and we have no good case for taking a third position that would judge their relative claims. However good or bad Sidgwick’s arguments may be, he begins from a holistic conception of practical reason that agrees with Plato’s belief that the rational part considers and pursues the good of the whole soul. Hence he accepts Platonic holistic concern without agreeing that justice is the destination. If we reject holistic concern, we need some further reason for rejecting it beyond its alleged association with the claim that justice is the destination. From this discussion of ethicized psychology and the Platonic conception of the will, we can draw some conclusions: 1. Holistic concern is plausibly attributed to the rational part of the Platonic soul. 2. It is not clear that this conception of the rational part is justifiably described as ‘ethicized’ (in the sense of relying on concepts that get their sense from ethics). 3. If, however, we construe ‘ethical’ broadly enough to make holistic concern an ethicized conception, we may doubt whether that is a good objection to it. 4. If, then, we want to know whether earlier Greek writers accept a Platonic conception of the will, we should look to see whether they recognize holistic concern, or they confine rational capacities to the task of finding ways to carry out desires. 6. Ethicized Psychology in Homer? Williams believes not only that Plato’s account of rational desire is illegitimately ethicized, but also that it is a regrettable innovation in Greek thought. In his view, this is one case in which the best of Greek ethical thought is not captured in Greek philosophy. Williams does not mean merely that pre-​Platonic Greeks did not [p. 57] 44 Terence Irwin formulate Plato’s description of reason and desire. He also means that their implicit views on reason and desire were better than Plato’s because they did not include Plato’s holistic view of practical reason. While they recognized that ‘rational capacities have been exercised in arriving at a course of action’, they did not attribute holistic concern to practical reason. In their view, then, there is nothing distinctively rational about holistic concern. Homer provides an example to show that how earlier (i.e. pre-​Platonic) Greek thought does without a conception of will. Williams argues that a Homeric agent has a conception of himself as a unified self distinct from a collection of psychological forces or organs (however thumos, phrenes, and so on are conceived).12 A Homeric agent is a unified self, and regards himself this way, and, in one respect at least, he expresses his will in his choices and actions. This will, however, is not the Platonic ethicized will, which essentially includes holistic concern. To evaluate Williams’s claim, let us recall that Plato cites Homer to illustrate the difference between the rational soul and other desires (Resp. 441bc; Phaedo 94de). He mentions Odysseus’ internal discussion and conflict to show how ‘what has reasoned about the better and worse’ is different from ‘what is unreasonably angry’ (Resp. 441c1–​2). There Odysseus, pondering in his spirit evil for the suitors, lay sleepless. And the women came out from the hall, those that had before been accustomed to lie with the suitors, making laughter and merriment among themselves. And the spirit stirred in his breast, and much he debated in mind and spirit, whether he should rush after them and deal death to each, or allow them to lie with the insolent suitors for the last and latest time; and his heart growled within him. And as a bitch stands over her tender whelps growling, when she sees a man she does not know, and is eager to fight, so his heart growled within him in his wrath at their evil deeds; but he struck his breast, and rebuked his heart, saying: ‘Endure, my heart; a worse thing even than this you once endured on that day when the Cyclops, irresistible in strength, devoured my stalwart comrades; but you endured until your wit got you out of the cave where you thought to die.’ So he spoke, chiding the heart in his breast, and for him in utter obedience his heart remained sternly enduring; but he himself lay tossing this way and that. And as when a man before a big blazing fire turns this way and that a paunch full of fat and blood, and is very eager to have it roasted quickly, so Odysseus tossed from side to side, pondering how at last he might lay his hands on the shameless suitors, one man against so many. (Od. 20.5–​30)13 12 These disputes arise from Snell’s view that a Homeric agent has no such conception of himself (Snell 1953). 13 ἔνθ’ Ὀδυσεὺς μνηστῆρσι κακὰ φρονέων ἐνὶ θυμῷ 5 /​κεῖτ’ ἐγρηγορόων: ταὶ δ’ ἐκ μεγάροιο γυναῖκες /​ἤϊσαν, αἳ μνηστῆρσιν ἐμισγέσκοντο πάρος περ, /​ἀλλήλῃσι γέλω τε καὶ εὐφροσύνην παρέχουσαι. /​τοῦ δ’ ὠρίνετο θυμὸς ἐνὶ στήθεσσι φίλοισι: /​πολλὰ δὲ μερμήριζε κατὰ φρένα καὶ κατὰ θυμόν, 10 /​ἠὲ μεταί̈ξας θάνατον τεύξειεν ἑκάστῃ, /​ἦ ἔτ’ ἐῷ μνηστῆρσιν ὑπερφιάλοισι μιγῆναι /​ [p. 58] Plato takes the passage to be so familiar that he does not quote the part that illustrates the difference between reasoning about better and worse and being unreasonably angry. Williams remarks that Plato read this passage ‘as showing the power of reason over the emotions (in particular, anger)’ (1993: 38). Did Plato read it correctly? We can infer Williams’s answer to this question from what he says later about the Platonic conception of the rational part. In the Platonic conception, as we saw, ‘reason operates as a distinctive part of the soul only to the extent that it controls, dominates, or rises above the desires’ (1993: 43). If Plato was right to suppose that the Homeric passage showed ‘the power of reason over the emotions’, it would show that reason controls, dominates, and rises above the emotions. Hence it would show that Homer accepted Platonic ethicized psychology. Since Williams believes that Homer does not accept Platonic ethicized psychology, he must believe that Plato read the Homeric passage wrongly, and that the passage does not show the power of reason over the emotions. To see whether Plato is wrong, we should ask whether this passage can be plausibly understood without any Platonic assumptions about rational desire. When Odysseus sees his slave girls misbehaving with the suitors, his spirit (thumos) is aroused, and he considers in his mind (phrȇn) and spirit whether to kill the slave girls immediately or to leave them alone for one last night. People in Homer often consider whether to do one thing or another. When they decide between the two options, Homer says that one option seemed better.14 In our present passage he implies, without saying so, that Odysseus has reached this conclusion, but his heart still protests within him. He reminds his heart that it had to bear worse than this, when he was in the cave of the Cyclops and had to watch some of his companions being eaten by Polyphemus. On that occasion Odysseus’s cunning (mêtis) led his heart out of the cave when it was sure he was going to be killed with his companions. Once his heart has fallen silent, Odysseus returns to his original question of how he can kill the suitors when he is one against many. Odysseus attributes personal attitudes to his spirit or heart; it remembers that it endured, was afraid, and none the less was saved.15 By reminding his spirit of all ὕστατα καὶ πύματα, κραδίη δέ οἱ ἔνδον ὑλάκτει. /​ὡς δὲ κύων ἀμαλῇσι περὶ σκυλάκεσσι βεβῶσα /​ ἄνδρ’ ἀγνοιήσασ’ ὑλάει μέμονέν τε μάχεσθαι, 15 /​ὥς ῥα τοῦ ἔνδον ὑλάκτει ἀγαιομένου κακὰ ἔργα: /​ στῆθος δὲ πλήξας κραδίην ἠνίπαπε μύθῳ: /​‘τέτλαθι δή, κραδίη: καὶ κύντερον ἄλλο ποτ’ ἔτλης. /​ἤματι τῷ ὅτε μοι μένος ἄσχετος ἤσθιε Κύκλωψ /​ἰφθίμους ἑτάρους: σὺ δ’ ἐτόλμας, ὄφρα σε μῆτις 20 /​ἐξάγαγ’ ἐξ ἄντροιο ὀϊόμενον θανέεσθαι.’ /​ὣς ἔφατ’, ἐν στήθεσσι καθαπτόμενος φίλον ἦτορ: /​τῷ δὲ μάλ’ ἐν πείσῃ κραδίη μένε τετληυῖα /​νωλεμέως: ἀτὰρ αὐτὸς ἑλίσσετο ἔνθα καὶ ἔνθα. /​ὡς δ’ ὅτε γαστέρ’ ἀνὴρ πολέος πυρὸς αἰθομένοιο, 25 /​ἐμπλείην κνίσης τε καὶ αἵματος, ἔνθα καὶ ἔνθα /​αἰόλλῃ, μάλα δ’ ὦκα λιλαίεται ὀπτηθῆναι, /​ὣς ἄρ’ ὅ γ’ ἔνθα καὶ ἔνθα ἑλίσσετο, μερμηρίζων /​ὅππως δὴ μνηστῆρσιν Translation mostly derived from Homer (1995). 15 This feature of Odysseus’ discussion with his spirit is emphasized by Gill (1996: 183–​90) in his helpful discussion of decisions in Homer. [p. 59] 46 Terence Irwin this, he calms it; hence it can be persuaded by recalling how it turned out better to keep quiet for a while. The spirit does not immediately think about consequences, but once Odysseus reminds it, it no longer insists on the course of action it had initially favoured. He has exercised several cognitive and desiderative powers in this address to his spirit. He recalls a similar situation in the past; he makes a retrospective judgement on the outcome in that past situation; he considers the consequences of acting on different options in the present and future; he applies his retrospective judgement to the present and future, so that he learns from experience; he makes a practical judgement about what is best in the present; and he chooses to act on that judgement. These different cognitive and desiderative powers combine to cause his eventual choice and action. They might not have combined in this way. Odysseus might have remembered his self-​restraint in Polyphemus’s cave, and thought, ‘That was so humiliating! I won’t take any more humiliation.’ Even if he remembered that things turned out better because of his self-​restraint, he might have thought, ‘So what? These slave girls have insulted me, and I’ll kill them now.’ He might have thought, ‘Admittedly it would be better to wait, but I’m so angry that I want to kill them now.’ He does not react in any of these ways. His cognitive and desiderative powers combine so that he tries to find out the better course of action, taking into account what matters most to him in the present and the future, and he chooses to pursue that course of action. We can now compare this description of the Homeric passage with Williams’s description. Homeric characters, then, are certainly capable of self control. Homeric ideas of self control, however, and related notions of endurance—​and to some extent the same goes for later Greeks as well—​are interestingly different from ours. The suffering of his heart is the suffering that Odysseus has to undergo when he cannot, for reasons of prudence, do what he would very much like to do and has good reason to do. Suffering is the cost of waiting until he can do what intelligence requires, and his endurance, in this case, is the capacity to sustain suffering that comes from an inner cause, though it is inflicted from outside. The painful character of what is going on indeed comes from outside, from what other agents are doing, together with Odysseus’s feelings about those things; but the need to wait and the length of the wait are what demand the endurance, and they come from his own mêtis, his own rational plan. This puts in a new light one of the standard descriptions of Odysseus, polutlas, ‘enduring’; indeed it associates this quality with what he even more famously is, polumêtis, ‘resourceful’. Odysseus has a will not only to endure what is inflicted on him: he has a will to endure the consequences of his own will to do. (1993: 38–​9) In Williams’s view, then, this passage describes Odysseus’ self-​control. [p. 60] This description of Homer’s view of Odysseus is true, but not the whole truth. Odysseus does not simply stick to his first intention despite the anger that he feels in his spirit. He might have done this because of mere stubbornness, which Aristotle distinguishes from continence (EN 1151b4–​17). But in fact he refuses to follow his spirit because he has thought about the consequences of following it and refusing to follow it, and has concluded that it is best to stick to his original intention. This is not mere stubbornness. Odysseus has formed an effective desire because of what he has learned from experience, what he thinks about his goals in the present and the future, and what he thinks about their relative importance of these goals. He persuades his spirit to be quiet after reminding it of all these things. At first, influenced by his spirit, he debates within himself about whether to attack at once. After he has decided it would be better not to attack at once, and has persuaded his spirit to be quiet for that reason, he still debates within himself, but on a different question, how best to deal with the suitors, given that he is one against many. Homer does not speak of reason or of a rational part here. He speaks of Odysseus addressing a part of himself.16 But he describes the formation of a rational desire that results from holistic concern and deliberation. According to Williams’s description of Plato’s ethicized psychology, ‘reason operates as a distinctive part of the soul only to the extent that it controls, dominates, or rises above the desires’ (1993: 43). This is an accurate description of Odysseus; something about him controls, dominates, and rises above his non-​rational desires, and this is a desire formed by reasoning based on holistic concern. This is the sort of desire that Williams attributes to Plato’s rational part, and takes to express an illegitimately ethicized psychology. One might object that my description of Homer substitutes ‘rational desire’ for Willams’s term ‘reason’. But we noticed earlier that Williams’s description of Plato is accurate only if we substitute ‘rational desire’ for ‘reason’. Homer attributes to Odysseus the sort of desire that Plato takes to belong to the rational part of the soul. In that respect Homer already agrees with Plato’s ethicized psychology, so that ethicized psychology is not a Platonic innovation or aberration. The passage we have discussed describes a situation that is not unusual in Homer, in which an agent considers two options and decides that one of them is better. In all these cases the agent thinks about the pros and cons of different options, and chooses the course of action that seems most likely to advance the aims that matter most to him, even though it has some immediate risks and dangers. Nestor, for instance, wonders whether he should join the battle immediately, or inform Agamemnon about the critical situation. He decides (being Nestor) to talk to Agamemnon. He takes a risk, because the Achaeans may be overwhelmed in the time it takes to decide what to do; but the risk is worth it, because Agamemnon 16 This and other instances of a conversation between the agent and his thumos are discussed by Sharples (1983). The whole subject is discussed fully by van Ackeren (2011: vol. 1, 212–​31). [p. 61] 48 Terence Irwin should be able to take more effective action than Nestor can take by himself (Iliad But not everyone who thinks about what to do takes account of everything that is relevant. When Hector thinks about whether to fight Achilles, he does not consider the options and follow the one he thinks best. He stops thinking about other aspects of the situation because he is ashamed at the prospect of being reproached by Polydamas for his mistakes (Iliad 22.99–​110). To express the difference between Hector and Odysseus, we need to say that Odysseus exercises reason in practice. Homer does not speak of reason; he simply says that the thoughts and desires that prevent Odysseus from following his angry spirit are the thoughts and desires of Odysseus himself. In saying this, Homer identifies the agent with his rational beliefs and desires. These beliefs and desires take account of Odysseus’s interests and concerns as a king, father, and husband, concerned for his family and dependants no less than for his own safety. Homer does not present a theoretical account of the different parts of the soul that are involved in this interaction between Odysseus and his spirit. On this point he differs from Plato. But he does not differ from Plato in having a less ethical conception of the self and its motives. Both in Homer and in Plato, the rational self displays holistic concern. We have no reason, then, to say that Plato innovates by introducing ethical terms into the analysis of motivation and action. 7. Ethicized Psychology in Thucydides? To support his view that Plato’s ethicized psychology is an innovation, Williams contrasts Plato not only with Homer, but also with Thucydides. He argues that in Thucydides ‘the psychology he deploys in his explanations is not at the service of his ethical beliefs’ (1993: 161). Thucydides’ conception of an intelligible and typical motivation is broader and less committed to a distinctive ethical outlook than Plato’s; or rather—​the distinction is important—​it is broader than the conception acknowledged in Plato’s psychological theories. The same is true, if less obviously, in relation to Aristotle. Thucydides does not express the view that justice is the destination of the rational part of the soul; we can agree that he lacks this aspect of Plato’s ethicized psychology. But it is less obvious that he lacks the Platonic belief in the holistic concern of rational desire. If Williams is right, we should find that in Thucydides ‘rational capacities have been exercised in arriving at a course of action’ (1993: 43), but we should not find any connection between reason and holistic concern. [p. 62] Thucydides often describes the instrumental use of reasoning to work out the means for achieving some aim. Themistocles, for instance, was especially good at devising ways to achieve his various aims. Cleon was capable of quick and bold thinking, when he found himself appointed to command the expedition to Pylos, which turned out better than his opponents thought it would (IV 28.7). But this is not all that Thucydides has to say about practical reason. Pericles displays his judgement (gnȏmȇ, ii 65.8) not primarily in planning campaigns or expeditions (though he certainly engages in this sort of planning), but in his grasp of a bigger picture (as Thucydides conceives it) and in forming his will in accordance with this bigger picture. He thinks he sees what Athens needs to do to survive as an independent state in the face of Peloponnesian hostility, and he devises a strategy with this overall aim in mind. To some degree he forms the same outlook in the Athenians, but he does not succeed completely, since they turn against him and his policy for a while. They are moved by anger (orgê, ii 65.3) to prosecute and fine him; then they display their fickleness (hoper philei homilos poiein, ii 65.4) in electing him as general again. Similarly, those who favour the Sicilian Expedition are said to be moved by an infatuation (erôs, vi 24.3). In the Mytilenean Debate, Diodotus suggests—​and Thucydides says nothing to contradict him—​that appeals to the moral views of one’s audience are simple attempts to substitute thoughtless emotion for judgement (iii 44.1–​2). When people are engaged in faction and civil war, they think they are displaying bravery in support of their own side, but in fact they merely exhibit unreasoning daring (tolmȇ alogistos, iii 82.4). When Thucydides describes these people as acting on infatuation, emotion, and foolhardy daring, he does not mean that they become incapable of using practical reason tactically. Pericles’ opponents may have been just as good at this exercise of practical reason as Pericles was. The reason that they are moved by infatuation and so on is that they lack Pericles’ capacity to form the appropriate longer-​term goals and to form his will to achieve them. Thucydides does not say, as Plato says, that some people are guided by the rational desires that belong to their rational parts. But when he describes people who are not guided by holistic concern, he uses terms that imply a failure of reason (infatuation, unreasoning daring, etc.). The outlook that opposes these failures of reason is the outlook of practical reason. Though Thucydides does not anticipate Plato’s theory of the tripartite soul, the desires that Plato attributes to the rational part are familiar to Thucydides. Since he recognizes rational desires that involve holistic concern, his psychology is no less ethicized than Plato’s. 8. Will and Freedom We have now found some reason to believe that Plato has a conception of the will, since he recognizes rational desire that involves holistic concern. We have also [p. 63] 50 Terence Irwin found that the Platonic conception states and articulates a view that is already present in earlier Greek sources (Homer and Thucydides). We have not yet vindicated Dante’s claim about the Greek philosophers, since we have not yet shown that the Platonic conception of the will (voluntas) has anything to do with freewill (liberum arbitrium). Dante asserts a connexion by claiming that deliberation about how to achieve the object of one’s will is the place where we find freedom. This connexion is affirmed by Aquinas (ST 1–​2 q6 a2 ad2). It does not imply that we are responsible only for those actions that result from deliberation. If the will consents to an action on passion without deliberation, it refrains from deliberation that the agent could have undertaken; if this consent contributes to an action, we are responsible for the action to the extent that it results from consent (ST 1–​2 q10 a3 ad1,2; q15 a4 ad 1,2). We can ask, then, whether Greek philosophers connect will and deliberation with freedom and responsibility, and whether pre-​philosophical sources anticipate this connection, explicitly or implicitly. As we saw earlier, Williams believes that questions about freedom and responsibility, as we understand them, are out of place in any discussion of Greek thought. We are especially mistaken if we suppose that Greek views about responsibility rest on metaphysical assumptions that might be taken to introduce questions about the compatibility of freedom and determinism. Such questions arise for us if we mistakenly attempt try to make the notion of the voluntary ‘profound’. The Greeks, however, ‘were not involved in these attempts; this is one of the places where we encounter their gift for being superficial out of profundity’ (Williams 1993: 68). If we begin from Plato and Aristotle, it is difficult to deny that rational desire and deliberation are relevant to responsibility, in so far as responsibility involves justified praise and blame. To show that actions that might appear involuntary—​for example, abandoning cargo in a storm—​are really voluntary, Aristotle argues that we choose them for the sake of some goal that we think we see good reason to pursue (e.g. safety in a storm; EN 1110a10). If rational desire and deliberation explain the action that is alleged to be involuntary, we have a reason to believe it is voluntary. If the action does not result from the agent’s will and deliberation, that is a reason to deny that it is voluntary. It is not always a decisive reason. We may be responsible for actions that result from negligence or from culpable ignorance. But these cases do not show that will and deliberation are not the basis of responsibility; they are cases where will and deliberation could have affected the action, but did not affect it. A similar concern with the agent’s aims and deliberation is found in discussions of blame and legal responsibility in law and forensic oratory. As Antiphon says, those who do or undergo something voluntary are responsible (aitioi) for what happens (Tetralogies ii 2.6). Demosthenes describes the various Athenian courts that deal with voluntary and involuntary homicide, mentioning the different penalties. The law of homicide distinguishes voluntary (hekousios phonos, Dem. 23.77; [p. 64] from involuntary killing. It is recognized as just that involuntary offences receive a lighter punishment than voluntary offences receive (Dem. 23.73). It is more difficult to say exactly how Athenian law conceives voluntary action. Laws about homicide and ‘wounding’ (i.e. violent assault with a lethal weapon) say that those who are liable to punishment are those who act from forethought (pronoia; Dem. 23.24, [Aristotle], Ath. Pol. 57.3). This term readily suggests premeditation and planning in advance. In that case the ‘fore’ in ‘forethought’ refers to some time before the occasion on which A killed B, and the ‘thought’ refers to some planning by A about what A will do when A meets B. We might also, however, understand ‘forethought’ so that it includes not only the sort of planning just described, but also the thought that precedes the action whenever we know what we are doing. Athenian assumptions about forethought include these different degrees of premeditation and planning.17 Whatever we think about the appeal to forethought, understood so broadly, as a basis for responsibility, such an appeal implies that Athenian law and Athenian juries take an agent’s will and deliberation to determine responsibility. Aristotle does not propose a wholly new condition for responsibility. His discussion of will and deliberation responds to beliefs about responsibility that are shared by his contemporaries, and by at least some of the legal systems he knew. Williams attends to some of this legal evidence, but he observes that the will and deliberation of agents did not determine everything that happened to them when they offended against laws and other norms. Laws and institutions assumed that homicide caused pollution (miasma), a collective danger to a city. Pollution needed some act of expiation (often by the exile of the killer). The pollution still needed to be removed even if a homicide was involuntary (Antiphon, Tetralogies This feature of pollution, however, does not refute the view that will and deliberation determine responsibility. Pollution is not a simple quantity that needs to be expiated, no matter what kind of homicide it may have been, and no matter what the motive or intention of the homicide may have been. If the homicide was justified (e.g. in tyrannicide), no pollution resulted from it (Dem. 23.55), and the degree of pollution varied with the character of the homicide. Though the distinction between voluntary and involuntary does not determine the presence or absence of pollution, it affects the type and degree of the pollution; the pollution caused by involuntary homicide was easier to remove.18 Williams sums up assumptions about pollution by saying ‘Miasma was incurred just as much by unintentional as by intentional killing’ (1993: 59). This is true if it means that (1) both intentional and unintentional homicide caused pollution. But it is false if it means that (2) just as much pollution was incurred by unintentional 17 I have discussed some of this evidence in Irwin (2024). 18 For these claims I rely on Parker (1983: esp. ch. 4 and appendix 5 on the ritual status of the justified killer at Athens), and on the convincing criticism of Williams on Oedipus by Harris (2013). [p. 65] 52 Terence Irwin as by intentional homicide. Since the second claim is false, facts about the agent’s will and deliberation affect not only praise and blame, but also the consequences of homicide that go beyond what the agent is blameworthy for. Agents cause pollution even through involuntary (and non-​negligent) homicide, but their degree of blameworthiness for the homicide determines the degree of the pollution that they are causally responsible for. This comparison between Aristotle’s views on responsibility and Athenian legal assumptions shows how one might argue in favour of Dante’s assertion that the Greek philosophers grasped the basis of freedom and morality through their discussion of will and deliberation. Contrary to Williams, we have found no gap between Greek philosophers and views about responsibility that are familiar to us. Moreover, we have found no gap between the philosophers and their predecessors and contemporaries. They share an ethicized psychology, and a conception of freedom and responsibility that relies on this ethicized psychology. One might object on Williams’s behalf that, if what I have said is right, we ought to expect Aristotle to discuss questions that (let us grant) 19 he does not discuss, about the compatibility of freedom with causal determinism. In answer to this objection, we have to recall the temporal limits of Williams’s discussion. Stoics and Epicureans discuss questions about freedom and determinism. Alexander argues in De Fato that Aristotle’s conception of responsibility requires causal indeterminism. If it is Aristotle’s doctrine that raises these metaphysical questions and debates, it is not exempt from the sorts of disputes that arise about modern views on responsibility. 9. Conclusion I have chosen a few instances in which one may reasonably doubt whether Williams is right to present us with a contrast between pre-​Platonic Greek thought and the views of the philosophers. In my last instance (will and responsibility), I have raised questions about the contrast he draws between Greek views (including those of Plato and Aristotle) and later philosophical views. I have suggested that, in these cases, we have no reason to congratulate pre-​Platonic Greeks for lacking the Platonic outlook. The Platonic outlook is better than Williams makes it seem. If it is implicitly present in earlier Greek thought, the earlier thought is all the better for agreeing with Plato20. 19 To decide whether or not we should grant this to Williams, we would need to discuss (e.g.) De Interpretatione 9, and Metaphysics vi 3. 20 I am grateful for useful comments by Marcel van Ackeren, which have helped me to improve an earlier version of this chapter. [p. 66] References Anscombe, G. E. M. 1965. ‘Thought and Action in Aristotle’. In New Essays on Plato and Aristotle. Edited by R. Bambrough, 143–​58. London: Routledge. Burnyeat, M. F. 2006. ‘Introduction’. In B. Williams, The Sense of the Past, xiii–​x xii. Edited by M. F. Burnyeat. Princeton: Princeton University Press. Dante Alighieri. 1970–​5. The Divine Comedy. Translated and edited by C. S. Singleton, 6 vols. Princeton: Princeton University Press. Gill, C. J. 1996. Personality in Greek Epic, Tragedy, and Philosophy. Oxford: Oxford University Press. Harris, E. M. 2013. ‘Is Oedipus Guilty? Sophocles and Athenian Homicide Law’. In Law and Drama in Ancient Greece, 122–​46. Edited by E. M. Harris, D. F. Leao, and P. J. Rhodes. London: Bloomsbury. Homer. 1995. Odyssey. Translated by A. T. Murray and G. E. Dimock. Cambridge, MA: Harvard University Press. Irwin, T. H. 1994. Critical Notice of Shame and Necessity. Apeiron 27: 45–​76. Irwin, T. H. 2007. Development of Ethics, vol. 1. Oxford: Oxford University Press. Irwin, T. H. 2024. ‘Aristotle on Legal and Moral Responsibility: Interpretation and Reform’. In Routledge Handbook of Responsibility, 97–​109. Edited by M. Kiener. London: Routledge. Long, A. A. 2007. ‘Williams on Greek Literature and Philosophy’. In Bernard Williams, Ch. 6. Edited by A. Thomas. Cambridge: Cambridge University Press. Parker, R. 1983. Miasma. Oxford: Oxford University Press. Sharples, R. W. 1983. ‘ “But Why Has My Spirit Spoken with Me Thus?”: Homeric Decision-​ Making’. Greece and Rome 30: 1–​7. Snell, B. 1953. The Discovery of the Mind. Oxford: Blackwell. Striker, G. 1993. ‘Are We Any Better?’ London Review of Books 15.16. van Ackeren, M. 2011. Die Philosophie Marc Aurels, 2 vols. Berlin: De Gruyter. Williams, B. 1981a. ‘Internal and External Reasons’. In Williams, Moral Luck, 20–​39. Cambridge: Cambridge University Press. Williams, B. 1981b. ‘Philosophy’. In The Legacy of Greece, 202–​55. Edited by M. I. Finley. Oxford: Oxford University Press. Williams, B. 1993. Shame and Necessity. Berkeley: University of California Press. [p. 67] Agamemnon at Aulis A Misfiring Example in Williams Sophie Grace Chappell 1. Williams’s Troublesome Example(s): Karenina, Gauguin, and Moral Luck It is an important philosophical skill to pick the right examples to make your argument (where ‘make’ often means ‘complement’, and sometimes means ‘constitute’). Bernard Williams was of course aware of the importance of this skill (1995: 217): Often, some theory has been under criticism, and the more particular material [e.g. Williams’s famous examples (Williams 1973a: 93–​100) of George and Jim] has come in to remind one of the unreality and, worse, distorting quality of the theory. The material . . . is itself extremely schematic, but . . . it at least brings out the basic point that . . . the theory is frivolous, in not allowing for anyone’s experience, including the author’s own. Alternatively, the theory does represent experience, but an impoverished experience, which it holds up as the rational norm—​that is to say, the theory is stupid. However, not all of the examples that Williams himself picked were as helpful or as illuminating as we might hope. One instance is in Williams’s classic essay on moral luck (Williams 1981).1 Williams’s thesis about moral luck is the thesis that when an agent pays a price in loss of value of one kind (typically, the morality-​system kind, or something close to it), for the sake of pursuing value of some other kind, then whether or not this price is worth paying can and sometimes does depend on the answers to questions that can only be asked much later, and retrospectively; in particular, whether the agent’s pursuit of her chosen kind of value has turned out successful or not. In his essay, it is central to Williams’s argument to take Tolstoy’s Anna Karenina and his own fictionalized Gauguin to be examples of moral luck in this sense. His suggestion is that Anna’s affair with Vronsky might have gone better than 1 My discussion of Williams’s examples of moral luck draws on Epiphanies 3.8 (Chappell 2022). Sophie Grace Chappell, Agamemnon at Aulis In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. [p. 68] disastrously had some other things turned out differently. In that case the risk that Anna took with (for instance) her son’s happiness might have been, as in fact it was not, justified. Not necessarily morally justified, but justified in the broader sense invoked when we say things like ‘Well, I can see how this made sense for you.’ Likewise, Williams suggests, his fictional Gauguin’s desertion of his family is justified, in this same broad sense, in so far as that Gauguin’s belief that he could become a great painter turned out, in the event, to have been correct; if it hadn’t, then he wouldn’t have been justified in deserting them. There are problems with Williams’s Gauguin, and they relate closely to the fictionality of the example. Williams claims explicitly to be saying nothing about the historical Gauguin, nor about any imaginable variant of Williams’s Gauguin. Yet it is hard for Williams, or for me, to talk solely about the very Gauguin that Williams describes without reference to the historical one. After all, Williams did call his stipulated character Gauguin. But Williams’s Gauguin is puzzling in ways that the Karenina example is not, in good part because the case is so thinly described by Williams. It is danger of being one of those schematic and unrealistic examples that Williams himself decries as quoted above; and it is hard to see legitimate ways of filling it in more that do not risk collapsing Williams’s Gauguin back into the historical Gauguin. As I say, what is meant to matter about Williams’s Gauguin is that he illustrates the proposition that a person can choose a life that involves him in stepping outside the morality system, and this choice can be justified (not morally but practically justified) if he succeeds in that life; but whether he will succeed in the way that justifies him, depends on factors that he cannot know about in advance. Whether or not that proposition is true, it is not clear that its truth is supported either by Williams’s Gauguin or by the historical Gauguin. Not by the historical Gauguin, because (as a look at a biography quickly shows) his actual life involved too much detail that does not at all give the proposition clear support. But not by Williams’s Gauguin either, because the life that is stipulated for him is too thin: without drawing tacitly on what we know about the historical Gauguin, we just don’t have enough detail to make the call. Of course, it matters that the Gauguin story is one about an exit from ‘the morality system’, so that we do not have the same kind of value on the two sides of the choice; nonetheless the effect of these two examples is to suggest that Williams’s moral-​luck point collapses into the somewhat less sophisticated claim that the outcome can justify the means taken to secure it, even though the outcome is beyond the agent’s control, provided the outcome is only good enough. Despite Williams’s protestations to the contrary, it then begins to sound like his Gauguin is being applauded for cold-​bloodedly taking a very unpredictable risk at the known expense of the happiness of others whom he ought to care about; while his Anna is being applauded for courting almost certain disaster. If there is such a thing as moral luck, then there is also such a thing as moral recklessness. The trouble seems to [p. 69] 56 Sophie Grace Chappell be that it is only too easy to see Williams’s Anna certainly, and Williams’s Gauguin possibly, as examples of this moral recklessness rather than of what Williams means by moral luck. 2. Agamemnon at Aulis My main concern here is not with Williams’s central would-​be examples of moral luck, but with his central would-​be example of tragic dilemma. Consider this, from Bernard Williams’s classic essay ‘Ethical Consistency’: One peculiarity of tragic cases is that the notion of ‘acting for the best’ may very well lose its content. Agamemnon at Aulis may have said ‘may it be well’ Agamemnon 217] but he is neither convinced nor convincing. The agonies that a man will experience after acting in full consciousness of such a situation are not to be traced to a persistent doubt that he may not have chosen the better thing but, for instance, to a clear conviction that he has not done the better thing because there was no better thing to be done. It may on the other hand even be the case that by some not utterly irrational criteria of ‘the better thing’, he is convinced that he did do the better thing: rational men no doubt pointed out to Agamemnon his responsibilities as a commander, the many people involved, the considerations of honour and so forth. If he accepted all this and acted accordingly: it would seem a glib moralist who said, as some sort of criticism, that he must be irrational to lie awake at night having killed his daughter. And he lies awake, not because of a doubt, but because of a certainty. Some may say that the mythology of Agamemnon and his choice are nothing to us, because we do not move in a world in which irrational gods order men to kill their own children. But there is no need of irrational gods to give rise to tragic situations. (1973b: 173) This passage is central to Williams’s argument in ‘Ethical Consistency’. His thesis is that our ethical beliefs, even at their most rational and well organized, need not and perhaps cannot always have the logical consistency that our other beliefs, especially our ‘scientific’ beliefs, are bound to have at all times, on pain of irrationality in the believer. Agamemnon at Aulis, a case from the first chorus of Aeschylus’s play Agamemnon of 458 bc, is supposed to illustrate this thesis because it is ‘a tragic case’, ‘a tragic dilemma’. Agamemnon’s tragic dilemma at Aulis was that he had to choose between the failure of the Greek military expedition to punish the Trojans for the abduction of Helen, and sacrificing his daughter Iphigeneia. This choice was forced on Agamemnon because the goddess Artemis was withholding a following wind for the fleet and, according to the soothsayer Calchas, could only be appeased by killing the princess. Hence, according to Williams, Agamemnon is caught in a [p. 70] situation where no available action is the best thing to do, the better thing to do, a good thing to do, or even a minimally permissible thing to do. We know from the play that Agamemnon goes ahead and kills Iphigeneia. But if Agamemnon then feels ineliminable agent-​regret, he is not mistaken, Williams says, to feel it: he would not ‘be irrational to lie awake at night having killed his daughter’, and what keeps him awake is not a doubt, but ‘a certainty’—​the certainty that he has done something terrible. That there are cases that have the structure that Williams has in mind here, I do not dispute. It certainly can happen that an agent who is virtuous, or at any rate well intentioned, finds himself in a situation where all his alternatives are such that taking any of them will leave him with good reason to feel ineliminable agent-​ regret: that is, pain or sorrow that he has been the agent who has done something appallingly bad, where this pain or sorrow is not to be soothed away by any rational consolation, such as the thought that the atrocious thing that he did was the right means to a necessary end. What I do dispute is that Agamemnon as Aeschylus portrays him is an agent like that. Given the thesis that Williams wants to argue in ‘Ethical Consistency’, his choice of Agamemnon as his key example for that thesis has struck me as a weird choice from the moment when, as an undergraduate, I first read ‘Ethical Consistency’. My point, to be clear, is not simply that Williams’s Aeschylean example falls apart the moment we examine it. If that were all, then a fair response would be ‘So pick a different example.’ But there is more to say, because the case of Agamemnon at Aulis is a deeply revealing one, in ways that might well have interested Williams (perhaps even should have interested him). Aeschylus is showing us something deep, dark, and important to our self-​understanding and our understanding of guilt, shame, regret, responsibility, and indeed dilemma; or so I shall argue. Still, on the face of it, we might well be tempted to say that the example does just fall apart straight away, because it rests on uninterestingly false beliefs. Williams, after all, wants us to see Agamemnon at Aulis as a central, and so one assumes particularly persuasive, case of tragic dilemma. But from a modern point of view, the first and most obvious thing to say about Agamemnon’s case is that his choice is a tension between a number of demands, at least two of which are patently ridiculous: his is not, in fact, a situation where he has good reason to feel forced to do the terrible deed. The reason-​giving force of parental obligations to a daughter is real enough, of course—​both in our time and in Agamemnon’s too. But if Agamemnon imagines—​a modern may say—​that he is under any kind of obligation to sail off to Troy, kill or rape and enslave the Trojans, and burn their city to the ground, then he is just plain wrong. Likewise—​the modern may continue—​if Agamemnon imagines that, in order to fulfil that obligation, he is under the further obligation that Calchas announces to him, to kill Iphigeneia in order to appease Artemis, then again he is just plain wrong. But without these two radically false beliefs, Agamemnon’s dilemma cannot even be set up. [p. 71] 58 Sophie Grace Chappell Does this objection betray a lack of imagination? Does it smack, even, of a certain kind of moralistic provincialism? We need not (it could be said) share a bronze-​age Greek chieftain’s beliefs to feel the dilemma, or other psychological pressures, that are faced by Agamemnon within Aeschylus’s drama; all we need is to broaden our minds a bit, to be appropriately open to ‘the relativism of distance’. After all (it might be added), even if we grant (with moralistic provincialism?) that the key beliefs that structure Agamemnon’s case are false, still there are plays where what we see is exactly the suffering that can be inflicted on agents or on a group of agents by their own false beliefs: that is true, for instance, of witchcraft beliefs in The Crucible, and capitalist cult-​of-​success beliefs in The Death of a Salesman. No doubt; but the Agamemnon is not that sort of play—​it is not a play that, in Arthur Miller’s polemical sort of way, exposes a whole false ideology as a source of human oppression. (Though maybe some ancient dramas are that sort of play; perhaps the Bacchae is, for instance.) At any rate Williams seems uncomfortably aware of the objection, as comes out at the close of the quotation above: as he truly, but rather lamely, says, ‘There is no need of irrational gods to give rise to tragic situations.’ But then the retort is irresistible: ‘So why make your prime example of a tragic situation one that does need irrational gods?’ If what Williams was after was a paradigm tragic case, then he should have found a case where we—​his readers, and Williams himself—​share the commitments that are in conflict in the case, or can at least see how we might have shared them, or come to share them. What is needed for persuasive purposes is a case where it is easy for the reader or audience to identify imaginatively with the protagonist—​to feel that we might occupy this person’s situation, including their relevant beliefs and values. The eponymous incident in William Styron’s novel Sophie’s Choice (1979) is that a Polish mother facing the ‘selection’ on arrival at Auschwitz is given this choice: she can either pick one of her children, Jan or Eva, to be gassed; or she can refuse to pick either, in which case both Jan and Eva will be gassed.2 Even though, fortunately, most of us have never been in anything like Sophie’s terrible situation, and never will be, her case naturally evokes from us a vivid, horrified, and sympathetic response. We find that we understand, at least to some extent, what it might be like to be in such a situation, and what we would feel if we were. (Regret seems much too mild a word for Sophie’s unbearable, lifelong, broken anguish; her eventual suicide in Styron’s novel is far from incomprehensible.) Most of us are a very long way from anything like this sort of empathetic response to Agamemnon at Aulis. If we try and take the case as a dilemma that we 2 Possibly that particular case was not available when Williams wrote ‘Ethical Consistency’, as Styron’s book was not published till 1979. But plenty of cases like it were; one (involving a Roma, not Polish, family) is described in Arendt’s Origins of Totalitarianism (1951). Or Williams could have contrived his own, as he did with ‘Jim and the Indians’. [p. 72] ourselves might be in, then what we are confronted by there is not so much our own provincialism as a case of imaginative resistance, a refusal to entertain a scenario that involves radically false ethical beliefs. (I have in mind beliefs such as ‘There is at least a prima facie reason for Agamemnon to kill Iphigeneia’ or ‘Agamemnon is under genuine moral and practical pressure to kill Iphigeneia’.) Indeed, if we really do set aside the relativism of distance, and try and imagine someone today who does not respond to the case like this—​someone who tells us, say, that they find Agamemnon’s allegedly forced choice between murdering his daughter and giving up his war of revenge an entirely understandable and relatable dilemma: someone who says this would not be giving us entirely reassuring information. For my money at least, they would most certainly be ruling themselves out of the babysitting rota. 3. Williams, Fraenkel, Hegel, Nussbaum: Agamemnon Is Innocent For a dilemma to be convincing, then, it needs to involve conflicting ethical pressures that we can find realistic, and that are not easily blocked by imaginative resistance. Ideally, a paradigmatically convincing dilemma also needs something else: it needs a more or less sympathetic protagonist—​‘an agent’ (as I put it above) ‘who is virtuous, or at any rate well intentioned’. This condition is met, in Styron’s novel, by Zofia Zawistowski. She is not a saint or a perfect human being, but she is good enough for our sympathy to track her; we find it easy and natural to identify with her imaginatively. Is the same condition met by Aeschylus’s Agamemnon? I shall be arguing that it is not. But it is extremely telling how keen commentators are to argue that it is met, when those commentators aim, like Williams, to present Agamemnon at Aulis as the innocent victim of a paradigmatic moral dilemma. One case of this is Williams himself, as quoted above. Williams does not explicitly say that he sees Agamemnon as essentially innocent and helplessly trapped, but that view is clearly implicit in what he does explicitly say. For Williams, Agamemnon is someone who experiences ‘agonies’ because he is ‘fully conscious’ of his situation, which is that ‘there was no better thing to be done’; criticism of him is ‘glib’; Agamemnon is properly aware of ‘his responsibilities as a commander, the many people involved, the considerations of honour and so forth’. (‘And so forth’? A trailing-​off like that is often a clue to rhetorical or logical weakness; one is reminded of the over-​use of ‘etc.’ in the lists that occur in revolutionary polemics like Lenin’s.) Above all, of course, the image of Agamemnon as ‘lying awake at night having killed his daughter’ is a particularly haunting one; even though, as we shall see, it is one without any textual basis in Aeschylus’s play. What was Williams’s source for this view of Agamemnon? As an undergraduate at Balliol, Williams was a pupil of Eduard Fraenkel, the great Aeschylus scholar [p. 73] 60 Sophie Grace Chappell from Corpus. And it is very plain from Fraenkel’s writings that he saw Agamemnon’s case as a tragic dilemma. Here is Fraenkel on Agamemnon at Aulis, in his magisterial two-​volume edition: In the narrative of the chorus the climax is reached with the monologue of Agamemnon (206 ff.) and the comments attached to it (218 ff.). The king is faced with the alternative of two deadly evils. Whichever course he takes is bound to lead him to unbearable hamartia. After a violent struggle he resolves to sacrifice his daughter, fully aware that what he is doing is an unpardonable sin and will have to be atoned for. His fatal step puts him under the yoke of compulsion; there can be no way back; on and on he must go, and the end, he knows as well as the Elders, will be utter ruin. Aeschylus, by using unmistakable language (220–​221 and ff., τόθεν τὸ παντότολμον φρονεῖν μετέγνω [literally: from then on he recognized the to-​dare-​everything thinking]) . . . makes it clear that all the evil that is to befall Agamemnon has its first origin in his own voluntary decision . . . [this is a] tragic conflict . . . [Aeschylus wanted] a moral dilemma . . . to be the fountain-​head of Agamemnon’s fate. (1950, vol. 1: 98–​9) Fraenkel adds: It is in merciless terms that Agamemnon describes, and implicitly condemns, what he is going to do. And yet he cannot avoid doing it. (1950, vol. 1: 121) On Fraenkel’s depiction too, Agamemnon’s plight is one in which two kinds of value meet and clash and remain, at least at this stage of the Oresteia, un-​sublated and unreconciled. He does act voluntarily, yet Agamemnon’s role, unfortunately for him, is simply to be the locus for this clash. There is something Hegelian about this picture; and indeed we can almost find a line of academic descent, via pupils and Doktorväter, that confirms the Hegelian pedigree. Just as Williams (1929–​2003) was a pupil of Fraenkel (1888–​1970), so Fraenkel was a pupil of Franz Bücheler (1837–​1908) and Bücheler in turn a pupil of Friedrich Ritschl (1806–​76). This nearly establishes an academic lineage back to Hegel himself (1770–​1831); but not quite, since so far as I know, though Ritschl and Hegel knew each other, there was never a formal pupil–​teacher connection between them. (However, entertainingly enough, there is a connection with Friedrich Nietzsche (1844–​1900), who was also a pupil of Ritschl’s. So Williams turns out to be Nietzsche’s great-​grandnephew academically speaking, and perhaps in other ways too.) With or without any direct connection with Hegel, we can say this: Fraenkel like Williams imagines Agamemnon as something decidedly like a Hegelian hero, caught through no fault of his own between two irreconcilable ethical imperatives. Fraenkel pictures an Agamemnon facing a ‘violent struggle’, a choice between [p. 74] ‘deadly evils’ for either of which there will be a terrible price; Williams, as we have seen, pictures a man in agonies who lies awake at night because of what he chose, and thought, rightly or wrongly, that he had no choice but to choose. And both with Williams and with Fraenkel, ‘imagines’ is the right word; for both of them, Agamemnon is someone they can imaginatively identify with. Both of them, in short, are firmly in the Agamemnon-​is-​innocent camp. As subject of a Hegelian dilemma, and as an agent whom we can empathize and imaginatively identify with, Agamemnon must in some sense be free of guilt. He may be faced with no blame-​free alternatives; but he is not to blame that that is his predicament. However he chooses at Aulis he will be at fault, because he is caught; yet it is not his fault that he is thus caught. At this second-​order level, Agamemnon is an innocent. Indeed, we might see him through the lens of another historically important schematism, Aristotle’s in the Poetics, and think of Agamemnon, though perhaps as no saint, still as essentially a good man, a suitable hero for a great tragedy. And so, indeed, Fraenkel explicitly accords Agamemnon the crucial Aristotelian honorific—​he calls him a megalopsychos, a great-​souled man (Fraenkel 1950, vol. ad 202 ff.). A third and even more emphatic advocate of Agamemnon’s innocence is Martha Nussbaum, in The Fragility of Goodness: No personal guilt of Agamemnon’s has led him into this tragic predicament. The expedition was commanded by Zeus (Ag. 55–​62) to avenge the violation of a crime against hospitality3 . . . Agamemnon is fighting in a just cause, and a cause that he could not desert without the most serious impiety . . . Agamemnon’s dilemma comes upon him as he is piously executing Zeus’s command . . . If Agamemnon does not fulfil Artemis’s condition, everyone, including Iphigeneia, will die. He will also be abandoning the expedition and, therefore, violating the command of Zeus . . . Both courses involve him in guilt . . . [up to this point Agamemnon, like Abraham, is] a good and . . . innocent man. (1986: 33–​5) 4. Aristotle’s Rejection of Agamemnon-​Is-​Innocent The pressure to see Agamemnon as innocent comes, I am suggesting, from interpreters’ desire to see Agamemnon as someone whom they, and we, can empathize with. We want the play to be as dramatically striking as possible; so (we assume) it must be possible for us to imaginatively identify with Agamemnon’s plight; and that in turn means that Agamemnon must be at least second-​order innocent—​not to be inevitably to blame for whatever he does at Aulis. 3 This claim can be supported by reference to, for example, Homer, Iliad 13.620–​39 (Menelaus denouncing the Trojans for their breach of the hospitality laws of Zeus). [p. 75] 62 Sophie Grace Chappell This assumption about how to read tragedies is a uniformity assumption—​an assumption that they can all (or for the most part?) be read in broadly the same way. One important source for the assumption is, as already noted, Aristotle’s Poetics. Aristotle says explicitly (Poetics 1452b35–​1453a27) that the protagonist of the ideal tragedy is not—​heaven forbid—​a villain; but he (yes, he) is not outstandingly virtuous either. He is an approximately good man, but with a fatal flaw: ὁ μήτε ἀρετῇ διαφέρων καὶ δικαιοσύνῃ μήτε διὰ κακίαν καὶ μοχθηρίαν μεταβάλλων εἰς τὴν δυστυχίαν ἀλλὰ δι᾽ ἁμαρτίαν τινά, τῶν ἐν μεγάλῃ δόξῃ ὄντων καὶ someone who is not specially distinguished for virtue or justice, and who does not fall into misfortune through wickedness and vice, but because of some mistake he makes; and he should be one of those who live in great honour and good fortune’ (1453a10–​12). Now there is plenty to find fault with in the Poetics. Aristotle’s analysis of Attic tragedy is at once wildly over-​schematic and also obviously inadequate to anything more than a small selection of the matter that it is supposed to apply to. It is strange that the philosopher whose approach to drama can often seem unnervingly reminiscent of his approach to shellfish (‘Find as many different kinds of specimen as you can, and then look for some generalizations that cover them all’) should also be the philosopher who at times seems to think that the only business of tragedy is always and everywhere to aspire to the condition of Oedipus Rex. As for the generalizations that the Poetics propounds, not even the Oedipus Rex itself seems really to uphold them. What goes wrong for Oedipus is not exactly to be summed up as ἁμαρτία, either in the classical sense of ‘a mistake, a going-​wrong’ or in the New Testament sense of ‘sin’; rather it is ἄτη, ‘ruinous delusion’. But more about these concepts a bit later. Still, we might expect Aristotle’s generalization about the tragic protagonist to be at least intended to cover Aeschylus’s Agamemnon as well as Sophocles’ Oedipus. And maybe it was. But there is some evidence the other way: some evidence that suggests that in fact Aristotle did not see Agamemnon that way. The evidence that I mean is at NE 3.1, 1110a26: ἔνια δ᾽ ἴσως οὐκ ἔστιν ἀναγκασθῆναι, ἀλλὰ μᾶλλον ἀποθανετέον παθόντι τὰ δεινότατα: καὶ γὰρ τὸν Εὐριπίδου Ἀλκμαίωνα γελοῖα φαίνεται τὰ ἀναγκάσαντα μητροκτονῆσαι. But some acts, perhaps, we cannot be forced to do, and ought rather to prefer to them death accompanied by the worst possible sufferings; thus, for instance, the things that ‘forced’ Euripides’ Alcmaeon to kill his mother seem ridiculous. We no longer have the play of Euripides to which Aristotle refers here; moreover, Euripides wrote a number of plays called Alcmaeon, one of them, it seems, a comedy. But in the best-​known version of the myth in question, what ‘forces’ Alcmaeon to kill his mother Eriphyle is the threats of his dead father. Eriphyle [p. 76] tricks her husband Amphiaraus into going into a battle in which she knows he will be killed; Amphiaraus then appears to Alcmaeon in a dream and tells him to avenge his death by killing Eriphyle. What Amphiaraus threatens Alcmaeon with is this: he tells him that the cost of not killing her is that horrible torments will be visited on Alcmaeon for his impiety in not avenging his father. So Alcmaeon does kill her—​and the cost of doing so turns out to be that he is pursued by the Erinyes for his impiety in killing his mother. More than one element in this narrative ought to sound pretty familiar to anyone who knows Aeschylus’s Choêphoroi. The inference seems clear. Aristotle would think (or possibly Aristotle thinks: maybe this is what he is getting at, when he issues this obiter dictum on Alcmaeon) that the threats that drive Orestes into killing his mother to avenge his father are equally ridiculous (even if Aristotle’s evident reverence for Aeschylus makes him less willing to use the word ‘ridiculous’ of him than of his bugbear Euripides). But then it seems likely that Aristotle’s attitude to the story of Agamemnon’s killing of his daughter will be equally dismissive—​just as, surely, Plato and Socrates too would have been highly unsympathetic to Agamemnon at Aulis. Strictly speaking, we do not know what Plato made of the story, because despite his extensive reference to Aeschylus in Republic II–​III, Plato never explicitly mentions the case of Aeschylus’s Agamemnon.4 But it is not hard to guess what Plato might have thought of Aeschylus’s tale—​if we leave aside Plato’s obvious emotional undertow of ambivalence about drama, an undertow that surfaces most clearly in the apparently temporary recantation of the Phaedrus.5 At least in the severe mood of his mature philosophy, Plato, like Socrates before him, must have seen the tale as a blasphemous, obscene farrago of voodoo and butchery, to which the best philosophical response is to pass over them in (what I take to be a disdainful) silence. (The nearest either Socrates or Plato gets to a direct response to such tragic material is, as Nussbaum points out (1986: 25), the Euthyphro.) 5. Agamemnon Is Guilty There seems, then, to be at least a chance that Aristotle, like Plato and Socrates before him, is not of the Agamemnon-​is-​innocent party, but of the opposite one. What the story of Agamemnon at Aulis prompts in him, on this reading, is the same imaginative resistance as he shows explicitly to Euripides’ Alcmaeon. Aristotle, then, does not see Agamemnon as an innocent caught in a predicament, but as a deluded murderer. And that, perhaps, means that Aristotle cannot see how to make sense 4 In the whole of the Republic, in fact, and so far as I know elsewhere in his works, Plato barely refers at all to any of the various Agamemnons in the Greek literary canon, though he does mention Palamedes’ comic Agamemnon at Resp. 522d, and (presumably) Homer’s at 620b. 5 On which see the fine paper in Nussbaum (1986). [p. 77] 64 Sophie Grace Chappell of Aeschylus’s play, any more than Plato or Socrates before him has a good way of making sense of it. The play falls apart for him aesthetically, because he cannot see how we could imaginatively identify with Agamemnon morally. Against this, I want to close by arguing that we do not have to be able to imaginatively identify morally with Agamemnon—​not, at least, if that means take up his moral position—​in order to make good sense of Aeschylus’s play. The play is not what it is taken to be by Williams and others of the Agamemnon-​is-​innocent school: it is not a depiction of a more or less noble person in a terrible situation. On the contrary, what the play depicts is someone of generally rather dubious character, who here and now, in front of us at the sacrificial altar at Aulis, drops off the edge of a moral cliff. In that moment, very suddenly and very easily, Agamemnon goes from being a rough, tactless, but no more than normally brutish soldier to being a moral monster; from being the king, husband, and father who left Argos at the beginning of the war, to being the atrocious butcher who returns at the end of it. (This before-​and-​after contrast is absolutely central to the structure of the play.) Agamemnon, then, is not an innocent, nor even a middlingly virtuous man; on the contrary, he is the epitome of murderous evil. But that does not mean that the play cannot work for us, because we cannot imaginatively identify with Agamemnon. In fact, it is all too easy to identify with him, to understand how easily he can slide into his terrible wickedness; just as it is with some of Shakespeare’s villains, with Richard III and Macbeth, for example. This kind of exercise of the imagination is both psychologically and morally disturbing: it is uncomfortable and alarming to see (and feel) such wickedness from the inside, and it is not hard to see why Plato might have wanted to suppress such imaginings, or thought of them as corrupting. But it is a kind of aesthetic achievement to make such visions of evil available to us. And engaging with them takes us a lot further than the simple antithesis that we began with, an antithesis that Fraenkel, Nussbaum, and Williams all buy into: the antithesis between finding some character in a drama to be basically innocent, therefore someone with whom we can imaginatively identify, and finding them to be basically wicked, therefore to be shunned—​and perhaps imaginatively resisted. In section 6 of this chapter I make out the case for this reading of the Agamemnon in closer textual detail. (The detail is not only from the text of Aeschylus.) 6. Agamemnon Is Guilty: The Receipts We may begin one significant, and perhaps underestimated, obstacle to the whole Agamemnon-​is-​innocent view that we have not mentioned so far: namely, that in most ancient depictions of him, Agamemnon is anything but innocent. This is true in Aeschylus’s play, for a start, where Agamemnon’s innocence is very far from being what Aeschylus emphasizes. In Aeschylus’s own text no one ever insists [p. 78] on Agamemnon’s innocence except Agamemnon himself. And as Aeschylus presents Agamemnon, the moment he arrives on stage he displays rude abruptness to Clytaemestra, telling her that her speech of welcome was too long; superstition, lack of resolution, and arrogance; not to mention the stunning tactlessness of showing up after nine years away from his wife with his mad-​barbarian-​princess concubine very prominently in tow in his baggage-​(or rather plunder-​) train. Indeed, we might look further back than Aeschylus here: to Homer, whose picture of Agamemnon is, I think, pretty strongly normative for the rest of the tradition including Aeschylus. Homer too very clearly and consistently presents Agamemnon as a clumsy, tactless, oafish, greedy, cowardly, brutal, and rather stupid blowhard. Anyone in Aeschylus’s audience would have been aware of the picture of Agamemnon that derives from Homer and others, as a brutal, overbearing, bullying, and often incompetent warlord. Moreover, what we have in the Agamemnon up to the point of Agamemnon’s crucial deliberating (and self-​exculpating) speech, 206–​17, is a picture of Agamemnon and Menelaus not as innocents, but as twin and inseparable forces of nature: dogs of war, pitiless eagles, bent irrevocably and remorselessly on a war of vengeance (see e.g. 108 ff., 122 ff.). And what we have in the rest of the Agamemnon is not in any way about Agamemnon as an innocent. It is about Agamemnon as a warlord, an agent of terrible destruction—​and it is about Agamemnon as so blinded by his own ἄτη, ruinous delusion, that he cannot see what the Watchman, Calchas, Cassandra—​and Clytaemestra—​can all see: his coming doom. Agamemnon is on an irreversible path to inexorable ruin: no image could tell us that more plainly than the purple woven way that Clytaemestra lays out for him on his return to her. (One of the first visual coups de théâtre known to us in the history of theatre; and still, perhaps, the greatest of them all.) But if we must speak naively of him as a figure who stands for something, it is hardly hapless innocence: in the Agamemnon there are no innocent characters, except the two women who get killed. He is the sacker of Troy and the rapist and enslaver of Cassandra. If Agamemnon must stand for any one thing, then what he ‘stands for’, I propose, is violation. And indeed, if there is any one thing that the whole play is about, then violation is what the Agamemnon is about. The key contrast that drives the whole action is the contrast so apparent in the chorus from which I have just quoted, the contrast between peace—​good rule, security, civilization, home, family, philia—​ and, well, war. It is the contrast between the well-​governed house and the disordered one with which the Watchman begins the play. It is the contrast between everything humans can build, and the nothing that is left when they choose, as Agamemnon does, to tear it all down. The pregnant hare too is a beautifully chosen image of this contrast. Helpless, soft, fugitive, she teems with life—​until the eagles, those emblems of death, seize her and disembowel her. If the huntress Artemis is angry, it is, apparently, at the unfairness of this contest; and Agamemnon against Iphigeneia too is no fair contest. [p. 79] 66 Sophie Grace Chappell The Agamemnon is about violation; and it is about Agamemnon as the agent of violation. To see his cause as just and pious, as Fraenkel and Nussbaum and Williams all suggest, is wildly out of line with what Aeschylus’s text actually tells us about him. It is not, for instance, true to say with Nussbaum, appealing to Ag. 55–​ that ‘the expedition was commanded by Zeus’;6 not at any rate if that is meant to suggest that we should picture Agamemnon as someone ‘piously executing Zeus’s command’. True, Aeschylus does write (60–​2): ‘Thus mighty Zeus the guest-​friend (xenios) sent (pempei) the sons of Atreus against Paris.’ But first, this is as consistent with saying that Zeus Xenios explicitly commands Agamemnon and Menelaus to make holy war on Troy as it is with saying that those warlords invoke Zeus Xenios as their patron, by which I mean their propaganda pretext, for a war that they choose to make because Menelaus has been insulted. The text is fully consistent with either reading; and the first is politically naive.7 And secondly, we should read on from lines 60–​2, to line 72. In these lines the Chorus do not, it seems to me, share Nussbaum’s faithful confidence in the justice of Agamemnon’s war, even if that war has the patronage of a most powerful divinity behind it. What they see is, first, a war of terrible suffering and destruction, and secondly, a war that is πολυάνορος ἀμφὶ γυναικὸς: all for the sake of one loose woman (62). For the Chorus throughout the play there is a recurring question about whether the whole war can possibly be worth it at all, given—​excuse their misogyny—​that it is only about one girl: we have already considered line 451, and compare the Chorus at 681–​781. The war/​peace contrast comes back here in modulated form: as the contrast between the individual and private injury that Menelaus suffers from Paris, and the public and universal calamity whereby he and Agamemnon avenge it. In the Agamemnon the Trojan War is not a just war; it is just a war. Just a war, without the moralism, but with everything that war means; which is, above all, injustice and the opposite of civilization. And Agamemnon is not a just warrior; he is just a warrior—​a man of blood and havoc, and a violator. The Chorus get this: from one end of the play to the other, it is what they sing about. And we get it too; unless we are taught not to get it, by readers of the play who are bent on seeing Agamemnon as an Aristotelian or Hegelian hero. What the first chorus of Aeschylus’s play tells us, and tells us quite explicitly, is that one key step—​perhaps the key step—​in the process that makes Agamemnon 6 And where Nussbaum gets the notion that if Agamemnon ‘does not fulfil Artemis’s condition, everyone, including Iphigeneia, will die’, I have no idea. Aulis (modern Paralia Avlidos) to Argos is about 50 miles. If the army get really hungry, they can always just spend two or three days walking home. 7 And after all, at Ag. 125 even the eagles who eat the pregnant hare are ‘the senders of the force’, pompous archas. Should we take this the natural way, as meaning merely ‘portents that set the expedition on its way’? Or are they too proclaiming a holy war, issuing a command that binds Agamemnon on pain of ‘the most serious impiety’? For Zeus xenios again, see 352, towards the beginning of the obviously propagandistic and triumphalistic choral ode W Zeu basileu kai Nyx philia, ‘O Zeus the king, O Night who fights for us . . .’. [p. 80] into a violator is the sacrifice of Iphigeneia. That is the crucial point about the narrative of Agamemnon at Aulis: not primarily that he is in a dilemma—​though no doubt he is, of a sort—​but how he responds to that dilemma; not primarily his self-​excusing speech τί τῶνδ᾽ ἄνευ κακῶν (‘Which of these options is without evils?’)—​though that is important too—​but what comes next. Which is this: ἐπεὶ δ᾽ ἀνάγκας ἔδυ λέπαδνον φρενὸς πνέων δυσσεβῆ τροπαίαν ἄναγνον One way to put this literally is: ‘When he put on the yoke-​strap of necessity, breathing an impious, unholy, unsanctified change of mind, then he came to recognize the thinking of all-​daring.’ Or as I have put it in verse:8 So with relief he gave it Necessity’s name. strophe 5 Once necked in his yoke, though, we smelt off him something unclean, something desanctified, something set free to defile; 220 new look in his face that said All is permitted for me. Agamemnon here puts on ‘the yoke-​strap of necessity’: he chooses it, his putting it on is not something that happens to him, but an action that he performs. In Williams’s words: [Denys Page] simply misrepresents the text. Aeschylus does not say that Agamemnon submitted to necessity. The word ἔδυ . . . is a straightforward verb of action, which means (as Page himself elsewhere translates it) ‘put on’, and Agamemnon is said to have put on the harness of necessity as someone puts on armour. (1993: 133) Now there is indeed, as many commentators (e.g. Nussbaum) have seen, something paradoxical about the idea that Agamemnon here should freely choose not to be free. But the paradox is deliberate, and it is Aeschylus’s paradox. The point is that Agamemnon is rationalizing. He is claiming that he must do something—​kill Iphigeneia—​that, in fact, he does not have to do at all. As indeed Clytaemestra points out after she has killed him (1415–​17), there is something routine about his killing of Iphigeneia—​as if she were just another lamb9—​that makes it very far 8 When they are not literal renditions, my translations mostly follow Chappell (2024). 9 Comparisons between killings of humans and of animals are indeed, as Nussbaum insists (1986: 32 ff.), pervasive in the Agamemnon. That does not justify Nussbaum in reading the people-​plenty of herds before [Troy]’ (Ag. 129) as meaning that Calchas ‘predicts only that the army, in laying siege to Troy, will slaughter many herds of cattle before its walls’, and erecting an exegetical puzzle on the basis of this reading: ‘No significant omen merely predicts a beef dinner.’ Indeed not, which is an excellent reason for not taking Calchas’s words Nussbaum’s way. [p. 81] 68 Sophie Grace Chappell from being a specially forced action. The routine of sacrifice is familiar, except that it is his daughter he now kills; but Agamemnon oversteps the inhibition of that detail with horrifying ease. ‘But isn’t Agamemnon forced to do it by Calchas’ prophecy?’ The short answer to that, as every Athenian knew, is that if you don’t like a prophet’s divine word, it is usually possible either to ignore him, or else to get another. (Another divine word, or another prophet: or indeed both, if necessary.) Prophecy was (and is) a political business, and Aeschylus knew it: nearly all his references to Calchas are loaded with a sinister irony and with sarcastic hints of self-​serving priestcraft. (Consider, in particular, 249, τέχναι δὲ Κάλχαντος οὐκ ἄκραντοι, ‘But Calchas does not scheme unenacted schemes’. Calchas is a man of schemes, a schemer. ‘Another way’, when he suggests one, is ‘another device’, ἄλλο μῆχαρ, 199.) Agamemnon’s choice to follow Calchas’ augury at Aulis is just that, a choice. Perhaps Aeschylus intends us to notice the contrast with Iliad Book 1, where Calchas again prophesies something difficult for Agamemnon—​and Agamemnon simply refuses to comply. Surely it is easier for Agamemnon to do what Calchas says and give up his trophy-​slave Briseis, than it is for him to do what Calchas says and kill his own daughter. Yet the Agamemnon who rebuffs the easy command in the Iliad obeys the hard command in Aeschylus’s play. Agamemnon, as Aeschylus portrays him, is not forced to kill Iphigeneia at all; he rationalizes that he is forced to, but he is not. As indeed the Chorus close-​to-​ explicitly say: βροτοὺς θρασύνει γὰρ αἰσχρόμητις τάλαινα παρακοπὰ Bad wisdom sets us out of reach of shame, then traps mortals in the oldest snares of all’ (223–​4). And all this goes to explain why, once Agamemnon—​freely—​‘puts on necessity’s yoke-​strap’, the soi-​disant witnesses who are the Chorus in the play immediately see a sinister and frightening change in him. He has stepped over the edge of the moral abyss; he has become someone for whom τὸ παντότολμον φρονεῖν—​seeing nothing as unthinkable—​has become easy and natural; he has lost all normal moral incapacity (cp. Williams 1992). In short, he has made a decisive move from the condition of peace towards the condition of war, which is also the condition of the rapist or the violator. Both Nussbaum and Williams, it seems to me, seriously misunderstand this change in Agamemnon. I think they misunderstand it because, bluntly, they just misread the text of Ag. 206–​27. Nussbaum (1986: 35–​6) is simpler to deal with, because (as Williams in effect remarks, 1993: 134) she is too plainly driven simply by the desire to see Agamemnon in terms of the wider schematism of Fragility’s Williams-​esque anti-​utilitarian polemic, as merely eliminating the ought that he does not act on. ‘Agamemnon seems to have assumed, first, that if he decided right, the action chosen must be right; and second[ly], that if an action is right, it is appropriate to want it, even to be enthusiastic about it’ (1986: 36). Here we are once more in Agamemnon-​is-​innocent [p. 82] territory, and once more I merely remark that Aeschylus goes out of his way from one end of his play to the other to make that territory unavailable to us, and to get us to see Agamemnon, instead, as a violator (though certainly also a violator who is himself violated by becoming a violator, and who can, for instance, weep at what is happening). As for Williams: in Shame and Necessity (1993: 134, cp. 208–​10n11) Williams tells us that ‘it is entirely clear what happens’ after the end of Agamemnon’s speech at Ag. 206–​17: ‘the father slaughters his daughter in a state of bloody rage’. But this is not only not entirely clear; and not only not clear at all; it is not even what happens. Lines 218–​27 contains no word that means ‘bloody’, and only one word, παρακοπὰ, that even can mean anything like ‘rage’—​though it can also mean what I translated it as meaning four paragraphs back, ‘infatuation’. And as we have already seen, what the lines suggest overall is not that Agamemnon is somehow suddenly carried away in a frenzy of bloodlust. Rather—​and just as Clytaemestra complains at 1415–​17—​what happens, perhaps even clearly happens, is that Agamemnon—​ once he has wiped away his tears—​quite coolly and calmly proceeds to perform what is a perfectly normal propitiatory sacrifice in every respect except one—​that it is the sacrifice of his daughter. Anyway, what could it mean, psychologically or dramatically, for Agamemnon suddenly to drop into a rage like the one that Williams foists on him? And why would it happen? On Williams’s misreading, Agamemnon’s behaviour simply becomes mystifying, as Williams himself obliquely admits: ‘This is not a text’, he writes (1993: 134), ‘that invites us very far into psychological interpretation.’ On the contrary, it seems to me, these lines most assuredly do invite us to interpret Agamemnon’s experience at the sacrificial altar of Aulis. But Aeschylus invites us to see him, not as suddenly filled with self-​induced random rage, but as undergoing, more or less with open eyes, an icy-​calm process of corruption. In this part of the text, the only thing that Aeschylus has anyone say about any mental state anything like rage is that it is themis to wish for Iphigeneia’s blood orga(i) periorgô(i), ‘with over-​passionate passion’ (Ag. 215–​17; Agamemnon’s words). But first, as Williams himself agrees (1993: 209), this does not have to be Agamemnon’s passion. With or without a textual conjecture that Williams rejects on the same page—​contrast Sommerstein, who accepts it (2008: 26–​7)—​it is possible to take it to be either the army’s (as I translate it myself ) or indeed no one’s in particular. Saying that a feeling is themis is different from saying that anyone in particular has it, or should have it. Secondly, even if it is Agamemnon’s own passion that he speaks of, a passion that you need to tell yourself it is themis to have does not sound like much of a passion. Certainly a ‘state of bloody rage’ is not something that Agamemnon will be able to whip himself up into, simply by reflecting that it would be themis to be in that state. How else could Williams read rage into this passage, as something that is not merely there, but entirely clearly there? I don’t know. Maybe he is led astray by [p. 83] 70 Sophie Grace Chappell Lloyd-​Jones (1962), who correctly identifies parakopa (Ag. 223) with atê, then translates both by ‘derangement’. But this does not get us to rage either. Atê is like English ‘delusion’: it can mean madness, but also blindness or misunderstanding; and either way it is an externally identified state: it is a state in which, by definition, things are not the way they seem from the inside. Compare Iliad 19.91, where Homer’s Agamemnon, apologizing for his own atê in the quarrel with Achilles in Book 1—​which patently wasn’t madness—​makes a folk-​etymology with aatai, ‘deceives’. 7. Conclusion What Agamemnon at Aulis shows us is not so much a classic moral dilemma in Hegel’s or Kierkegaard’s or even William Styron’s sense. There is, of course, a dilemma of a kind before Agamemnon, but it is not Aeschylus’s main intention to display an innocent man facing an impossible dilemmatic choice and thereby incurring ineliminable agent-​regret. It is to describe, in detail and from the inside, the psychological processes whereby someone becomes capable of extreme and horrifying evil. Maybe Agamemnon does feel ineliminable agent-​regret; but given that his regret is about ‘needing’ to sacrifice his daughter in order to keep his war of pillage going, there is an awkwardness about using him as an example of that phenomenon. Maybe Himmler felt ineliminable agent-​regret too; we are still likely to see Himmler as an apter example of corruption than of regret. And the same applies to Agamemnon: the most that can be said of him, as an example of agent​regret, is that it just goes to show how corrupted he is, that this should be the content of his agent-​regret. What Aeschylus shows us is how a man, who is a husband and a father as well as a king and warrior, can be—​and very quickly—​so taken over by his war-​making role as to become the destroyer and enemy of everything that is involved in his family roles. What Aeschylus is showing us, in short, is how a man can become a violator. In this dramatic masterpiece his achievement is to draw us psychologically even into the experience of someone who is in the very throes of being irreversibly and irremediably corrupted. The Agamemnon is, then, another case—​alongside famous Shakespearean cases like Richard III and Macbeth—​of the power of narrative art to get us its audience inside all sorts of ‘views from somewhere’: even into the head of someone seriously evil. To presuppose that we cannot enter imaginatively into someone’s life and experience unless they are, at the very least, not seriously evil is the key presupposition of the Agamemnon-​is-​innocent school, including Williams. But I have argued that this presupposition should be rejected. Its key flaw is, in the end, just a lack of imagination. [p. 84] References Arendt, Hannah. 1951. The Origins of Totalitarianism. New York: Schocken Books. Chappell, Sophie Grace. 2022. Epiphanies. Oxford: Oxford University Press. Chappell, Sophie Grace. 2024. Aeschylus’ Agamemnon: A New Translation. Durham, UK: Ellipsis Imprints. Fraenkel, Eduard. 1950. Aeschylus’ Agamemnon. Oxford: Clarendon Press. Lloyd-​Jones, Hugh. 1962. ‘The Guilt of Agamemnon’. Classical Quarterly 12.2: 187–​99. Nussbaum, Martha. 1986. The Fragility Of Goodness. New York: Cambridge University Press. Sommerstein, Alan. 2008. Aeschylus’ Oresteia. Cambridge, MA: Loeb Classical Library. Styron, William. 1979. Sophie’s Choice. New York: Random House. Williams, Bernard. 1973a. ‘A Critique of Utilitarianism’. In Williams with J. J. C. Smart, Utilitarianism: For and Against, 75–​150. Cambridge: Cambridge University Press. Williams, Bernard. 1973b. Problems of the Self. Cambridge: Cambridge University Press. Williams, Bernard. 1981. Moral Luck. Cambridge: Cambridge University Press. Williams, Bernard. 1992. ‘Moral Incapacity’. In Proceedings of the Aristotelian Society, New Series, Williams, Bernard. 1993. Shame and Necessity. Berkeley: University of California Press. Williams, Bernard. 1995. ‘Replies’. In World, Mind, and Ethics: Essays on the Ethical Philosophy of Bernard Williams. Edited by J. E. J. Altham and Ross Harrison. Cambridge: Cambridge University Press. [p. 85] Bernard Williams on Truth and Plato’s Republic on Justice What Are Genealogical Arguments Good For? Catherine Rowett In his Truth and Truthfulness (Williams 2002; hereafter T&T), Bernard Williams sets out to explain our commitment to virtues of sincerity, truthfulness, and accuracy in human discourse and behaviour. He recommends the ‘genealogy method’ as a good (or the best) way to approach such a question: ­chapter 2 of T&T is devoted to explaining that method, how it works and what it is good for. He had also devoted an earlier paper, ‘Naturalism and Genealogy’ (Williams 2000; hereafter N&G), to the same issue. In this chapter, I first (section 1) review what Williams has to say in T&T about the virtues of genealogy as a philosophical tool. I consider how it accords with naturalism in philosophy, in the sense that Williams has in mind. In section 2, I argue that Plato’s Republic fits this description, using a naturalistic genealogy to vindicate the value of justice. In section 3, I suggest that Plato uses his narrative to answer three questions: what justice is, why or how it came to be valued, and whether it is beneficial. By contrast, Williams, in T&T, does not use genealogy for the ‘what is it?’ question, but only for the other two. I then consider some potential risks that arise in using genealogy for the explanatory or evaluative questions: I engage with some challenges that have been raised against Williams’s defence of truthfulness, and outline some possible advantages of using genealogy to address the ‘what is it?’ question, as Plato does. Finally, in conclusion, I briefly consider why Williams did not include Plato’s Republic among his examples of naturalistic genealogies. 1. Bernard Williams on the Genealogy Method In this section I explore Williams’s enthusiasm for genealogy as a method in philosophy. Our main text is ­chapter 2 of his T&T (Williams 2002), but the issues are also discussed in N&G (Williams 2000), to which I shall also refer. Catherine Rowett, Bernard Williams on Truth and Plato’s Republic on Justice In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. DOI: 10.1093/​9780191966361.003.0005 [p. 86] Vindicatory Genealogy By ‘genealogy’ Williams means a philosophical inquiry that proceeds by telling an imaginary story about how a certain human practice developed in a primitive society, or how it came to be respected or treated as a virtue. Such narratives are quasi-​historical in character, but are typically a kind of fiction or legend. Most of the cases that Williams considers are ‘State of Nature’ narratives: that is, stories that start by imagining human beings in a pre-​social or pre-​culture condition. They then tell a story about the developments that might have led from that pre-​social state to the situation that we have now. Thus they sketch developments that seem plausible or likely, given the nature and likely choices of the players in the story (Williams 2002: 32–​4). Using this narrative, the philosopher considers how such values and practices contribute to society; one can show how, with those institutions in place, a sophisticated society could have emerged—​as it evidently has; one can offer a plausible explanation for why people would embrace those values and practices. It is irrelevant that these are not historical facts: a genealogical account explains why, at certain stages of the narrative, agents in that situation would be inclined to adopt or promulgate those values or practices. This is what Williams calls a ‘vindicatory genealogy’ in its naturalistic form. A vindicatory genealogy aims to show that the specified value or practice makes good sense, because the resulting social order serves society well.1 If all the stages can be explained by appealing to natural tendencies, and the explanation appeals solely to natural causes, it counts as naturalistic, and thereby gives a naturalistic vindication of the practice.2 It vindicates the practice, by showing that rational agents would naturally find it desirable, and would naturally want to endorse it, for good reasons (Williams 2002: 36–​7). Naturalism and Naturalistic Explanation Williams favoured genealogy partly because he saw it as a naturalistic method, in the sense just explained: a method that bases its explanations on an examination of human nature, its natural conditions and needs, taking ‘human nature’ to be part of the rest of nature, as examined by the natural sciences (Williams 2000: 148–​ A ‘State of Nature’ genealogist bases the account of the original 1 There can also be debunking genealogies that aim to show that a practice or institution has no natural value or rationale. Nietzsche’s Genealogy of Morals seems to be an example of that kind. See Williams (2002: 12–​19). But see Queloz and Cueni (2019) who argue that Nietzsche is not himself debunking but rather debunking the debunking. Nietzsche’s legacy regarding the value of truth also provides the provocation for Williams’s defence of truth and truthfulness (Williams 2002: 12–​13). 2 There can also be non-​naturalistic genealogies, where, after failing to find a naturalistic explanation of this kind, one resorts to some other kind of causation. An example can be found, for instance, in Protagoras’s Great Speech in Plato’s Protagoras. See Rowett (2022), and below, section 1.2. [p. 87] Catherine Rowett condition on an implicit account of what humans are like by nature, before culture and other interventions get going, and then imagines the subsequent stages by considering what creatures of this nature, with these natural desires or rational capacities, would do. Their natural tendencies are used for explaining why they choose these social, political, or ethical constraints, and establish or uphold institutions that preserve them; and granted the story may stray from historicity, because events could not have been quite like that,3 yet the story renders the outcome comprehensible, because it appeals to facts about human nature. The outcome is explicable within the existing laws of nature. However, as Williams notes in T&T there is a difficulty about stabilizing the notion of ‘nature’ to which this idea of naturalism in philosophy implicitly appeals.4 In both N&G and T&T, Williams discusses how the resulting narratives constitute a ‘naturalistic’ explanation, as he understands it (2000: 148–​9; 2002: 22–​3), but he offers hardly any examples of explanations that count as ‘non-​naturalistic’, by his typology. In T&T his one example, taken from science rather than philosophy or ethics, is ‘vitalism’ in biology (: 23). It is less clear what would make an ethical stance ‘non-​naturalistic’. However, in N&G he offers divine-​command ethics as an example (2000: 148), and there are hints elsewhere in T&T that he took Plato to be a non-​naturalist in ethics: he comes close to saying so explicitly in T&T (2002: 61), where he appears to reference Plato’s transcendent Forms as a non-​naturalist account of value. These hardly provide a general account, but it seems that an ethical stance would count as ‘non-​naturalistic’, (1) if it invokes supernatural or divine causation; if it appeals to external metaphysical foundations or truths beyond the scope of natural science; or (3) if it suggests that some practices, institutions, or values are intentionally established by artifice to override natural human tendencies (e.g. natural tendencies towards evil or natural sexual desires) so as to retrain our desires or replace them with other inclinations (e.g. towards ‘virtue’, as something unrelated to, or in conflict with, natural impulses). However, this does not mean that an explanation invoking the Form of the Good would, ipso facto, count as non-​naturalistic, for—​as Williams himself grants (2002: 61)—​Plato conceives of human beings as naturally fitted for recognizing the forms and being appropriately motivated thereby. Indeed, any attempt to show that human morality is not natural, just because it appeals to standards independent of human nature and desires, would potentially collapse into naturalism.5 On that basis, Plato’s middle period ethics would be willy-​nilly naturalistic. Nevertheless, this question remains: can our ability to access the Forms during 3 That is, the story is fact-​defective, (Williams 2002: 31). The terminology is from Nozick (1974) referring to Hempel (1965). 4 See below, note 5. 5 This is part of what Williams means by the inherent and systematic instability of the idea of nature and ‘naturalism’, (as mentioned above): (Williams 200: 148–​9; 2002: 22). [p. 88] this life, and to appreciate the value of virtues such as justice, in private and public life, be explained in entirely naturalistic terms? That they can, and should, be so explained is, I submit, what Plato tries to show in the Republic, using a naturalistic genealogy to do so. 2. Plato on Justice: A Genealogy? A Naturalistic Genealogy? The Republic as a Genealogy of Justice In this section I argue that the ‘founding’ of the city in Republic Books II–​IV is a vindicatory genealogy narrative of the kind that Williams had in mind.6 Socrates tells a story about the origins of justice in a primitive community to discover what justice is and ‘whether it pays’.7 The central and later books of the Republic also address other questions, using an array of other methods, some of which I have discussed elsewhere.8 My main focus here is on the first four books, though I shall suggest that the genealogy method is still at work in later sections too. Socrates begins by imagining the formation of a small primitive settlement in which individuals gather to share produce and tasks. He then traces the emergence of the so-​called ‘city of pigs’, and subsequently proceeds to build a more complex community with more luxurious institutions and values. He asks when and where in this process we should see justice emerging as a virtue or practice of the community, and he explores what that justice consists in and whether it is intrinsically desirable and valuable, and seen as such by the community. A few extracts from Republic Book II should suffice to support these claims. First, at 369a–​e, Socrates explains the idea and how it will work as an inquiry into justice. He and his companions will reconstruct, in imagination, an emerging community so as to watch a case of justice emerging in the political sphere. This will allow them to observe the development of justice: T1 Well, I said, if we could, in our discussion, watch a polis-​community in formation, we’d be able to observe its justice and its injustice developing, wouldn’t we? —​Probably, yes, said Adeimantus. 6 As far as I know, no one has proposed this before. Cross and Woozley (1964: 80–​2) toy with it, when comparing Plato’s treatment to social contract theories, but they conclude that ‘this seems an unlikely interpretation of Plato’, on the grounds that there is no social contract in Plato’s first city, and that Plato is not claiming to write history. But, as we have seen, genealogy neither aims at historicity, nor confines itself to contract explanations. 7 The narrative is vindicatory, because it sets out to show (against debunking rivals such as Thrasymachus and the ‘immoralists’) that the virtue of justice is a genuine virtue and choiceworthy: that it pays to be just. On the ‘immoralists’, see below, conclusion. 8 Especially Rowett (2018: Part III). [p. 89] Catherine Rowett And we can expect to see what we’re looking for more easily, once it’s emerged. a6–​b2)9 The idea is to examine the emergence of a certain way of thinking and behaving in the primitive community, and work out what makes it happen. The story is overtly naturalistic, in that Socrates derives the first collective on the basis of our natural need for food, shelter, etc.: T2 Well, I said, this is how I think a polis-​community comes to be formed: it’s because as things are, each of us is not self-​sufficient as an individual, but is in need of many things. . . . So with people needing lots of things, and exchanging them with each other, for this need or that, and when many gather together into one dwelling place to be companions and helpers to each other, it’s to that common habitation that we give the name polis. Is that right? —​Absolutely. (369b7–​c5) He then explains that the practice of exchange emerges because we are naturally needy, not self-​sufficient: T3 Right, let’s create our polis from scratch in words. It’s our neediness that will create it, apparently. —​Quite so. (369c9–​11) Plato cites our neediness as the source of a solution, not just the vulnerability (by contrast with, for instance, David Hume and Protagoras in Plato’s Protagoras).10 Plato sees our lack of self-​sufficiency as a natural prompt, causing us turn to each other for mutual support. The story begins from a minimal community. Just the basics are covered. A genealogy need not always be about early man, and Socrates does not say that he is describing the first ever polis, but equally nothing tells against that: T4 Let’s see, I said. How big a community will suffice for providing for that much [sc. food, shelter and clothing]? We’ll need one farmer, won’t we, and one builder and one weaver? Or shall we add a shoemaker, or something else to do with the care of the body? —​Yes, indeed so. So the minimal subsistence community is composed of four of five people? —​It appears so. (396d6–​e1) 9 All translations are my own. 10 On Protagoras, see Rowett (2022); on Hume, see Williams (2002: 33–​6). [p. 90] He then considers the advantages of specialization and division of labour, noting that people are not all the same, but are naturally suited for different activities. After a while the community will have to ‘expand’ somewhat: T5 Well Adeimantus, we need more than four citizens for the provisions that we identified. Or the farmer will not, it seems, be making his own plough for himself, if it’s to be a fine plough, nor his hoe or the rest of his implements for farming. Nor will the builder make his own implements: he will need a lot of such things too. And the same for the shoemaker and the weaver. —​That’s true. So carpenters, blacksmiths, and a whole lot of other craftsmen of that kind will be joining in as participants in our teeny weeny mini-​state, and making it rather crowded. (370c8–​d7) Socrates expands the community stage by stage, adding more expert craftsmen and traders, further institutions for exchange—​a market place, coinage—​and a distinction between merchant and producer. Finally he adds unskilled labourers hired for their physical strength, and at that point he declares the community complete (371e8–​9). It is ready, he suggests, for answering the justice question: T6 Where exactly in this community do we find its justice and its injustice? And with the arrival of what development did they materialize? —​I’ve no idea, Socrates. Unless it’s in some intercourse that they have with one another. Perhaps you’re right, I said. We’ll need to investigate, and not shirk the task. So first we must investigate what manner of life the people who are thus provided will live. (371e11–​372a5) However, when they examine the primitive community, its plain diet, its simple amusements, Glaucon complains that it is fit only for pigs (372d5). He suggests adding furniture, and some more elaborate cuisine. This prompts Socrates to continue the story, imagining the community introducing various kinds of luxury. There is much to puzzle about, in the sequence just described, but the relevant point for now is that Socrates is sketching a quasi-​historical narrative, allowing us to watch the progression from primitive to more developed, and noting when the community would establish the relevant institutions. My point is that this part of Republic Book II can be read as a philosophical genealogy for justice. The narrative is constructed: it is not intended to be accurate history, nor does it place the primitive city in any particular place or time. The idea is to imagine how a polis would develop, and to explain why people so placed would rationally choose the institutions that we now have and would find them worth having. The participants choose to engage in commerce with each other because the results are rationally [p. 91] Catherine Rowett choiceworthy for the participants (369c6–​7). We are surely to infer that the subsequent developments are similarly motivated. Practices evolve and become institutionalized because they are found to be choiceworthy. I turn now to two questions, to be addressed in this section. These are whether Plato’s genealogy of justice is naturalistic, in Williams’s sense (section and (2) how Plato addresses the ‘what is it?’ question alongside the question about the value of justice (section 2.3). Is Plato’s Genealogy Naturalistic? Is Plato’s genealogy naturalistic, in Williams’s sense, as explained above? By contrast with the Great Speech in the Protagoras,11 the Republic narrative does not require divine intervention to enable the inhabitants to cooperate. Instead, as we have seen, Plato supposes that their natural needs suffice to prompt them to collaborate. Socrates pictures human beings as naturally needy, collaborative, and rational, capable of seeing how to improve their lot by collective settlement and sharing of tasks. On that basis, the first community develops naturally, without help from the gods. Unlike Aristotle, Socrates makes no mention of the nuclear family and does not begin with family groups.12 In the Republic, justice requires that people are assigned roles according to natural aptitudes, not in virtue of their sex, age, or position in a family. Strikingly, Socrates’ embryonic community comprises single adult genderless citizens taking appropriate roles as individual contributors to a collective. So while his approach is similar to Aristotle’s—​both take the state to be a natural extension of basic cooperative relationships—​Plato’s basic cooperative relationship is a working commune in which all participants are equals, while Aristotle’s is a hierarchical, monarchic, patriarchal household with slaves. In the first stages of Plato’s simple city, the division of labour flows from the natural variations in individuals’ aptitude and interests: T7 That’s not terribly strange, by Zeus, I said. I was just thinking, as you said that, firstly that we’re not born exactly alike, each and every one of us, but with differing natures, one fitted for the activities of one task and another for another. Don’t you think so? —​Yes, that seems right. 11 Plato Protagoras 320c8–​323a4. 12 Aristotle approaches his task biologically, and begins from the pairing of male and females for reproduction (Aristotle, Politics I.2, 1252a24–​31). Compare Hume’s account of the origin of justice in Treatise, Book III, Part II, Section 1. [p. 92] So it follows that one will get more, finer, and easier produce when one person is engaged in tasks in accordance with their nature and at the right time. (370a7–​c6) This proposal appeals to nature, but not a single essence of what it is to be human, nor a list of universal human characteristics or ‘human nature’, but to the idea that we naturally differ. Precisely because people are different, a division of labour is natural, and it naturally follows that a collaborative community is naturally advantageous. For we get better shelter, better shoes, and better food if we don’t try to make all of them ourselves, given that not all of us are good at all those things. This first stage, leading to the ‘city of pigs’,13 does not envisage the invention of any statutory institutions of justice, since that small egalitarian community works through a simple collaborative distribution of productive tasks among equals, with tasks assigned according to their natural aptitudes. Scholars have puzzled over the matter, but it seems to me that we should not infer that such a community is devoid of justice, or that justice has not yet arrived and is found only in the more complex ‘fevered’ city. A city that does not require any institutions to promote or enforce justice need not be a city in which justice is not practised or promoted. Such a city may promote and practise a virtue without the need for institutional mechanisms. And indeed in this city we see already the practice of distributing tasks on the one-​ man-​one-​task principle, which is what Socrates will later find is the key principle of justice. On that basis, the city of pigs already illustrates the emergence of (an embryonic form of ) justice, as a natural by-​product of the distribution of labour by aptitude.14 If that community falls short of justice, perhaps it is only in the fact that it allows no space for those whose gifts are in intellectual and artistic fields, since in the city of pigs, all the collaborators are apparently engaged in utilitarian productive work. The Luxurious City: A Naturalistic Genealogy Continued We might ask about the move from the ‘city of pigs’ to the luxurious city. Is that naturalistic too, or does it appeal to anything supernatural or artificial? Here again, I submit, the argument is naturalistic, and here again Plato focuses on natural differences, not some universal human nature. When Glaucon complains about the pigsty conditions in the simple city, Socrates agrees that ‘those things won’t be enough for some people’ (372e8–​373a1). Some people want fancy stuff; but not everyone feels like that: it is not an essential characteristic of human nature, but for those who have such desires, it is in their nature to be dissatisfied with a frugal 13 Resp. 372c. 14 This could raise issues about whether what one finds in a primitive community is continuous in any meaningful way with whatever current practice is to be explained by genealogical enquiry. On these issues and other questions about the validity of genealogical enquiries, see Queloz (2020). [p. 93] Catherine Rowett lifestyle. Not everyone would be delighted with the simple uncluttered life of the first city, though Socrates would. Others, like Glaucon, delight in a multitude of cultural paraphernalia. That too is a way of being human. Clearly, then, the luxurious city too is a natural development. It develops because people are naturally disposed to choose it. Such desires are not unnatural, nor the work of, say, the devil (or some such), as in narratives from other traditions about the origin of sin and corruption. Rather, Socrates describes the natural emergence of a more complex community, based on the natural pleasure that some people take in luxuries. Socrates uses this development to reflect on how wars start, once the people will not be satisfied with the produce of their own region (373e). As a result this luxurious city will need an army, and this opens the question of whether a professional army is required, with expertise in warfare, according to the principle of distributing tasks by natural aptitude. This initiates a discussion of the requisite education and training (for skills and character) to secure a guardian class that is both passionate about defence and also gentle and cultured towards fellow citizens, and with a love of learning (374b–​376c). Throughout all these developments, including the nurture of young citizens, the task is always to fit each citizen’s role to his or her natural abilities, and to develop those abilities and characteristics to full maturity, by techniques that awaken, confirm, and cement their natural capacities—​a form of training to deliver what we call ‘second nature’. There is no attempt to divert people from what they are naturally cut out to be. It is into this naturalistically developed city that Socrates and his companions look, in order to discover where and how its virtues are instantiated (427c6–​d7). Many scholars assume that Socrates offers a definition of justice when he identifies the distribution of roles in the community as constituting its justice (432d). In Rowett (2018: ch. 7) I argued that this is not so. What is discovered at 432d is not a definition but a singular case of intra-​state justice (the justice in that exemplary city, its justice).15 With further work, to examine whether justice is similarly instantiated in personal morality, Plato develops a fuller picture of ‘what justice is’ as a generic virtue across multiple domains. However, although the genealogical narrative constructs just one imaginary city, it does provide answers to more generic questions: when we ask ‘In a city such as this, what is there, that we could call justice?’ (Socrates’ question at 432d), we secure a first answer to the ‘what is it?’ question about justice. And, as with all genealogical narratives, although we answer it for the specific city that Socrates has constructed, the answer is also iconic in a way, because the exemplary primitive city, laid bare for us from its earliest 15 To capture the sense correctly, the definite articles in the Greek at 432b5, b10, and 433b4 need to be retained in the English, not omitted as in all current English translations of the text. [p. 94] beginnings, is representative of how we all got to where we are now, from some condition such as this, and for reasons such as this. The founding of the city is the founding of an exemplar.16 In all this, Socrates presents entirely naturalistic explanations. While the most basic things are traced to a generic human nature—​our neediness, our willingness to collaborate, etc.—​the more complex developments reflect people’s divers natural gifts, aptitudes, and inclinations. What emerges is a diverse and inclusive community where people contribute according to their own distinctive interests and talents, rather than a uniform population competing for the same jobs and the same rewards. In sum, it is a naturalistic genealogy, in which Plato explains the development of a sophisticated society by appealing to human nature (not, as in most modern genealogies, a nature common to all humans, but rather the very fact that we differ). Plato’s story does not end at the founding of the luxurious city. In the subsequent books, where Plato explains how justice will be understood, valued, and preserved in the ideal version of this city, he continues to appeal to the differences between human beings, as much as, or more than, to what they all share. The roles that individuals take in establishing and preserving justice in the city vary from person to person, and indeed what it is to be a just citizen differs accordingly. In fact, justice is identified as the practice of giving people roles and responsibilities that accord with their natural aptitudes: to deliver a just and fitting distribution of power and resource, yielding good outcomes for all. As a result, to explain why a certain action is just, when it is just, and who should do it, it is always necessary to mention the individual’s natural abilities and inclinations. There are no general truths about what is just for everyone. And since their education is designed to nurture citizens in accordance with, not against, their natural aptitudes, and to open their eyes to the true values built into nature, those who fully realize their capacity for objective knowledge will also be equipped to discover truths and values that are not merely human values, but also natural.17 Thus it seems that even with Forms in the picture, and in a situation where some individuals base their decisions on knowledge of absolute truths that transcend human society and its practices (not on discovering instrumental benefits that would accrue), it remains true that Plato’s account of the developed structures and virtues in Kallipolis is entirely naturalistic. He sets out to show that this is what would happen naturally, if we let people follow their natural talents, and did not mess things up by having people do what they are not naturally fitted to do. 16 On the idea of iconic exemplars and the methodology of the Republic, see my fuller treatment in Rowett (2018). 17 On this, see Rowett (2016). [p. 95] Catherine Rowett 3. What Questions Can We Answer Using a Genealogy? Williams favours genealogy as a method for addressing two questions about a practice or virtue: a ‘why do we have it?’ question, and a value question (‘is it worth having?’). In Plato’s Republic, we typically express the value question with the formula ‘does it pay?’.18 Correspondingly, in T&T, Williams is asking ‘whether truthfulness pays’, by considering whether human beings in the primitive society would choose it, and would stand to gain benefits or advantages thereby. That is the value question. To answer it, the genealogical narrative imagines a community inventing such a practice and finding it advantageous, and considers how such people could discover those advantages. These moves not only address the value question but also help to answer the ‘why do we have it?’ question—​or at least Williams seems to suggest that they do. While these two questions are distinct, Williams tends to conflate them in his own explanation and justification of truthfulness. We have such institutions, he implies, because they are beneficial in some ways and were seen to be so. We can contrast those two questions with the ‘what is it?’ question about a virtue or value, which is a question that Socrates often claims we must ask first. So there are, in all, three questions: (1) ‘what is it?’; (2) why does it occur?; and (3) does it pay? Plato starts the Republic with ‘does it pay?’,19 but at the end of Book I (354b–​c), Socrates chides himself for moving too soon to that question, when one should answer his ‘what is it?’ question first. In Book II, Socrates starts again, asking ‘what is justice?’. It is striking that this is the point at which Socrates embarks on genealogy (as just described), in order to look for (as he puts it) ‘justice writ large’ in a community instead of an individual.20 Plato is using his genealogical narrative not only to find the explanation and value, but also for the ‘what is it?’ question. By contrast, Williams does not address the ‘what is it?’ question by means of genealogy: his ‘what is truth?’ and ‘what is sincerity?’ questions are not addressed until Chapters 4 and 5 of T&T, and by then he has moved away from narrative, and offers instead a non-​genealogical account of what it is. Indeed, he claims that one should not try to give a history or genealogy of truth, on the specious grounds that the concept of truth is always and everywhere the same, and hence there can be no history of the concept of truth (Williams 2002: 61). This thought is specious because the lack of variation in the concept of truth across cultures and times (if true) is irrelevant to whether the genealogical narrative tracing its role in an emerging society could 18 For this formulation of the question, see Resp. 444e6–​445a1. 19 The discussion with Cephalus in Republic Book I leads to this question—​because Cephalus suggests that wealth has helped him to live justly, implying that living justly is choiceworthy. This assumption is challenged by others who think that the just life puts you at a disadvantage, and that injustice is preferable for the individual, but proscribed with sanctions by society. 20 Resp. 368c5–​369a4. [p. 96] help us to see what it is —​which surely it could, because what truth is shows in the way we and others use the concept. Indeed, in his ‘Endnote’ Williams (2002: 271–​ 7) discusses the vocabulary of truth in the ancient world, and there, despite his claims to the contrary, his treatment clearly reveals that we can look at the history of this and other related concepts to find out about what truth is, including how it functions in societies other than our own, and how it relates to other concepts or practices. These explorations are similar to what Socrates does for justice and the other virtues in the Republic. In principle, a ‘what is it?’ question can be addressed independently of the other questions. One can ask ‘what is it?’ about a practice without considering how we come to have it, or whether it is worth having. Yet although these questions are distinct, they can all (I submit) be answered by genealogy, and one can clearly address more than one from the same narrative. Equally one can choose to address some questions by a different method, which is what Williams chooses to do, for the ‘what is it?’ question. In the Republic, Socrates uses his genealogy to answer all three questions. First, in Book II, he uses it to ask ‘what is justice?’—​using the imaginary community as a ‘writ large’ example, to see justice in a big version first. But throughout the narrative Socrates offers reasons for why various practices (both ethical and practical) would arise at each stage, thereby providing non-​evaluative explanations, answering the ‘why do we have it?’ question. Then, the narrative completed, he uses the same resource for the third question, ‘does justice pay, and if so how?’—​by considering whether its virtues thus identified yield benefits for the participants and community. On this score, I would say that Williams underestimates the scope of the genealogy method. He takes it to be a method for answering the explanation and value questions, but he does not seem to see that one can also use it to address the ‘what is it?’ question, as Socrates does. Using Genealogy to Address the ‘What Is It?’ Question: (a) Avoiding the Risk of Essentialism Is genealogy a good way to address that ‘what is it?’ question? Does it have advantages over the conceptual analysis approach which Socrates tries (unsuccessfully) in the series of failed definitions explored in Republic Book I, and which Williams deploys in his Chapters 4 and 5? My hunch is that genealogy works better for several reasons. My first thought is that genealogy offers a way to discover ‘what it is’ without falling into essentialism—​the presupposition that there is a single common factor—​or unified definition—​of the kind that perhaps we typically associate with Socrates and his ‘what is the F?’ questions. Instead of positing necessary [p. 97] Catherine Rowett and sufficient conditions, the genealogist looks at the phenomena naturalistically and imaginatively; but by starting from a very simple community, at an early stage of development, one considers a pared-​down example, though without claiming that this exemplar is the only form that such a virtue or practice can take, or that the features of this one would be exactly replicated in every case.21 One claims only that one imagines that one typical or paradigmatic example might be like this. One can then explore how the practice as found in that typical case relates to adjacent practices that would be found in that imaginary society. Again, in so doing, one would not claim that every society develops this practice in exactly the same way. Nor would one conclude that everyone in the society must do the same thing, or behave in exactly the same way, to count as complying. Indeed, in Plato’s community, people do very different things. We see that they are all practising justice, not by discovering something that they all have in common, but by viewing each practice as part of a whole, and each as presenting a perspicuous case of how one might be just. Secondly, although genealogy is, like conceptual analysis, an armchair method, it also provides some quasi-​empirical grounding. Instead of engaging in intuitive conceptual analysis of a term in language, we attend to the practices and values of communities of social beings: we view the institutions emerging naturally in such a community; our attention is directed to a practice, not a theory, and to something recognizable as the thing by which people live, in real life. Seeing that Plato approaches his question in this way, we find ourselves discarding some common assumptions about Plato’s philosophical outlook, such as the idea that Plato is committed to unitary definitions, or to non-​naturalistic methods and ideals. Equally, it seems to me that Williams could have considered the ‘what is it?’ question about truthfulness in the same genealogical way, thereby grounding it in the practice of an imaginary human community reconstructed in his narrative. Using Genealogy to Address the ‘What Is It?’ Question: (b) Avoiding the Risk of Reductive Explanations In response to Glaucon’s myth about Gyges’ ring, with its accompanying thought-​ experiment wherein the unjust man reaps the rewards of justice while the just man loses every reward that society can confer,22 Socrates needed to show that a rational person would still choose justice for itself, even if the rewards or penalties were removed. In T&T Williams’s task is comparable: to show (by a genealogical 21 Compare Wittgenstein’s use of imaginary language games and primitive societies pared down to a bare minimum in the Philosophical Investigations. 22 Resp. Book 2II, 359b–​362c. [p. 98] narrative) that truthfulness is not just a functional practice, without which society would be crippled, but also a virtue with intrinsic value. Williams is aware that functional genealogy tends to reduce the moral to the non-​moral. He notes that Plato is also alert to that risk, in the Republic.23 But Williams himself has incurred criticism for this very fault, and for failing to establish that truthfulness has the intrinsic value he claims for it. The challenge is most clearly posed by Colin McGinn, writing in the New York Review of Books, who observes that Williams’s genealogical defence of truth and truthfulness resembles the indirect utilitarianism that Williams had always opposed (McGinn Showing that truthfulness yields pragmatically useful results for society, and would have been rationally chosen and promoted for those results, does not show that it has intrinsic value. And this is true even if the society and its members come to love the virtue as if it were of intrinsic value. It remains true even if the virtue produces its functional benefits only if the members value it in that way. Of course, one might say that pragmatically, that practice of valuing it is vindicated in a sense, if that is what makes it work. But then what is vindicated is a practice of valuing something, valuing it as if it were of intrinsic value. And it is ‘vindicated’ not by being true—​correctly appreciating some genuine value—​but by its usefulness, such as the survival advantages that flow from so valuing it, regardless of the truth of the matter. The fact that we need to make truth a virtue, and that we need people to love for its own sake, in the belief that it matters in itself, does not entail that it does, in reality, matter in itself, or that they are right to love it for its own sake (except in the debased sense that they are right to do so if they want prosperity). In some recent papers, Matthieu Queloz has sought to defend Williams from these challenges raised by McGinn and others. In Queloz (2018) he asks what separates the instrumental vindication of the practices of truth-​telling from indirect utilitarianism, given that Williams’s genealogy focuses on the advantages of treating truth as intrinsically valuable. Queloz responds that truthfulness manifests what he calls ‘self-​effacing functionality’. This is when a practice has functional and instrumental value only if our valuing practices are not done for the sake of the functional value. They work—​they function to deliver instrumental advantages—​only because we do not do them for the sake of those advantages but for some kind of perceived intrinsic value. There are indeed many things that function in this way: love and trust are examples. If you try to love someone in order to make yourself happy, it will not work. We can’t decide to trust someone so as to 23 T&T (2002: 36): ‘It might be said, further, that the motivations of justice have a tendency to resist a functional account, on the grounds that it will represent as instrumental and “lower” what is properly seen as intrinsic and “higher” (This idea is very prominent in Plato’s Republic).’. Note 18 (T&T links this to Socrates’ debate with Thrasymachus in Republic Book I, Glaucon’s contract theory in Republic Book II (also discussed in Williams (1997)), and Plato’s rejection of those accounts as reductive. He does not consider Plato own genealogy, which is intended to avoid that problem. [p. 99] Catherine Rowett be able to rely on their word, and we can’t rely on them if they say what they say in order to be trusted. Clearly, truth-​telling practices fall into this category. Williams is citing the instrumental value of the practice of valuing truth as a reason for having it. He suggests that, in order to have that instrumental value, it needs to be a practice of valuing truth for its own sake, not for its instrumental benefits. This raises an obvious worry: might that requirement, that we value it for its own sake, be satisfied if we mistakenly value it for its own sake or if we value it as if it had intrinsic value, the issue of whether it does being entirely irrelevant. If the advantages flow from the practice of valuing it as if it mattered, then so long as we value it as if it mattered we shall gain the advantages, which would be sufficient to explain why a practice of valuing truth as if it mattered would arise and stick. This would be a paradoxical outcome, since it would suggest that the truth of our values is irrelevant as long as we treat the truth of our beliefs and utterances as hugely important and not irrelevant. It seems, then, that a genealogical account of this kind, if it looks solely for explanations in terms of instrumental gains, cannot distinguish between the instrumental value of valuing truth as if it had inherent value, and the instrumental value of valuing truth because it really is inherently valuable. The outcome in terms of utility is the same. Proving Rorty wrong by his own lights (Queloz 2018: 2) is all very well as an ad hominem response to Rorty, if the question concerns why we should bother, in pragmatic terms, to act as if we value the truth: what would be lost if we didn’t? But if our question is whether truth matters in itself, and whether the practice of valuing it is intrinsically worth preserving, in its own right, even supposing it had no further pragmatic pay-​off, then it seems that it will not help to show a pragmatic pay-​off attainable only from imagining that it is worth preserving. Williams is aware of this risk and mentions it at Williams (2002: 90), as Queloz notes. Queloz is right that vindicating it in terms of its pragmatic pay-​off does not entail that it has no intrinsic value. It does not entail that its apparent value is illusory. But equally it does not show that it is not illusory. It shows only that we must not think it illusory (not if we want to keep the benefits of believing in it). One might make the same argument about the pragmatic value of religious beliefs.24 Williams’s own response to worries on these lines can be found in his chapter ‘Plato’s Construction of Intrinsic Goodness’25 in The Sense of the Past. There, Williams considers what pursuing justice ‘for its own sake’ might be, and why Plato thinks it important to do so. He argues that Plato is not talking about moral value; rather his question is about what contributes to the best life, a life worth living 24 See the genealogical explanations of religious practices offered by René Girard (Girard 1977). At T&T (2002: 92) proposes two conditions that he considers sufficient, though not necessary, for something to count as an intrinsic good, but neither condition seems to suffice for the item to be an intrinsic good, since they are really conditions for counting as valuing something intrinsically (i.e. treating something as important in itself ). 25 Williams (2003) reprinted as ­chapter 8 in Williams (2006). [p. 100] (Williams 2006: 121). Challenging ideas in Korsgaard (1983), he concludes that we should reject the idea that intrinsic goodness is self-​explanatory, and say rather that ‘something is intrinsically good if we need to value it as intrinsically good, and we can make sense of our doing so’ (Williams 2006: 136). This seems to confirm the impression that Williams is, in effect, reducing intrinsic goodness to a practice of valuing x as intrinsically good, leaving no room for asking whether the value we attribute to x itself actually exists. There is, it seems, no room for a non-​pragmatist account of truth left, when it comes to values. Does Plato encounter the same problem in relation to the intrinsic value of justice? I think not, because although Plato does suggest that a society will function better if it is characterized by a just distribution of work (which supports his claim that justice is instrumentally advantageous), he also defends the view that it is choiceworthy for its own sake on the grounds that it is the natural, healthy, sound condition for a soul or society. He thinks that once we see that, we should need no further consequential reasons. Having our minds, our appetites, and our social order in good condition, rather than disordered, is, he supposes, something that we would prefer for its own sake, without requiring any further benefits. Would a person, or community, continue to choose this if it really brought no beneficial consequences but many painful effects? Although Plato does not explicitly ask this question about his imagined society, the question is implicit in the challenge that Glaucon had posed initially, at 360d8 to 361d4, asking Socrates to show that the just man who is punished as if he were wholly unjust is still right to choose the life of justice. The task is therefore to show not just that sticking to justice as if it mattered a lot pays off in the rewards it brings, but that it is worth choosing even if all those rewards are replaced by negative effects. Implicitly, Plato is aiming for that answer in his account of why justice is worth choosing, and we should seek a similar defence for truthfulness if it is to be considered intrinsically desirable. Could truth be valuable in that way? It seems to me so—​as, for instance, one might prefer to be lucidly aware and able to resist, or at least protest against, being drugged or treated with mind-​altering techniques to make one feel unnaturally and inappropriately happy under conditions of great evil, or against being told untruths designed to allay fears. This response, it seems to me, would apply to questions about truth and truthfulness too, as well as justice.26 Does Plato’s distinctive approach help with finding intrinsic value of this kind, and not equivocating between intrinsic value and the instrumentally valuable effects of treating it as intrinsically good? My original hypothesis, in coming to this question, was that Plato’s approach, using genealogy to answer the ‘what is it?’ 26 Does this response merely show that we do value truth intrinsically, without showing that, or why, we should do so? Do we need more reasons for why being awake is preferable to drugged, and aware is better than being deceived? Plato frequently contrasts dreaming lucid awareness, and considers lucid awareness intrinsically preferable. [p. 101] Catherine Rowett question, assists him here too.27 On further reflection, I still think that is probably correct, though with reservations. The alternative (as practised by Williams), whereby one asks ‘what is it?’ by conceptual analysis, and uses genealogy only for the other two questions, implies or suggests that the practice must be explained not for its own value, but in terms of some pragmatic advantages of inventing or perpetuating it, so that one inevitably loses sight of any intrinsic value that the practice might have in itself. For it would seem that no practice could survive unless it brought some utilitarian value or survival advantages. The method seems to presuppose that nurturing or admiring a virtue is rational only when some instrumental gain is identifiable. By contrast, Plato’s narrative is not focused on explaining or justifying the virtue. Rather he lets the society’s virtues emerge silently, alongside the emergence of a functioning collaborative community based on natural needs and aptitudes. In the second luxurious city (but also to some extent in the city of pigs)28 he allows practices to develop whose sole purpose is to satisfy human desires for what is pointless but beautiful or stylish, or the desire for having more than enough (desires which, as we saw, are natural, though they vary from person to person). Watching various practices and institutions develop in an emerging society, without prejudice as to whether they are functional or dysfunctional, by-​products of the evolution of social organization, or even bad habits, makes it more likely that we could discover something with no functional role, something purely accidental, something that appeals because of its purely intrinsic value, and so on. In the course of this narrative, the practices that Socrates later identifies as virtues emerge under no particular description, though their emergence was in each case explicable and understandable, recognizable as the effects of habits and inclinations characteristic of human nature or second nature, at the time that they arrived. Then Socrates turns, only later, once these institutions are already established, to ask whether they have any positive value, and whether they are valued for their consequences or for no other reason. But, we should ask, does this really avoid the problem? Is the intrinsic value that Plato thinks he has revealed actually the value that he needed to find for justice, or is it a fudge? For sure, he thinks that a healthy soul is worth having, regardless of whether it brings a good outcome. The remaining worry, however, is that this means that justice is chosen because it is healthy and fine: that is, because it is ‘good for me’ or ‘good for my society’. It looks as though we have found a kind of aesthetic appreciation: the idea that a nicely ordered structure keeps everything in its place and does not leave us with any mental health problems. But how, if at all, is a healthy or well-​ordered system connected with virtue, especially where the virtue in question is one that commits us to altruism, to self-​sacrificial 27 That is, in addition to avoiding essentialism, on which see above, section 2.2. 28 For example, Resp. 372b. [p. 102] commitment, to acting rightly even at a cost to one’s own health and well-​being? Somewhere, Plato needs to ensure that the reason for valuing the soul’s healthy condition is not primarily our own mental and social well-​being (even though that is indeed intrinsically worth having) but also the well-​being and equal treatment of others, of those people to whom we shall act justly and virtuously. And it is not sufficient if the altruistic virtue is an accidental by-​product of our desire for self-​centred benefit, because that does not justify valuing the other-​regarding virtue in itself. It would be parasitic on valuing the self-​regarding benefit, which happens incidentally to deliver an altruistic virtue. So we have this remaining worry, that the intrinsic value quest has been somehow diverted, away from the virtue and onto something else, albeit something valuable in itself, that is accidentally correlated with the virtue. This worry can be answered, I believe, and I have tried to answer it elsewhere, as have others.29 Although this is not the place to complete that task again, we might observe that the answer has to do with being a well-​ordered soul—​which means successfully orienting one’s cognition, decisions, intentions, and actions, towards things that are of real value, instead of mistaken values (e.g. a misplaced pursuit of honour, money, and so on). Pursuit of those misplaced values is the corruption that blights an unbalanced soul or society. With the right understanding of value, a soul or city has no difficulty in preferring to sacrifice money and status for the true good, which is the life of virtue honestly lived. 4. Conclusion In conclusion, then, I think that Williams’s discussion of how genealogy can be used for vindicatory purposes in exploring ethics and political issues is enlightening. It has been enlightening to apply this also to the story told by Socrates in the Republic, and see that Plato’s practice conforms to the method. Plato turns out to be an antecedent of Hume—​someone who thinks that human beings, left to themselves, would and could develop a just society (even though in practice our society is typically corrupted by perverse values that privilege wealth, birth, or ignorance as a criterion for influence, rather than wisdom).30 It is enlightening to see that Plato traces the development of a good and just society that institutionalizes true values 29 The challenge is related to (but not identical to) that posed by Sachs (1963). A reductive answer that eliminates the moral value is offered by Kosman (2007). My own answer in Rowett (2018: 128–​34) takes a similar line to the detailed treatment by Kamtekar (2010). 30 Note that as the narrative continues by charting the decline of societies into timocracy, plutocracy, oligarchy, tyranny. etc. these too are shown to be natural developments, the result of human nature when it pursues disproportionately things that in their natural proportions were indeed good and choiceworthy. [p. 103] Catherine Rowett to the natural capacities of human beings, and specifically to the varied interests, passions, and abilities of its citizens. These are important results for understanding Plato’s political philosophy, and for appreciating his method in the Republic. In relation to evaluating the work of Bernard Williams we may also draw some interesting conclusions. It is intriguing that Williams saw ancient precedents for his approach in the proposals from what he called the ‘immoralist’ interlocutors, such as Thrasymachus in Republic Book I,31 and in the proto-​contract theory put forward as a thought experiment by Glaucon and Adeimantus—​two ideas that try to give explicitly reductionist and debunking accounts of morality. Oddly, Williams does not explore Socrates’ own genealogical narrative in response to those challenges, which also deploys the narrative method, and without the reductionist aim. Why did Williams see anticipations only in the reductionist opponents—​in the debunking thinkers against whom Plato was hoping to preserve some intrinsic value for justice? Can we trace this blind spot to Williams’s own intellectual biography? Certainly Williams came to Plato by way of a twentieth-​century Oxford education which typically assumed that Plato’s ethics (by contrast with Aristotle’s) was non-​naturalistic, and that Plato ignored natural human capacities or inclinations in explaining the virtues, but rather took the virtues to be aspirations towards a transcendent good unconstrained by any reference to human desires or needs. Such assumptions about Plato have been standard for generations, especially among neo-​Aristotelians in ethics. That deeply rooted interpretative error has dogged modern philosophy for a century, disseminating and perpetuating a misunderstanding about Plato’s project in the Republic. In reality, the idea that goodness transcends human nature, and is not good because we pursue it (but rather the reverse), is entirely compatible with supposing that we naturally pursue it, because it is good, and because we have a natural capacity to recognize and pursue what is genuinely good. Once we recognize that Plato’s position is the latter, and that he engages in naturalistic genealogy to explain and justify his political philosophy, many obstacles to understanding Plato’s views on intrinsic value fall away. While it is clear that Williams was at least partially aware that Plato does not fit into that box, it seems that he had not noticed that the Republic is offering a naturalistic genealogy of justice. It has also been enlightening, I believe, to observe that Plato answers the ‘what is it?’ question by genealogy, and to reflect on whether this helps to avoid some potential risks of genealogy as a tool in ethics and political philosophy. I have suggested that it might help to avoid essentialism; to avoid begging the question; to avoid exclusively reductive and eliminative explanations, that explain away any intrinsic value; to avoid equating intrinsic value with the belief, on the part of 31 Williams (1997). [p. 104] those within the practice, that what they are valuing is intrinsically valuable; to avoid leaving no space for any practice that emerged accidentally, for no purpose or advantage, or where it arrived as a by-​product of other things: of aspects of human or animal nature or of the disinterested love that we feel for what is supremely beautiful, good, and perfect. In such cases, it seems, the emergence of such practices as a key part of a good society might have nothing whatever to do with any consequential benefits or survival advantages and would be no less explicable if there were none.32 References Cross, R. C., and A. D. Woozley. 1964. Plato’s Republic: A Philosophical Commentary. London and Basingstoke: Macmillan. Girard, René. 1977. Violence and the Sacred. Translated by Patrick Gregory. Baltimore: Johns Hopkins University Press. Hempel, C. G. 1965. Aspects of Scientific Explanation. New York: The Free Press. Kamtekar, Rachana. 2010. ‘Ethics and Politics in Socrates’ Defense of Justice’. In Plato’s Republic: A Critical Guide. Edited by Mark L. McPherran, 65–​82. New York: Cambridge University Press. Korsgaard, Christine. 1983. ‘Two Distinctions in Goodness’. Philosophical Review 92: 169–​95. Kosman, Aryeh. 2007. ‘Justice and Virtue: The Republic’s Inquiry into Proper Difference’. In The Cambridge Companion to Plato’s Republic. Edited by G. R. F. Ferrari, 116–​37. New York: Cambridge University Press. McGinn, Colin. 2003. ‘Isn’t It the Truth?’. New York Review of Books L (6). Nozick, Robert. 1974. Anarchy, State, and Utopia. Oxford: Oxford University Press. Queloz, Matthieu. 2018. ‘Williams’s Pragmatic Genealogy and Self-​ effacing Functionality’. Queloz, Matthieu. 2020. ‘How Genealogies Can Affect the Space of Reasons’. Synthese Queloz, Matthieu, and Damian Cueni. 2019. ‘Nietzsche as a Critic of Genealogical Debunking: Making Room for Naturalism without Subversion’. The Monist 102.3: 277–​97. Rowett, Catherine. 2016. ‘Why the Philosopher Kings Will Believe the Noble Lie’. Oxford Studies in Ancient Philosophy 50: 67–​100. Rowett, Catherine. 2018. Knowledge and Truth in Plato: Stepping Past the Shadow of Socrates. Oxford: Oxford University Press. Rowett, Catherine. 2022. ‘Another Platonic Method: Four Genealogical Myths about Human Nature and Their Philosophical Contribution in Plato’. In New Perspectives on Platonic Dialectic: A Philosophy of Inquiry. Edited by Kristian Larsen, Vivil Valvik Haraldsen and Justin Vlasits, 213–​32. New York: Routledge. Sachs, David. 1963. ‘A Fallacy in Plato’s Republic’. Philosophical Review 72: 141–​58. Williams, Bernard. 1997. ‘Plato against the Immoralist’. In Platons Politeia. Edited by Otfried Höffe, 55–​67. Berlin: Akademie Verlag. Williams, Bernard. 2000. ‘Naturalism and Genealogy’. In Morality, Reflection and Ideology. Edited by Edward Harcourt, 148–​61. Oxford: Oxford University Press (cited as N&G). 32 This chapter is one of a number of descendants of a paper I prepared for a conference on the work of Bernard Williams in 2016. I am grateful to Sophia Connell and Nakul Krishna for inviting me to present at that event. I am also grateful to Janosch Prinz, Matthieu Queloz, and Marcel van Ackeren for comments on earlier versions. [p. 105] Catherine Rowett Williams, Bernard. 2002. Truth and Truthfulness: An Essay in Genealogy. Princeton: Princeton University Press (cited as T&T). Williams, Bernard. 2003. ‘Plato’s Construction of Intrinsic Goodness’. In Perspectives on Greek Philosophy: S. V. Keeling Memorial Lectures in Ancient Philosophy 1991–​2002. Edited by Robert W. Sharples, 1–​18. London: Ashgate. Williams, Bernard. 2006. The Sense of the Past: Essays in the History of Philosophy. Princeton: Princeton University Press. [p. 106] The Invention of the Humanistic Discipline Williams on Plato on Philosophy Marcel van Ackeren 1. Williams and Plato—​Only a Troubled Relationship? According to famous one-​liner, everyone has the Plato they deserve. This slogan holds true for Bernard Williams in a peculiar way. Throughout his entire career—​ that is, from his own training as a classicist to his late writings as one of the leading philosophers—​Williams wrote extensively about Plato, and it feels safe to assume that no other philosopher has a more prominent role in Williams’s thinking. However, this lifelong engagement with Plato poses the question of what it actually, in terms of philosophical positions, led to in Williams. When asked about Williams’s relation to Plato, most will find it is easy to come up with a list of things that he did not like in Plato. Take, for starters, one of his very early writings from 1959, a review of Plato Today by Richard Crossman, in which Williams characterizes the ideal state in the Republic as ‘a sclerotic monstrosity, high principled in intent, but ultimately—​and explicitly—​based on oligarchic deceit and a contempt for much legitimate aspiration and human diversity a system of coercion’ (Williams 2015: 4). Then, in two of his major monographs, Williams makes the criticism of Plato part of his own central themes. In the opening chapters of Ethics and the Limits of Philosophy (Williams 1985; hereafter ELP), Williams starts by paying tribute to Plato, considering the ‘Socratic Question’ a good starting point for doing ethics, especially if compared to the narrower questions that were taken to be fundamental by later ethicists (see van Ackeren 2019: 55–​8; Crisp 2020). Yet, Williams then quickly moves on to criticize every aspect of Plato’s own answer: Plato is said to be overly confident about philosophy’s capability to provide a compelling answer to the Socratic Question, and Plato’s own answer is rejected because it involves an unacceptable moral cognitivism, intellectualism, and arguments against moral scepticism and relativism that Williams takes to be unattainable to us. According to its Preface, ELP is about the ‘criticism of present philosophy’ (vii), but Williams assumes that things in moral philosophy are not as they should be, because the continuation of the mistakes of ancient philosophers has led to the present predicament. Thus, readers are not surprised when Williams concludes the Marcel van Ackeren, The Invention of the Humanistic Discipline In: Bernard Williams on Philosophy and History. Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. DOI: 10.1093/​9780191966361.003.0006 [p. 107] Marcel van Ackeren three opening chapters on ancient philosophy by stating that the ancient assumptions have ‘collapsed’ (ELP: 53). One might object to this reading that Williams concedes that ‘some extension of ancient thought, greatly modified, might be able’ to meet the ‘demands of the modern world’ (ELP: vii) because ‘in its general outline the description of the ethical self we have recovered from the ancient writers is correct’ (ELP: 59). However, it is important to note that he sees no silver linings in Plato, for only Aristotle is credited for ‘the only colourable attempt to provide a foundation for ethics’ (ELP: 59). Jenkins, who wrote an excellent book on Williams, considers this the reason why one might skip Williams’s account of Plato and start with Aristotle, as he states in his chapter on Williams and ‘the ancient world’ (2006: ch. 7, esp. 150). Later on, in Shame and Necessity (Williams 1993; hereafter SN), the criticism of Plato focuses on ‘ethicized psychology’, amounting to a suggestion that can be considered even more radical than that in ELP: Williams now claims that if the Greeks have to offer any alternative that is better than the modern morality system, it can only be found in pre-​Socratic authors like Homer, Aeschylus, Sophocles, and Thucydides, but not in Plato and the other philosophers, because they are still part of the enduring problem they have created. What can be concluded from all this criticism of Plato? Even if one does not want to go so far as to assume that Williams is the type of critic who, like Heidegger, thinks that with Plato the philosophical evolution had already taken the decisive wrong turn in the primordial soup, it is very tempting to regard the relation as entirely dismissive. At least, this is the impression that readers of ELP and SN might have. Or is Plato just a cardboard figure for Williams, allowing him to highlight his own positions by presenting them against some historical and contrasting background? If this were the case, it would strengthen an already existing misrepresentation of Williams as a purely negative author who had nothing positive to say, either for himself or about his historical predecessors. The aim of this chapter is to show that such a pejorative characterization of Williams’s relation to Plato neglects an important point: Williams, in other writings, advances some surprisingly positive and worthwhile opinions about Plato’s more general conception of philosophy, which also help to shed some light on his own views about philosophy and what it can and cannot achieve. Besides, or rather despite, all the apparent criticism, Williams does not regard Plato as someone who pursued a fundamentally misguided kind of philosophy as, for example, Kant did in Williams’s view. I will try to show that he is also a very charitable interpreter of the Platonic dialogues, who highlights aspects of Plato’s philosophy that other commentators have overlooked or downplayed. These aspects are condensed in Williams’s fundamental conviction, expressed in his introduction to Plato, Plato. The Invention of Philosophy (see The Sense of the Past (Williams 2006: 148; hereafter SP)), that ‘Plato invented the subject of philosophy [p. 108] as we know it’. Both parts of the sentence carry weight, because Williams does not only want to make a historical observation, he also wants to express that some aspects of Plato’s invention are worthwhile today. Williams stops short of calling Plato the ‘greatest philosopher of the world’, but only because this is ‘pointless’ as ‘there are many different ways of doing philosophy’. Williams praises Plato in a way that might be surprising to the readers of ELP and SN: we can say what the various qualities of great philosophers are: intellectual power and depth; a grasp of the sciences; a sense of the political, and of the human destructiveness as well as creativity; a broad range and fertile imagination; an unwillingness to settle for the superficially reassuring; and, in an unusually lucky case, the gifts of a great writer. If we ask which philosopher has, more than any other, combined all these qualities—​to that question there is certainly an answer, Plato. (SP: 180) There is more than one reason to explore the more positive side of Williams’s take on Plato. First and most importantly, Williams provides an account of Plato that is a freestanding contribution to our understanding of Plato. ‘Freestanding’ is supposed to mean that his interpretation of Plato is valuable even if we ignore that it is provided by Williams, an otherwise influential philosopher. The other reason why Williams’s Plato is worth analysing is the fact that his interpretation of Plato helps us understand Williams’s own positions: Williams finds some rather fundamental views about philosophy in Plato’s works that mirror his own. I will focus on two of these views: namely, the reluctance to establish a systematic theory in one’s own name and a keen awareness of the limits of philosophy—​that is, of what it can achieve and how this can be achieved. Thus, Williams’s image of Plato is not only a historical account of Plato, it also philosophy. I will proceed as follows: each of the next three sections will single out a key aspect of ‘Plato: The Invention of Philosophy’ (SP: ch. 10), which provides a surprisingly sympathetic image of Plato, which was overshadowed by the more critical works, such as ELP and SN. Section 2 will attend to the dialogue form, which is not simply a vessel for content that could also be communicated in a different way. According to Williams, the dialogue form also allowed Plato to maintain a distance from the arguments and views that are exchanged in the dialogues and present not only the results of arguments, but also the practice of philosophy and how and why it might fail in this world. Section 3 elaborates on Williams’s observation that philosophy in Plato is not only about the ascent to the transcendental realm of Forms. Williams depicts a Plato who, unlike the Plato in Raphael’s famous painting The School of Athens, does [p. 109] Marcel van Ackeren not only point upwards. Williams’s Plato is also committed to the quality of the earthly and real human life, including the love for individual persons. Section 4 will discuss three limits of philosophy in Plato. For one thing, Williams provides a compelling interpretation of Plato’s discussion of writing philosophy and what philosophy can communicate. For another, Williams’s Plato, very much like Williams himself, thinks that philosophy can achieve what it intends to achieve not by establishing theories and presenting results of philosophical argumentation, but only through practising philosophy: that is, through active engagement. And this finally means that, according to Williams and Plato, the success of philosophy depends on lucky circumstances. In the concluding section, I will try to answer two questions that arise from this more positive image of Plato that Williams presents: How are Williams’s critique of Plato, on the one hand, and his more sympathetic views, on the other, related? And does the more positive side just happen to be a list of things that Williams finds to his liking or is his positive interpretation based on some methodology? 2. The Dialogue Form Plato made the dialogue (from dialegesthai, meaning to converse with each other) his form of philosophy, and it plays four different roles: the dialogue form is (1) a literary genre and (2) the form of thinking itself; it also serves two important functions for Plato: (3) it allows him to distance himself from the arguments and (4) to present real discussions and not only ideal arguments. Most obviously, the dialogue form is a literary genre, which is ‘a tribute in writing to Socrates’ style of life and talk’ (SP: 149). But Plato was not the only one who was impressed by Socrates and therefore started to write so-​called Socratic Talks, Sokratikoi logoi. Unlike Antisthenes, Phaedo, Eucleides, Aristippus, Aeschines, and, of course, Xenophon, Plato not only used this specific form, he also regarded it as having special relation to its content: the ‘dialogue form is not, for the most part, just an artful way of his telling one something. It is an entry and an invitation to thought’ (SP: 148–​9). This raises two questions: what kind of thinking does Plato invite the reader to and in what way does the dialogue do more than simply tell the reader something? As mentioned above, I will distinguish four different roles of the dialogue form, which will help us answer these two questions. The second function of the dialogue form is part of an answer to the first question regarding the kind of thinking; the third and fourth functions of the dialogue form will help answer the second question. The thought that Plato invites the reader to and advocates is dialogical itself: the dialogue form is the modus operandi of the reasoning and inquiry that are known as the Socratic elenchos and Plato’s dialectics. In two of Plato’s later [p. 110] dialogues, dialectics is described as the ‘dialogue of the soul’ (Theaet. 189e–​190a; Soph. 263e–​264b). Williams remarks that the ‘idea (which no doubt came from the historical Socrates himself ) is that only through question and answer is it possible to construct and follow a logical argument, which will actually prove or disprove a definite conclusion’ (SP: 156). We can distinguish two ways in which Williams’s remark can be considered correct. For it is one thing to say that thought itself comes in the form of question and answer; and it is another to say that good philosophical thinking and inquiry needs an interpersonal dialogue. In the Theaetetus, Socrates calls himself a midwife, whose cross-​examination is necessary when other persons are pregnant with an argument. Discussing their argument with him is required for either the birth or the abortion of the idea and argument, depending on whether they are worthwhile or not. That Plato regards the interpersonal dialogue as an essential method of inquiry distinguishes his approach from other well-​known texts and philosophical accounts in the tradition, such as Marcus Aurelius’s Meditations, Augustine’s Soliloquy, and, of course, the inquiry by Descartes in his Meditations. All of these texts have a special form as well, and they express the idea that thought is dialogical; yet, for different reasons, none of them makes use of or presents an interpersonal exchange. Thus, the dialogue form is characteristic for the thought itself and the interpersonal conversation is the method that is meant to shape our reasoning in the right way. With that said, the two further features of the dialogue form, which help us to see why Williams stresses that the dialogues are about more than just conclusions, can be addressed. The dialogue form allows Plato himself to keep a distance to thoughts, arguments and conclusions made in his works, simply by not making himself one of the interlocutors. Although his contemporary readers certainly knew that he was among Socrates’ followers and supporters, he never appears as a character that joins the discussion; he is not even present as a passive listener in his dialogues. Plato stresses the distance between himself and the arguments in the dialogue: for example, by making the narrator of the conversation of the Phaedo remark that ‘Plato, I think, was ill’ (Phaed. 59b) and that he therefore did not join the others when they met with and spoke to Socrates for the last time. Plato thereby reminds the readers that his dialogues are not minutes or historically truthful witness accounts, but works of his imagination. That Plato does not speak in his own name leads to the question of how interpreters can attribute any claim to Plato. This problem is especially pressing for those readers who are interested in the results of logical arguments, because even if the exact meaning of the claims and the structure of the argument can be made sufficiently clear, this does not settle the question of whether the conclusion is also Plato’s. Furthermore, this raises the question of whether an argument, especially in the latter dialogues, should be read as an attempt by Plato to substantiate, evolve, [p. 111] Marcel van Ackeren correct, or dismiss one of his previous arguments. Before coming back to these more fundamental philosophical implications in section 4, which deals with the limits of philosophy, we need to assess different interpretational strategies to cope with the fact that Plato never makes a statement of his own. Most interpreters assume that Socrates, whenever he is the main speaker, is Plato’s speaker, and that if Socrates assumes something, then Plato assumes it as well. Williams opposes this view: Even when an authoritative figure in a dialogue, usually Socrates, seems to leave us with a conclusion or theory to be taken away from it, we should not necessarily suppose that this is what Plato is telling us to believe. Not everything asserted in a dialogue, even by Socrates, has been asserted by Plato: Socrates asserting may be Plato suggesting. Because Plato is an immensely serious philosopher, who indeed set philosophy on the path of claiming to address our deepest concerns by means of argument, orderly enquiry, and intellectual imagination, and because we project on to him images of seriousness which are drawn from other philosophy and from later experience, we may well underestimate the extent to which he could combine intensity, pessimism, and even a certain religious solemnity, with an ironical gaiety and an incapacity to take all his own ideas equally seriously. (SP: 149) By now, we have only established what Plato, according to Williams, deliberately and artfully does not want to give us: the dialogues are not simply a vehicle for presenting research outcomes, and Plato does not aim at presenting study results, a theoretical system of his own. As has been said, we will come back to the distance that Plato puts between the arguments of the dialogues and himself in section 4 of this chapter, which deals with the limits of philosophy. For now, we need to consider the question of what we can, according to Williams, positively get from the dialogue form beyond conclusions. The dialogue form is used by Plato to show us the practice of philosophy, and Williams highlights three features: the use of fallacious arguments, that arguments have a force, and finally that human beings do philosophy and that their personhood heavily influences the form and results of an investigation. Although Plato incorporates lengthy reflections on and examples of the right form of philosophical conduct in his works, he also uses the dialogue form to present not only the ideal form of research, but also real cases, which notably include showcases of how and why discussions fall short of the ideal. That Plato deliberately uses fallacies was not recognized by all commentators, some of whom struggled with some of the arguments in dialogue. Once we accept the idea that they are not necessarily Plato’s, we can also see that he details how and why an argument goes wrong or must be regarded as weak. Plato thereby challenges his readers. In the Phaedo, for instance, the interlocutors present more [p. 112] than one argument for the claim that the soul or its rational part is immortal. The first argument, which, in fact, is quite a poor argument, gets approval from the interlocutors. This way Plato makes the readers think for themselves. Moreover, the readers are also led to wonder why all the interlocutors of the dialogue agree that other arguments are needed after the first, although no one doubted its quality. In another very prominent case, Socrates’ argument against Thrasymachus in Book I of the Republic, the situation is more straightforward. Although Thrasymachus first accepts Socrates’ argument, Plato later makes one of the other interlocutors, who happens to bear the name of his brother, namely Glaucon, complain about the argument: ‘Socrates, is it your desire to seem to have persuaded us or really to persuade us?’ (Resp. 357a). So, Plato quite obviously uses fallacies.1 Yet, unlike Sprague, who argues that Plato has consciously made use of fallacies in order to set forth his own ideas by pre-​empting possible lines of attack against his own position, Williams views many of these cases differently. To him, Plato’s ambitions go beyond presenting ideal arguments, and thus he depicts arguments made in real discussions, not ideal ones, which necessarily include cases in which something goes wrong. Plato’s works are about arguments, but they also argue about some important feature of arguments that Williams calls ‘the force of argument’: we should not assume that ‘the force of argument’ is an entirely fixed and determinate notion. It is not so anyway, and it is less so in Plato, for the special reason that he more or less invented the idea. What one sees in his dialogues is a process, of his seeking in many different ways to distinguish sound argument from the mere power of persuasive speech. (SP: 156) This leads to the last point that is considered in this chapter to be a lesson to be learned from the dialogue form. Although good arguments, according to Plato, might refer to transcendental objects of knowledge, the arguments are neither pursued in this realm nor do transcendental objects need or exchange arguments. Arguments are put forward by persons who are quite unlike these objects. In Plato’s terms, we can say that only a specific part of these persons—​namely, the rational part of their soul—​is similar to the Forms, while the rest is not. And other features of these persons shape the argumentation and how the arguments are received: But as well as explicit philosophical content, Plato lays before us in his dialogues various attitudes which also belong to philosophy, such as curiosity, puzzlement, and intellectual surprise. His characters may be excited, bored, confused, or 1 See the influential study by Sprague (1962). [p. 113] Marcel van Ackeren impressed; they take up, in such moods, all sorts of analogies, metaphors, stories, and models (for instance, of the human mind); they learn something from them, and often then throw them away. (SP: 84) The dialogues depict non-​ideal discussions, which are shaped by more than logic. Michael Frede, one of Williams’s interlocutors in Oxford, stressed that this tells an even more important story about the core of Platonic Philosophy: Only individuals with a certain character, a general outlook, a certain social position, certain interests enter the debate and this background notably colours their views. By the artful characterization of the dramatic context of the arguments the dialogues show us in an unsurpassable way how philosophy is tied to real life, to forms of life, to character and behaviour. (Frede 1992: 220) Frede is in line with Williams, but he also bridges the discussion of the dialogue form and an account according to which the core of Plato’s philosophy is not merely about ascent, but also concerns the real messy human world, which is the next central feature of Williams’s picture of Plato. 3. Better for This World One of the most intriguing aspects of Williams’s picture of Plato is that it corrects a dominant image of Plato, which was depicted by Raphael in his School of Athens: a Plato pointing upwards. This image suggests that the core of his philosophy is about an ascent that leaves the mortal life behind and strives towards a transcendental realm of timeless and unchanging entities, the (in-​)famous Forms. This view became prevalent due to neo-​Platonic and Christian transformations of Plato’s thought, which nonetheless insisted that they represented the real and original Plato. This even highly influenced interpreters who would not consider themselves neo-​Platonic or religiously inspired, precisely because of the long-​lasting dominance of these representations of Plato, which are in fact misrepresentations. In order to appreciate the way in which Williams provides a worthwhile correction of the traditional view, it is crucial to see that he does not simply deny that the ascent to the realm of Forms plays an important role. He speaks of Plato’s ‘highest metaphysical ambitions’ and of the ascent to ‘a reality which in some sense lies beyond everyday experience’ (SP: 166). Williams acknowledges that this kind of metaphysical speculation is a key feature, especially in dialogues such as the Republic, but he also sees that it does not represent the entire Plato. Williams points to two aspects in Plato’s dialogues that provide a more well-​rounded image of him: Plato envisaged good philosophy to be better for this world and this includes the love of individuals. [p. 114] First, even in the Republic, the highly demanding metaphysical ambitions should not be read as advocating transcendental escapism, which is overly prominent in Plotinus and other representatives of Platonisms. Williams stresses that the entire argument, even that regarding the highest knowledge about the Form of the Good, is about philosophy being able to improve the world instead of getting away from it: ‘Plato thinks that it is better that the just and wise should rule unwillingly, rather than that those who actually want power should have it. But that must mean better for the world’ (SP: 174). It is hard to overstate the importance of Williams’s observation here, because his claim that, in Plato, the point of philosophy is to improve this world holds true in many significant ways. Plato assumes that philosophy is beneficial for those who pursue it, and this holds true even for those philosophers who return to the cave and are eventually murdered. This is not a surprising claim, at least not in the Platonic framework, according to which the aim is not to live a long life, but a good one: that is, the good and just life (see Cri. 48). And returning to the cave is the just thing to do for those who acquired the highest knowledge (see Resp. 521e). According to Plato, knowledge is virtue and virtue is the sole constituent of the good life (see Euthyd. One might think that, according to Plato, only few people can acquire the highest knowledge and thus become virtuous and enjoy the hight pleasures, and that therefore philosophy is good only for the very few philosophers but not for the majority of the population. But Plato makes it clear that philosophers, if they become rulers, also improve the quality of life of those who are ruled. This, of course, sparked the criticism that Plato’s political views are highly paternalistic, and Williams shared this view, as the first quote at the beginning of this chapter indicated. However, the paternalism complaint presupposes that Plato aims at improving our individual and collective practices and institutions, because the complaint is about Plato accepting coercion in order to achieve better political affairs in this world. So, even though knowledge has to do with objects that are not part of this empirical world, including human conduct, the aim of knowledge is not to escape this world; rather, knowledge is seen as necessarily improving actions with regard to personal and political matters and thereby the well-​being of all members of a society (see Resp. 519c–​d). In what way is philosophy good? It would be an oversimplification to assume that Plato’s guiding idea that philosophy is an improvement is about an instrumental relationship between metaphysical knowledge and the improvement of the lives of philosophers and others. Regarding the philosophers who have been taken out of the cave, Williams notes ‘that returning to the cave is good for them only because it is the good thing to do’ (SP: 174; emphasis added). This needs explanation, and it can be found in other works by Williams: namely, ‘Plato against the Immoralist’ and ‘Plato on Intrinsic Goodness’ (SP: chs 6 and There, Williams discusses Glaucon’s Challenge at the beginning of Resp. [p. 115] Marcel van Ackeren II: Socrates’ task is to show that virtue guarantees the good life, even if reputation is not taken into account. This passage has become a challenge for interpreters, too, because it contains the puzzling claim that justice is a good that is pursued for its own sake and its results (Resp. 358a). Williams’s interpretation is pivotal to his understanding of Plato’s ethics: Socrates claims that justice ‘has no value (really), unless it has value when it is pursued for its own sake’ (SP: 102). Accordingly, we can understand the claim concerning the goodness of returning to the cave as follows: if the philosophers return to the cave only for the sake of doing the just and good thing, then it will contribute to their good life as well.2 Williams interestingly notes that Plato’s argument does not have to be accepted by us now in order to deserve our interest.3 This is so, because engaging with Plato’s argument and pinpointing where and why one disagrees will help our own ethical thinking: We need not accept Plato’s outlook, and not many people do. But one needs to go on a long ethical journey to find what exactly it is that one rejects, and it will be important, for each person, to decide what this will be. Certainly, the claim that justice has to be valued in itself, or for its own sake, is not a mere moralistic assumption of Plato’s, and he has not simply set himself a gratuitously difficult task of demonstrating it. It is deeply rooted in the whole enterprise. One important question we are left with now, reading the Republic in a very different, and differently democratic, time, is not so much whether we agree with Socrates or with Glaukon, but how far they might both be right. What is it to admire and practice justice ‘for its own sake’? Perhaps the lesson of Glaucon’s argument is just this, that precisely because we need justice as an instrument, we need to admire it for its own sake; and that what we need to do is to learn how to do this, while not forgetting why we are doing so. (SP: 107) Williams’s discussion of this matter reveals two more general merits of his account of Plato. Williams manages to combine an awareness of the historical distance with an acknowledgement of Plato’s lasting impact as well as his argumentative rigour and doxographical subtleness. 2 There is another related problem that I can only mention here. Williams’s explanation, I think, does help to explain why the philosophers are reluctant to go back to the cave in the first place. Socrates provides an argument why it would be rational and just to overcome their reluctance, but why are they reluctant? If they are philosophers and have an orderly soul, they do not have irrational desires. I would like to thank Roger Crisp for bringing this up. 3 A corollary of this can be found in M. Queloz’s account of genealogy in Williams, for he argues that this argumentative strategy—​that justice is good for just agents if they act as if justice is an intrinsic good—​also holds true for truth and truthfulness. He argues that Truth and Truthfulness explains an account of the value of truth as something that, just because we need it as an instrument, we need to admire for its own sake (see Queloz 2021: 181). [p. 116] Several interesting points can now be made regarding the second feature of Williams’s account that provides a correction of the dominant image of Plato just pointing upwards. The Symposion is well known for its scala amoris, the ladder of love, according to which lovers make philosophical progress if they advance from loving one beautiful body to the love and vision of the abstract and unearthly Form of Beauty. Williams objects to a certain aspect of this understanding of love: in his reading, this advancement from loving one empirical object or person to the Form of Beauty is not an escape from one world to another, separate world, which is the one real world. Williams’s point is not simply that there is no complete separation between the Forms and their images in this world. The separation cannot be complete, because of the presence (parousia) of the Form of Beauty in a beautiful body, causing its beautiful aspects (see Phaid. 100c). Williams makes two points about this: first, he notes that the beautiful body is to be loved in its own right—​that is, for the beauty it has, though not unqualifiedly so. Secondly, he also notices that the first step on a ladder is necessary for the next, and that it is not merely overcoming an error, precisely because it enables progress: Diotima’s account of this progress or ascent does not imply, as some have thought, that no one ever really loves a particular person, but only the beauty in that person, or beauty itself. On the contrary, one can love a particular person in any of the various ways that count as bringing something to birth in the presence of that person’s beauty. The account directly denies, in fact, that all love is love of beauty. Moreover, it does not suggest that the particulars, sights and sounds and bodies, were only seemingly or illusorily beautiful. They are not unconditionally, or unqualifiedly, or absolutely, beautiful, which is what the item of the final vision is. Indeed, Diotima can say that from the vantage point of the vision, colours and human bodies and other such things are merely ‘mortal nonsense’, but that is only by comparison with the vision, and it does not imply that the mortals who thought that those things were beautiful were simply mistaken, or that they were mistaken to have pursued them. The undertaking she teaches is something like a growth in aesthetic taste, from kitschy music, say, to more interesting music. It does not deny the point or the object of the earlier taste, and indeed the earlier taste is a condition of the process, which is a progress rather than the mere detection of error or the elimination of a misunderstanding. (SP: 177) Much can be said in favour of Williams’s reading, which is focused on the Symposion, but I can only hint at another dialogue, the Phaedrus, in which Plato comes back to the relation of eros, love, and ascent. The simile of the cave might suggest that the ascent of the prisoner can be completed and maintained as a permanent and superior form of existence, because the former prisoner seems to be able to stay above the cave and not to go back. However, the Phaedrus clarifies that this is not the case, as the ascent of the tripartite soul is limited in three significant [p. 117] Marcel van Ackeren ways. First, only in very rare cases, precisely when a philosopher has a particularly distinguished soul, is it possible for its rational part to get a glimpse of the transcendental realm of Forms. Secondly, even during that time of apprehending the Forms, the rest of the philosopher—​that is, the other parts of his soul and his body—​continue to live in the earthly realm. Thirdly, Plato also stresses that the time of contemplating the Forms is very short and that even philosophers necessarily spend the majority of their lifetime among their fellow human beings. Plato’s metaphysical speculations gained a dominating influence in western thinking, but he carefully includes less influential hints concerning the limits of philosophical ambitions, which will be the topic of the next section. 4. Limits of Philosophy in Plato Perhaps it is not surprising that Williams, being famous for a book called Limits of Philosophy, also discusses the limits of philosophy in Plato. Though Williams opposes the assumption that Plato’s ethical views can be part of a contemporary ethical outlook, he seems to think that quite a few of Plato’s more general views about philosophy still hold true. Williams discusses three different limits: the limits of writing philosophy, the limits of philosophical ambitions in terms of building systematic theories, and, finally, the limits of philosophy’s effect in this world. Plato reflects on writing in a passage in the Phaedrus (275c–​277a), claiming that written philosophy has several disadvantages compared to true philosophy pursued in interpersonal conversation. In most cases, a text cannot explain or defend its arguments. Also, an author of a text has no control over the application of his ideas by others. Williams comments on this criticism of writing: There has been discussion of why Plato, after this, should have gone on writing. But even if we take Socrates’ remarks (a little stolidly, perhaps) entirely at their face value, they do not mean that Plato should not write—​they give him a reason to write, and that reason is obviously only one among similar reasons we might imagine. This passage does not mean the end of philosophical writing. But it does express an important idea about the limitations of philosophical writing, an idea which, I shall suggest, is important in relation to the spirit in which Plato wrote his works and the spirit in which we should read them. (SP: 176) Much depends on the spirit in which Williams thinks we should read the dialogues. He steers a middle course between two more extreme interpretations. According to a few (mostly older German) scholars, the passage is a clear indication of Plato’s reluctance to discuss the most important doctrines in the written dialogues, sharing the valuable ones only orally with his inner circle at the Academy. The so-​called [p. 118] ‘Tübingen School’ thus favoured an esoteric interpretation, which devalues the dialogues, but at the same assumes that Plato had an ambitious and systematic theory, the ‘unwritten doctrine’. The other interpretation of this passage, often assumed only implicitly, is extreme in so far as the critique of writing in the Phaedrus was simply ignored. Mainly analytically orientated scholars treat the dialogues as—​more or less—​systematic treatises that can be mined for arguments leading to definite conclusions, and thereby consider the dialogues to contain a theory held by Plato. Williams rejects the idea that Plato established a systematic theory, whether in the written dialogues or in unwritten oral ones. Before returning to this issue, we need to occupy ourselves with the spirit in which Williams wants us to read the dialogues, because so far, I have only argued what kinds of reading he rejects. To Williams, Plato made a ‘great inquiry’, but the ‘results are never in the text before us’ (SP: 179). This has to do with Plato’s style and what it wants to achieve. As has already been noted, Plato uses the dialogue form to distance himself from the arguments in the texts by not making himself one of the interlocutors. Also, the dialogues depict an exploration rather than its results: Plato allowed himself to be ‘driven forward by his ideas, curious at any given point to see what will happen if some striking conjunction of them is given its head’ (SP: 154). Plato’s writings are not meant to inform the reader of some conclusion drawn by the author. So, Plato writes philosophy that differs in style and intent from the well-​known and highly typical style of many texts in analytic philosophy which seeks precision by total mind control, through issuing continuous and rigid interpretative directions . . ., this style tries to remove in advance every conceivable misunderstanding or misinterpretation or objection, including those that would occur only to the malicious or the clinically literal-​minded. (Williams hereafter PhHD) The difference can be put in a slogan: Plato makes the readers think for themselves! The now common style tries to force the reader to think only what the author wants.4 Plato’s emphasis on the practice of philosophy rather than on the results—​ written or not—​brings us to the second limit regarding the ambitions of philosophy. Williams stresses that there are limits to the theoretical and systematic ambitions of written philosophy, as a comparison with the approach of Julia Annas 4 Although Williams did not write dialogues, his style shares some functional aspects with that of Plato. ‘Williams wrote in a highly compressed and elusive manner, a feature that became more pronounced in his later work’ (Russell 2018: 1). This is not necessarily a complaint. Reading Williams means that the reader has to do the thinking, which in turn explains that ‘the interpretation of his work continues to be a matter of considerable debate’ (2018: 1). Williams also notes that ‘clarity and argument’ are philosophical virtues, but says that they are virtues of any academic discipline (PhHD: 181). [p. 119] Marcel van Ackeren highlights. After many eminent and influential publications on Plato, including a landmark commentary on the Republic, Annas excluded Plato from her great study The Morality of Happiness (1993) because of his deliberate use of the dialogue form and his consequent rejection of systematic discussion of ethical theory. There is no Platonic dialogue which is ‘about ethics’ in the way that the Nicomachean Ethics is about ethics. Each dialogue has its own theme, or group of themes, which often do not answer neatly to our subject divisions. Even the dialogues which are most obviously relevant to ethics, like the Republic or the Philebus, are also, and equally, about metaphysical and epistemological issues. And even in the dialogues which are most clearly about ethics, Plato in using the dialogue form deliberately refrains from identifying and laying out his own position. It would be extremely naive simply to identify Plato’s position with that which is allotted to Socrates in a given dialogue. And even if we do, we find immensely different positions occupied by Socrates in different dialogues; the structure of Platonic ethical theory is greatly underdetermined by what we find in the dialogues. (Annas 1993: 18) In many regards Annas is right, and she and Williams agree that Socrates or some other speaker need not express Plato’s view. Yet, the crucial difference between Annas and Williams is that Annas considers the ‘systematic discussion’, including the divisions established in later philosophy, to be most important and the measure for answering the question of whether or not Plato should be discussed in a monograph on ancient ethics. So, when she finds that the dialogues present something ‘underdetermined’, she finds something that diminishes the value of what Plato wrote. Williams, in contrast, as has already been indicated in the discussion of the dialogue form, finds other worthwhile lessons in the dialogues and their peculiar relation of form and content. To him, the dialogues contain a specific idea about the limits of philosophy and the right way to approach it—​and not simply a less valuable form of a latter approach: namely, that of systematic philosophy. Williams acknowledges that Plato had some systematic convictions, but did not write in order to establish a fixed theory or even a system (see SP: 83, 154). Williams thinks that expecting to find a systematically laid out theory in Plato is an anachronistic confusion that is linked to another misunderstanding: It is a weakness of scholars who study philosophers to think that philosophers are just like scholars, and it is particularly a mistake in the case of Plato. . . . From the public space on the edge of Athens in which Plato carried on his discussions, it was called the Academy, and in this way, Plato gave the word ‘academic’ to the world, but it is an irony that he should have done so. We should not be trapped into thinking of him as a professor. (SP: 149–​50) [p. 120] Williams here obviously combines his own views about philosophy—​for example, that the concept of philosophy is not necessarily coextensive with the concept of scholarship or with having a theory or even a system—​with a description of Plato. Williams, unlike Annas, who thinks that the dialogue form is an obstacle, does not only find something in Plato’s dialogues that differs from the later ambition to establish theories or systems, he actually appreciates Plato precisely because of this difference. One notable consequence of this spirit in which Williams want us to read Plato concerns Plato’s discussion of Forms. Williams denies that the many passages in Plato evoking Forms should read as being about a ‘Theory of Forms’ or—​in turn—​ about giving it up in later dialogues; instead, he argues that it ‘is more helpful to see Plato as having a general conception . . .; having also a set of philosophical questions; and as continually asking how such objects might contribute, in various ways, to answering those various questions’ (SP 149–​50). It is important to note that the view that philosophy is not about establishing, defending, or revising a systematic theory does not imply an opposition to arguments. Williams here stresses that Plato does not focus on establishing a theory, but on answering various questions. But this is not simply a view that Williams attributes to Plato; it is also his own. Above I noted that Plato uses the dialogue form to distance himself from the doctrines that are discussed in the dialogues. And if we take this distance together with the limited ambition to build theories, we have found another important parallel between Williams and Plato. In an interview Williams says: You are quite right in saying that my philosophy isn’t full of positions and theses. But there’s more unity to it than just in its areas of concern. . . . It’s not an accident that my work isn’t full of positions and theses. I think one of the things that I acquired from my formation, and haven’t lost, is my suspicion of philosophical theory. I’m much less suspicious of theory in some other areas of philosophy than I am in ethics, and some areas certainly demand theory. But I’m still not somebody who naturally expresses himself by coming up with a lot of theoretical positions. (Williams 1994: 4) Williams refers to his ‘formation’ and in the interview he spoke about his time as a classics student in Oxford immediately before giving the answer I just quoted. But my argument about the parallel does not have to involve a biographical explanation; it is more about a metaphilosophical agreement and alignment. From the aspirations of philosophy, we can now move to the final limitation of it, which concerns its helpfulness or success. Williams observes that Plato can be very optimistic about the power of knowledge, but—​at the same time—​very pessimistic about the success of philosophy, which is not only due to the fact that philosophers, like his teacher, can get killed. This is, of course, an important point in Plato, who [p. 121] Marcel van Ackeren describes the fate of the philosopher and prisoner who, after returning to the cave, fails to enlighten the other prisoners and gets killed by them. This is an extreme reaction to philosophy, but according to Williams, Plato is also worried about more daily limits of philosophy. Williams, as noted above, attributes an important original idea to Plato: namely, the ‘force of arguments’ (see SP: 156). To Plato, the helpfulness of philosophy is limited because even the best arguments only have limited force: opposing and irrational views, fallacious arguments, or different and less argumentative forms of speech—​for example, speeches rather than questions and answers—​can have a greater impact on certain people than arguments with plausible premisses and sound logic. In the Gorgias, Socrates describes a competition between the arguments of philosophy and the persuasion of rhetoric, which are compared to the competition of a physician and a cook (Gorg. 464b–​466b). Often, more people are willing to follow the cook, because what they offer is more pleasing than the therapies of the physician. With that, Plato highlights that bad arguments in the form of speeches often gain more acceptance with the majority of people because the weaker arguments often flatter and please the addressees (see also Resp. 488a–​e). The quality of a philosophical argument does not guarantee its success. This limited force of philosophy also comes into view when Williams comments on the question of whether virtue is knowledge and whether and how it can be taught. He points out that ‘the most striking instance of someone who failed to teach his virtue is Socrates himself. Socrates had a pupil and lover, Alcibiades, who was very talented and, it seems, very beautiful. His life was a disaster: vain and petulant, he betrayed Athens and others as well, and died a ruined man. The case of Alcibiades was a reproach to Socrates as a teacher’ (SP: 161). In consequence, Williams views Plato’s recurring and intensive discussion of the relation between virtue and knowledge as an ‘ongoing apology’ (SP: 161). Besides worries concerning the effect of philosophy on our private and public lives, there is another aspect concerning the philosophy of philosophy. According to Williams, Plato makes qualifications regarding the core of this philosophy that are so significant that the qualification becomes part of the central message: The limitations of writing do not apply only to writing. Rather, Plato seems to have thought that the final significance of philosophy for one’s life does not lie in anything that could be embodied in its findings, but emerges from its activities. Two aspects are especially relevant here. For one thing, Williams’s Plato thinks that philosophy has something to offer, and the offer clearly comes in the form of arguments and reasoning. Yet, what can be drawn from philosophy is not simply a statement, but something that follows from philosophical activity and engagement. The other thing is a consequence of the previous point: if philosophy can [p. 122] achieve something worthwhile only when it is pursed and practised, then its success depends on ‘luck and favourable surroundings’. No theory or argument can overcome what Plato described as ‘distrust and even the rejection of the empirical world’ (SP: 178). Philosophy needs luck. Williams repeatedly stresses how different the ancient world was from ours, and that Plato’s ethical views will not help us today; but he also asks why Plato’s work remains so vivid. My take is that Williams found an answer to this by attributing a view to Plato according to which the significance of philosophy does not lie in results that are outlined in defined theories and texts, but only in what follows from philosophical activities. And this requires luck. This view is one of his own core beliefs regarding philosophy’s uncertain effect on us: The dialogues preserve a sense of urgency and of the social and psychological insecurity of the ethical. Plato never forgets that the human mind is a very hostile environment for goodness, and he takes it for granted that some new device, some idea of or imaginative stroke, may be needed to keep it alive there and to give it a hold on us. (SP: 178) This, of course, can be considered a purely doxographical account of what Plato meant to say, and as such it is a highly original way of interpreting Plato, which has not received the attention in the literature on Plato that it deserves. But it is also, to a considerable degree, an account that Williams himself thought to be correct: expressions such as ‘ethical insecurity’ or the question of how philosophy can be a tool for establishing goodness in us are hallmarks of Williams’s own stance. Williams famously criticized the ‘morality system’, but this critique was not about moral thinking only; it was also about philosophy and expressed his more general opposition to theorizing, especially when combined with idealizing, and concerned his view that the effect of philosophy on our lives is very limited indeed. He also found these views in Plato. I have tried to indicate that Williams found views in Plato that match some of his own core beliefs about philosophy, but there is one important and fundamental difference: Plato, and Greek philosophy in general, had no sense of or for history: [T]‌he Greeks did not evolve any theoretical conception of men’s categories of thought being conditioned by the material or social circumstances of their time, nor did they look for systematic explanations of them in terms of history. This type of historical consciousness is indeed not present in all philosophical thought of the present day, but its absence from Greek philosophy is certainly one thing that marks off that philosophy from much modern thought. (SP: 5–​6) Maybe it is historical irony that, in this regard, Williams himself turned out to be a distinctly modern thinker, being indebted in his thinking to philosophers like [p. 123] Marcel van Ackeren Hegel, who made the historical perspective an essential part of philosophy and, at some time, advocated what Williams rejected most vividly: namely, systematizations and idealistic accounts. 5. Conclusions My concluding remarks will address two questions that arise from the more positive image of Plato in Williams, which was described in the previous sections. Recall that I had started by observing that most readers of Williams will characterize his stance towards Plato as negative. So, the first question concerns the relation of Williams’s positive picture of Plato to the well-​known criticism contained in works such as Ethics and the Limits of Philosophy and Shame and Necessity. Secondly, we can ask whether the positive picture just happens to be a list of things that Williams liked in Plato or if it springs from some methodology and thus can be explained by some more fundamental views held by Williams. In an interview Williams was asked, ‘Were you influenced by philosophers from the past, either negatively or positively?’ He answers: A figure in the background who has always exerted a continuous pressure in my philosophical consciousness, although I do not agree with any of his doctrines, is Plato. At heart, one is either a Platonist or an Aristotelian. There is no such thing as a half-​time Platonist. That does not mean that I agree with his answers. In fact, I don’t believe in that sort of transcendental metaphysics at all. But Plato’s agendum is powerfully present. (Williams 1983: 41) Williams here rejects each and every doctrine by Plato, but nonetheless Plato is not a philosopher who influenced him only negatively. Williams’s remark about the power of Plato’s agendum is surely meant to be positive, but he says here neither what the agenda is nor how it relates to all the doctrines he cannot agree to. To Williams it is a ‘fact’ that Plato’s ethical and political suggestions ‘are far removed from any that could serve us now’ (SP: 178). I take it that Ethics and the Limits of Philosophy and Shame and Necessity can be read as elaborating this statement. However, if Plato’s views cannot be considered useful today, what is the point of reading Plato—​are we just reading his works to learn that his ethical and political views are outdated? The previous sections tried to explain what the agenda might be that Williams views as still powerful today. Now, in order to see how the critical and more affirmative aspects of Williams’s take on Plato are linked, we can recall that Williams has two kinds of argument for his claim that Plato’s ethical and political views are not helpful to us. First, there is historical relativism: the ancient [p. 124] theories have ‘collapsed’ because we cannot share their theoretical presuppositions anymore, but also because we cannot simply act according to these theories in a social and political world that has changed so radically since antiquity. The second reason why Plato’s ethical and political theory cannot help us nowadays does not have to do with the actual theories in question, but with philosophy itself. To Williams, many have overestimated the potential power of philosophy over our lives, but philosophy becomes a humanistic discipline if we acknowledge its limits. Above, I have argued that Williams produces a quite surprising key witness for this latter argument: Plato himself. At least in Plato (see SP, ch. 10) he does not focus on what he takes to be an unattainable ethical suggestion, but finds key aspects of his own views about philosophy’s limits in his works. So, the point of reading Plato is not only to learn about historical relativism by devoting oneself to outdated ethical aspects that cannot help us today; the point is also to learn about a concept of philosophy, including a different attitude towards (one’s own) theories and texts, which notably entails a keen awareness of philosophy’s fate in the real world: Plato did think that if you devoted yourself to theory, this could change your life. He did think, at least at one period, that pure studies might lead one to a transforming vision. But he never thought that the materials or conditions of such a transformation could be set down in a theory, or that a theory would, at some suitably advanced level, explain the vital thing you needed to know. So, the dialogues do not present us with a statement of what might be most significantly drawn from philosophy, but that is not a peculiarity of them or of us; nothing could present us with that, because it cannot be stated anywhere, but can only, with luck and in favourable surroundings, emerge. (SP: 179) So, one point of reading Plato concerns the content and what cannot be expected to be the content of this, or any other, writing. Now, the second question regards methodology: is the more positive view on Plato’s philosophy the outcome of a certain methodology? In a famous passage, Williams remarks: Paul Grice used to say that we ‘should treat great and dead philosophers as we treat great and living philosophers, as having something to say to us’. That is fine, so long as it is not assumed that what the dead have to say to us is much the same as what the living have to say to us. Unfortunately, this is probably what was being assumed by those who, in the heyday of confidence in what has been called the ‘analytic history of philosophy’, encouraged us to read something written by Plato ‘as though it had come out in Mind last month’—​an idea which, if it means [p. 125] Marcel van Ackeren anything at all, means something that destroys the main philosophical point of reading Plato at all. (PhHD: 181) Here, Williams refers to his idea that the study of historical texts yields philosophy before history if it leads to an alienation effect (see his more elaborate account in SP: ch. 17): ‘One way in which the history of philosophy can help to serve this purpose is that basic and familiar one of making the familiar seem strange, and conversely’ (SP: 259). This methodology was later overshadowed by his genealogy, but I suggest keeping the alienation effect distinct from genealogy (see van Ackeren 2018, 2019). The idea of the alienation effect has been best summarized by A. W. Moore: The contribution [of the history of philosophy to philosophy] was not, as philosophers in the analytic tradition used to think, to indicate voices of yore which could be heard as participating in contemporary debates: precisely not. It was to indicate voices of yore which could not be heard as participating in contemporary debates, and which thereby called into question whatever assumptions made contemporary debates possible. (A. W. Moore, quoted in Patricia Williams’s Foreword to Williams 2006: IX) In this context, we should understand Williams’s suggestion not to read Plato as though he was recently published in MIND as part of a general account in which ‘Plato’ just serves as a placeholder for any historical author. However, the aim of this chapter was to show that the statement is particularly true in the case of Plato and bears significant additional meaning: the point of reading Plato is that it evokes an extraordinary strong and fruitful alienation effect. What Williams highlights in Plato are those aspects that make Plato seem strange from the viewpoint of mainstream philosophy: the dialogue form, his lack of interest in establishing a systematic theory in his own name, an unscholarly approach to philosophy, and a keen awareness that philosophy should aim at improving lives while, at the same time, being mindful of its own limits. One final caveat: I have focused on Williams’s more positive image of Plato and have only hinted at the relation to his own metaphilosophical views. By pointing to this relation, I do not mean to claim that there is a simple causal relation in which the reading of Plato’s work is the cause and his own view of philosophy the effect. Nor do I wish to claim that it is the other way round. I am more inclined to assume that this is a showcase of how doing philosophy and doing history of philosophy philosophically can be intertwined in such a way that they become one philosophical enterprise.5 5 I would like to thank Roger Crisp and Lee Klein for helpful comments. [p. 126] References Annas, Julia.1993. The Morality of Happiness. Oxford: Oxford University Press. Crisp, Roger. 2020. ‘Are We Climbing the Same Mountain? Moral Theories, Moral Concepts, Moral Questions’. Zeitschrift für Ethik und Moralphilosophie 3.2: 269–​78. Frede, Michael. 1992. ‘Plato’s Arguments and the Dialogue Form’. In Oxford Studies in Ancient Philosophy, Supplementary Volume, 201–​20. Oxford: Oxford University Press. Jenkins, Mark P. 2006. Bernard Williams. Chesham: Acumen Publishing. Queloz, Matthieu. 2021. The Practical Origin of Ideas: Genealogy as Conceptual Reverse-​ Engineering. Oxford: Oxford University Press. Sprague, Rosamond R. 1963. Plato’s Use of Fallacy: A Study on the Euthydemus and Some Other Dialogues. New York: Barnes and Noble. Russell, Paul. 2018. ‘Bernard Williams: Ethics from a Human Point of View’. Times Literary Supplement. van Ackeren, Marcel. 2018. ‘Philosophy and the Historical Perspective: A New Debate on an Old Topic’. In Philosophy and the Historical Perspective (Proceedings of the British Academy, vol. Edited by Marcel van Ackeren, 1–​17. Oxford: Oxford University Press. van Ackeren, Marcel. 2019. ‘Bernard Williams (on) Doing History of Philosophy: A Case Study on Ethics and Limits of Philosophy’. In Ethics beyond the Limits: New Essays on Bernard Williams’ Ethics and the Limits of Philosophy. Edited by Sophie Grace Chappell and Marcel van Ackeren, 53–​71. Abingdon/​New York: Routledge. Williams, Bernard. 1983. ‘The Uses of Philosophy: An Interview with Bernard Williams’. The Center Magazine, November/​December: 40–​9. Williams, Bernard. 1985. Ethics and the Limits of Philosophy. Cambridge, MA: Harvard University Press. Williams, Bernard. 1993. Shame and Necessity. Berkeley: University of California Press. Williams, Bernard. 1994. ‘An Interview with Bernard Williams’. Cogito 8.1: 3–​19. Williams, Bernard. 2006. The Sense of the Past: Essays in the History of Philosophy. Princeton: Princeton University Press (cited as SP). Williams, Bernard. 2008. Philosophy as Humanistic Discipline. Princeton: Princeton University Press (cited as PhHD). Williams, Bernard. 2015. Essay and Reviews. Princeton: Princeton University Press. [p. 130] Pure Enquiry, the Absolute Conception, and Convergence Bernard Williams in Dialogue with Descartes John Cottingham 1. Why Descartes? Bernard Williams published his classic study Descartes: The Project of Pure Enquiry in 1978, at the age of forty-​nine. In many ways Descartes might seem an unlikely philosopher to have attracted his attention. Although Williams’s philosophical concerns were wide ranging, a glance down the list of his publications is enough to show that his predominant interest was in the field of moral philosophy—​not an area to which Descartes is normally thought of as a prime contributor. What is more, Williams’s general philosophical outlook could scarcely be more at variance with the principles that informed Descartes’ world view. In the first place, Descartes was a committed theist, while Williams was deeply unsympathetic to what he called Descartes’ ‘religious metaphysics’ (D: 147).1 In the second place, Descartes was par excellence a system builder, who aimed to construct a tightly integrated edifice on solid foundations,2 whereas Williams’s philosophical approach can broadly be described as anti-​systematic. And thirdly, Descartes saw his enterprise as regulated by what he called the ‘natural light’ of reason, a God-​given human endowment which, when used properly, could not err. Williams, by contrast, in line with his Nietzsche-​inspired attraction to the genealogical perspective, was ever alert to how the supposedly universal voice of reason might turn out to speak in ways that were pervasively conditioned by a particular history and culture. Nevertheless, Williams chose to devote an intricate full-​ length study to Descartes, built around what he took to be a central Cartesian concern: namely 1 D refers throughout to Williams, Descartes: The Project of Pure Inquiry (2015 [1978]); see References for details. 2 Descartes, Meditations, AT VII 17: CSM II 12. An even stronger metaphor, the organic one of a ‘tree of philosophy’, is found in the Preface to the French translation (1648) of the Principles of Philosophy AT IXB 14: CSM I 186. See References for abbreviations used throughout to refer to standard editions of Descartes. John Cottingham, Pure Enquiry, the Absolute Conception, and Convergence In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. DOI: 10.1093/​9780191966361.003.0007 [p. 131] John Cottingham (as the book’s subtitle has it), the ‘project of pure enquiry’. In highlighting the character of Descartes’ epistemological project in this way, it seems plausible that Williams was aiming to further the goal Nietzsche had articulated in his The Gay Science: namely, making the familiar strange and vice versa.3 As Williams interprets it, the Cartesian meditator, or ‘Pure Enquirer’, is one who adopts a very special philosophical perspective, in which ‘there is no worth but the worth of truth-​pursual’ (D: 33). Though it may strike many as perfectly natural and obvious that philosophers should devote themselves exclusively to the pursuit of truth (Descartes himself had used the phrase ‘La Recherche de la Vérité’ as the title of a dialogue he wrote to dramatize his philosophical quest),4 Williams sets himself to bring out why such a perspective is, by the standards of ordinary human life, actually a rather strange one. Here, as so often in his work on Descartes (as in his discussions of other canonical philosophers such as Plato and Nietzsche), we see Williams’s approach to the history of philosophy as profoundly dialectical. He aims not just to expound but to challenge, and to challenge by bringing his own, and our, philosophical concerns into confrontation with those of the author under scrutiny. This is all of a piece with how he conceived of the sort of history of philosophy that he took to be fundamentally worth doing—​one that, in contrast with the history of ideas, aims to be (as he put it) ‘philosophy before it is history’(D: xv). This does not at all imply a disregard for the historical and cultural context in which a text was composed (in his study of Descartes, Williams pays close attention to Descartes’ exchanges with his contemporary critics); but it does require there to be, as Williams put it, ‘a cut-​off point, where authenticity is replaced as the objective by the aim of articulating philosophical ideas’ (D: xvi). It is typical of Williams’s philosophical approach here that confronting the Cartesian project of ‘pure enquiry’ turns out to be a subtle and nuanced undertaking, which criticizes while at the same time bringing out the enduring interest of certain aspects of what is criticized. So, while he underlines certain difficulties that beset the notion of pure enquiry (e.g. its reliance, in its Cartesian form, on the first-​person perspective (D: 54)), Williams also lets us see the potential appeal of the project, in so far as it aims to reach a special kind of truth—​the truth about how things are independently of our own local ways of conceiving them. This in turn leads him to articulate one of the most fertile ideas he took from his study of Descartes, that of the ‘absolute conception’ of reality: One might say that what God has given us, according to Descartes, is an insight into the world as it seems to God, and the world as it seems to God must be the 3 Williams published an edition of Nietzsche’s The Gay Science in 2001; see Nietzsche 2001 [1873–​ Book 5, §355. See also Nietzsche’s Untimely Meditations (Nietzsche 1977 [1873–​6]: Essay 2, p. 98). 4 Descartes, La Recherche de la Vérité par la lumière naturelle, AT X 495: CSM II 400. The date of the dialogue is uncertain, but it may well have been composed in Stockholm during the final months of Descartes’ life. [p. 132] PURE ENQUIRY, THE ABSOLUTE CONCEPTION, AND CONVERGENCE 119 world as it really is. God is thus, on the Cartesian construction, deeply involved in our having . . . an ‘absolute conception’ of reality—​a conception of reality as it is independently of our thought, and to which all representations of reality can be related. (D: 196) The Cartesian project, so understood, is supposed to give us the kind of knowledge that is free from the relativity arising from the preconceptions of the local cultural context in which we operate, and even free from the particular perspective of our human standpoint (e.g. our human modes of sensory awareness). The closest one might come to expressing this notion in secular terms might perhaps be as a conception of reality that could be agreed upon by any rational inhabitant of the cosmos if they were able to abstract completely from their own location and culture and perceptual equipment. The conception has affinities with that described by Williams’s contemporary Thomas Nagel: namely, the conception of a ‘view from nowhere’ (Nagel 1986: ch. 5). 2. The Absolute Conception Williams’s idea of the ‘absolute conception’ has a peculiar resonance, since it calls to mind a problem that in one form or another has recurred throughout the history of western thought. In its broadest shape, it is about the contrast between appearance and reality—​a contrast that gets an intuitively clear purchase from all sorts of specific situations in which we are able to escape from local fluctuations and distortions in our perceptual field to a more stable and reliable grasp of some object, but which on reflection rapidly morphs into a general problem about how we are to have any knowledge of the world whatsoever. Williams put it this way: [I]‌f knowledge is what it claims to be, then it is knowledge of a reality which exists independently of that knowledge, and indeed (except for the special case where the reality known happens to be some psychological item) independently of any thought or experience. Knowledge is of what is there anyway. (D: 48; original emphasis) But what seems so basic and straightforward, the idea of ‘being there anyway’—​ basic and straightforward because we use it in quite ordinary non-​philosophical contexts (I know the milk is in the fridge anyway, whether or not I open the door and look at it)—​turns out on philosophical reflection to be surprisingly elusive. As Williams sees it, if ‘being there anyway’ refers merely to whatever it is that our various representations represent, it does not provide us with any solid conception of what an independent reality is; on scrutiny, such purportedly independent reality ‘slips out of the picture, leaving us only with a variety of possible [p. 133] John Cottingham representations’. Our dilemma is that ‘we have no independent point of leverage for raising [any one representation] into the absolute representation of reality’ (D: 50; emphasis added). The longing for an ‘independent point of leverage’ seems tantalizing because we at once recognize it to be almost by definition out of reach of our human cognitive resources—​a line of thought that leads us forward to Kantian conclusions about the unknowability of the ‘thing in itself ’. Yet Williams invites us, in effect, to consider what shape the problem takes if we bracket off these later developments and consider the matter from the earlier, in some ways more philosophically naive, Cartesian perspective. If self-​reflection leads directly, as it does for Descartes, to knowledge of God, then can we not help ourselves to precisely the independent point of leverage we looked for? For reality as it is ‘anyway’ is now construed as reality as it is perceived by God, and ‘the world as it seems to God must be the world as it really is’ (D: 196). Yet in putting the matter this way, Williams seems to have allowed his distaste for ‘religious metaphysics’ to distort his grasp of the nature and role of the deity in Descartes’ system. It is certainly true that Descartes’ God is that which provides the grounding for objective truth, and in this sense underwrites the idea of a reality that is ‘there anyway’. But there is something very unCartesian about explicating the idea of reality ‘as it is anyway’ by reference to the idea of reality ‘as it is perceived by God’. The picture this latter phrase conjures up is of God as a kind of spectator of the cosmos, so that what we see is an imperfect view of what is seen perfectly by God. But Descartes was far too steeped in Thomistic metaphysics to be tempted by anything that smacked in this way of an anthropomorphic conception of the divine. When we think of a human being perceiving reality, we think of our perception being dependent on or caused by an ‘external permanency’ (to borrow a phrase from C. S. Peirce’s 1877 essay ‘The Fixation of Belief ’; Peirce 1966: 107). Yet in the Cartesian cosmos there can be no ‘external permanency’ in the sense of something independent of the divine power that continuously sustains it in being.5 What is more, our notion of a humanly perceived external reality is essentially bound up with the idea of something that constrains our will (so that in the case of a real object it is not up to us whether to see it or not when it is there in front of us). But for Descartes’ God, like that of Aquinas, there is no real distinction between perceiving or understanding on the one hand and willing on the other. Descartes explicitly endorses the Thomistic doctrine of divine simplicity according to which, as he put it in his magnum opus, the Principles of Philosophy, ‘there is always a single identical and perfectly simple act by means of which God simultaneously understands, wills and accomplishes everything’ (Principles of Philosophy, Part I, art. 23 (AT VIIIA 14: CSM I 201)). 5 For Descartes, there is no real distinction between creation and conservation—​they are only conceptually distinct. See Descartes, Meditations, Third Meditation (AT VII 49: CSM II 33). [p. 134] PURE ENQUIRY, THE ABSOLUTE CONCEPTION, AND CONVERGENCE 121 It is, as Descartes later admitted, very hard for us to understand what it would be to understand or perceive something yet simultaneously and in the same mental act to will it to be so (Conversation with Burman, AT V 166: CSMK 348). But we can say at any rate that the metaphysical picture suggested here must be utterly removed from the idea of objective reality as something ‘out there’ that God perceives (somewhat after the fashion of the popular caricature sometimes invoked in discussion of Berkeley, of God keeping an eye on the tree in the quad when there is nobody else there to observe it).6 Instead, objective reality is for Descartes a rationally intelligible structure that is permanently held in being by the divine will. And our human grasp of this structure, though limited by the finitude of our minds, is nevertheless accurate as far as it goes, in virtue of the divinely bestowed natural light, which, when properly focused, illuminates so brightly as to constrain our assent (Meditations, Fourth Meditation, AT VII 59: CSM II 41). It is this, for Descartes, that allows for, and indeed guarantees, the possibility of human scientific knowledge, knowledge of a reality that is ‘there anyway’ in the sense that it is part of the system of the intelligible structures ordained by God. Or, as Descartes puts it in a well-​known passage in the Discourse on the Method: I noticed certain laws which God has so established in nature, and of which he has implanted such notions in our minds, that after adequate reflection we cannot doubt that they are exactly observed in everything which exists or occurs in the world (AT VI 41: CSM I 131). How does this picture relate to the desire for a ‘point of leverage’, in Williams’s phrase, from which we can catapult ourselves to knowledge of things as they really are? I would suggest that Williams’s formulation, of our having an idea of how things are perceived by God, is less than helpful, for the reasons just explained: we have no real idea of how things are perceived by a being in whom understanding and willing are an identical simple act. What does open the possibility of objective knowledge, for Descartes, is rather that our minds are, to borrow a phrase of Thomas Nagel, ‘instruments of transcendence that can grasp objective reality’ (Nagel 2012: 85). And what this comes down to in Cartesian terms is that the cosmos is a rationally intelligible system, describable in the quantitative language of mathematics, such that when we focus on the mathematical properties of which we have a clear and distinct perception, our will is spontaneously led to assent to the truth so revealed. Yet it is perhaps worth adding that these elements of the Cartesian system, though they may provide us with an accurate enough picture of how Descartes understood the idea of objective scientific knowledge, still do not quite get us 6 The popular caricature is, of course, a travesty since, for Berkeley, there is no tree ‘out there’ in the quad, that which God perceives being, for Berkeley, an idea in the divine mind. [p. 135] John Cottingham as far as explicating the notion of a reality that is ‘there anyway’. For a conception of something as ‘being there anyway’ must include not just a conception of what it is (independently) like, of its essential nature, but also a conception of its really existing. Now we might be tempted to say on Descartes’ behalf that the world ‘as it is anyway’ is simply the vast and immensely intricate system of mathematically characterizable properties ordained by God. But it is still not clear what it amounts to for such a complex actually to exist, except in the abstract and formal sense in which mathematical structures exist (compare Dummett In Cartesian terms, what ‘being there’ amounts to seems rather to be a matter of God’s actualizing or making physical the quantitative system of which we have some grasp through our divinely bestowed mathematical intuitions. But the nature of such actualizing—​the actualizing of a real extended cosmos by an infinite, non-​extended divine mind—​remains mysterious, as Descartes himself expressly remarked, when he compared the divine creation of the world to the mystery of the Incarnation (Early Writings, AT X 218: CSM I 5). Such mysterianism may not seem to sort very well with the image of Descartes as a ‘rationalist’ committed to a transparent system of clear and distinct ideas; but in fact, for all his appeals to reason, there are many passages in Descartes that emphasize the limits of human comprehension, and it is a fundamental principle in Descartes that the infinity that is God cannot be comprehended by the finite human mind.7 Yet this does not preclude God’s having an absolutely central place in the Cartesian system, not just for underwriting the possibility of knowledge, but for accommodating the very idea of a world that is ‘there anyway’ (albeit in a way that falls short of making it fully comprehensible to the human mind). Be that as it may, Bernard Williams’s own reflections in this area led him in a very different direction, one which involved explicating the ideas of objective knowledge and the absolute conception in terms of the notion of convergence. To this we will now turn. 3. Convergence If we accept Williams’s basic thesis that having a conception of a reality that is ‘there anyway’ is ‘basic to the notion of knowledge itself ’ (D: 49), it becomes clear why he was fascinated by Descartes’ attempts to build a ‘religious bridge’ (D: 146) to such an absolute conception via the proofs of the existence of God (proofs which Williams dissects in minute detail in a central chapter of his study of Descartes). While he was in no doubt that the Cartesian strategy fails in this respect, I think it is fair to say that Williams implicitly acknowledges that if the strategy had 7 See, for example, Descartes’ letter to Mersenne of 28 January 1641 (AT III 293: CSMK 172). [p. 136] PURE ENQUIRY, THE ABSOLUTE CONCEPTION, AND CONVERGENCE 123 succeeded, it would indeed have secured the desired prize: in a cosmos of the kind envisaged in Descartes’ theistic metaphysics, the possibility of genuine knowledge is in principle secured. But the corollary of this, for Williams, is that the ‘collapse of the religious bridge’ means that philosophy after Descartes is ‘driven to search for alternative ways’ of securing the possibility of knowledge (D: 146–​7). The alternative that Williams set out to develop turns out to be deeply indebted to the pragmatist approach to truth and reality articulated by C. S. Peirce. In a paper of 1871, to which Williams briefly alludes, Peirce had observed that All human thought and opinion contains an arbitrary, accidental element, dependent on the limitations in circumstances, power and bent of the individual; an element of error, in short. But human opinion universally tends in the long run to a definite form, which is the truth. This final opinion then, is independent, not indeed of thought in general, but of all that is arbitrary and individual in thought; is quite independent of how you, or I, or any number of men think. Everything, therefore, which will be thought to exist in the final opinion is real, and nothing else.8 There are two main ideas in respect of which Williams seems to have been influenced by Peirce here. The first is in his (Williams’s) idea of the absolute conception as something independent of the peculiarities arising from local perspectives and particular modes of apprehending the world (what Peirce calls the ‘arbitrary, accidental element’). The second idea, which is rather trickier to unravel, concerns Peirce’s idea of the truth as that to which ‘human opinion universally tends’; in Williams’s terms, the conception of reality as it is anyway is to be understood as a function of the ‘point of convergence’. These two ideas (convergence, and abstraction from local peculiarities) are explicitly linked by Williams as follows: The scientific representation of the material world can be the point of convergence of the Peircean enquirers precisely because it does not have among its concepts any which reflect merely a local interest, taste or sensory peculiarity. In short, the scientific representation of reality, which in Descartes was the quantitatively describable cosmos ordained by God, is now explicated as the quantitatively describable cosmos on which scientific inquirers tend to converge, once they abstract from the peculiarities of their particular perspectives. 8 From ‘Critical Review of Berkeley’s Idealism’, published in the North American Review of 1871, reprinted in Peirce (1966: 81–​2). The first clause of the penultimate sentence is quoted by Williams [p. 137] John Cottingham The question of how far Williams’s ‘absolute conception’ is to be interpreted in the pragmatist spirit of Peirce is nevertheless not easy to settle. The correct interpretation of Peirce’s views is itself far from straightforward, but on the standard reading he is to be thought of as giving a deflationary account of truth, so that instead of looking for some metaphysical account of an independent reality to which our theories and beliefs are supposed to correspond, we should simply accept that the term ‘true’ is to be cashed out in terms of what is arrived at by the consensus of scientific inquiry, or what would be so arrived at were such inquiry to continue indefinitely. In practical terms, this means for Peirce that ‘to have a true belief is to have a belief that is dependable in the face of all future challenges’ (Capps 2019). This kind of practically oriented conception of truth, while it may indeed capture a great deal of what we are doing when we describe our beliefs as true, clearly tends to move us firmly away from seeing the scientific enterprise in terms of a realist metaphysics. In the case of Descartes, by contrast, there is, as Williams underlines, a fundamental commitment ‘to realism, and to an absolute conception of the world, which includes a conception of matter given by a realistic physical science’ (D: 232). But what of Williams’s own position? It would be a clear mistake to construe his appeals to Peirce as implying an endorsement of an anti-​realist view of science. As Williams interprets convergence, it implies ‘an ideal limit of certainty as the end of scientific inquiry’ (D: 231). But that there is such an ideal end, and that we have already progressed some way towards it, Williams firmly asserts: ‘theories that have the powers that our theories have . . . could not fail to represent in some way how the world really is’ (D: 232). Williams readily admits that convergence may not be as straightforwardly cumulative or inevitable a process as some of Peirce’s formulations suggest (D: 237n17), but to suppose that the idea of convergence is an illusion, or that the notion of an absolute conception is baseless, will, Williams argues, face major difficulties when it comes to explaining the success of our scientific theories. He also argues against the deflationary view that scientific theories are simply a cultural product, which could never be freed from local relativities, pointing out that such a view is going to be hard put to establish a stable conception of the natural world ‘in relation to which it can understand cultural phenomena such as science and its own view of science’ (D: 232). These passages make clear the extent to which Williams, in his conception of scientific inquiry, is committed to a robustly realist view of the world investigated by science, and in so doing, they reveal his philosophical stance as having striking affinities with the Cartesian outlook, despite all the specific points of divergence which we began by noting at the start of this chapter. Discarding Descartes’ theistic framework, Williams nevertheless feels strongly drawn to his absolute conception of reality, and seeks to hold onto it by other means. But is he successful? [p. 138] PURE ENQUIRY, THE ABSOLUTE CONCEPTION, AND CONVERGENCE 125 4. Convergence without God? Williams’s strategy of relying on convergence to defend a realist view of scientific inquiry seems to me vulnerable to a certain kind of difficulty—​one that may be approached by comparing his position with that of Descartes. In expounding Descartes’ view of the absolute conception, Williams rightly points to the role played by the mathematically describable properties (the so-​called ‘primary’ qualities, though this is not a Cartesian term), as against the secondary properties ascribed on the basis of our sensory faculties. As it emerges in the Sixth Meditation, Descartes’ position is that our sensory faculty is, to put it schematically, configured in terms of functional utility for survival, while our intellectual faculty (the one that deals, for example, with the clear and distinct perceptions of logic and mathematics) is configured in such a way as to track the truth. This crucial difference provides Descartes with his building blocks for the achievement of genuine scientia—​including knowledge of intellectual essences and also knowledge of the nature and workings of the physical world in so far as it can be described in logical and mathematical language (AT VII 71: CSM II 49). It also enables Descartes to propose a corrective in cases where, inevitably, the deliverances of the senses will sometimes lead us astray: as he says in the Sixth Meditation, we can keep in mind that the job of the senses is not to track the truth but simply to deal with matters ‘conducive to the preservation of the healthy human being’ (AT VII 87: CSM II and, in cases of conflicting sensory evidence, we can invoke considerations of consistency, and in general employ our (reliable and truth-​tracking) intellectual faculty to sort out confusions and conflicts in our sensory beliefs (AT VII 90: CSM II 62). In short, the Cartesian inquirer, though certainly not possessed of any easy or guaranteed fast track to the truth, is at least established as equipped in principle for the pursuit of truth. The world view in terms of which this all makes sense is, of course, that presupposed in the ‘religious metaphysics’—​the central core of the Cartesian system which Williams so strongly rejected. By contrast, Williams’s own outlook, evidenced in many of his writings, is pervasively informed by what he called the ‘first and hardest lesson of Darwinism’, according to which there is no divine teleology, no benign ‘orchestral score’ overseeing our human actions, or guiding our beliefs (1995: 109–​10). Yet this now risks calling into question the basis of Williams’s faith in scientific convergence. For what Descartes had argued to be the case with respect to our sensory faculty—​namely, that it is configured simply in terms of functional utility for survival—​w ill, I take it, be what the Darwinian naturalist has to maintain not just with respect to our sensory faculty, but with respect to all our human faculties. And this in turn raises the question of why we have any reason to suppose that any of our cognitive faculties are reliable. Alvin Plantinga, in a series of writings culminating in a paper [p. 139] John Cottingham entitled ‘Against Naturalism’, has put the point as follows, focusing on the neural structures associated with a given belief content (e.g. that there is a predator in the vicinity): As a result of having that neuronal event . . . the creature in whom this event is to be found also believes a certain proposition. But what reason is there to think that proposition is true? Granted, the structure in question helps cause adaptive behaviour. But that doesn’t so much as slyly suggest that the content that gets associated with the structure is true. As far as its causing the right kind of behaviour is concerned, it simply doesn’t matter whether the content, that associated proposition, is true or false. At this point, as far as the truth or falsehood of the content that arises, natural selection just has to take potluck. (Plantinga and Tooley Against this, one might argue that the beliefs of organisms such as ourselves about the environment cannot be isolated states, but must form a complex interlocking web that enables us to map our place in the world; and if too many of these beliefs turn out to be false, the whole structure of the web is likely to unravel, with disastrous effects for the organism’s success in the struggle for survival. Nevertheless, Plantinga’s worry does, it seems to me, succeed in highlighting something disquieting about the naturalistic picture of our human cognitive predicament: on the naturalist picture, there is no systematic and principled linkage between our possession of equipment that has turned out to be beneficial in the evolutionary struggle for survival and our supposed capacity to track the truth with respect to the most evident intuitions of our intellects, or what Descartes calls the ‘clear and distinct’ perceptions of the mind.9 For Descartes, hasty judgements and perceptual errors can be corrected by the carefully focused intuitions of the intellect, which are the touchstone for assessing and, if necessary, revising judgements depending on the use of the less reliable faculties. So, the whole sense of ourselves as in principle equipped to inquire into the truth depends on faith in the reliability of our most basic and fundamental logical intuitions. This is exactly the situation as described by Descartes in his replies to his critics: In the case of our clearest and most careful judgements . . . if such judgements were false they could not be corrected by any clearer judgement or by means of any other natural faculty. In such cases, I simply assert that it is impossible for us to be deceived. (Meditations, Second Replies, AT VII 144: CSM II 103) 9 The argument in this section draws on material from Cottingham (2013). [p. 140] PURE ENQUIRY, THE ABSOLUTE CONCEPTION, AND CONVERGENCE 127 Descartes’ reason for making this confident pronouncement rests on the view that God cannot ultimately be a deceiver when it comes to our simplest and most basic intellectual intuitions; in other words, it depends ultimately on the theistic world view that undergirds his entire philosophy. By contrast, anyone who, like Williams, subscribes to a Darwinian account of how our human faculties developed seems to have no reason for any principled confidence that our scientific inquiries will tend to converge on the truth, or that the mathematical reasoning that figures in the resulting ‘absolute conception’ will be reliable. The underlying point is that it seems impossible for any philosopher to characterize our human situation with respect to the truth—​the ways in which we have fallen short, the ways in which we are able to correct our mistakes—​without implicitly taking a stance which assumes that we are indeed, when the chips are down, creatures who are in principle equipped to undertake the search for truth, so that the human mind is, in Nagel’s phrase, an ‘instrument of transcendence, able to grasp objective reality’ (Nagel 2012: 85). And as Nagel—​himself no theist, to say the least—​concludes, it is not clear that this is an assumption that can be underwritten via the resources of evolutionary naturalism. 5. Science and Ethics Williams returned to the idea of convergence some seven years after the publication of Descartes: The Project of Pure Enquiry, in one of his most brilliant and widely admired works, Ethics and the Limits of Philosophy (1985). In this later work, he deploys the idea of convergence in order to make a sharp distinction between the scientific and the ethical domains: In a scientific inquiry there should ideally be convergence on an answer, where the best explanation of the convergence involves the idea that the answer represents how things are; in the area of the ethical . . . there is no such coherent hope. The optimistic view of the prospects for convergence in the scientific sphere and the pessimistic view of such prospects in the ethical sphere are, on Williams’s view, closely linked. For, in the scientific case, he argues, convergence is ‘guided by the way things actually are’ (1985: 136). Williams describes this as a ‘compelling’ picture of the world as ‘already there’, ‘helping to control our descriptions of it’ something for which he insists there is no plausible corresponding picture in ethics. This conception of reality as it is anyway, ‘guiding’ or ‘helping to control’ our descriptions of it, is a striking one. It presents a picture of reality such that our proper relation to it is in an important sense passive: authentic science is not a matter of our [p. 141] John Cottingham being autonomous controllers or determiners of the truth, but rather of our being responsive to something beyond ourselves that constrains our choice. For anyone familiar with the Cartesian picture of our relation to the truth, this will strike an immediate chord. To be sure, the Cartesian spirit of inquiry often seems to emphasize the independence and autonomy of the inquiring mind, and we see this especially in Descartes’ own method of doubt, his determination to ‘bend his will in the opposite direction’, as he puts it in the First Meditation (AT VII 22: CSM II 15), so as to cautiously withhold assent from anything that might lead him astray. But as the Meditations progress, and the meditator learns to ‘gaze in wonder’ at the ‘immense light’ irradiating the mind on its journey towards God (Third Meditation, final paragraph), the stance shifts decisively, from one of detached scrutiny to one of submissive assent. As Descartes observes in the Fourth Meditation, the clarity of certain evident truths produces a spontaneous and irresistible assent in the will: ex magna luce in intellectu magna consequuta est propensio in voluntate (‘from a great light in the intellect there followed a great propensity in the will’; AT VII 59: In his study of Descartes, Williams had drawn attention to this famous phrase in the Fourth Meditation in order to discuss Descartes’ account of the will and his theory of mental assent. But it is significant that the examples Williams gives of the will’s assent following the perceptions of the intellect are all drawn from the theoretical propositions of mathematics. What Williams fails to see, or at any rate to comment on, is that Descartes’ doctrine of the assent of the will following the clear perceptions of the intellect applies equally to the domain of ethics. For Descartes, whether you focus on a transparent mathematical truth or a transparent moral truth, you have no choice but spontaneously to declare ‘yes, it’s true—​to be affirmed!’ or ‘Yes, it’s good—​to be pursued!’ It is striking, especially perhaps for the modern reader, to see Descartes explicitly asserting that the scope of the ‘great light’ in the intellect encompasses the domain of ethical as well as of mathematical truth: in the Fourth Meditation, Descartes considers the ratio veri (‘reason of truth’) and the ratio boni (‘reason of goodness’) both together in the same breath (AT VII 58: CSM II 40). A common assumption of much philosophical inquiry in the aftermath of David Hume has been that the domains of fact and of value are sharply separated; but in Descartes’ metaphysical world view, they are part of one and the same objective, divinely grounded reality. (To avoid misunderstanding, it should perhaps be added that we are speaking here of the simplest and clearest ethical truths that are manifest to the natural light—​perhaps truths such as ‘generosity is good’, or ‘cruelty is bad’—​truths analogous in their simplicity and clarity to a simple mathematical truth such as ‘two plus three makes five’. Descartes’ thesis about the role of the natural light in ethics does not commit him to the idea that the answer to every detailed practical question about the ordinary conduct of life can be deduced from first principles, any more than [p. 142] PURE ENQUIRY, THE ABSOLUTE CONCEPTION, AND CONVERGENCE 129 he is committed to holding a deductivist view of all scientific truth (see further Cottingham 2017).) One fairly obvious upshot of this parallelism between the theoretical and the moral domains in Descartes is that, in a Cartesian picture of reality, one may expect the same kind of convergence in the ethical sphere as in the scientific. Ethical properties, if we were to express it in Williams’s terminology, would turn out to be just as much a part of the reality that is ‘there anyway’ as the physical properties that are the object of scientific inquiry. It is perhaps a pity that, in discussing the Fourth Meditation in his book on Descartes, Williams seems to have overlooked the Cartesian parallelism between reasons of truth and of goodness, since attending to it might have led him to explain how deeply uncongenial such a view was to his own philosophical stance. For a core tenet of Williams’s view of ethics, expressed in a famous paper published three years after his Descartes, was the insistence that there can be no ‘external’ reasons requiring us to act: that is, no reasons that are not derivable in principle from the motives, the internal motivational set, of the agent (Williams 1980). And this in effect makes the whole idea of responding to a reality that is ‘there anyway’, irrespective of the motivational stance of the agent, something of a non-​starter. In Ethics and the Limits of Philosophy, Williams went so far as to consider the possibility of a more objectivist account of ethical reality. For this to be plausible, he reasoned, one would have to be able to give some account of what it would be for us to have propositional knowledge of ethical truths, which in turn would require some account of what it would be for our beliefs to ‘track the truth’, or be responsive to ethical reality. But reflecting on such notions leads Williams to a swift and uncompromising conclusion that it could not work: ‘I cannot see any convincing theory of knowledge for the convergence of reflective ethical thought on ethical reality in even a distant analogy to the scientific case’ (1985: 152). Ruling out any return to a theistic world view, in which reasons of truth and reasons of goodness are equally grounded in the nature of an ultimate divine reality, the only possible kind of ethical objectivism Williams could envisage was one based on some agreed idea of the most satisfactory life for human beings, given certain general features of human nature. But here, again influenced by a Darwinistic perspective on the highly contingent nature of our origins and development as a species, Williams was inclined, when reflecting on these matters some ten years later, in his Making Sense of Humanity, to see human nature as a disparate amalgam of genetically determined and culturally inherited propensities and dispositions, with no reason to think that there is one clearly right way of living for any given individual, let alone one that will suit everyone: [The] most plausible stories now available about [human] evolution, including its very recent date and also certain considerations about the physical characteristics of the species, suggest that human beings are to some degree a mess, and that [p. 143] John Cottingham the rapid and immense development of symbolic and cultural capacities has left humans as beings for which no form of life is likely to prove entirely satisfactory, either individually or socially. (Williams 1995: 109; emphasis added) Convergence becomes highly unlikely on this distinctly pessimistic conception of the human predicament, and indeed the seeds of such pessimism are already present in Ethics and the Limits, where Williams emphasized the widely disparate character of the various projects and goals that a human being may adopt, with no reason to think they will all ‘fit together into one harmonious whole’ (1985: 153). To set in perspective Williams’s dismissal of the prospects for convergence in ethics, it may help to contrast it with the position arrived at by a fairly close contemporary, the equally distinguished moral philosopher Derek Parfit, whose magnum opus, On What Matters, argued for a clear convergence in their practical implications between two seemingly radically different ethical theories, Kantianism and Consequentialism.10 Parfit’s ethical outlook differed from that of Williams not only in respect of convergence, but also, and partly connectedly, as regards meta-​ethics, where he took a strongly objectivist view of ethical truth. In line with the title of his book, Parfit maintains that some things really, objectively matter. How we treat people really matters; whether we look after our planet so that humanity survives really matters. These are genuine moral truths. As Parfit puts it, ‘In believing that some things matter, I am believing that there are some irreducibly normative truths’ (2011, vol. 2: 464; emphasis added). In other words, such moral truths have objective authority over us and give us decisive reasons to act in certain ways. But what grounds this objectivity? What makes the truths true? The Cartesian answer to such questions would be, as we have seen, that there is an objective moral order, grounded ultimately in the nature of God, independent of our mere contingent inclinations and desires, and that this order exerts a normative power or authority over us, providing us with ‘reasons of goodness’, conclusive and compelling reasons to behave in certain ways. But, along with Williams and the majority of contemporary philosophers, Parfit rejects the theistic answer: he cannot accept God as the reality underlying the objective moral order. And, again like Williams, he subscribes to the standard materialist-​naturalist paradigm: only the natural world exists, or, as Parfit puts it, there are no ‘strange’ parts of reality (2011, vol. 2: 487). So what grounds the objective truths of morality? The arguably baffling answer that Parfit offers is: nothing. Although he insists they are true, ‘as true as any truth could be’, he holds that there is no underlying reality that makes them true: they have ‘no ontological status’. According to his view, ‘for such claims to be true, the reason-​involving properties need not exist either as natural 10 Derek Parfit was born thirteen years after Williams; the first two volumes of his mammoth On What Matters appeared in 2011, the third in 2017. [p. 144] PURE ENQUIRY, THE ABSOLUTE CONCEPTION, AND CONVERGENCE 131 properties in the spatio-​temporal world, or in some non-​spatio temporal part of reality’ (2011, vol. 2: 486). As we have seen, a principal reason for Williams’s scepticism about objectivity and convergence in the ethical sphere was his assumption that there could be no coherent account of an ‘ethical reality’ analogous to the reality studied by science. Parfit’s position could be seen as trying to cut this gordian knot by denying that objective ethical truths need be made true by any ethical reality. Though there is no space to argue the point in any detail here, it is possible to see the respective positions of Williams and Parfit as illustrating the dilemma that arises once a theistic underpinning for ethics is abandoned. In the case of Williams, there is no recourse but to abandon the hope of objectivism in the ethical sphere. In the case of Parfit, objectivism is retained, but at the cost of baldly asserting that there can be irreducible ethical truths without any truth-​makers. Many may feel that the two positions are, in different ways, equally unsatisfactory (see Cottingham 2014: ch. 4). 6. Conclusion Whatever one’s stance on these difficult questions of objectivity in science and in ethics (and clearly, in both domains, there are many more aspects to these questions than can be touched on here), there is little doubt that Williams’s encounter with the Cartesian system was extraordinarily fruitful. In unravelling Descartes’ ‘project of pure inquiry’, he was able to articulate the idea of an ‘absolute conception of reality’, the idea of a reality that is ‘there anyway’, in so fertile a manner that those addressing many of the enduring problems in contemporary philosophy seem likely sooner or later to have to confront this idea in ways that will be indebted to how Williams formulated it. To take one example, the intractable problems of consciousness (including the so-​called ‘hard’ problem, which so much exercises today’s philosophers and cognitive scientists) seem partly to hinge, as Williams put it, on ‘what kind of thing is added to the physical picture of things when consciousness is added to it’, and ‘what sort of facts, from an objective or absolute standpoint, can psychological facts be?’ (D: 231). Another example concerns the still unresolved question of the status of ethical inquiry, and how far it can coherently be seen in terms of a search for ethical knowledge; and how far, in turn, such knowledge can be understood in terms of a reality that guides and helps shape our ethical conceptions. And finally, by choosing to study the ‘father of modern philosophy’ who was at the same time a committed theist, Williams often seems implicitly to be setting himself the task of seeing how much of the Cartesian system can shed light on the philosophical concerns of a secularist epistemology and metaphysics, thereby setting up challenges that the protagonists of our contemporary naturalist world view still wrestle with. By raising such questions in the context of expounding and criticizing the work of one of the world’s great canonical philosophers, Williams [p. 145] John Cottingham succeeded in bringing the history of philosophy vividly to life, and showed with consummate skill how what he called the ‘humanistic’ discipline of philosophy can itself be significantly enriched by bringing the ideas of the great thinkers of the past into confrontation with our present-​day perspectives.11 References Capps, John. 2019. ‘The Pragmatic Theory of Truth’. In The Stanford Encyclopedia of Philosophy (Summer 2019 Edition). Edited by Edward N. Zalta, . Cottingham, John. 2013. ‘Descartes and Darwin: Reflections on the Sixth Meditation’. Proceedings of the Aristotelian Society, Supplementary Volume 88: 239–​77. Cottingham, John. 2014. Philosophy of Religion: Towards a More Humane Approach. New York: Cambridge University Press. Cottingham, John. 2017. ‘Context, History and Interpretation: The Religious Dimension in Descartes’s Metaphysics’. In Essays in Honour of Desmond Clarke. Edited by S. Gaukroger and C. Wilson, 42–​53. Oxford: Oxford University Press. Descartes, René. Citations employ the standard abbreviations for editions of Descartes, as follows: ‘AT’ refers to Œuvres de Descartes, 12 vols, revised edn. Edited by C. Adam and P. Tannery. Paris: Vrin/​CNRS, 1964–​76. ‘CSM’ refers to The Philosophical Writings of Descartes, vols I and II. Translated by J. Cottingham, R. Stoothoff and D. Murdoch. Cambridge: Cambridge University Press, 1985. ‘CSMK’ refers to vol. III, The Correspondence, by the same translators and A. Kenny. Cambridge: Cambridge University Press, 1991. Descartes, René. Conversation with Burman [1648]. Translated by J. Cottingham. Oxford: Clarendon Press, 1976. In AT V 146–​79: CSMK 332–​54. Descartes, René. Discourse on the Method [Discours de la méthode, 1637]. In AT VI 1–​78: CSM Descartes, René. Early Writings [Cogitationes privatae, 1619–​22]. In AT X 215–​19: CSM I 2–​5. Descartes, René. Meditations on First Philosophy [Meditationes de prima philosophia, 1641]. In Descartes, René. Principles of Philosophy [Principia philosophiae, 1644]. In AT VIII 5–​329: CSM I 193–​291. French translation [Les Principes de la philosophie, 1648] in AT IXB 1–​325. Dummett, Michael. 2006. Thought and Reality. Oxford: Clarendon Press. Nagel, Thomas. 1986. The View from Nowhere. Oxford: Oxford University Press. Nagel, Thomas. 2012. Mind and Cosmos. Oxford: Oxford University Press. Nietzsche, Friedrich. 1977 [1873–​6]. Untimely Meditations [Unzeitgemässe Betrachtungen]. Translated by R. Hollingdale. Cambridge: Cambridge University Press. Nietzsche, Friedrich. 2001 [1882–​7]. The Gay Science [Die Fröhliche Wissenschaft]. Edited by Bernard Williams. Cambridge: Cambridge University Press. Parfit, Derek. 2011. On What Matters, vols I and II. Oxford: Oxford University Press. Peirce, Charles. 1871. ‘Critical Review of Berkeley’s Idealism’. North American Review 93: 449–​72. Peirce, Charles. 1877. ‘The Fixation of Belief ’. Text available at . Peirce, Charles. 1966. Selected Writings. Edited by P. Wiener. New York: Dover. Plantinga, Alvin, and Michael Tooley. 2008. Knowledge of God. Oxford: Blackwell. Williams, Bernard. 1980. ‘Internal and External Reasons’, repr. in Williams 1981: 101–​13. Williams, Bernard. 1981. Moral Luck. Cambridge: Cambridge University Press. 11 See one of the last essays that Williams published, his ‘Philosophy as a Humanistic Discipline’ in Williams (2005: 180–​99). [p. 146] PURE ENQUIRY, THE ABSOLUTE CONCEPTION, AND CONVERGENCE 133 Williams, Bernard. 1985. Ethics and the Limits of Philosophy. London: Collins/​Fontana. Williams, Bernard. 1995. Making Sense of Humanity. Cambridge: Cambridge University Press. Williams, Bernard. 2005. Philosophy as a Humanistic Discipline. Edited by A. W. Moore. Princeton: Princeton University Press. Williams, Bernard. 2015 [1978]. Descartes: The Project of Pure Enquiry. Abingdon: Routledge. Cited by the abbreviation D. [p. 147] Getting Round the Cartesian Circle Gerald Lang In the Meditations, Descartes envelops his meditator in successive waves of sceptical doubt, and then describes how the meditator manages not only to defeat the scepticism, but to arrive at a new understanding of the foundations of his knowledge. These aims are not unrelated. As Bernard Williams remarks, the Cartesian meditator ‘does not merely encounter scepticism as an outside force and survive it, like the knightly hero of some romance’ (2006: 232). Rather, the meditator is engaged from the very beginning in the task of ‘deploying a scepticism which he has intelligibly shaped to his own purpose . . . of satisfying the desire for knowledge’ Having defeated the scepticism, we should not expect the epistemology to remain undisturbed, or indeed the metaphysics to which the epistemology grants us access. This is a philosophical audit that is meant to secure certain results. On the face of it, this sort of exercise might seem blameless. If scepticism is a challenge to which we are properly answerable, at least when we are doing fundamental philosophy, we should be able to learn something about the structure of our knowledge by seeing what stresses and strains it is forced to endure when we tame the sceptical challenge. Accompanying this exercise should also be a readiness to amend our conception of knowledge, and to revise our account of what we know. Nonetheless, I want to show that the relationship between the scepticism-​defeating exercise and the reconstruction exercise, at least in Descartes’ hands, comes closer to something like insider trading. The sceptical strategies deployed by the meditator are intimately connected to the sort of world, and the sort of access to that world, which the meditator expects to get out of them. Had those sceptical strategies been different, the results would be different as well. The crux of the meditator’s sceptical crisis is realized in the problem of the ‘Cartesian Circle’, and most of this chapter will be focused on this thorny problem, and on how the meditator can resolve it. The chapter will unfold as follows. A brief introduction to the Cartesian Circle is undertaken in section 1. In sections 2 and 3, I will outline and appraise Williams’s influential treatment of the Cartesian Circle. Though I will be critical of Williams’s efforts, his approach contains several insights that will not go unheeded when I begin to construct my own solution, which I will refer to as the ‘Recognition Interpretation’, across sections 4–​9. Section 10 Gerald Lang, Getting Round the Cartesian Circle In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. [p. 148] returns to the thoughts about the scepticism-​defeating exercise and the reconstruction exercise which I raised above, and applies some suspicions to Williams’s entirely secular version of the reconstruction exercise. 1. The Problem of the Cartesian Circle The problem of the ‘Cartesian Circle’ was raised, in a pleasingly understated manner, by Arnauld in the Fourth Objections:1 I have one further worry, namely how the author avoids reasoning in a circle when he says that we are sure that what we clearly and distinctly perceive is true only because God exists. But we can be sure that God exists only because we clearly and distinctly perceive this. Hence, before we can be sure that God exists, we ought to be able to be sure that whatever we perceive clearly and evidently is true. (Fourth Objections, The basic story is a familiar one. Descartes’ meditator, in the First Meditation, decides to treat as false every belief for which he can adduce the slightest reason for doubt, for the purposes of establishing orderly and stable foundations for scientific knowledge. This is the ‘Method of Doubt’. Soon, under the pressure of sceptical hypotheses, cumulating in the possible existence of a deceiving God or a malicious demon, he is left with almost nothing in which to believe. All his previous beliefs about the external world are treated as false, as are his abilities to latch on to simple arithmetical truths. In the Second Meditation, the meditator reaches—​or appears to reach—​his first existential certainty, concerning his own existence. (This is the cogito.) From this rather modest-​looking beachhead of certainty, the meditator derives the so-​called ‘Truth Rule’ at the beginning of the Third Meditation, according to which ‘I now seem to be able to lay it down as a general rule that whatever I perceive very clearly and distinctly is true’ (CSM II 24). Almost immediately, however, the Truth Rule seems to be in jeopardy, as the meditator reflects on his failure to dismiss the possibility that he is being continuously deceived by an infinitely powerful being: [W]‌henever my preconceived belief in the supreme power of God comes to mind, I cannot but admit that it would be easy for him, if he so desired, to bring it 1 I will cite Descartes (1985), abbreviated as ‘CSM’. Page numbers will be preceded by the volume in question, and (where needed) the relevant section of the text (First Meditation, Second Replies, etc.). 2 Arnauld’s query is foreshadowed in the Second Objections (CSM II 89). See below, section 9. [p. 149] Gerald L ang about that I go wrong even in those matters which I think I see utterly clearly with my mind’s eye. (Third Meditation, CSM II 25) Now the meditator is prepared to affirm his psychological certainty about many of his beliefs (CSM II 25). While experiencing them, he cannot doubt them. But he knows that he might, for all that, be deceived, because psychological or subjective certainty is not sufficient for metaphysical certainty. So what he needs to do now is to prove that a non-​deceiving God exists, a God who would prevent him from being deceived in this way. But how is he to argue for the existence of a non-​deceiving God? He can only use his clear and distinct perceptions, which are what the Truth Rule authorizes him to use. The purpose of proving God’s existence, however, was to vindicate his clear and distinct perceptions. As things stand, they are not completely certain. The problem of the ‘Cartesian Circle’, then, is this: Descartes’ meditator needs to be metaphysically certain that a non-​deceiving God exists in order to be able to be certain of his clear and distinct perceptions, but he also needs to be metaphysically certain of his clear and distinct perceptions in order to be certain that a non-​ deceiving God exists. Before proceeding further, it will be helpful to enumerate three central challenges that any serious solution to the Cartesian Circle must address. First, solutions to the Circle need to be grounded in Descartes’ own writings. This is the Textual Challenge. Now the Textual Challenge can be satisfied to a greater or lesser extent, and we should be prepared to appraise each solution in a scalar way. A number of passages have to be somehow negotiated and brought under the fold of any particular interpretation, and this task is not made any easier by the fact that these passages do not always appear to be pointing in the same direction. For each of these interpretations, some passages will be more recalcitrant than others, and it may be that no interpretation wholly satisfies every one of them. That one’s favoured interpretation makes a decent fist of the Textual Challenge, however, is still a sensible desideratum. Secondly, it has to be shown why the clear and distinct perceptions involved in the proof of God’s existence stand in no need of the epistemic resources provided by the outcome of that proof. Why are the clear and distinct perceptions relevant to proving God’s existence, but not other clear and distinct perceptions, already certifiable as reliable? This is the Exemption Challenge. The third challenge, which I will call the Redundancy Challenge, can be regarded as forming a second horn of a dilemma with the Exemption Challenge on the first horn of it. According to the Redundancy Challenge, it has to be shown exactly how the proof of God’s existence is supposed to help with the initial problem. If clear and distinct perceptions lead him to certainty about God, and the meditator can trust those perceptions, then God’s role seems redundant, at least in terms of combating scepticism. The meditator can simply rely on his clear and distinct [p. 150] perceptions instead. God’s proof adds nothing to his epistemic credentials. How, then, does the proof of God’s existence avoid this kind of redundancy? Here is another way of putting some of these points, focusing in particular on the epistemic value of clear and distinct perceptions. A plausible constraint on solutions to the Circle is that the epistemic value of clear and distinct perceptions be low enough to benefit from the proof of God, so that the Redundancy Challenge is tackled, but high enough to be able to deliver the proof of God, so that the Exemption Challenge can be dealt with. It is taxing, to say the least, to identify what sort of epistemic value clear and distinct perceptions must have in order for them to pull off this double feat. 2. Williams’s Solution to the Circle To locate Williams’s contribution to the discussion of this problem, it is worth paying greater attention to Descartes’ basic description of the Method of Doubt in his First Meditation ([A]‌and [B] are my additions): [A]‌Reason now leads me to think that I should hold back my assent from opinions which are not completely certain and indubitable just as carefully as I do from those which are patently false. [B] So, for the purpose of rejecting all my opinions, it will be enough if I find in each of them at least some reason for doubt. What is the connection between [A]‌and [B]? Here, I believe, we arrive at a fundamental fork in the road for interpretations of Descartes’ philosophy, which will, in turn, have significant implications for the appropriate approach to the Circle. On one view, [B]‌carries priority over [A]. The withdrawal of assent from one’s opinions will be determined by the truthful apprehension that one has a reason for withdrawing it, and nothing else. (That reason for doubt will be constituted by a sceptical hypothesis whose truth he is not yet in a position to rule out.) Call this view Epistemic Priority. On an alternative view, [A]‌carries priority over [B]. What makes something a reason for doubt is determined by whether it in fact leads the meditator to withdraw assent from his opinions. Reasons for doubt are simply defined, in part, in psychological terms, by their actual tendency to dislodge the meditator’s assent to propositions that, up to now, he has confidently taken to be true. Call this view Psychological Priority. The two-​stage statement of the Method of Doubt leaves the priority between [A]‌and [B] underdetermined. Some writers embrace Epistemic Priority,3 while 3 See, for example, Feldman (1975), Van Cleve (1998), Kenny (1993), and Della Rocca (2005). [p. 151] Gerald L ang others are more favourable to Psychological Priority.4 Within this broad division of approaches to the Circle, Williams cuts an interestingly ecumenical figure. While his basic allegiance appears to lie with Epistemic Priority, much of the substance of his approach draws on features that are especially prized by Psychological Priority. Williams acknowledges the abundant textual evidence in the Meditations and elsewhere that, while being experienced, clear and distinct perceptions compel our assent. They are in this sense irresistible. Of course, as Williams also concedes: ‘Irresistibility does not entail truth’ (PPE: 187).5 This is why a doubt about clear and distinct perceptions can arise only when we are not directly experiencing them. Clear and distinct perceptions’ vulnerability to doubt is not best tackled by seeing the meditator as being especially concerned with the fragility of our memories (PPE: 195–​8, building on arguments in Frankfurt 1970). The possibility of faulty or deceitful memory simply flushes to the surface the underlying fundamental concern: why should psychological certainty sustain metaphysical certainty, when we might be the playthings of a malicious demon? Even if one’s memory were faultless, there would be no guarantee that what one takes to be true is genuinely true, even if one is incapable of doubting the veracity of clear and distinct perceptions while experiencing them. One of Williams’s thoughts here—​and a defensible thought as far as it goes—​is that Descartes never disinvests entirely in the reliability of clear and distinct perceptions. As we have already seen, Descartes’ project is to defeat sceptical challenges in order to arrive at stable foundations for knowledge: the type of scepticism the meditator engages with can thus be regarded as ‘pre-​emptive scepticism’ (PPE: 62). Scepticism is treated as the barrier to the acquisition of any such knowledge, and Descartes takes these challenges seriously. There is no guarantee that they can be met, and the story is not guaranteed to end happily. But our very ability to acknowledge the force of the sceptical challenges presupposes our ability to treat reasons for doubt as reasons for doubt. Reasons for doubt can be defeated by reasons for withstanding those doubts, and it is the meditator’s aim to uncover these reasons. Perhaps the meditator cannot do that, because his attempts to defeat the sceptical challenges will be unsuccessful, but that will not be due to his very inability to recognize the reasons for doubting that they enjoin on him. This sort of cognitive fuel for the journey ahead of him is simply taken for granted. The sceptical challenge is an intellectual challenge. The meditator at no point challenges our ability to reach conclusions through reason, and to maintain those conclusions unless he can identify better reasons for overturning those conclusions. 4 See Loeb (1990, 1992). Bennett (1990) is also attracted to this broad approach, though more equivocally so than Loeb. 5 I abbreviate in-​text references to Williams (1990) to PPE. See also Williams (2006) for a useful guide. [p. 152] If we can be counted on to be sensitive to reasons for belief and for suspending belief, then we must be able, psychologically, to detect such reasons. On the psychological side, our grasp of such reasons, if they are to be grasped at all, will be embodied in clear and distinct perceptions. True, non-​occurrent clear and distinct perceptions can be doubted, at least until the hyperbolic sceptical scenario can be laid aside. But not even the malicious demon depicted in the hyperbolic sceptical scenario is capable of disabling the meditator’s ability to recognize that the demon’s possible existence gives him a reason to suspend his ordinary beliefs. The malicious demon could not, for example, get the meditator to overlook the threat to his knowledge posed by the undiscounted possibility of the demon’s existence; the demon could not trick the meditator into settling for a form of epistemic complacency that would convince him he was rigorously applying the Method of Doubt when he was not. In this sense, the meditator’s approach to scepticism is tailored to his project of achieving knowledge. Taking scepticism seriously also licenses the meditator to take his own intellectual powers seriously. If it is reason he is trying to vindicate, Descartes knows that he cannot dispense with reason in the process of doing so. This best explains his cursory dismissal of certain sceptical possibilities, such as the ‘madness possibility’ briefly aired in the First Meditation (CSM II 13), that are incompatible with his possession of the cognitive resources that enable him to recognize, and respond to, the reasons he encounters under the Method of Doubt. Still, the meditator’s ability to recognize reasons for doubt when he encounters them does not, in and by itself, provide him with the resources for tackling hyperbolic scepticism. He needs the proof of a non-​deceiving God to dispel the last epistemic scruple attaching to his clear and distinct perceptions. So how does he do that? Williams thinks that the meditator can avoid circularity by intuiting the clear and distinct perceptions involved in the arguments for God directly. (These are the cosmological argument in the Third Meditation, and the ontological argument in the Fifth Meditation. I will not comment on the adequacy or otherwise of these arguments, compelling though they may have seemed to Descartes.6) While intuiting the clear and distinct perceptions for God’s existence, or while these perceptions are occurrent, they are irresistible. The meditator cannot doubt them while experiencing them. These momentary certainties are of limited use to the meditator, however, in constructing more robust foundations for knowledge, because assembling a wider body of knowledge is incompatible with continuous and single-​minded absorption in the occurrent clear and distinct perceptions relating to the arguments for God’s existence in particular. So the meditator’s challenge, as Williams interprets it, is to transform the momentary certainties that 6 For a frank assessment of them, see PPE: ch. 5. Williams is not remotely convinced by them. [p. 153] Gerald L ang a non-​deceiving God exists into a usable basis for knowledge that allows him to enrol his other clear and distinct perceptions. What will fit the bill? The meditator can earn the benefits of these intuitions of divinity by adopting the following general acceptance-​rule for beliefs: (A) Accept as ongoing beliefs just those propositions which are at any time clearly and distinctly perceived to be true. (PPE: 203) Once (A) is accepted, the meditator will be in a position to accept (1), which will then allow him to accept (2), followed by (3): God exists and is no deceiver. What one clearly and distinctly perceives to be true is true. A) as an acceptance-​rule for a truth-​seeker is a sound rule. (PPE: 203) Is the Circle avoided if the meditator accepts (A)? Williams concedes that the Circle would still be unsolved if the meditator has to admit that the sole basis for his knowledge of (1) is (2) and (3), and vice versa (PPE: 203). But the meditator can avoid this form of circularity, Williams thinks, because he does not have to agree that his knowledge of (1) presupposes his knowledge of (2) and (3). The meditator intuits (1), directly, through an occurrent clear and distinct perception that delivers the content of (1), and he can use that to buttress (2) and (3) (PPE: 203–​4). Thus the Circle is broken. The intuition that (1) is true comes first, and is used to fortify the credentials of (2) and (3). Serious science now awaits, together with a reconstructed understanding of how our world is physically organized. 3. Problems with Williams’s Solution Williams wishes to avoid interpretations of the Method of Doubt which ‘would make his project trivially hopeless from the start’ (PPE: 204), and which would involve acquiescing to the ‘demand . . . that [the meditator] can never deploy any consideration until he has validated that kind of consideration’ (PPE: 204). By refusing this demand, Williams thinks the Cartesian meditator can secure (1), through direct intuition, without having to demonstrate prior knowledge of (2) and (3). I argued in section 2 that Williams is within his rights to suppose that the Cartesian meditator can simply help himself to the intellectual wherewithal to execute the Method of Doubt. If the meditator encounters a non-​dismissible sceptical possibility, he knows that he has reason not to dismiss it. But on pain of hollowness, the meditator’s possession of this intellectual wherewithal should not be allowed to guarantee victory against the sceptic. And there are signs, as I see it, of hollowness in Williams’s suggested route to victory. [p. 154] Non-​occurrent clear and distinct perceptions can be significantly doubted. Why so? It is because the hyperbolic doubt does not exempt them from doubt: the malicious demon may thwart the operation of the meditator’s clear and distinct perceptions, or add unknown faculties to him that would lead him into error: But what about when I was considering something very simple and straightforward in arithmetic or geometry, for example that two and three added together make five, and so on? Did I not see at least these things clearly enough to affirm their truth? Indeed, the only reason for my later judgment that they were open to doubt was that it occurred to me that perhaps some God could have given me a nature such that I was deceived even in matters which seemed most evident. (Third Meditation, CSM II 25) Just as the [spontaneous] impulses . . . seem opposed to my will even though they are within me, so there may be some other faculty not yet fully known to me, which produces these ideas without any assistance from external things . . . (Third Meditation, CSM II 27) Such passages will be revisited later, through a new interpretive lens. Their significance for now is to remind us that the Redundancy Challenge, at least, has ready answers. The meditator needs a non-​deceiving God because otherwise he will be stuck with these sceptical doubts. The price of dealing so comfortably with the Redundancy Challenge is that the Exemption Challenge now risks being unanswered. How do we get to repose trust in the intuition that God exists? The fact that it is an intuition, and the fact that it compels assent while being experienced, cannot make all the difference. In fact, its assent-​compellingness is a distraction: it renders us unable to assess its reliability. We might be taken in by a clear and distinct perception while experiencing it, even if it is unreliable. If that is so, then how does the particular occurrent clear and distinct perception that God exists and is no deceiver surmount the sceptical challenge? There is nothing special about this particular clear and distinct perception. It is vulnerable to the same sceptical uncertainty as all his other clear and distinct perceptions, at least in their non-​ occurrent form. In defence of his approach, Williams writes: It is not . . . as though the sceptic could complain to Descartes that that concentration on irresistible propositions might be like sniffing some hallucinogenic gas, and Descartes’s procedure like someone who, when a doubt was raised in the absence of the gas, proposed to dispel the doubt by inviting one back to the gas. [p. 155] Gerald L ang But why isn’t it like that? Yes, the meditator reposes trust in reason. Clear and distinct perceptions provide the most secure access to reason he can think of. But if clear and distinct perceptions, or what he takes to be clear and distinct perceptions, were good enough as they stood, the Redundancy Challenge would be unmet. The meditator would not have to embrace anything more than the Truth Rule, because clear and distinct perceptions would already have been given a clean bill of epistemic health. But if God is required to dispel the slight doubts that attend even the perceptions he takes to be clear and distinct, then the belief that God exists while intuiting the arguments for God’s existence will rest on grounds that, by the meditator’s own lights, are insufficient for metaphysical certainty in the absence of the proof of God’s existence. In other words, the Exemption Challenge will not have been tackled, and we will be back with the Circle.7 4. A New Approach? We need a new approach. In Williams’s hands, and in the hands of most other commentators on the Cartesian Circle, the challenge is to supplement the momentary certainties afforded by clear and distinct perceptions while we are actually intuiting them to bestow a similarly high status on our non-​occurrent clear and distinct perceptions. Now I do not dispute that the epistemic credentials of clear and distinct perception are what, at bottom, the meditator must rely on. But we must characterize them in such a way that both the Redundancy Challenge and the Exemption Challenge can be satisfied. This will require us to pay greater attention to the sources of irresistibility in the meditator’s occurrent clear and distinct perceptions. At the same time, this new approach will uphold the meditator’s entitlement not to question his basic epistemic powers, which Williams rightly prizes, while avoiding the pretence that the meditator is standing within touching distance of the finishing line at the very outset of his journey. I will provide a quick preview of the argument before citing textual evidence and progressing it in a more detailed way. My claim, to put it roughly, is that prior to the proof of God’s existence, the sceptical worry can be treated as the worry that the class of assent-​compelling perceptions is wider than the class of truth-​guaranteeing perceptions: genuine clear and distinct perceptions, which are both assent-​ compelling and truth-​guaranteeing, may be competing with fraudulent clear and distinct perceptions, which are merely assent-​compelling. The meditator’s challenge, then, is to recognize which of the perceptions he takes to be clear and 7 Stubbs (1980) also presses a form of this line of argument against Williams. [p. 156] distinct really are clear and distinct. I call this the ‘Recognition Interpretation’, and the problem to which it responds the ‘Recognition Problem’. As I see it, Descartes holds that clear and distinct perceptions are self-​ guaranteeing, before and after God’s existence has been proved. Before God’s existence is proved, however, we cannot tell which of our perceptions are clear and distinct. After God’s existence has been proved, we are assured that the perceptions which seem to be clear and distinct really are clear and distinct. Furthermore, and unusually, we do not need the reassurance that our clear and distinct perceptions, or what seem to be our clear and distinct perceptions, are truth-​guaranteeing in order to be able to prove that God exists. Even in outline, the Recognition Interpretation will surely raise many doubts. Even if it tackles the Redundancy Challenge, how does it deal with the Exemption Challenge? This particular question will be pursued in section 6. Before that, I will round up some evidence relating to the Textual Challenge, and progress the story further. 5. Hidden Faculties and the Recognition Problem It is worth drawing attention to a certain pattern of wording in the following passages, some of which we have already encountered: But what about when I was considering something very simple and straightforward in arithmetic or geometry, for example that two and three added together make five, and so on? Did I not see at least these things clearly enough to affirm their truth? Indeed, the only reason for my later judgment that they were open to doubt was that it occurred to me that perhaps some God could have given me a nature such that I was deceived even in matters which seemed most evident. (Third Meditation, CSM II 25; emphasis added) [W]‌henever my preconceived belief in the supreme power of God comes to mind, I cannot but admit that it would be easy for him, if he so desired, to bring it about that I go wrong even in those matters which I think I see utterly clearly with my mind’s eye. (Third Meditation, CSM II 25; emphasis added) Yet when I turn to the things themselves which I think I perceive very clearly, I am so convinced by them that I declare [them to be true]. (Third Meditation, CSM II emphasis added) For I can convince myself that I have a natural disposition to go wrong from time to time in matters which I think I perceive as evidently as can be. (Fifth Meditation, CSM II 48; emphasis added) [p. 157] Gerald L ang The second [general] reason for doubt was that since I did not know the author of my being (or at least was pretending not to), I saw nothing to rule out the possibility that my natural constitution made me prone to errors even in matters which seemed to me most true. (Sixth Meditation, CSM II 53; emphasis added) In each of these passages, the meditator distinguishes between what he seems to perceive clearly and distinctly and what he actually perceives clearly and distinctly. These passages suggest that Descartes has not allowed clear and distinct perceptions—​the real ones—​to be encompassed by the hyperbolic doubt. The problem instead is that, without a suspension of the hyperbolic doubt, the meditator is unable to distinguish clear and distinct perceptions from perceptions that merely masquerade as clear and distinct. The meditator does not yet know how to eliminate perceptual imposters whose phenomenal properties might coincide with clear and distinct perceptions.8 Now the Recognition Interpretation certainly accounts for why clear and distinct perceptions seem encompassed by the hyperbolic doubt. In this connection, the last sentence in the fourth paragraph of the Third Meditation is crucial: For if I do not know [that there is a non-​deceiving God], it seems that I can never be quite certain about anything else. (CSM II 25)9 The Recognition Interpretation can accommodate this sentence, provided that we paraphrase it in one of the two following ways—​either as: For if I do not know [that there is a non-​deceiving God], it seems that I can never be quite certain, with respect to any particular perception, whether it is, in fact, clear and distinct. Or as: For if I do not know [that there is a non-​deceiving God], it seems that I can never be quite certain, on any occasion on which I take myself to have experienced a clear and distinct perception, whether I have, in fact, experienced a clear and distinct perception. 8 Alan Nelson and Lex Newman (1999: 397n10) acknowledge the existence of this pattern of wording, but treat it as insignificant because clear and distinct perceptions will also seem clear and distinct. This provides no explanation of why Descartes would have consistently chosen to express himself in the way he does, and neither does it establish that the entailment runs in both directions. 9 James Van Cleve (1998: 112n30) declares this sentence to be ‘an embarrassment for almost any interpretation of Descartes’, so the Recognition Interpretation’s easy accommodation of it is a notable advantage. [p. 158] As the argument in the Third Meditation proceeds, Descartes draws a distinction between ‘spontaneous impulses’ (or ‘natural impulses’) and the deliverances of the ‘natural light’ (CSM II 26–​7). Deliverances of the natural light are simply clear and distinct perceptions. Spontaneous or natural impulses are portrayed as having a variety of sources. They may, for example, reflect the entrenchment of our childhood prejudices, or over-​confidence in the verdicts yielded by our sensory experience. There are more exotic possibilities too: Just as the [spontaneous] impulses . . . seem opposed to my will even though they are within me, so there may be some other faculty not yet fully known to me, which produces these ideas without any assistance from external things . . . (Third Meditation, CSM II 27) And despite the fact that the perceptions of the senses were not dependent on my will, I did not think that I should on that account infer that they proceeded from things distinct from myself, since I might perhaps have a faculty not yet known to me which produced them. (Sixth Meditation, CSM II 53–​4)10 The meditator’s problem, as it can be detected from these passages, is that his clear and distinct perceptions, and his spontaneous or natural impulses, may be conflated. They both compel assent while being experienced. They are indistinguishable with respect to their immediate phenomenal properties. But that, in fact, is just the problem. The meditator cannot yet distinguish the deliverances of the natural light from merely natural or spontaneous impulses, and neither can he do so until he knows that a non-​deceiving God exists. Clear and distinct perceptions—​the genuine ones—​are thus not merely ideas which seem clear and distinct, and which command the meditator’s immediate assent. They also have a further truth-​fixing function. The meditator simply cannot tell, out of all those ideas which might be clear and distinct, in virtue of appearing clear and distinct, and commanding his immediate assent, which of them really are truth-​guaranteeing. The overall picture, as it has been presented so far, might strike us as disappointingly complacent, given the evidence for Descartes’ refusal to question the reliability of the unimpeded operation of the natural light. But the charge of complacency is at least mitigated by the dangers he is prepared to register in respect of the successful operation of the natural light. There may be obstacles—​perhaps immoveable obstacles—​to its proper operation. This suspicion is what makes Descartes endorse what I will call Modified Epistemic Priority, rather than Epistemic Priority. Though Descartes is prepared 10 Note also, as a variation on these worries, the meditator’s reference to the non-​existence of ‘a faculty specially bestowed on me by God [for going wrong]’ in the Fourth Meditation, CSM II 38. (By this point in the Meditations, the meditator can dismiss the possibility of such a faculty.) [p. 159] Gerald L ang to make high claims on behalf of reason, or the natural light, he also feels that he cannot afford to assume from the outset that we are straightforwardly well-​oiled, cognitive machines, able to respond to genuine reasons for belief and to discount prejudices. For the meditator is deeply troubled by the following possibility: what if, given his inability to enumerate all the sources of his beliefs, or enumerate and inspect all his faculties, he is incapable of withdrawing assent to certain propositions from which assent should be withdrawn? Although the meditator’s aim is to interrogate ‘the basic principles on which all my former beliefs rested’ (First Meditation, CSM II 12), he also knows that he is not presently in a position to disclose a full list of those principles. They include, but may not be restricted to, sense perception and half-​forgotten folklore. Although the Cartesian meditator never seriously doubts that the natural light is reliable, he is aware that its operation may be derailed by the recalcitrance of our prejudices, or by the existence of faculties unknown to us, or because a malicious demon might have endowed us with a special faculty for making mistakes. Each of these various undiscounted possibilities voices the basic worry that the natural light, which can be assumed to be perfectly reliable in and by itself, may be obstructed, interfered with, or undermined by, the arrival of other ideas with an uncertain provenance. This is the essence of Modified Epistemic Priority, which concurs with Epistemic Priority’s high antecedent evaluation of the natural light, but takes seriously the doubt that the natural light may be in continuous competition with some other, less reliable faculty, in such a way that our judgements are systematically distorted.11 6. Tackling the Exemption Challenge As the meditator approaches the problem of proving God’s existence, the most pressing challenge is this. How is he supposed to be in a position to trust the perceptions which are mobilized in proving that God exists? He takes them to be clear and distinct. But perhaps they only seem to be clear and distinct. If Descartes is nonetheless prepared to rely on particular perceptions to prove God’s existence—​ as he surely must—​then presumably those perceptions will be, at the point of deployment in the cosmological argument, unratified. If, moreover, they are good enough to be used in the cosmological argument in the Third Meditation, then the meditator faces a familiar dilemma. Either the perceptions are not useable, which will provide no answer to the Exemption Challenge; or they did not need to be 11 See also Clark (2019) for a recent discussion that pays attention to the role of faculties of knowledge in Descartes’ project. My treatment is broadly complementary with his, though our emphases are different. [p. 160] upgraded in the first place, in which case there was no need to go to the trouble of proving that God exists, and the Redundancy Challenge will be unmet. This challenge is entirely just. For the Recognition Interpretation to succeed, we need to secure a special exemption for the perceptions used in the cosmological argument. We have to find a way of ensuring that they actually are clear and distinct. The problems here are exacerbated by the fact that the meditator’s confidence in these perceptions is frequently dismissed by Descartes’ commentators as amounting to little more than an opportunistic bit of forelock-​tugging to scholastic philosophy. This is where the road starts to get bumpy for the Recognition Interpretation. We have to explain why the ‘general causal principle’, or the ‘causal adequacy principle’, as it is sometimes called, is immune to the Recognition Problem. And there is no getting round the fact that Descartes is certain of the causal adequacy principle. It is, as he says, ‘manifest by the natural light’ (Third Meditation, CSM II 28). The same certainty extends to the application of this general principle to his ideas. (This principle says that the cause of any idea has as much formal reality as the idea has objective reality.) Has Descartes’ argument simply run aground here? I believe not. Start with the question: what is required for the Recognition Problem to present itself as a general reason for doubt? What explains its content? The meditator’s problem lies in not knowing where his ideas come from. They could be transmitted to him by a deceiving god, or by unknown faculties whose cognitive reliability is poor, or which interfere with the judgements of other faculties that are cognitively reliable. His basic problem is that he is unaware of the provenance of his ideas, and of the processes that produce them. Notice, however, that the causal adequacy principle, as applied to ideas, is conservative with respect to all these possibilities. It does not, by itself, resolve them. More importantly, the causal adequacy principle, as applied to ideas, helps to explain how any of these sceptical possibilities might actually get to be instantiated. The meditator’s idea of X can be caused by X itself, or by something else. It might be caused by the meditator himself, in an unexpected way—​by the operation, for example, of an unknown faculty; or it might be caused by a deceiving God; or it might be caused by a malicious demon. What matters is that no existential commitment to X itself is possible at this stage. The causal adequacy principle, as applied to ideas, depends on little more than the point that ideas need thinkers, or at least substances, to cause them. Thus, this principle is not condemned by the meditator’s worries about the Recognition Problem. Instead, it explains why the Recognition Problem is a problem. It supplies the essential metaphysical background against which any of these sceptical possibilities can be so much as entertained by the meditator. Of course, it may be complained at this point that this sort of reasoning does not justify the meditator in going beyond the bare claim that ideas have causes, as [p. 161] Gerald L ang opposed to helping himself to the much stronger claim that the cause of any idea has as much formal reality as the idea has objective reality. But, again, even this stronger claim is conservative with respect to the particular sceptical possibilities, involving deceiving Gods and so forth, that the meditator has adumbrated; and, as I have been suggesting, those particular possibilities are deeply revealing of the scepticism-​countering assumptions which Descartes is prepared to deploy against them. So even though Descartes’ grounds for certainty in the causal adequacy principle may strike us as being less than invincible, their attribution to him is consistent with the wider strategy of Modified Epistemic Priority that I have attributed to him, which at least renders his confidence in the causal adequacy principle intelligible. That is, Descartes’ very high level of confidence in the causal adequacy principle is explained by the same sorts of considerations that explain his very high level of confidence in the unimpeded operation of the natural light—​and those same considerations are, moreover, implicitly betrayed by the particular sceptical hypotheses sketched in the First Meditation. It should be conceded that there might be other sceptical possibilities that would subvert the causal adequacy principle. Crucially, however, Descartes does not entertain them. He is only genuinely interested in a certain range of sceptical possibilities, all of which are consistent with the truth of the causal adequacy principle. However, another weighty challenge now needs to be tackled. Even if it is granted that the truth of the causal adequacy principle is conservative with respect to the instantiation of the various particular sceptical hypotheses mentioned in the First Meditation, it does not follow that the causal adequacy principle cannot itself be attacked by those same sceptical hypotheses. That is to say, even if the actual instantiation of, for example, the malicious demon possibility would not be inconsistent with the truth of the causal adequacy principle—​since the event of the malicious demon’s seeing to it that the meditator believed in the causal adequacy principle would actually conform to the content of that principle—​it would surely undermine the meditator’s grounds for believing in it, since his belief would be the product of the malicious demon’s activity. In reply to this challenge, however, Descartes can say that he requires only the truth, not the secure grounding, of the causal adequacy principle. If the belief in the causal adequacy principle is the product, in the non-​sceptical world, of the natural light, there is no problem: both its truth and secure grounding are in place. If it is the product of the instantiation of one of the sceptical hypotheses sketched in the First Meditation, that particular grounding for the principle will be lost, but its truth will not be. For it will be true, in each of those sceptical worlds, that ideas are caused in such a way as to conform to the content of the principle. Therefore the warrant for believing in the principle would not be lost even if one of the sceptical scenarios were instantiated. It is actually difficult to think of a more secure grounding for the causal adequacy principle than that—​a grounding so secure that even a putatively faulty grounding counts simply as a further confirmation of it! [p. 162] Now it must be acknowledged that my explanation of the meditator’s entitlement to employ the causal adequacy principle in his reasoning differs, or appears to differ, from the meditator’s own estimation of it. The meditator merely appeals to the natural light, not to the more complicated argument that the availability of the causal adequacy principle, as a premiss in the cosmological argument, would be unthreatened in both the non-​sceptical and the sceptical worlds. Since one of my aims is to make sense of Descartes’ own views, the difference between my explanation of the security of the causal adequacy principle and the meditator’s explanation may seem problematic. In this respect, the Textual Challenge is only dubiously satisfied. The difference in explanations is far from disastrous, however. For it is still possible that what explains the peculiar security possessed by the causal adequacy principle—​a security that has traditionally been the source of puzzlement to Descartes’ commentators—​is precisely the sort of reasoning I have offered on Descartes’ behalf. This reasoning might then explain why the meditator is so confident in this particular affirmation of the natural light. On this reading, what the natural light will affirm is that the truth of the causal adequacy principle does not require a secure grounding in order to be counted as true. 7. Is Assent-​Compellingness Part of the Solution? On the Recognition Interpretation, assent-​compellingness is not a distinctive property of all and only clear and distinct perceptions, for the meditator is not in a position to dismiss the possibility of assent-​compelling imposters. Assent-​ compellingness is, in fact, part of the problem, not part of the solution, contrary to the thrust of many other solutions to the Circle. But it is tempting to think that it cannot be a promising sign of any solution to the Circle that it fails to make the assent-​compellingness of clear and distinct perceptions part of the solution. My view is that the assent-​compellingness of clear and distinct perceptions is not a redundant part of Descartes’ system, because it provides the solution to a separate problem in that system.12 Assent-​compellingness comes into its own—​it becomes epistemically productive—​when we move past the Circle, to the concerns of the Fourth Meditation, where the meditator is attempting to establish appropriate boundaries between cognitive overambition and divine deception. In this part of the Meditations, the meditator’s principal problem is to decide what would count as an act of deception by God. Having limited cognitive faculties does not, in itself, reflect badly on God—​human beings’ visual inability to detect small objects at a distance of half a mile, for example, does not make God 12 Clearness and distinctness of perception also, for Descartes, determines the objective reality of mental content; it is a guide to real possible existence (Third Meditation, CSM II 30). [p. 163] Gerald L ang a deceiver. But this consideration, when properly digested, would surely incline the conscientious meditator to suspend judgement, rather than commit himself to the truth of any judgement. This is where assent-​compellingness can help. Assent-​ compellingness provides the meditator with a way of eschewing this putative Pyrrhonian policy of suspension of belief. For if we perceive something clearly and distinctly, we cannot help forming a belief about it. We are compelled to have this belief. It then follows that God would be a deceiver if he did not produce a world to match the representations of it—​the geometrical bits of it, roughly—​revealed through the beliefs sustained by our clear and distinct perceptions. In summary, assent-​compellingness gives us a means for establishing the proper boundary between making overambitious, presumptuous, or reckless cognitive verdicts, whose falsity would in no sense reflect badly on God, and making verdicts, which, if false, would properly entitle us to conclude that God is a deceiver. To put it another way, it gives us a method of contracting God into producing a world that is truthfully represented by some of our perceptions of it, by ensuring that those perceptions are caused by corporeal substance, just as they seem to be.13 8. Tackling the Redundancy Challenge How does the God’s existence give us a method for distinguishing between clear and distinct perceptions, and their false surrogates? The proof of God’s existence helps, but indirectly. The meditator now knows that his normal cognitive faculties have no shadowy, error-​producing competitors. The materials for this sort of deception are simply not in place. We are not being bamboozled by a malicious demon. We do not have hidden faculties whose deliverances obstruct, or interfere with, the deliverances of the natural light. So even though the proof of a non-​deceiving God does not deliver a recipe or a method for distinguishing between real clear and distinct perceptions and perceptions that are merely imposters, that does not matter. Because the upshot of the proof of God is that he is no deceiver, and that we have no hidden faculties that produce error or obstruct the operation of the natural light, the systematic possibility of these perceptual imposters has now melted away, leaving clear and distinct perceptions without any relevant competition. Dangers may still exist, of course, with respect to the application of the Truth Rule. Descartes should be happy to admit that inattentiveness and so forth may lead subjects to repose trust in perceptions which are less than trustworthy. It then follows that, on some occasions, when we are careless or inattentive, or when we have been manipulated by others, it is possible that we will continue to take certain 13 This the conclusion drawn by the meditator in the Sixth Meditation, CSM II 54–​5. [p. 164] perceptions as clear and distinct, though they are not clear and distinct, and are, in fact, false. But none of this impairs the Recognition Interpretation. These mistakes are, in principle, detectable, and they are therefore both explicable and remediable for investigators who are sufficiently careful. What exercised Descartes was the possibility of a systematic mismatch between our most reliable perceptions, formed in apparently optimal circumstances, and the truth, due to the systematic possibility of ideas with an undetected truth-​subverting provenance. 9. Mersenne’s Challenge Before I complete my case for the Recognition Interpretation, I wish to examine a couple of further, notoriously problematic, passages from the Second Replies. Descartes is replying here to Mersenne’s very stark presentation of the problem of the Circle (Second Objections, CSM II 89). The dilemma for Descartes, as Mersenne sees it, is that either clear and distinct perceptions are all doubtful, including even the cogito-​perceptions, in which case the meditator will have nothing to go on to prove that God exists (or anything else for that matter, including his own existence); or they are in perfectly good shape already, in which case there is no need to prove that God exists, and from which it would also follow that an ‘atheist geometer’, no less than theists, can be in the possession of knowledge. This is part of Descartes’ reply to Mersenne’s challenge, which deals with the second horn of the dilemma, concerning the atheist geometer: The fact that an atheist can be ‘clearly aware that the three angles of a triangle are equal to two right angles’ is something I do not dispute. But I maintain that this awareness of his is not true knowledge, since no act of awareness that can be rendered doubtful seems fit to be called knowledge. Now since we are supposing that this individual is an atheist, he cannot be certain that he is not being deceived on matters which seem to him to be very evident (as I fully explained). (Second Replies, CSM II 101) If the Recognition Interpretation is correct, it is difficult to see how the atheist’s clear and distinct awareness of the triangle’s geometrical properties could fail to amount to knowledge. Yet his awareness does not amount to ‘true knowledge’, because it ‘can be rendered doubtful’. What is going on? Are clear and distinct perceptions exempt from doubt, or not? The Recognition Interpretation can explain Descartes’ reservations about the atheist geometer. To do so, we need to interpret the phrase ‘no act of awareness that can be rendered doubtful seems fit to be called knowledge’ as being equivalent in meaning to ‘no occasion on which there is an act of awareness whose identity as a genuine clear and distinct perception can be rendered doubtful seems fit [p. 165] Gerald L ang to be called knowledge’. This reading exempts clear and distinct perceptions from doubt, but not non-​veridical, easily confused imposters. Since the atheist himself is not in a position to dismiss the possibility that he experiences such imposters, it follows that he does not possess true knowledge. But this source of doubt is consistent, in turn, with the perfect reliability of the natural light itself, whether the natural light is hosted by an atheist or a theist. This is Descartes’ response to the first horn of the dilemma: When I said that we can know nothing for certain until we are aware that God exists, I expressly declared that I was speaking only of knowledge of those conclusions which can be recalled when we are no longer attending to the arguments by means of which we deduced them. (Second Replies, CSM II 100)14 The worry voiced by this Second Replies passage seems to lead us away from the Recognition Interpretation: it concerns the difference between occurrent and non-​occurrent clear and distinct perceptions. It is not so difficult, however, for the Recognition Interpretation to accommodate the significance of this particular distinction. First, clear and distinct perceptions are beyond doubt: they are metaphysically certain, whether those perceptions are occurrent or non-​occurrent. Second, given their assent-​compellingness, doubts over clear and distinct perceptions, or the perceptions we take to be clear and distinct, can arise only when they are non-​occurrent. The reason for the meditator’s doubts when his would-​be clear and distinct perceptions are non-​occurrent is because he can then (and only then) entertain the possibility that they may not be genuinely clear and distinct, given undiscounted possibilities of divine deception or the existence of hidden unreliable faculties. So it is non-​occurrent clear and distinct perceptions, after all, construed as the non-​occurrent perceptions we take to be clear and distinct, that are the problem. There is no tension between the Recognition Interpretation and Descartes’ investment in this distinction, just as long as we understand the references to non-​occurrent clear and distinct perceptions as references to the non-​ occurrent perceptions we take to be clear and distinct. A World without Peculiarities I have argued that Williams’s treatment of the Circle is exposed to the dilemma that I set out in section 1. Either every clear and distinct perception is already more broadly usable, in which case the Redundancy Challenge is unsatisfied, or his solution cannot tell a convincing story about the status of the particular clear and distinct perceptions used to prove God’s existence, in which case the Exemption Challenge goes unanswered. 14 For similar considerations, see the Fifth Meditation, CSM II 48–​9. [p. 166] The Recognition Interpretation contends instead that God’s existence can be facilitated within the Cartesian system by safeguarding the presuppositions of the particular sceptical waves of doubt which Descartes enrols into his argument. These include, crucially, the causal adequacy principle, which the meditator requires if he is to uphold his certainty in God. Unlike Williams’s interpretation, the Recognition Interpretation returns satisfying answers to both the Exemption Challenge and the Redundancy Challenge, while still making a decent fist of the Textual Challenge. The Recognition Interpretation thus leans heavily into the sceptical apparatus that Descartes deploys—​more heavily, in fact, than Williams himself—​while accommodating Williams’s central insight that scepticism is often the creature of a more productive epistemic project, in which the aim is to safeguard or prioritize certain forms of knowledge over others. Now a willingness to battle against scepticism is often taken to be a healthy and courageous opening move in a conscientious philosophical audit of our sources of knowledge. Descartes is the trail-​ blazing hero of this sort of story. We should be warier of it than this quasi-​heroic picture of it suggests, given the healthy possibility that the scepticism-​defeating investigator has her thumb on the scales. There are different directions in which this suspicion can be taken. I will focus, in conclusion, and in an admittedly gestural way, on just one of them, in which there are well-​known shared commitments between Descartes and Williams.15 For Williams, the ‘absolute conception of reality’16 can be understood as a conception of the world as it really is, or ‘a conception of the world as it is independently of all observers’ (PPE: 241). In greater detail: The idea of the world as it really is involves at least a contrast with that of the world as it seems to us: where that contrast implies, not that our conception of the world is totally unrelated to reality, but that it has features which are peculiar to us . . . the world as it really is is contrasted with the world as it peculiarly seems to any observer—​that is to say, as it seems to any observer in virtue of that observer’s peculiarities. (PPE: 241; original emphasis) A corollary of the absolute conception of reality is that the world ‘as it seems to us’ will include our secondary quality ascriptions, regarding colours, sounds, smells, and tastes, as well as our ascriptions about value. The world as it is independently of the observer’s peculiarities, by contrast, will include primary quality ascriptions, of shape and size and motion. This is, roughly speaking, the world described by physics and the hard sciences. One further expectation of the absolute conception of reality is that it should 15 PPE: ch. 10 is particularly relevant, together with Williams (1985: ch. 8). See also Cottingham (this volume, Chapter 6) for a fuller discussion of these issues. 16 It is sometimes described as the ‘view from nowhere’: see Nagel (1986). [p. 167] Gerald L ang make it possible to explain how the more local representations of the world can come about—​it is this that would enable us to relate them to each other, and to the world as it is independently of them. (PPE: 245–​6) In the Cartesian system, the being with a flawless, comprehensive, and unobstructed access to reality is, of course, God. God is the measure of reality, and our differences from God are sources of divergence from that reality. The Cartesian meditator’s scepticism-​defeating exercise revolved around the need to prove God, and the reconstruction exercise is again tailored to God. This is a conception of reality inherited, in its secular version, by Williams. In this sense, there is a theological residue of Descartes’ overtly theological picture of reality in Williams’s thought. The claim that reality in itself is that portion of reality that could be detected by investigators without local ‘peculiarities’ is the product, in part, of an acceptance of a picture of reality that can be disclosed to an observer without such peculiarities. What interest is served by such a picture? For Descartes, this was God’s world. Its interest, for his immediate readers, was self-​evident. Williams needs other answers, since for him there is no observer who, without any peculiarities, is observing the world as it is in itself. Moreover, ‘peculiar’ is a loaded term. Either ‘peculiar’ is defined pejoratively at the outset, in such a way that the deployment of modalities that are peculiar is just another way of saying that, in deploying them, we do not learn about the world as it is in itself; or else ‘peculiar’ is characterized, more neutrally, simply as a way of saying that these are modalities possessed only by some but not all creatures who navigate through the world. Perhaps we might think that it is a discourtesy to those investigators who lack certain modalities or powers that they should be denied access to reality as it is in itself. But why should that be? It would be odd to invoke something like a charge of epistemic injustice as a decisive consideration at a level as rarified as this. Alternatively, perhaps there is no principled way of favouring the verdicts of some modalities over others, if the world appears differently to differently endowed investigators, and we are after a univocal characterization of the world. But why shouldn’t there be provision for some harmless relativity in at least some of these matters? However these difficult issues are dealt with, we should not stop thinking about why we have a fundamental interest in the view from nowhere and why, and how, other views of reality supposedly fall short.17 17 An earlier version of this chapter was drafted several years ago and reshaped for this collection. For written comments on the earlier draft, I am grateful to Gonzalo Rodriguez-​Pereyra, and for useful discussions along the way, I thank Corine Besson, Alix Cohen, John Cottingham, James Harris, Christopher Hookway, Oliver Pooley, and Nico Silins. I am also grateful to audiences in Oxford and Leeds, and at the White Rose Centre for the History of Philosophy. Finally, I thank the editors of this volume, Marcel van Ackeren and Matthieu Queloz, for their helpful comments on a penultimate draft. [p. 168] References Bennett, Jonathan. 1990. ‘Truth and Stability in Descartes’ Meditations’. Canadian Journal of Philosophy, Supplementary Volume 20: 1–​108. Clark, Matthew. 2019. ‘Faculties, Knowledge, and Reasons for Doubt in the Cartesian Circle’. Della Rocca, Michael. 2005. ‘Descartes, the Cartesian Circle, and Epistemology without God’. Philosophy and Phenomenological Research 70.1: 1–​33. Descartes, René. 1985. The Philosophical Writings of Descartes. Translated by John Cottingham, Robert Stoothoff, and Dugald Murdoch, vols I–​III. Cambridge: Cambridge University Press. Feldman, Fred. 1975. ‘Epistemic Appraisal and the Cartesian Circle’. Philosophical Studies Frankfurt, Harry. 1970. Demons, Dreamers, and Madmen: The Defence of Reason in Descartes’ Meditations. Indianapolis: Bobbs-​Merrill. Kenny, Anthony. 1993. Descartes: A Study of his Philosophy. Bristol: Thoemmes Press. (Originally published in 1968.) Loeb, Louis. 1992. ‘The Cartesian Circle’. In The Cambridge Companion to Descartes. Edited by John Cottingham, 200–​35. Cambridge: Cambridge University Press. Loeb, Louis. 1990. ‘The Priority of Reason in Descartes’. Philosophical Review 99.1: 1–​43. Nagel, Thomas. 1986. The View from Nowhere. Oxford: Oxford University Press. Nelson, Alan, and Lex Newman. 1999. ‘Circumventing Cartesian Circles’. Noûs 33.3: 1–​404. Stubbs, A. C. 1980. ‘Bernard Williams and the Cartesian Circle’. Analysis 40.2: 1–​8. Van Cleve, James. 1998. ‘Foundationalism, Epistemic Principles, and the Cartesian Circle’. In Descartes. Edited by John Cottingham, 1–​31. Oxford: Oxford University Press. Williams, Bernard. 1985. Ethics and the Limits of Philosophy. London: Fontana Press/​Collins. Williams, Bernard. 1990. Descartes: The Project of Pure Enquiry. Harmondsworth: Penguin. (Originally published in 1978 and abbreviated as ‘PPE’ in the main text.) Williams, Bernard. 2006. ‘Descartes’ Use of Scepticism’. In The Sense of the Past: Essays in the History of Philosophy. Edited by Myles Burnyeat, 1–​45. Princeton: Princeton University Press. [p. 169] A Humean Williams and a Williamsian Hume Lorenzo Greco 1. Introduction In his review of Ethics and the Limits of Philosophy, Simon Blackburn (1986) notes that the great absentee in the book is David Hume. In his reply, Bernard Williams accounts for this absence by commenting that Hume’s main flaw is his overconfidence that our philosophical analysis will allow us to come up with an unambiguous description of human nature, thus lacking awareness of the integral part played in that description by history, psychology, and the social sciences. Some aspects of Hume’s thought, Williams continues, are certainly to be welcomed, particularly those that connect Hume with classical thinkers such as Aristotle—​in this regard, Williams mentions Hume’s treatment of free will and his resistance to making a clear distinction between virtue and other admirable traits of human beings. However, Hume would not understand that the variety of human practices inevitably invalidates the ambition to explain human nature in general and uniform terms: ‘my problem with Hume’, Williams concludes, ‘is not . . . that he fails to be an ancient thinker, but that he is not a modern one’ (1986: 206). In spite of Williams’s refusal to ‘join the Church of Hume’ (Blackburn as Blackburn aptly puts it in a recent essay in which he again addresses Ethics and the Limits of Philosophy—​there are still some who perceive the presence of the great Scot in Williams’s writings, and continue to wonder how much Williams was influenced by him (in addition to Blackburn 1986, 2019, see Greco Russell 2021; Sagar 2013; Taylor 2015: ch. 5). On the other hand, it is Williams himself who seems to encourage his own readers to pigeonhole him within a particular tradition of thought. From time to time, he has shown himself to be close to the theses of Aristotle (in Ethics and the Limits of Philosophy, at least in part), and indeed Hume (especially in some of Williams’s early works). At least until Ethics and the Limits of Philosophy, in fact, Williams’s attitude towards Hume was more accommodating. But it did not last very long. As Williams himself acknowledged, ‘I once had a great admiration for Hume. Now I think that he suffered Lorenzo Greco, A Humean Williams and a Williamsian Hume In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. [p. 170] from a somewhat terminal degree of optimism’ (1999: 256). The late Williams showed an increasing fondness for Nietzsche, especially with regard to his psychological theory (Williams 1995b) and to his genealogical method (Williams 2002; see Queloz 2021: ch. 7). According to Paul Sagar (2013), it is this Nietzschean influence which accounts for Williams’s general moral perspective, especially at the expense of his early preference for Hume. Yet, as Martha Nussbaum observes, this might not actually be the case: He [Williams] relied on a sharp distinction between the methods and aims of science and those of ethics, which he saw as a human discipline with its own characteristic modes of inquiry, which ought to include a close attention to the complex realities of life. Thus, though something of a rebel in British academic society, he was also quintessentially British. If Nietzsche became a passion, he remained closer to Hume. (Nussbaum 2003) Nussbaum’s observation is insightful, especially the way she juxtaposes Hume with Williams regarding the focus on ethics as a ‘human discipline’ that must pay ‘close attention to the complex realities of life’. In what follows, I shall argue that Hume looms large in Williams’s philosophy, far beyond the occasions when the presence of Hume is most visible. Hume and Williams are closer than Williams himself wished to admit or publicly declare; moreover, Hume turns out not to be the kind of optimist that Williams thought he was, while Williams himself proves to be more of an optimist than he imagined himself to be. 2. The Amoralist’s Challenge Williams explicitly devotes only one short essay to Hume: ‘Hume on Religion’. Notwithstanding the title, Hume’s treatment of religion appears to be just a pretext for Williams to tackle other aspects of Hume’s thought. First, Williams discusses Hume’s use of irony in his critique of Christianity—​an irony that would also largely characterize Williamsian philosophical method. If one then looks at how Williams proceeds in addressing the religious phenomenon, it becomes clear that he is primarily concerned, indeed like Hume, with the working of the human mind—​what are the psychological dynamics underlying religious belief, and, above all, what are the practical consequences of it in terms of religious superstition and fanaticism? This reference to the mechanics of human nature allows Williams to shift his focus to Humean moral theory and to distance himself from Hume. An account of ethics such as Hume’s, that appeals to the natural sympathy and benevolence of human beings, would not explain the hatred and persecution that religion can produce, and this precisely because of Hume’s overly simplistic idea of human nature: [p. 171] Lorenzo Greco Now this is obviously a very limited and inadequate account of the effects of religious belief—​just as, we may add, the story about men’s fears of the unknown is an inadequate account of its origins. In both cases, the limitations lie in the general body of Hume’s philosophy: in the one case, in his moral psychology, in the other, in a limited empiricist theory about the origins of belief. Hume, like Bertrand Russell in our own time, is too amiable and optimistic a man really to understand religion. (Williams 2006: 270–​1) This is a rather strong claim, since it is made against one of the philosophers who have largely determined the way we understand some of the central problems of religious belief today, such as the argument from design or the nature of miracles. Be that as it may, let us stick to the two points that Williams highlights: Hume’s great limitations would be evident in his moral psychology as well as in his overly empirical explanation of the origin of belief. As early as the 1960s, therefore (‘Hume on Religion’ was written in 1963), Williams considered Hume’s attitude to be too optimistic. However, if we examine Williams’s production up to Ethics and the Limits of Philosophy, his ‘great admiration for Hume’ clearly emerges as well. In Problems of the Self, which collects Williams’s essays from 1956 to 1972 (but ‘Hume on Religion’ is not included), Hume is mentioned in ‘Personal Identity and Individuation’, in relation to the role of memory for personal identity, and in ‘Deciding to Believe’, again in relation to belief. Finally, Hume is present in the book’s concluding essay, ‘Egoism and Altruism’, where Williams contrasts the Humean approach to the Kantian one in relation to the motives that allow us to move from the personal level to that of morality: Kantians make the distinction between motivations which are emotional, grounded in desire, and particular, as against motivations which are universal, rational and of principle. The step to morality involves a discontinuous step to the latter. For Hume, there are desires of different degrees of generality, involving objects with different degrees of remoteness and independence from the subject: the slide towards morality is a slide along these continua. Connected with this is the point from which we started, that for Kant impersonal and intrapersonal altruism is grounded in the structure of practical reason; for Hume, it is grounded in a special sort of desire or sentiment. (1973: 260) The reference to Hume in opposition to Kant represents a precise philosophical move that reveals one of Williams’s standards to account for the phenomenon of morality: the idea that ethics does not start from a hypothetical practical rationality and is not to be formulated in terms of universal principles but arises from the personal motivations, of an exquisitely passionate and non-​rational nature, of individuals. ‘Egoism and Altruism’ concludes with a statement that sounds like a manifesto: [p. 172] [B]‌oth in moral theory and also in moral psychology, it is not the Kantian leap from the particular and the affective to the rational and universal that makes all the difference; it is rather the Humean step—​that is to say, the first Humean step—​from the self to someone else. (1973: 265) This strategy is the same one that Williams adopted in 1972’s Morality to checkmate the amoralist. The one who refuses to recognize or give importance to moral distinctions, Williams tells us—​unless he is sick or depressed, in which case he needs to be medically treated—​is ‘a parasite on the moral system’ (1972: 5), someone who depends on participation in the morality game for the satisfaction of his needs, and who possesses the capacity to grasp the value of morality and to behave accordingly, even though in the end he may not be persuaded to act morally. Note the starting question that, for Williams, represents the amoralist’s challenge: ‘Why is there anything that I should, ought to, do?’ (1972: 3). The amoralist asks why he should abide by the precepts of morality, and more generally why he should abide by any precept whatsoever. The question is, as always for Williams, posed in the first person; it is necessary to present reasons that are valid for the person who must respect them, reasons that that person can recognize as reasons for them and thus motivate them to action—​in this case, therefore, reasons must be presented that are capable of convincing the amoralist. Now, Williams points out, the fact that reasons must be put forward does not mean that the person has to be persuaded through reasoning; in other words, the need to put forward a reason does not mean that rationality necessarily has to be called in. Rationality, for Williams, has nothing to do with someone like the amoralist: ‘I do not see how it could be regarded as a defeat for reason or rationality that it had no power against this man’s state; his state is rather a defeat for humanity’ (1972: 4). The observation that we would be faced with ‘a defeat for humanity’ is reminiscent of what Hume states, in Section 9 of An Enquiry concerning the Principles of Morals, about the ‘sensible knave’: that is, he who in particular incidents, may think, that an act of iniquity or infidelity will make a considerable addition to his fortune, without causing any considerable breach in the social union and confederacy. That honesty is the best policy, may be a good general rule; but is liable to many exceptions: And he, it may, perhaps, be thought, conducts himself with most wisdom, who observes the general rule, and takes advantage of all the exceptions. (Hume 1998: 9.22)1 As with Williams, so with Hume it is not a matter of convincing the sensible knave by appealing to rationality; if the sensible knave does not realize that his conduct 1 I shall quote both Hume’s An Enquiry concerning Human Understanding and An Enquiry concerning the Principles of Morals by referring to the section and the paragraph. [p. 173] Lorenzo Greco is wrong—​that is to say, if he lacks the sensibility that puts him in touch with the common humanity of those with whom the sensible knave interacts—​then there is little that can be done with him: I must confess, that, if a man think, that this reasoning much requires an answer, it will be a little difficult to find any, which will to him appear satisfactory and convincing. If his heart rebel not against such pernicious maxims, if he feel no reluctance to the thoughts of villany or baseness, he has indeed lost a considerable motive to virtue; and we may expect, that his practice will be answerable to his speculation. But in all ingenuous natures, the antipathy to treachery and roguery is too strong to be counterbalanced by any views of profit or pecuniary advantage. Inward peace of mind, consciousness of integrity, a satisfactory review of our own conduct; these are circumstances very requisite to happiness, and will be cherished and cultivated by every honest man, who feels the importance of them. We are facing a defeat for mankind in this case, too—​surely a defeat for all those who decide to adopt a knavish conduct, who will soon realize that they themselves are, in the end, the greatest dupes, and have sacrificed the invaluable enjoyment of a character, with themselves at least, for the acquisition of worthless toys and gewgaws. How little is requisite to supply the necessities of nature? And in a view to pleasure, what comparison between the unbought satisfaction of conversation, society, study, even health and the common beauties of nature, but above all the peaceful reflection on one’s own conduct: What comparison, I say, between these, and the feverish, empty amusements of luxury and expence? These natural pleasures, indeed, are really without price; both because they are below all price in their attainment, and above it in their enjoyment. The considerations of the two philosophers against someone who is defiant to morality are of the same tenor. Back to Williams, there are several approaches that allow us to show the amoralist how he is already within morality, despite everything. Not only does he depend on the system of morality to satisfy his needs, but above all it is a fact that he has needs that involve other people, however minimal one might imagine them to be. That is, if the amoralist is not a psychopath, ‘one might picture him as having some affections, occasionally caring for what happens to somebody else’ (Williams 1972: 10). If this is true, Williams continues, then it will be possible to stimulate his imagination and lead him also to consider the needs of others. As in ‘Egoism and Altruism’, here as well Williams identifies a path in which there is no break between the interests of the individual and the dimension of morality: ‘To get him [the amoralist] to consider their situation seems rather an extension of [p. 174] his imagination and his understanding, than a discontinuous step onto something quite different, the “moral plane”.’ And, Williams concludes, ‘if we could get him to consider their situation, in the sense of thinking about it and imagining it, he might conceivably start to show some consideration for it: we extend his sympathies’ (1972: 11). The appeal to the extension of sympathy is a clear reference to Hume, who is explicitly mentioned on the next page by Williams: in Morality, too, we are presented with the ‘first Humean step’. In fact, if we read what Hume says in Book 2 of A Treatise of Human Nature, the argument is the same; it is worth quoting it in full: But however we may look forward to the future in sympathizing with any person, the extending of our sympathy depends in a great measure upon our sense of his present condition. ’Tis a great effort of imagination, to form such lively ideas even of the present sentiments of others as to feel these very sentiments; but ’tis impossible we cou’d extend this sympathy to the future, without being aided by some circumstance in the present, which strikes upon us in a lively manner. When the present misery of another has any strong influence upon me, the vivacity of the conception is not confin’d merely to its immediate object, but diffuses its influence over all the related ideas, and gives me a lively notion of all the circumstances of that person, whether past, present, or future; possible, probable or certain. By means of this lively notion I am interested in them; take part with them; and feel a sympathetic motion in my breast, conformable to whatever I imagine in his. If I diminish the vivacity of the first conception, I diminish that of the related ideas; as pipes can convey no more water than what arises at the fountain. By this diminution I destroy the future prospect, which is necessary to interest me perfectly in the fortune of another. I may feel the present impression, but carry my sympathy no farther, and never transfuse the force of the first conception into my ideas of the related objects. If it be another’s misery, which is presented in this feeble manner, I receive it by communication, and am affected with all the passions related to it: But as I am not so much interested as to concern myself in his good fortune, as well as his bad, I never feel the extensive sympathy, nor the passions related to it. (Hume 2007: 2.2.9.14)2 Williams was thus well aware of Hume’s approach to the problem of access to the moral dimension and follows in his footsteps. Wanting to be more precise, Williams follows Hume as he expresses himself in the Treatise (the reference to ‘extensive sympathy’ is replaced in the second Enquiry, 9.5, with a more generic ‘sentiment of humanity’), thus showing that he is familiar with Hume’s philosophy in detail. For the Hume of the Treatise, too, one can hope to push individuals to 2 I shall quote Hume’s Treatise by referring to the book, the part, the section, and the paragraph. [p. 175] Lorenzo Greco take an interest in the fate of others, and thus to develop reasons that can be said to be moral, by moving from the individual’s present sentimental condition and their present desires, working on their sympathetic capacity, and broadening the perimeter of what the individual is able to imagine in order to consider the suffering of others. This is an idea that Williams will develop further in another essay: ‘Internal and External Reasons’. 3. A Reason for Someone In ‘Internal and External Reasons’, Williams advocates the thesis that, in order to identify a reason to act for someone, one must start from elements already present in the ‘agent’s subjective motivational set’ (1981a: 102).3 All reasons for acting turn out to be internal: that is, in Williams’s words, the only statements that allow us to identify a reason for acting are those of the form ‘ “A has a reason to f” . . . (where “f” stands for a verb of action)’ (1981a: 101). There are, therefore, no ‘external’ reasons of the form ‘ “There is a reason for A to s” ’ (1981a: 101): that is, there is not any reason for acting that can be said to be such for someone unless one can connect that reason through a ‘sound deliberative route’—​as Williams specifies in ‘Internal Reasons and the Obscurity of Blame’, a later restatement of his internalist proposal (1995a: 35)—​to personal motivations that someone already possesses.4 The point Williams makes is the same one we have already seen at work in ‘Egoism and Altruism’ and in Morality: reason or rationality alone proves to be ineffective when it comes to moving individuals to action since it is of no help in getting these individuals to recognize that they have a reason to act. The accusation of irrationality has no hold; it is rather, once again, a matter of working on their imagination, starting from sentimental elements that are already within them. Hume’s influence is also evident here: just consider what he says about the fact that ‘morality is . . . supposed to influence our passions and actions, and to go beyond the calm and indolent judgments of the understanding’ (Hume 2007: 3.1.1.5). If this is true, ‘reason is perfectly inert, and can never either prevent or produce any action or affection’ (Hume 2007: 3.1.1.8). It is no coincidence that Williams calls his model ‘the sub-​Humean model’ (1981a: 102), where the prefix ‘sub’ is only due to the fact, Williams tells us, that his scheme would be simpler than Hume’s perspective. In ‘Internal Reasons and the Obscurity of Blame’, Williams confirms that eventually we have to remain faithful to its ‘Humean origins’ (1995a: 36) against an interpretation either in purely prudential or in moral terms of the sound deliberative route. 3 In Moral Luck, where ‘Internal and External Reasons’ is collected, Hume is also quoted in ‘Wittgenstein and Idealism’ about the impossibility of recognizing a ‘knowing self ’ as an object in the world (Williams 1981b: 145). 4 Williams (2001) returns a third time to the question of the internal nature of the reasons for acting. [p. 176] And in ‘Formal Structures and Social Reality’, Williams reaffirms his Humeanism regarding the nature of motivation, advancing what he calls ‘Hume’s axiom’, according to which ‘[o]‌nly motivations motivate’. He then points out that [t]‌his might be thought to be a tautology, and that is, just about, what I take it to be. Its content, relative to the question under discussion, is that if someone acts to further some collectively desirable outcome, then his doing so must be explained by some disposition or desire that he has (for instance, to do just that). His so acting cannot be explained merely by the fact that it furthers that outcome, or merely by his knowing that it does. It follows from this that if one wants people to pursue such outcomes, one will have to see that appropriate motivations do actually exist to produce that result. (Williams 1995e: 117) That means that at the turn of the 1980s and the 1990s (‘Formal Structures and Social Reality’ was initially written in 1988 and ‘Internal Reasons and the Obscurity of Blame’ in 1989, and then both were collected in Making Sense of Humanity in Williams still believed that Hume had something valid to say. In the publications we have been focusing on, Williams refers to Hume to attack, from an anti-​ Kantian perspective, the idea that deliberation can be the result of the appropriate use of practical rationality. Williams is criticizing the idea that there is a rational way of structuring morality and is defending the thesis that one must always start from the desires of individuals. His endorsement of individualism turns out to be programmatic. It states that a correct understanding of ethics ‘does require . . . that there be individuals with dispositions of character and a life of their own to lead. . . . In one sense, the primacy of the individual and of personal dispositions is a necessary truth’ (Williams 1985: 201). From a methodological point of view, the presupposition of this ‘substantial’ (1985: 201) individualism is a ‘formal individualism’ according to which there are ultimately no actions that are not the actions of individual agents. One may add to this the truth that the actions of an individual are explained in the first place by the psychology of that individual: this means, for instance, that to the extent those actions are intentional, they are explained in the first place by the individual’s intentions. (Williams 1995c: 85–​6) Specifically, formal individualism consists of the two truths that ‘deliberative or practical questions are radically first-​personal, where this means that they are individually first-​personal’, and ‘what an individual does is often explained by the individual’s deliberation, and, to the extent that his or her action is intentional, it can be explained in terms of a deliberation that the individual could have conducted’ (Williams 1995f: 125). Taking these two truths together, it follows that ‘intentional action can always be explained by reference to a consciousness which the [p. 177] Lorenzo Greco agent at least could have had and in many cases did have, and which refers to the agent’ (1995f: 125). Some might observe—​Blackburn does so in his review of Ethics and the Limits of Philosophy—​that this focus by Williams on the deliberative aspect of the individual would reveal an Aristotelian rather than Humean spirit, which is based ‘on the importance of finding a fundamentally egoistic justification for ethics, and the anxieties which we face if we cannot’ (Blackburn 1986: 194). On the contrary, Blackburn continues, reference to Hume would have paved the way for a ‘non-​egoistic, social, theory of morality’ (1986: 194). Blackburn’s concern with Williams on this point, however, is perhaps excessive; that Williams was after ‘a fundamentally egoistic justification for ethics’ is, in fact, debatable. Discussing his alleged Aristotelian stance, and the consequent egoistic answer to the question ‘how should I live?’, Williams comments that ‘[i]‌f the search is for a practical justification of the ethical life, from the ground up, the answer will necessarily give reasons to the agent who asked them’ (1986: 205). This reference to the necessity of providing reasons to the agent does not necessarily imply egoism but only that the agent recognizes that those reasons—​whether selfish or altruistic—​ are indeed reasons for them. On the other hand, it may not be Aristotle who stands behind Williams’s reasoning. Indeed, even if Hume does not state it so explicitly, his perspective can be understood in a light that brings it very close to the individualism Williams advocates. In his reply to Blackburn, Williams continues as follows: That is why an account of the ethical life which is an explanation and only an explanation—​one that represents morality as a socially evolved answer to a coordination problem, for instance—​cannot, whatever its other merits or interest, answer the question; for the agent can always ask ‘And why does that give me a reason?’ (1986: 205) Now, in Book 2 of the Treatise there is a famous passage (or an infamous one, depending on how one sees it) in which Hume makes the following statements: ’Tis not contrary to reason to prefer the destruction of the whole world to the scratching of my finger. ’Tis not contrary to reason for me to chuse my total ruin, to prevent the least uneasiness of an Indian or person wholly unknown to me. ’Tis as little contrary to reason to prefer even my own acknowledg’d lesser good to my greater, and have a more ardent affection for the former than the latter. A trivial good may, from certain circumstances, produce a desire superior to what arises from the greatest and most valuable enjoyment . . . In short, a passion must be accompany’d with some false judgment, in order to its being unreasonable; and even then ’tis not the passion, properly speaking, which is unreasonable, but the judgment. (Hume 2007: 2.3.3.6) [p. 178] In his trenchant critique of the role of reason in the actions of human beings, Hume, as Williams would later do, declares that at the bottom there are always people’s desires. Although these desires can certainly be criticized and possibly blamed, they are to be traced back to the deliberation and intentions of individuals and are valid as such. Besides, it is Hume’s conviction that the dimension of ethics, for it to be literally identified as such, entails a practical dimension: that is, it must move individuals to action. ‘Extinguish all the warm feelings and prepossessions in favour of virtue’, says Hume, ‘and all disgust or aversion to vice: render men totally indifferent towards these distinctions; and morality is no longer a practical study, nor has any tendency to regulate our lives and actions’ (Hume 1998: 1.8). And the lives and actions affected by ethics are inevitably the lives and actions of particular people. True, in the second Enquiry Hume states that in order to pave the way for such a sentiment [some internal sense or feeling, which nature has made universal in the whole species], and give a proper discernment of its object, it is often necessary, we find, that much reasoning should precede, that nice distinctions be made, just conclusions drawn, distant comparisons formed, complicated relations examined, and general facts fixed and ascertained. Some species of beauty, especially the natural kinds, on their first appearance, command our affection and approbation; and where they fail of this effect, it is impossible for any reasoning to redress their influence, or adapt them better to our taste and sentiment. But in many orders of beauty, particularly those of the finer arts, it is requisite to employ much reasoning, in order to feel the proper sentiment; and a false relish may frequently be corrected by argument and reflection. There are just grounds to conclude, that moral beauty partakes much of this latter species, and demands the assistance of our intellectual faculties, in order to give it a suitable influence on the human mind. (Hume 1998: 1.9) And yet, this admission of a co-​participation of reason and feeling in the identification of moral distinctions seems rather to be a concession on Hume’s part to his readers that morality is not resolved in the blind expression of personal whims that cannot be shared but is the result of a sympathetic involvement of all those who take part in the moral discussion. ‘Our intellectual faculties’, however important, only provide ‘assistance’. In the end, it is always a sentiment that has the last word, a sentiment that, however widespread among human beings and ascribable to their nature, remains nevertheless an individual sentiment that people may or may not possess. Furthermore, in the first appendix of the second Enquiry Hume makes another statement that can be dovetailed with the ‘scratching of my finger’ passage. Again in contrast to an interpretation of human conduct as being guided by reason, he observes that, in the chain of justifications we can provide for our actions, we will each eventually come up against a desire—​either for the pursuit of pleasure or for the avoidance of pain—​that corresponds, for each of us, to ‘an ultimate end’ [p. 179] Lorenzo Greco (Hume 1998: app. 1.18) which, as such, cannot be further rationally questioned. Hume can be interpreted as an individualist as much as Williams: for both, human conduct finds its explanation as much as its justification in the intentions of individual persons. Williams probably did not have Hume in mind here; and yet the two, whether Williams was aware of it or not, proceed in a similar direction. 4. The Science of Man The appeal to individualism for Williams is set against a background in which human nature matters (1985: ch. 3; 1995g: 194–​202; 1972: 55–​62). If there is any hope of establishing a foundation for ethics, it lies in a convincing conception of human nature, and this involves having a clear idea of what we mean by naturalism in relation to the prescriptions and evaluative judgements of which ethical thought is composed. Naturalism in ethics, Williams explains, ‘consists in the attempt to lay down certain fundamental aspects of a good human life on the basis of considerations of human nature’ (1995d: 101). That is, it is ‘the project of thinking out, from what human beings are like, how they might best and most appropriately live’ (1995d: 109). Such a project, according to Williams, cannot fail to take into account the revolution brought about by Charles Darwin, which has wiped out any attempt to read into human nature some final cause: ‘The first and hardest lesson of Darwinism, that there is no such teleology at all, and that there is no orchestral score provided from anywhere according to which human beings have a special part to play, still has to find its way fully into ethical thought’ (1995d: 110). Williams’s relationship to the truth of Darwinism, however, is multifaceted. As it happens, in fact, ‘evolutionary biology is not at all directly concerned with the well-​ being of the individual, but with fitness, which is the likelihood of that individual’s leaving offspring’ (Williams 1985: 44). Williams, however, given his formal as well as substantial individualism, is interested in people as unique individuals; if ethical reflection is to have any hope of being able to say anything meaningful about human existence, it must start from a consideration of individuals, their deliberations, and the characters on which their deliberations depend. In turn, individuals must be able to understand ethical reflection not so much as a list of more or less general obligations but as considerations that they can recognize as important to them, in the sense that those considerations must have for them, as they are those specific individuals and not others, a practical significance that leads them to act in one way or another (Williams 1985: 182–​3). Given the truth of evolutionary biology, then, ‘[i]‌f any science is going to yield conclusions that are for each person, . . . it will be some branch of psychology’ and psychology is fully within evolutionary biology. However, ‘[w]hat is true is that each action is explained, in the first place, by an individual’s psychology; what is not true is that the individual’s psychology is entirely explained by [p. 180] psychology. There are human sciences other than psychology, and there is not the slightest reason to suppose that one can understand humanity without them’ (Williams 1995c: 86). In this sense, when one wants to account for the ethical life of human beings, and one wants to do so within a naturalist context, this in no way implies that we should adopt a radically reductionist approach: ‘It is important, once again, to recognize that what is at issue here is not a mechanistic reductionism, an attempt to represent ethical thought in terms supposedly appropriate to the natural sciences’ (Williams 1995g: 204). There is more to it than this. Although it is true that human behaviour is dependent on biology as it emerges after the Darwinian revolution, nevertheless, biology . . . requires ethology and . . . the ethology of the human involves the study of human cultures. . . . There is no reason at the moment, as I understand the situation, to suppose that patterns of development are independent of history and culture; South Korea and Victorian England are by no means the same place. But however that may be, these are indisputably matters for historical and cultural understanding. (Williams 1995c: 87) When it comes to explaining what being human means, and thus what it means for people to have an ethical life, the factors that come into play are therefore manifold. As Williams observes, ‘ “Humanity” is, of course, a name not merely for a species but for a quality’ (1995c: 88); human nature can indeed be explained in terms of how human creatures have developed evolutionarily, but it also has aspects that can only be understood by taking into account the fact that human beings are creatures that have a culture and are able to think reflectively about themselves. It is only with this in mind that it is possible to speak of good or bad, right or wrong, correct or incorrect behaviour. Provided that human beings as a species depend on biological processes that can be explained evolutionarily, to believe that when speaking of humanity one can limit oneself to human biology alone is, for Williams, just wishful thinking: the notion of humanity includes descriptive but also evaluative aspects and is the result of a process that is necessarily also cultural and historical. Now, if we look at how Hume sets up his ‘science of man’ (Hume 2007: Intro.: 0.4), we come across something very similar. Already in the Introduction to the Treatise, in outlining his empirical and experimental method of investigating human nature, Hume advances the systematic thesis that the experiments valid for such an investigation are of a very different kind from those that characterize natural science: When I am at a loss to know the effects of one body upon another in any situation, I need only put them in that situation, and observe what results from it. But should I endeavour to clear up after the same manner any doubt in moral [p. 181] Lorenzo Greco philosophy, by placing myself in the same case with that which I consider, ’tis evident this reflection and premeditation would so disturb the operation of my natural principles, as must render it impossible to form any just conclusion from the phænomenon. We must therefore glean up our experiments in this science from a cautious observation of human life, and take them as they appear in the common course of the world, by men’s behaviour in company, in affairs, and in their pleasures. Where experiments of this kind are judiciously collected and compared, we may hope to establish on them a science, which will not be inferior in certainty, and will be much superior in utility to any other of human comprehension. (Hume 2007: Intro 0.10) Hume’s science of human nature is never given in a vacuum, nor can it be translated into exclusively biological terms, but it only finds its full expression if the object of investigation—​human beings—​are conceived as agents moving in an evaluatively and normatively dense reality: that is, in a reality that is socially, culturally, and historically distinguished. If we then look at what Hume argues in Section 1 of An Enquiry concerning Human Understanding, the matter is the same: ‘Man is a reasonable being’, Hume tells us, but at the same time he is also ‘a sociable . . . being’ as well as ‘an active being’ (Hume 1999: 1.6); only by observing human beings in this perspective will it be possible to provide a truly comprehensive profile of creatures whose nature eludes us if we do not observe them as always involved in a practical dimension. According to Paul Sagar, ‘Hume did not think morality vulnerable to the same critique of religion—​Humean ethical practice could not be established the way religion could, because living within ethics is a necessary part of fully realized human nature for Hume’ (2013: 22). Also because of this belief in a morality dependent on a fully realized human nature, Hume would err on the side of optimism, which would distance him from Williams’s more thoughtful Nietzschean pessimism.5 However, Sagar’s reading of Hume misses the point of the Humean appeal to human nature. There are various things for Hume—​beliefs, customs, habits—​that appear to be ‘natural’ for human beings, which means that these things enter into the description we give of human beings as natural creatures after careful observation. That being the case, both religion and ethics are elements we register when we describe human beings as natural beings; in Hume’s view, religion is as natural to human beings as ethics. However, Hume is not Aristotle; he does not believe that it 5 I will not go into the merits of Williams’s debts to Nietzsche’s thought here because that would take me too far from my purposes. Concerning the question of his Nietzschean pessimism, see Krishnan and Queloz (2023, esp. 234–​40). As they observe, Williams’s pessimism, in line with Nietzsche’s, can be defined as a ‘pessimism of strength’: that is, ‘a form of pessimism . . . that . . . involves being open-​eyed about the horrors of the world; but . . . a pessimism of strength in that it aspires to face these appalling truths truthfully, without aesthetically idealising or distracting from them, and without rendering them more bearable through redemptive myths and illusions’ (2023: 239). [p. 182] is possible to identify an end proper to human nature. There is no doubt that ethics for Hume is an integral part of the way in which for him human nature unfolds; human nature is given through the analysis of human practices, but these practices can vary and do not conform to any ideal of full realization of human nature. On the contrary: as the History of England shows—​which can be read as an application to history of those principles of human nature that Hume had identified in his theoretical works—​the passions and sentiments of human beings find expression in the most varied contexts. These passions and sentiments find substance in these contexts, are defined by these contexts, and the idea of human nature that is derived from them responds to a uniformity that inevitably depends on the innumerable ways in which human beings operate and relate to each other throughout their history. 5. A Reasonable Optimism There is nothing naively optimistic on Hume’s part in his confidence in human nature so understood, nor does this confidence inevitably distance him from Williams’s perspective. Indeed, the latter, very much like Hume, identifies constants in the multitude of ways in which human beings live that allow us to recognize each other and give each other the respect and relevance we deserve. Consider the way Williams justifies the importance of our ‘common humanity’ in The Idea of Equality: The tautology [that human beings are equal] is a useful one, serving as a reminder that those who belong anatomically to the species homo sapiens, and can speak a language, use tools, live in societies, can interbreed despite racial differences, and so forth, are also alike in certain other respects more likely to be forgotten. These respects are notably the capacity to feel pain, both from immediate physical causes and from various situations represented in perception and in thought; and the capacity to feel affection for others, and the consequences of this, connected with the frustration of this affection, loss of its objects, and the like. The assertion that people are alike in the possession of these characteristics is, while indisputable and (it may be) even necessarily true, not trivial. For it is certain that there are political and social arrangements that systematically neglect these characteristics in the case of some groups of people being fully aware of them in the case of others; that is to say, they treat certain people as though they did not possess these characteristics, and neglect moral claims that arise from these characteristics and which would be admitted to arise from them. (2005: 99) As with Hume, for Williams, too, the characteristics that define human beings have nothing essential but are normatively and evaluatively charged, and are found in [p. 183] Lorenzo Greco specific circumstances in which these characteristics take hold and thrive; likewise, they can also be denied and repressed by the hand of other human beings. When this happens, Williams gives voice to his pessimism regarding the fact that the human condition is perpetually dependent on luck and at risk of being violated and exposed to pain and distress. At the same time, however, Williams’s considerations do not give way to hopeless despair, for just as human beings can suffer, given their nature, because of other human beings, they can also, for the very same reason, be happy and prosper. Hume is no different. For him also, in fact, ethics comprises a reflective dimension that can develop precisely because human beings are only conceivable in relation to one another within a dimension that is cultural and historical: It requires but very little knowledge of human affairs to perceive, that a sense of morals is a principle inherent in the soul, and one of the most powerful that enters into the composition. But this sense must certainly acquire new force, when reflecting on itself, it approves of those principles, from whence it is deriv’d, and finds nothing but what is great and good in its rise and origin. Those who resolve the sense of morals into original instincts of the human mind, may defend the cause of virtue with sufficient authority; but want the advantage, which those possess, who account for that sense by an extensive sympathy with mankind. According to the latter system, not only virtue must be approv’d of, but also the sense of virtue: And not only that sense, but also the principles, from whence it is deriv’d. So that nothing is presented on any side, but what is laudable and good. This ‘extensive sympathy with mankind’ is a form of reflexivity or ‘reflective endorsement’, to use an expression of Christine Korsgaard (1996)—​and it is no coincidence that she classified both Hume and Williams (and John Stuart Mill) under this rubric. For Hume, through reflection we can explain and justify why adopting an ethical conduct is something good for us, given our shared human nature. In fact, reflective endorsement represents the final answer for someone who seriously questions why we should abide by ethics. If this is the case, then Hume appears to be less of an optimist than Williams (and Sagar) might have believed. The Humean kind of reflective endorsement leaves space both for the knave who does not respect the rules of justice and for the wicked person who is pleased in adopting a morally odious behaviour; when dealing with such individuals, the only thing one can do is to show them how terrible it is to act the ways they do, given the amount of suffering that such actions may cause. Although there is a lot to say in favour of the moral stance after reflection, this does not prevent the possibility that other people with different dispositions will not comply. Reflexivity in Hume does not prevent immorality, nor does it prove the superiority of the moral stance once and for all. Reflexivity will not [p. 184] silence the knave, but this does not mean that its conclusions lose their importance because of that. Williams adopts a very similar strategy—​a strategy that he would maintain well beyond his early writings and well into his later ones (one thinks of the essays collected in In the Beginning Was the Deed as well as the last chapters of Truth and Truthfulness). Note, in this regard, what Williams says in response to Charles Taylor in World, Mind, and Ethics. Against ‘the morality system’, Williams argues, we need to focus on what is important for us: that is, on ‘securing the protection of important interests’. Williams continues thus: ‘This was not meant to be a novel idea. It is broadly in the spirit of Hume, and, like Hume’s account, it deals in everyday, but still important, kinds of importance—​the importance to each person of not being killed or despoiled’ (1995g: 205). Given that this is ‘the spirit of Hume’, how can he be accused of suffering ‘from a somewhat terminal degree of optimism’? In ‘Lonely in Littlemore’, Blackburn recalls Williams’s reasons for disappointment with Hume: ‘Hume had a naïve Enlightenment confidence in a single human nature’, Blackburn says, ‘which was no use to Bernard, in his explorations of the ways that different histories and cultures shape different ways of living and their associated mores and ethics’ (2019: 28). However, from what Williams himself argues, it would seem that he sides with Hume in identifying what it is that really matters when human nature is called into question. This is something that Blackburn also notes in his defence of the Humean perspective: ‘A form of life deserves confidence insofar as it does well in avoiding those things: insecurity, war, poverty, displacement, lawlessness, random violence, and misery, to name just a Williams observes that the problem with Hume for his modern readers is that he is ‘too comfortable’ (1986: 206). If this was what Williams himself believed, then we might dare say that he misunderstood Hume’s aspiration to articulate a science of human nature. Hume’s idea of a ‘single human nature’ emerges from the generalizations that can be made by looking at the different histories and cultures in which Williams himself was so interested. The singularity of the results of these generalizations corresponds to those aspects that appear to be constant and steady in the behaviour of human beings—​aspects that Williams underlined as well when he noted the capacity human beings possess to flourish or languish because of the actions of other people, given that human beings are endowed with sentiments and passions, and with an imaginative capacity to conceive, and take part in, other people’s condition. This is the ‘single human nature’ Hume had in mind. It is not a supposed essence that remains rigidly the same despite the variability of the ways human beings can live; on the contrary, human nature as conceived by Hume emerges from this very variability as what remains constant and steady—​a constancy and steadiness that are not presupposed nor assessed at a metaphysical level but are the results of empirical observation. Far from putting forward a ‘ “purely philosophical” understanding’ (Williams 1986: 207) of what ethical life and ethical [p. 185] Lorenzo Greco thinking are, ‘Hume himself was fascinated by human diversity, and devotes the dialogue at the end of EPM as well as several essays to what he calls the “moral causes” that underlie it—​moral causes being exactly the kind of social and cultural environmental variations that interested Bernard’ (Blackburn 2019: 28). 6. Concluding Remarks Of course, Williams elaborated an extremely sophisticated philosophical perspective, and to describe him only as an heir of Hume is certainly reductive and ultimately incorrect. However, there is more of Hume in Williams than Williams was ready to grant, and the Hume with whom Williams found similarities was not that optimistic philosopher that Williams accused him of being. According to Paul Russell, ‘[Williams’] investigations begin, substantially, with materials that were provided by Hume. The Humean outlook, we may say, is the principal point of entry for the trajectory of Williams’s ethical thought’ (2021: 279). And given that for Williams ‘the very foundations of ethical life must be viewed as historically and culturally contingent, with no basis for privileged authority’ (Russell then for Russell ‘[w]‌hile Williams’s ethical outlook has a Humean point of entry, it exits with a Nietzschean set of reflections and concerns—​and these do not encourage any easy optimism about our ethical predicament’ (2021: 280). Nevertheless, Russell concludes that Williams never repudiated his Humean commitment; Nietzsche, for Williams, represents ‘a development from within the Humean tradition . . . Hume’s “terminal optimism”, although real, is in many respects superficial. If this is correct, then the distance between Hume and Williams is not as great as Williams took it to be’ (2021: 281). Russell is right: what Williams can appreciate in Hume is the idea that morality is based in a human nature which is explained by seeing human beings in action and in relation to one another. For both Hume and Williams, human nature is characterized in ethical terms from the start since it is the result of interactions among human beings and emerges from the innumerable descriptions of how human beings can be happy or miserable because of others. The happiness and misery being discussed consist, for both, in the happiness and misery of persons considered in their uniqueness; reflections on human nature and ethics find their beginning as well as their end, for both Hume and Williams, in the individual. So, is Hume’s position too optimistic when compared to Williams’s? I would rather define it as realistically reasonable, or reasonably realistic. The fate of human beings, for Hume, does not inevitably point to the best. What he could derive ‘from a cautious observation of human life’ was a picture of human nature as presenting good as well as bad aspects. What interested Hume was the capacity human beings possess to tune in with each other and feel what other people feel. Human beings are capable of sympathizing with each other, but given the very same principle [p. 186] of sympathy they can compare themselves with each other and react to this comparison in morally questionable ways. Human nature is not good in itself for Hume; Hume is no Rousseau. Hume was not any more optimistic than Williams; on the contrary, Hume shared Williams’s awareness that the human condition is fragile and open to pain and suffering as the result of intentional actions of human beings. If all this is the case, then it may be fair to say that Williams did pray, although quietly and from time to time, in the church of Hume.6 References Blackburn, Simon. 1986. ‘Making Ends Meet’. Philosophical Books 27.4: 193–​203. Blackburn, Simon. 2019. ‘Lonely in Littlemore: Confidence in Ethics and the Limits of Philosophy’. In Ethics beyond the Limits: New Essays on Bernard Williams’s Ethics and the Limits of Philosophy. Edited by Sophie Grace Chappell and Marcel van Ackeren, 27–​36. Abingdon and New York: Routledge. Greco, Lorenzo. 2007. ‘Humean Reflections in the Ethics of Bernard Williams’. Utilitas Greco, Lorenzo. 2019. ‘Humanism and Cruelty in Williams’. In Ethics beyond the Limits: New Essays on Bernard Williams’s Ethics and the Limits of Philosophy. Edited by Sophie Grace Chappell and Marcel van Ackeren, 84–​103. Abingdon and New York: Routledge. Hume, David. 1998. An Enquiry concerning the Principles of Morals. Edited by Tom L. Beauchamp. Oxford and New York: Oxford University Press. Hume, David. 1999. An Enquiry concerning Human Understanding. Edited by Tom L. Beauchamp. Oxford and New York: Oxford University Press. Hume, David. 2007. A Treatise of Human Nature: A Critical Edition. Edited by David Fate Norton and Mary J. Norton. Oxford: Clarendon Press. Korsgaard, Christine M. 1996. The Sources of Normativity. Cambridge: Cambridge University Press. Krishnan, Nikhil, and Queloz, Matthieu. 2023. ‘The Shaken Realist: Bernard Williams, the War, and Philosophy as a Cultural Critique’. European Journal of Philosophy 31.1: 226–​47. Nussbaum, Martha C. 2003. ‘Tragedy and Justice: Bernard Williams Remembered’. Boston Review, 1 October. Queloz, Matthieu. 2021. The Practical Origins of Ideas: Genealogy as Conceptual Reverse-​ Engineering. Oxford and New York: Oxford University Press. Russell, Paul. 2021. ‘Hume’s Optimism and Williams’s Pessimism: From “Science of Man” to Genealogical Critique’. In Russell, Recasting Hume and Early Modern Philosophy: Selected Essays, 265–​82. Oxford and New York: Oxford University Press. Sagar, Paul. 2013. ‘Minding the Gap: Bernard Williams and David Hume on Living an Ethical Life’. Journal of Moral Philosophy 11.5: 1–​24. Taylor, Jacqueline A. 2015. Reflecting Subjects: Passion, Sympathy, and Society in Hume’s Philosophy. Oxford and New York: Oxford University Press. Williams, Bernard. 1972. Morality: An Introduction to Ethics. Cambridge: Cambridge University Press. Williams, Bernard. 1973. ‘Egoism and Altruism’. In Williams, Problems of the Self: Philosophical Papers 1956–​1972, 250–​65. Cambridge: Cambridge University Press. 6 I would like to thank Roger Crisp, Eugenio Lecaldano, Dan O’Brien, Matthieu Queloz, and Marcel van Ackeren for reading and commenting on an earlier version of this chapter. [p. 187] Lorenzo Greco Williams, Bernard. 1981a. ‘Internal and External Reasons’. In Williams, Moral Luck: Philosophical Papers 1973–​1980, 101–​13. Cambridge: Cambridge University Press. Williams, Bernard. 1981b. ‘Wittgenstein and Idealism’. In Williams, Moral Luck: Philosophical Papers 1973–​1980, 144–​63. Cambridge: Cambridge University Press. Williams, Bernard. 1985. Ethics and the Limits of Philosophy. London: Fontana. Williams, Bernard. 1986. ‘Reply to Simon Blackburn’. Philosophical Books 27.4: 203–​8. Williams, Bernard. 1995a. ‘Internal Reasons and the Obscurity of Blame’. In Williams, Making Sense of Humanity and Other Philosophical Papers 1982–​1993, 35–​45. Cambridge: Cambridge University Press. Williams, Bernard. 1995b. ‘Nietzsche’s Minimalist Moral Psychology’. In Williams, Making Sense of Humanity and Other Philosophical Papers 1982–​1993, 65–​76. Cambridge: Cambridge University Press. Williams, Bernard. 1995c. ‘Making Sense of Humanity’. In Williams, Making Sense of Humanity and Other Philosophical Papers 1982–​1993, 79–​89. Cambridge: Cambridge University Press. Williams, Bernard. 1995d. ‘Evolution, Ethics, and the Representation Problem’. In Williams, Making Sense of Humanity and Other Philosophical Papers 1982–​ Cambridge: Cambridge University Press. Williams, Bernard. 1995e. ‘Formal Structures and Social Reality’. In Williams, Making Sense of Humanity and Other Philosophical Papers 1982–​1993, 111–​22. Cambridge: Cambridge University Press. Williams, Bernard. 1995f. ‘Formal and Substantial Individualism’. In Williams, Making Sense of Humanity and Other Philosophical Papers 1982–​1993, 123–​34. Cambridge: Cambridge University Press. Williams, Bernard. 1995g. ‘Replies’. In World, Mind, and Ethics: Essays on the Ethical Philosophy of Bernard Williams. Edited by E. J. E. Altham and Ross Harrison, 185–​224. Cambridge: Cambridge University Press. Williams, Bernard. 1999. ‘Seminar with Bernard Williams’. Ethical Perspectives 6.3–​4: 243–​65. Williams, Bernard. 2001. ‘Some Further Notes on Internal and External Reasons’. In Varieties of Practical Reasoning. Edited by Elijah Millgram, 91–​7. Cambridge, MA: MIT Press. Williams, Bernard. 2002. Truth and Truthfulness: An Essay in Genealogy. Princeton and Oxford: Princeton University Press. Williams, Bernard. 2005. ‘The Idea of Equality’. In Williams, In the Beginning Was the Deed: Realism and Moralism in Political Argument, 97–​114. Princeton and Oxford: Princeton University Press. Williams, Bernard. 2006. ‘Hume on Religion’. In Williams, The Sense of the Past: Essays in the History of Philosophy, 267–​73. Princeton and Oxford: Princeton University Press. [p. 188] Recasting Responsibility Hume and Williams Paul Russell Few would deny that Bernard Williams has made enormously important and influential contributions to the subject of free will and moral responsibility. His views on this subject have been presented in a number of different works, including several papers published in the 1970s and 1980s, along with Shame and Necessity (1993).1 Ethics and the Limits of Philosophy (1985), which is arguably Williams’s most ambitious contribution to moral philosophy, has broader reach but it shares a central theme with his other writings, the critique of ‘the morality system’ (Williams 1985/​2011: esp. ch. 10). What Williams has to say about free will and moral responsibility is fundamental to that critique. One way to understand and assess Williams’s views on free will and moral responsibility is to consider them in relation to his most significant predecessors. Two figures who stand out in this regard are Kant and Nietzsche. Kant is the most prominent representative of ‘the morality system’ and Nietzsche is identified by Williams as its greatest critic, and someone whose work Williams increasingly drew from and cited over the years (Williams 1985/​2011: 194).2 There remains, nevertheless, a third important figure who should not be overlooked in this context, David Hume. Williams not only regarded Hume’s work in moral philosophy as ‘manifestly important’, he also mentions him as someone he greatly admired and who impressed him early on (1983: 41–​2; 1999: 256). Williams also states that Hume’s work ‘matters a great deal to some of my [i.e. Williams’s] concerns: in his treatment of free-​will, for instance, and in his resolute rejection of the assumptions of what I call “morality”, in particular of the idea that there is some deep difference between virtues and other forms of admirable human quality’ (1986: 206). The discussion that follows will examine the significance of this, not only for their respective views, but also for our own understanding of how we might best make sense of moral responsibility and how it relates to the free will problem. 1 The papers I have particularly in mind are: ‘Moral Luck’ (1976); ‘How Free Does the Will Need to Be? (1985)’; ‘Internal Reasons and the Obscurity of Blame’ (1989); ‘Nietzsche’s Minimalist Moral Psychology’ (1993). 2 ‘It is certain, even if not everyone has yet come to see it, that Nietzsche was the greatest moral philosopher of the past century’ (Williams 2014: 256). Paul Russell, Recasting Responsibility In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. [p. 189] Paul Russell Morality’ and the Free Will Problem Williams’s critique of the morality system is a central theme of his influential contribution to contemporary ethics, and, as already noted, his critical views concerning moral responsibility and blame are essential to that critique.3 According to Williams, the most important and distinguishing feature of the morality system is its (peculiar) conception of obligation and, closely associated with this, its understanding of voluntariness and blame (1985/​2011: 193–​8). Moral obligations, on this view, are grounded in reasons that are available to all rational agents—​what Williams refers to as the ‘the universal constituency’ (1985/​2011: 16). Agents who voluntarily violate the demands of morality are subject to blame and retribution (Williams 1985/​2011: 200; 1993a/​1995a: 72–​4). In these circumstances the agent deserves to suffer as a matter of justice (Williams 1985/​2011: 214; 1993a/​1995a: 72; According to this (idealized) view, all moral agents belong to a community of rational free agents—​‘the notional republic’—​governed by moral demands and backed by sanctions that apply equally to all. In order to support this idealized picture of the moral community and moral agency, we need to look well beyond ‘the ordinary materials of psychological explanation’, such as belief, desire, deliberation, intention, and decision (Williams What is required is that we ‘deepen’ and ‘refine’ the idea of the voluntary in order to make it ‘profound’ (Williams 1993: 68). This pursuit leads to efforts to secure some sort of ‘limitless freedom’ or ‘total control’ (Williams 1985/​2011: 63–​5, 215–​17; 1985/​1995: 16; The ‘metaphysical fuel’ required for this is, Williams argues, impossible to satisfy or make sense of (2002/​2004: 86). Lying behind these features of ‘morality’ is the aspiration to ‘ultimate justice’ (Williams 1985/​2011: 43, 216–​17; b/​1995a: 241–​4). In order to achieve perfect equality of moral opportunity we must be able to somehow ‘transcend luck’ and ‘contingency’. It is clear, nevertheless, that the metaphysical and psychological claims needed to support this are ‘untruthful’ about our human ethical predicament and, consequently, ‘morality’ collapses under the weight of its own extravagant assumptions and aspirations. We are, Williams maintains, ‘better off without it’ (1985/​2011: 193). How, then, does this critique relate to ‘the free will problem’? Given Williams’s remarks concerning the ‘illusions’ and ‘fantasies’ relating to the aspiration to ‘total control’, he might be taken to be specifically targeting libertarian metaphysics and arguing for some form of orthodox compatibilism. That would, however, be a mistake. The aspirations of libertarian metaphysics to secure some form of ‘ultimate’ agency or ‘limitless freedom’ are, no doubt, motivated by the aims and aspirations 3 For a general overview of Williams’s views on this subject, see Russell (2022) and Queloz (2022). 4 The crucial element missing from this list, according to ‘morality’ (and the ‘progressivist’ view associated with it) is ‘the will’. [p. 190] b/​1995a: 242–​4; 1994: 7–​9). Williams makes clear, nevertheless, that he also wants to distance himself from the ‘reconcilers’ and ‘old compatibilists’ who are, he claims, no less committed to ‘moral responsibility’ as morality understands it (1985/​1995: 6–​7, 19; 1995b: 578). Compatibilists of this ilk are no less confident than their libertarian counterparts that our existing concepts of responsibility and blame are generally in good shape and that we can ‘leave everything where it was’ (Williams 1976/​1981: 39; 1985/​1995: 19; 1995b: 578). The compatibilism that Williams defends, by contrast, maintains that we need to radically ‘recast our ethical conceptions’ (1985/​1995: 19).5 If we are to ‘recast’ our ethical conceptions relating to moral responsibility, we need to turn to the methods of genealogy, which draws on history and literature, as well as philosophy.6 With this in mind, Williams encourages us to turn back to the ancient Greeks (1993: esp. ch. 1; 1994: 18. These genealogical resources, he argues, make it clear that we have no reason to accept unqualified or categorical scepticism about moral responsibility. Williams is sceptical not so much about (the concept of ) moral responsibility as about the demands that ‘morality’ places upon it.7 Contrary to the sceptical view, Williams maintains that we can provide a convincing vindicatory genealogy of moral responsibility—​this being the fundamental task of Shame and Necessity.8 Although any recognizable form of human ethical life will involve some shared basic elements, such as our concern with an agent’s intentions and state of mind, there is no single ‘correct’ or ‘ideal’ concept of moral responsibility (Williams 1993: 55–​6).9 It is a mistake encouraged by ‘morality’, Williams argues, to assume that we have evolved to arrive at one correct conception of responsibility. Assumptions of this kind are not only fundamental to the ambitions and expectations of ‘morality’; they have served to generate the free will problem. The fundamental point that Williams is making is ‘that our conceptions of freedom, responsibility and blame are often not what they seem, and are variously exaggerated, self-​deceiving, sentimental or vindictive’ (1985/​2011: 216; 1993a/​ a: 72; 1995b: 578). ‘We have’, as he puts it elsewhere, ‘fooled ourselves into believing that we have a more purified notion of moral responsibility than we have’ One particular danger of ‘morality’s’ misconceptions regarding moral 5 ‘We need’, Williams says, ‘to recast our ethical conceptions. But that is not in order to escape or adjust ourselves to determinism or naturalistic explanation. We need to do so in order to be truthful even to what we know already about our psychology, and to much of our ethical life . . .’. 6 Although Williams employs genealogical methods most extensively in Shame and Necessity, he does not discuss them in detail in that context. See, however, his extended remarks on genealogical methods (Williams 1993a/​1995a: esp. 75–​6n12; 2000; and, at greater length, 2002: ch. 2). 7 See, in particular, Williams (1993: ch. 3, esp. 55–​6, 67–​8,153, 158; 1985/​2011: 43, 63–​5, 216–​17). 8 See, in particular, Williams (1993: ch. 3), which is titled ‘Recognizing Responsibility’. 9 Other core (or universal) elements that Williams mentions include our concern with any harm or injury caused by an agent’s actions and forms of compensation when this occurs. [p. 191] Paul Russell responsibility is that it generates the ‘free will problem’, where this is understood in terms of the task of ‘saving’ its illusory conception of moral responsibility. Among other things, this entire dynamic pushes us towards the precipice of a misguided, unqualified scepticism.10 Williams’s concern is not to ‘solve’ the free will problem, but to show that it is generated by the illusions of ‘morality’.11 2. Making Sense of Hume’s Compatibilism Throughout much of the twentieth century, the established view of Hume was that he is a founding figure of the classical compatibilist tradition. This is a tradition that begins with Hobbes and runs through Hume to twentieth-​century figures such as Schlick and Ayer.12 In line with much of the modern debate, classical compatibilists believe that what is fundamental to understanding moral responsibility is a credible account of free will. Compatibilists deny that morally free action requires the falsity of determinism or that responsible action cannot be causally necessitated by antecedent conditions. Whether an action is free or not depends not on the absence of causation and necessity but on the type of cause involved. Free, responsible action is caused by the agent’s desires and willings. Action that is produced by the agent’s willings is not compelled or forced to occur through external causes of some kind. If an action were entirely uncaused, it could never be attributed to any agent; it would be a capricious and random event. Rewards and punishments secure valuable social benefits (i.e. conformity to the law etc.) only because they cause agents to act differently than they would in their absence. These are the familiar central arguments of the classical compatibilist position. So understood, they are concerned with the logic of the relevant concepts in question (freedom, necessity, etc.).13 According to the naturalistic interpretation, the classical account overlooks Hume’s crucial concern with the role of moral sentiment for understanding these arguments. Holding a person responsible is, according to Hume, a matter 10 The danger, related to this, is that once these illusions of ‘morality’ are exposed, and discarded, we will go on to conclude that ‘there can be no coherent ideas of social justice, but only efficiency, or power, or uncorrected luck’ (Williams 1985/​2011: 218). 11 See, for example, Williams (1993: 67–​8, 94–​5, 152, 158). 12 An influential statement of this orthodoxy is presented in Stroud (1977: ch. 7). 13 In both the Treatise (T 2.3.1–​2) and Enquiry (EU 8) accounts ‘Of Liberty and Necessity’, Hume makes clear that his most original contribution to ‘the free will controversy’ rests with his ‘new definition of necessity’ (TA 34). It is because we confuse necessity with some form of force or compulsion that we mistakenly suppose that liberty requires the absence of causation. When we recognize that causation involves nothing more than a regular succession of objects or constant conjunction, this confusion is eliminated. Although classical compatibilists coming after Hume endorse some version of this argument, the emphasis remains on distinguishing two kinds of ‘liberty’, one that requires the absence of force or constraint and the other requiring the absence of causation and necessity. For details relating to the classical interpretation of Hume, see Russell (1995: esp. chs 1–​3; 2007; 2015/​2021). This section and the section that follows draw from this earlier work. [p. 192] of regarding that person as an object of moral sentiments of approval or disapproval. These sentiments are themselves (calm) forms of love or hate.14 It is virtues and vices, or our enduring pleasurable or painful qualities of mind, that give rise to moral sentiments. We infer a person’s character or qualities of mind from their actions. This inference requires some relevant (causal) regularity or ‘constant conjunction’, without which no moral sentiment would be produced. The presence of such regularities between action and character is all that causation and necessity involve. If actions were uncaused or caused by external causes of some sort (i.e. other than intentions and motives indicative of the agent’s character), it would be psychologically impossible to hold any agent responsible. It is in this way that responsibility actually requires causation and necessity. The naturalistic interpretation makes clear that the classical account significantly misrepresents Hume’s compatibilist strategy. This can be gauged by considering the significant parallels between Hume’s views and P. F. Strawson’s ‘Freedom and Resentment’ (1962/​2013), a highly influential contribution to the contemporary debate. The most notable resemblance between them is their shared ‘naturalistic’ approach, which appeals to the role of moral sentiments or reactive attitudes. The naturalist maintains that describing the psychology of our moral sentiments serves to discredit scepticism about moral responsibility arising from worries about ‘necessity’ or ‘determinism’. Both compatibilists and (libertarian) incompatibilists, Strawson argues, ‘over-​intellectualize’ the issue of moral responsibility (1962/​2013: 81).15 The classical compatibilist appeals to a ‘one-​eyed utilitarianism’ that fails to account for deserved praise and blame, rewards and punishments. The libertarian tries to fill this gap by appealing to an incoherent ‘contra-​causal freedom’ (Strawson 1962/​2013: 81). What they both overlook are moral sentiments, understood as natural emotional responses to the agent’s attitudes and intentions as manifest in their conduct. These emotions are part of our essential human make-​up and are naturally aroused in relevant circumstances. While these responses are under rational control, and subject to relevant excusing and exempting considerations, there is no question of us entirely abandoning or altogether suspending them (Strawson 1962/​2013: 71–​3). Whereas the classical interpretation presents Hume as a notable and obvious target of Strawson’s criticisms, the naturalistic account presents Hume as broadly anticipating a number of the basic features of Strawson’s approach. Both Hume and Strawson are agreed that moral responsibility has to be explained and described with reference to psychological facts relating to human emotions and the circumstances under which they are aroused or inhibited. We cannot, they argue, 14 On this aspect of Hume’s system, see Árdal (1966: ch. 6). 15 Strawson refers to compatibilists as ‘optimists’ and calls (libertarian) incompatibilists ‘pessimists’ because they suppose that the truth of determinism threatens our attitudes and practices associated with moral responsibility. [p. 193] Paul Russell understand issues of ‘freedom’ and ‘necessity’, as they concern moral responsibility, unless we consider their relevance to the operation and functioning of our moral sentiments. According to the naturalistic interpretation of Hume, it is this aspect of his discussion that constitutes its principal interest and significance for the contemporary debate. This brings us back to the question concerning the relevance of Hume’s views for Williams’s critique of the morality system and its (problematic) understanding of moral responsibility and blame. 3. Hume Against ‘Morality’: Virtue, Voluntariness, Luck Although Williams regarded Hume’s work in moral philosophy as ‘manifestly important’, his various remarks about Hume also make clear that his admiration was not unqualified and that it waxed and waned over time.16 This blend of admiration and criticism is apparent in Williams’s remarks (in reply to Simon Blackburn) explaining why Hume is ‘absent’ from his discussion in Ethics and the Limits of Philosophy. The reason for this, Williams suggests, is that in Ethics and the Limits of Philosophy, he is primarily concerned with issues arising from ‘moral diversity’. Hume is unhelpful with respect to this, because he is committed to the view that there is ‘a uniformity in the general sentiments of mankind’ and is, in this sense, an insufficiently ‘modern’ thinker (Williams 1986: 206; citation from Hume, T n1, Hume’s emphasis). However, Williams prefaces this remark with the observation, cited above, that Hume’s work on free will shows a ‘resolute rejection of the assumptions of . . . “morality” ’. Elsewhere, Williams repeats his view that Hume displays a ‘striking resistance to some central tenets of . . . “morality” ’ and, again, draws attention to Hume’s refusal to ‘take seriously’ the distinction between ‘virtue and talents’ (1985/​1995: 20n12). Williams recognizes that Hume is ‘in some ways an archetypal reconciler’ or ‘old compatibilist’, but also notes some important points of difference. These include Hume’s refusal to present blame ‘simply as an instrument for social control’ (Williams 1985/​1995: 14–​16, 20n12).17 The more fundamental difference, lying behind this, is that Hume is ‘less indebted to ideas of free will of an overambitious kind’. In this respect, he ‘is very consciously operating in a pagan perspective’ that positions him closer to the ancient Greeks (Williams 2002/​2004: 86). Clearly, then, Williams’s brief and passing remarks suggest that the classical reading, with its emphasis on utilitarian instrumentalism, fails to capture important features of Hume’s 16 There was a period in the early 1990s when Williams became more critical of Hume and distanced himself from him. During this same period, Williams became more heavily influenced by Nietzsche (see e.g. Williams 1994: 9). Nevertheless, in later reviews and works, Williams continues to refer to Hume in respectful terms (e.g. 2002/​2004: 86). This is particularly apparent in Truth and Truthfulness, where Williams discusses genealogy and presents Hume as anticipating Nietzsche’s methods (2002: 35–​8). 17 See also Williams (1985/​2011: 197–​8). [p. 194] position. Although Williams does not explicitly endorse the naturalistic interpretation, the observations he makes fit neatly with it. The details of this, however, still need to be unpacked. The morality system encourages us to understand responsibility in terms of obligations, as laid down by rules or principles (laws), where violations constitute a wrong that is liable to blame and punishment. This view of responsibility places heavy weight on the role of choice, will, and voluntariness. Classical compatibilism accommodates this understanding of responsibility and fits it neatly. According to classical compatibilism, responsibility is a matter of free action, where the agent acts according to her own will and desires. This does not require indeterminism or any special form of ‘moral causation’. Nevertheless, both the classical compatibilist and incompatibilist libertarian are agreed that responsibility is essentially a matter of free action (under some contested interpretation). Hume, on the other hand, rejects this doctrine, which we may call ‘voluntarism’.18 For Hume, as we noted, an action must be indicative of durable qualities of mind if a person is to be held accountable for it.19 This requirement reflects his commitment to virtue ethics and its focus on traits of character rather than individual actions. While voluntary action plays an important and significant role in the assessment of character, there are other channels by which a person’s character and qualities of mind may be expressed. These include feelings, desires, sentiments, and even our gestures and deportment (Hume, T 2.1.11.3/​317; EU 8.9/​ In light of this, it is clear that our moral qualities may be betrayed in ways that we have limited or even no control over. It is a mistake, therefore, to suppose that Hume endorses the view that responsibility is simply a matter of free action or that it is confined to our interest in an agent’s voluntary, intentional conduct. Hume rejects the doctrine of ‘voluntarism’ and the understanding of moral responsibility that goes with it.20 A second issue, closely related to this but distinct from it, concerns the way in which moral character is acquired and to what extent this depends on our own (free) choices. Hume makes clear that while our qualities of character are expressed primarily through our choices and decisions, for the most part our character is not acquired this way. By and large our character is conditioned and determined by factors independent of our will, such as age, sex, bodily condition, occupation, social situation, and so on (Hume, T 2.3.1.5–​10/​401–​3; EU 8.7–​15/​83–​8). Even when character is expressed through voluntary action, the character that it expresses is one that the agent herself does not choose or could have (fundamentally) altered or changed. Our conduct and character is no less subject to ‘absolute fate’ and ‘the 18 On this theme in Hume, see Russell (1995: 6, 14, 171, 176–​81). 19 See Russell (1995: ch. 7). 20 We might understand ‘voluntarism’ in terms of what J. L. Mackie has labelled ‘the straight rule of responsibility: an agent is responsible for all and only his intentional actions’ (1977: 208). [p. 195] Paul Russell bonds of necessity’, on Hume’s account, than physical bodies or material objects Hume’s views concerning the natural abilities can be better understood in terms of this account of the relationship between virtue and voluntariness (Russell Since our moral character may be acquired without our choice or consent, and it may be expressed independently from our voluntary conduct, any pleasurable or painful quality of mind will, on Hume’s principles, naturally arouse moral sentiments. It follows that the natural abilities (e.g. intelligence, imagination, wit, memory, etc.) stand on the same footing, in these respects, as the moral virtues more narrowly understood (Hume, T 3.3.34.1/​606; EM, App. 4.2, Certainly any distinction that we may draw between the natural abilities and moral virtues cannot, on this account, be based on the suggestion that the former have been acquired involuntarily. There is, nevertheless, some relevant distinction to be drawn here. The relevance of the voluntary/​involuntary distinction in this context concerns our interest in the regulation of conduct in society. Unlike natural abilities, the moral virtues and vices, or the particular actions that flow from them, ‘may be chang’d by the motives of rewards and punishments, praise and blame’ (Hume, T 3.3.4.4/​609). By means of rewards and punishments we can, in some measure, influence and change the voluntary actions of agents. For the most part, however, this is not true of the natural abilities. The significance of the ‘invented’ distinction between the moral virtues and natural abilities, Hume maintains, rests largely with considerations of this kind.22 It nevertheless remains a mistake to suppose that the boundary of moral concern rests with the voluntary/​involuntary distinction. On Hume’s principles, it would distort and truncate the nature and foundations of moral concern to limit or confine our views about responsibility in these terms.23 Hume’s arguments along these lines, relating to virtue and voluntariness, are of considerable relevance to Williams’s critique of the morality system. According to ‘morality’, when agents voluntarily violate their obligations they do wrong and are liable to blame and some measure of retribution. This engages what Williams calls ‘the blame system’. The blame system, with its focus on particular acts, requires ‘a voluntariness that will be total and will cut through character and psychological and social determinism, and allocate blame and responsibility on 21 See also Hume’s remark that ‘the fabric and constitution of our mind no more depends on our choice, than that of our body’ (Hume, ESY I, 141/​168—​‘ The Sceptic’). 22 It should be noted that Hume also associates the origin of this distinction with the (‘warping’) influence of theology, which treats ‘all morals, as on a like footing with civil laws, guarded by the sanctions of reward and punishment’ (Hume, EM, App. 4.21/​322). 23 It is worth pointing out that we may reject Hume’s (controversial) claims about responsibility for our natural abilities and still accept his (independent) claim that responsibility for our moral qualities does not depend on their being voluntarily acquired. What is implausible about Hume’s position is his (utilitarian-​oriented) conception of moral virtue, not his understanding of moral responsibility. On this, see Russell (1995: ch. 9). [p. 196] the ultimately fair basis of the agent’s own contribution, no more and no less’ (Williams 1985/​2011: 216). The reason why the morality system is under pressure to ‘deepen’ the voluntary and ‘make it profound’ is that it aspires to show that morality—​ specifically moral responsibility and blame—​ somehow ‘transcends luck’ (Williams 1985/​2011: 217; 1993: 68, 158; 1993a/​1995a: 72–​3; 1993b/​ a: 241–​4). This is required to ensure that blame is distributed in a way that is ‘ultimately just’ (Williams 1985/​2011: 216–​74). Despite the significant challenges involved, compatibilists have attempted to satisfy these aspirations of the morality system. They do this by describing powers of rational agency that ensure that we are self-​controllers who are capable of guiding and shaping our own practical identity and moral trajectories. This suffices to provide for free and responsible agency, insulated from fate and luck, within compatibilist constraints.24 In contrast with this, Hume’s naturalistic arguments do not attempt to satisfy these aims and aspirations. The particular form of moral luck that is immediately relevant to Hume’s views concerning the involuntariness of (moral) character may be described as ‘constitutive luck’.25 Constitutive luck concerns the sort of person we are. It includes our desires and feelings, abilities and dispositions, as well as our deliberate actions. It is Hume’s view that we are all subject to moral luck in this respect.26 That is to say, although we have little or no control over the kind of person we turn out to be (in these respects), we are still liable to be praised and blamed, rewarded and punished, on this basis. According to Hume’s naturalistic principles, our (moral) interest in an agent’s conduct is guided by our more fundamental interest in her qualities of mind or character traits. Depending on whether these are found pleasurable or painful, an agent is liable to be praised or blamed on this basis—​which is to say they will be held responsible. This is true despite the fact that, for the most part, agents have no (final or ultimate) control over their fundamental character, since this is almost entirely conditioned and determined by external factors of various kinds. There is, from Hume’s perspective, nothing ‘unfair’ about this so long as our beliefs about the agent are accurate and our responses are suitably calm and measured (i.e. as governed by the ‘general point of view’) (Hume, T 3.3.1.14–​17/​580–​4; EM 5.2/​229). These features of Hume’s system make clear that he rejects what Nagel has called ‘the condition of control’. According to this principle, no one can be justly held responsible—​subject to (moral) praise and blame—​for what they do not control.27 Not only does Hume reject this principle, he maintains that it is directly at odds with ordinary human experience and observation. Since the condition of control, 24 See, for example, Dennett (1984: esp. chs 4 and 5); Wallace (1994: esp. ch. 7). 25 Nagel (1976/​2013: 37): ‘Kant was particularly insistent . . .’. 26 On this, see Russell (1995: ch. 9, esp. 128–​33). Where a significant aspect of what a person does depends on factors beyond his control, yet we continue to treat him in that respect as an object of moral judgment, it can be called moral luck . . .’ [p. 197] Paul Russell in some form or other, is essential to ‘the morality system’, Hume plainly rejects this fundamental component of ‘morality’.28 In this respect, as Williams suggests, Hume’s views about freedom and moral responsibility show that he is ‘very consciously operating in a pagan perspective’ that displays some strong affinities with the ancient Greeks (2002/​2004: 85–​6). 4. The Blame System and Moral Address Blame is, Williams argues, ‘the characteristic reaction of morality’ (1985/​ Closely related to this is morality’s concern with ‘guilt’, which is ‘the characteristic first-​person reaction’ 1985/​2011: 197).30 These conceptions, as suggested by morality, are ‘purified’ to meet the standards ‘demanded by moral justice’. This, as we have noted, pushes blame towards the ideal of ‘the absolutely voluntary act’ and ‘the peculiar psychology of the will’ (Williams 1985/​1995: 16; conceived in these ‘purified’ terms, requires ‘complete control’ and a source of ‘ultimate authorship’—​this being the point at which the will is introduced (Williams a/​1995a: 72–​3; 1993: 36, 40, 46). For the purposes of this conception, what we are blamed for cannot be a matter of luck, as otherwise the very justice of our ‘moral’ reactions would be placed in doubt. One important point of agreement between Hume and Williams, in opposition to the assumptions of ‘morality’, is that not every failure to comply with moral norms—​or to do the right thing—​is a product of ‘deliberative failure’ or ‘irrationality’ (Williams 1985/​1995: 16; 1989/​1995a: 42–​3; 1985/​2011: 213).31 The problem we face, in these circumstances, is that blame may be in order even though it is not true that there was a reason for the agent to have acted otherwise or do the right thing. There are, as Williams points out, people who are ‘part of our ethical world’ who are subject to various vices (malicious, selfish, brutal, etc.), and arouse ‘various negative reactions’, but who are evidently not unreasonable 28 For a more detailed examination of Hume’s views on moral luck and the limits of control, see Russell (1995: esp. 130–​3). It may be argued that, although Hume rejects the condition of control, he is careful to limit the scope of moral luck. His various remarks on this subject suggest that the moral spectator can filter out the extraneous influence of ‘fortune’ as it may affect the particular circumstances we are required to act in, and also the variable and unpredictable consequences that our actions may have in given circumstances. To this extent it seems clear that his position on the subject of moral luck is complex and allows for some degree of ambiguity. 29 See also Williams’s related remarks about morality’s tendency to flatten and constrain ‘moral’ reactions into binary judgements of approval/​disapproval and guilt/​innocence. This excludes ‘non-​moral’ reactions such as resentment, contempt, and other ‘such minor revelations of ethical life’ (Williams 30 See also Williams (1993: 88–​95). 31 Williams’s (Humean) views about the limits of practical reason are especially relevant here (Williams 1980/​1981). [p. 198] It is these agents who we need to ‘recruit into our deliberative community’. Blame can serve to ‘recruit people into the deliberative community’, Williams suggests, by appealing to the ‘fiction’ that the agent being addressed had a reason to act differently (1985/​1995: 16; 1985/​2011: 214–​15; 1989/​1995a: 40). Even in circumstances where the blamed agent lacks the relevant motivation to do the right thing, they may still be ‘reached’ as long as they have ‘a disposition to have the respect of other people’—​in particular, other people whom they respect and care about. This is not just a matter of wanting to avoid the hostility of others but of having ‘a desire to be respected by people whom, in turn, one respects’ (Williams a: 41). This involves what Williams labels the psychology of ‘the proleptic mechanism’. For the purposes of ‘morality’, considerations and processes of these kinds are insufficiently ‘pure’: To the extent that the institution of blame works coherently, it does so because it attempts less than morality would like it to do . . . [The blame system] is surrounded by other practices of encouragement and discouragement, acceptance, and rejection, which work on desire and character to shape them into the requirements and possibilities of ethical life. (Williams 1985/​2011: 215–​16) These are social and psychological resources that morality neglects and seeks to replace with a form of total (rational) control that bypasses contingencies of character and psychological or social determination, in order to ensure that responsibility and blame are distributed on an ‘ultimately fair basis’. Almost everyone knows, says Williams, that it is an ‘illusion to suppose that this demand can be met' (Williams !985/​2011: 216. See also 1985/​1995: 17). Nevertheless, as long as we remain wedded to this illusion, we are discouraged from employing the ‘impure’ psychological and social materials that the proleptic mechanism relies on for dealing with ‘deviant members’ of the community.32 So how, then, should we understand Hume’s views about blame in relation to Williams’s concern with the ‘proleptic mechanism’? The classical reading of Hume closely associates him with the ‘reconciling strategy’ and its (utilitarian) emphasis on the efficaciousness of blame and punishment as instruments of social control. This is not, as we have noted, how Williams reads Hume. In contrast with this, the naturalistic account places considerable emphasis on the role of the ‘proleptic mechanism’ that Williams describes, taking this to play a significant part in Hume’s account of moral responsibility.33 The basis of this is the mechanism of the indirect 32 These are issues that are further developed in Williams’s discussion of ‘Shame and Autonomy’ (Williams 1993: ch. 4), particularly as this concerns the shame/​guilt contrast. A fundamental objective of this chapter is to show that ‘the Greeks’ understanding of shame . . . was strong and complex enough to dispose of the familiar criticism that an ethical life shaped by it is unacceptably heteronomous, crudely dependent on public opinion’ (Williams 1993: 97). 33 See Russell (1995: ch. 11). [p. 199] Paul Russell passions and sympathy in sustaining moral motivation and conduct. It is, more specifically, the motivations offered by pride and humility, and love and hate, that serve this purpose.34 It is a fundamental feature of Hume’s system of ethics to emphasize the social context and importance of sympathy in motivating and sustaining virtuous conduct and character. Granted that we care about and value the opinion of others, and believe that their evaluations are well founded, we will be directly affected and influenced by praise and blame (Hume, T 2.1.9.9–​13/​320–​ Virtue enables a person to take pleasure in herself, just as vice makes it painful for the mind ‘to bear its own survey’ (Hume, T 3.3.6.8/​620; EM 9.10–​11/​276). Hume and Williams are clearly agreed about the importance of ‘proleptic mechanisms’ as means of ‘reaching’ people who fail to act in appropriate ways. Blame, when expressed and communicated to the wrongdoer, has this ‘function or point’ (Williams 1985/​1995: 14–​15). It aims to ‘recruit’ others into our moral community and to encourage them to share our ethical values. Beyond this, however, there are some important differences between Hume and Williams concerning the nature of blame. In his discussion of blame, Williams makes clear that when we blame agents for some specific act or omission (i.e. ‘focussed blame’), we treat the person blamed ‘like someone who had a reason to do the right thing but did not do it’ (1989/​ a: 42). Proleptic mechanisms provide the blamed agent with a ‘sound deliberative route’ by which they may arrive at a different conclusion than might otherwise be available to them (1989/​1995a: 42). There are, however, some individuals whom we cannot reach at all because they ‘[lack] any general disposition to respect the reactions of others’ (1989/​1995a: 43). Faced with these ‘hard cases’, the fiction that we can ‘recruit’ them into a deliberative community of shared values (i.e. through available rational channels) evaporates. In these circumstances, Williams suggests, ‘we cease to blame’ and instead regard those concerned as simply ‘hopeless’ and ‘dangerous characters’ who are beyond blame (1989/​1995a: 43). ‘Hard cases’, understood this way, become exempted from blame, as blame loses its (practical) function and point in respect of them.36 Taken this way, Williams’s account of blame retains the view, which is fundamental to ‘morality’, that moral address is an essential feature of blame. Blame functions as a form of communication (‘advice’) with a view to bringing the wrongdoer (back) into the moral community. In 34 See, in particular, T 2.1 and 2; esp. 2.1.7 and 11; 2.2.5 and 6; and EM 5 and 9. 35 See also Russell (1995: 156–​8). 36 Williams does not cite any examples of ‘hard cases’ but a relevant example is provided by Robert Harris, as described in a well-​known paper by Gary Watson (1987/​2013). Harris was a vicious murderer who was eventually executed in California in 1992. Watson makes clear that the problem with Harris is not that he lacked moral understanding—​as we might find with a child or a mentally ill person—​but that ‘his heart is frozen’ (1987/​2013: 97) Watson goes on to observe: ‘To be homicidally hateful and callous in Harris’s way is to lack moral concern, and to lack moral concern is to be incapacitated for moral community. However, to exempt Harris on these grounds is problematic. For then everyone who is evil in Harris’s way will be exempt, independently of facts about their background . . .’. (1987/​2013: 100). [p. 200] the absence of any capacity to respond to such overtures, however, blame lacks any point and cannot be justified, and the wrongdoer is exempted. It is evident that Williams wants to retain the connection between blame and moral address. This accounts for the importance of ‘proleptic mechanisms’ with respect to agents who (currently) do not have any reason to do the right thing but may not be ‘unreasonable’—​since we can still ‘reach’ them through these alternative channels. However, when an agent is ‘beyond any such mechanism’, then, Williams claims, ‘we cease to blame’ (1989/​1995a: 43). Hume, however, does not share this view. According to Hume, blame is targeted at persons (qua objects of blame) on the basis of their virtues and vices, which are taken to be pleasurable or painful qualities of mind. This includes, as we noted, the natural abilities as well as moral qualities more narrowly conceived. Neither our natural abilities nor our moral qualities are acquired voluntarily, as both are generally unchosen. What distinguishes the moral qualities from the natural abilities is that they ‘may be chang’d by the motives or rewards and punishments, praise and blame’ (Hume, T 3.3.4.3–​ Blame, expressed and communicated to the wrongdoer, may serve as a form of proxy-​punishment or sanction, with the intent of influencing the agent’s conduct and character. It does not follow from this, according to Hume’s system, that blame would cease to be ‘appropriate’ unless we are able to secure these (aimed at) benefits or pragmatic ends.38 Any proleptic function that (expressed) blame may perform, important as this may be, is not, on Hume’s account, essential to blame being justified or appropriate. Nor, related to this point, does Hume grant that we exempt agents from blame where they do not satisfy the condition of being open to moral address or rational persuasion such that they may come to share our values.39 On a Humean analysis, it is a mistake to assimilate blame too closely with any ‘proleptic’ function or any role that it may play (e.g. qua some form of proxy-​ punishment). The psychological structure and rationale of blame and punishment are, although intimately connected, quite different. Among the prominent features of (standard cases of ) punishment are the following:40 1. Punishment must be voluntarily and intentionally administered (by someone other than the wrongdoer or offender). 37 See also Hume, EU 8.28/​97–​8; EM, App. 4.2, 4.20–​1/​313, 321–​2. 38 For Hume, of course, the ‘proleptic mechanism’ cannot be effectively employed to alter a person’s natural abilities. However ashamed we may be of our faults of this kind, our ability to alter them is limited. 39 For an influential statement of (moral) blame understood in these terms (qua ‘the morality system’), see, for example, Darwall (2006: esp. 27–​8 and chs 2 and 3). 40 The list provided here draws from Hart (1959/​1968: 4–​6). Hart is specifically concerned with legal punishment. He is especially concerned to offer a ‘definition’ of punishment that does not confuse the (distinct) issue of the justification of punishment. Nevertheless, even unjustified punishment must take a certain form and structure. [p. 201] Paul Russell 2. Punishment must involve some sort of harsh or unpleasant treatment of the offender. 3. Punishment must be imposed in a public, open manner. The imputation of harsh treatment is not properly punishment if it is purely private or concealed (e.g. private revenge, even when motivated by wrongdoing, is not punishment). 4. In the case of punishment, the wrongdoer or offender is made aware that the harsh treatment is imposed on them because of their offence or wrongdoing. To this extent, properly speaking, punishment is a form of ‘communication’ directed at the wrongdoer.41 5. Punishment is a voluntary practice; it can be suspended—​either in the particular case or as a (social or legal) institutional practice. 6. One ground that may be proposed for suspending or abandoning punishment is that it serves no practical or worthwhile end or purpose (i.e. it is pointless in these terms). 7. Those who impose or administer punishment must have some relevant standing or authority. The imposition of harsh treatment on the ground that the agent has violated some (legal or ethical) normative standard is not properly punishment in the absence of any relevant authority. Although the features described above may not always be satisfied and there are ‘non-​standard’ cases of punishment, these are features that serve to identify and distinguish punishment from other forms of ethical response available to us.42 Hume’s views about punishment are both scattered and thin, but all of the above features are entirely consistent with his own views.43 These features do not, however, apply to his views about blame. We may blame an agent without imposing any harsh or unpleasant treatment. Blame need not be publicly expressed, and when it is, such expression may be voluntary or involuntary. We may, indeed, find it difficult in some circumstances to conceal the (overt) expression of blame even when this is what we aim or intend to do. Moreover, even when we intend to express our blame, we may or may not intend to communicate with the offender or wrongdoer—​our intended audience may be others. In some cases, we may want or wish to communicate our sense of blame to the wrongdoer, but we cannot (e.g. because they are dead or beyond our reach). We can, nevertheless, still blame them (in contrast with punishment). Even in cases where blame lacks any practical point or value, and we may prefer not to dwell on it or entertain it for that reason, we may still find that it (naturally) forces itself upon us. To the extent that we are 41 For an influential discussion of this aspect of punishment, see Duff (2001: esp. ch. 3). 42 Hart cites collective or vicarious punishment as an example of a ‘non-​standard’ case (1959/​ 43 An account of Hume’s views on punishment is provided in Russell (1995: ch. 10). [p. 202] (sincerely) normatively committed to certain values and standards, there are psychological limits to the extent to which we can (voluntarily) suspend the attitudes and feelings involved in blame.44 Finally, blame need not be confined to those who stand in some relevant relationship of authority over the offender (although the overt, public expression of blame may be deemed inappropriate when we lack relevant moral standing). The moral psychology of blame is, in all these ways, fundamentally different—​both on Hume’s account and in fact—​from what we find with punishment. How, then, does this ‘gap’ between blame and punishment relate to Williams’s ‘proleptic’ account of blame and the role that it assigns to moral address? According to Williams, when we cease to blame because the ‘hard case’ lies beyond reach of the proleptic mechanism, we will simply regard the person concerned as a ‘hopeless and dangerous character’ (1989/​1995a: 43). Although this is consistent with the attitude that ‘morality’ would encourage us to adopt, Williams’s own critique provides reason to resist this view. The fact is, as Williams suggests elsewhere, we encounter people who are rational and normatively competent, and ‘part of our ethical world’, who are still beyond reach and cannot be (rationally) persuaded to arrive at ‘a different conclusion’ (1989/​1995a: 42). They are bad but they are ‘not necessarily behaving irrationally or unreasonably’ (Williams 1985/​2011: 213). This is, at any rate, Hume’s view. Expressing or communicating blame to them may well be pointless and of no value—​but entertaining blame towards them may still be entirely justified and appropriate. Moral address is not, for Hume, an essential or a fundamental feature of blame (however significant a role it may still play in ethical life). There is no basis for exempting an agent from blame simply on the ground that they are not open to moral persuasion or ‘correction’ (e.g. via the proleptic mechanism). Our practical interest in moral persuasion or ‘correction’ is irrelevant to the justification of blame itself. The general Humean view is that a person may (fully) satisfy the requirements of moral competence and still not be someone we can reach through ‘moral address’ of any kind.45 That they do not and cannot share our reasons does not imply that they are like a child, or a mentally ill person, or someone who lacks any capacity to understand our moral norms and the responses that they give rise to (such as blame).46 It may be that, even when moral address and blame are pointless in 44 On the natural foundations of moral sentiments (reactive attitudes) see, for example, Hume, EM App. 1.10; App. 4.21/​273, 289, 322. The same general (‘naturalistic’) theme is presented in Strawson 45 Suffice it to note that Hume’s views on moral capacity and competence are not sufficiently detailed or robust, as argued in Russell (1995: 91–​3, 179–​81). This does not, however, compromise his view that normative competence (e.g. as based on moral understanding and an appreciation of the moral responses of others etc.) does not presuppose that the agent has a capacity to come to share or endorse our values and norms. 46 What matters, in other words, for this (evolved and amended) Humean view, is that the person we blame is able to understand our moral reactions and the reasons we have for them—​but this does not require a capacity to (come to) share or endorse them. [p. 203] Paul Russell these terms (i.e. with regard to our hopes for moral persuasion and conversion), entertaining and expressing blame towards the wrongdoer still serves to recognize and acknowledge that the agent concerned is not only a human agent but remains an ethical agent. Our own ethical responses are, to this extent, constrained by this recognition and this serves to constrain the ways in which we may treat them—​ however ‘hopeless and dangerous’ they may be. 5. Responsibility Realism: Naturalism, Genealogy, Pessimism It is clear from the account provided that Williams’s assessment of Hume as a prominent philosophical figure who rejects the assumptions of ‘morality’ is well founded. There are, in fact, some respects in which Hume’s position is even more radical than Williams’s in this regard. Lying behind Hume’s and Williams’s shared opposition to the presuppositions of ‘morality’ is their ‘naturalistic’ understanding of human moral psychology and the materials that this involves (or does not involve). What a ‘naturalistic’ moral psychology commits us to is, as Williams points out, open to different interpretations (1993a/​1995a: 67–​9, 74–​5). With regard to Hume there are two particularly important strands to his ‘naturalistic’ account. The first is rooted in his wider project of a ‘science of man’, where this involves applying the observational or empirical (i.e. ‘experimental’) methods of the natural sciences to moral subjects (Hume, T, Intro, 4–​10, 2.4.7.13, 2.4.7.14/​xix–​x xii, 271, 273). What is crucial to this approach is that it excludes any appeal to non-​empirical (i.e. ‘metaphysical’) entities or agents (gods, immaterial souls, modes of ‘free will’, etc.) It begins its investigations with a descriptive account of what (actual) moral agents and their capacities are like. The description offered aims to explain as much of human ethical life as possible in terms of psychological and social features that can be independently identified. A naturalism of this kind aims to find universal features of human nature that best account for the attitudes and practices that we seek to understand.47 Secondly, and related to this, Hume’s naturalism emphasizes, as we have noted, the importance of ‘feeling’ and emotion in human life—​especially as this concerns ethics. In highlighting the role of moral sentiment in the sphere of moral responsibility, Hume’s naturalistic approach pursues a theme that runs throughout his moral philosophy. One particularly important aspect of this is that, in the words of P. F. Strawson, it discourages any effort to ‘over-​intellectualize the facts’ as they concern the attitudes and practices involved in responsibility (1962/​ Williams’s attitude to these two modes of naturalism is entirely sympathetic. Much of his own moral psychology draws on material that could well be described 47 On this, see Russell (1995: 173–​5). [p. 204] as broadly ‘Humean’.48 It is, nevertheless, Nietzsche whom Williams refers us to as providing the most reliable ‘route towards the naturalization of moral psychology’ (1993a/​1995a: 75).49 The reason for Williams’s preference for Nietzsche over Hume, in this respect, is that the approach that he endorses needs not only to be more than ‘naturalistic’ in the terms that Hume suggests; it needs to be ‘realistic’ as well (Williams 1993a/​1995a: 68). As Williams sees it, a naturalistic moral psychology can fail either because it includes or excludes too much. It includes too much if it fails to exclude aspects of our self-​image that are illusory and distort our nature. In this respect Hume’s naturalism is evidently sufficiently ‘suspicious’ of non-​natural, metaphysical entities and forms of self-​deception that this encourages. It includes too little, on the other hand, ‘if it tries reductively to ignore culture and convention; this is misguided even on a scientific basis, in the sense that to live under a culture is a basic part of the ethology of this species (Williams a/​1995a: 67).50 This concern with the importance of ‘culture’—​and cultural variation and diversity—​returns us to a point of criticism, mentioned earlier, that Williams raises against Hume. Hume’s naturalism is primarily concerned with the universal psychological features that help us to make sense of moral responsibility and blame. What he is insufficiently concerned with, according to Williams, is the significance of cultural variation in this context (1986: 206). It is a matter of fundamental importance for Williams’s critique of ‘the morality system’ that there are ‘different possibilities for ethical life: some of them possibilities for different cultures, others for our own culture’ (1985/​2011: 55). Related to this, Williams argues that human nature underdetermines ethical life (1985/​ What gives any particular form of ethical life its identity is its culture, history, and language (1985/​2011: 53–​5, 106–​7, 115–​16, 123–​5, 129–​30, It is in these terms that we need to understand that ‘morality’ is a product of a culture with a particular history and is in this sense ‘local’ and ‘contingent’—​not an inescapable, inevitable feature of human nature (Williams 1993a/​ a: 74; 1993b/​1995a: 242). These considerations apply to ‘morality’s’ conception of responsibility and blame. Faced with contingencies of this kind, we require something more than the methods of a naturalism that is concerned only with universal features of human nature (Williams 1995c: 203). What we need, methodologically speaking, to move beyond ‘morality’, is a (Nietzschean) genealogy (Williams 1993a/​1995a: 73). It may well be that Williams’s suggestion that Hume is insufficiently concerned with the problem of ethical variation (and the associated conflicts that this presents us with) is unfair to Hume. These are, after all, matters that Hume draws attention 48 See Williams (1995c: 203–​5). This is, perhaps, most apparent in his earlier writings (e.g. Williams 49 See also the related discussion in Williams (2000). 50 See also Williams (1991/​1995a: 86), where this theme is discussed at some length (‘What is true is that each action . . .’ ); and Williams (2000: 151) (‘A special case that interests us . . .’ ). [p. 205] Paul Russell to and seeks to explain in the context of his (genealogical) account of justice and in many other contexts.51 It remains true, nevertheless, that Hume is primarily interested in emphasizing the ‘uniformity in the general sentiments of mankind’ and the possibility of some shared ‘standard or morals’ or ‘common point of view’ (Hume, T 3.3.1, 3.3.3.2/​574–​91, 603; Hume, EM 5.42/​228–​9). For Williams, however, this is the starting point of his (more sceptical) investigations. We need to consider our particular (i.e. culturally embedded) ethical concepts, and the assumptions and aspirations that they carry, in historical and genealogical terms. The naturalist program, as advanced by Hume, is too general and universal in its orientation for this purpose. This difference between Hume’s naturalism and Williams’s genealogy is reflected in their approach to responsibility and their divergent understandings of the problems that we are faced with. The gap between Hume and Williams can be described in terms of three significant issues where they present their concerns in very different terms. On the interpretation provided, Hume plainly rejects the assumptions and aspirations behind ‘morality’s’ conception of responsibility—​and to this extent, as we have noted, he anticipates important features of Williams’s critique. Williams, however, is careful to emphasize that this requires us to ‘recast our ethical conceptions’, which are tainted and distorted by ‘morality’ (1985/​1995: 19). Taking this step suggests that, ethically speaking, we cannot ‘leave everything where it was’. In particular, contrary to what the ‘old compatibilists’ or ‘reconcilers’ have supposed, we cannot claim to vindicate our ‘ordinary’ views or assume that our current (modern, western) understanding of these matters is all in good order In contrast with this, Hume’s presentation of his views on ethics, as they concern responsibility, are less openly radical or revisionary. The general tenor of his remarks is to suggest that his account is more or less consistent with our ordinary or common-​sense views about such matters.52 To a considerable extent, the differences between Hume and Williams in this respect proceed ‘more from the manner than the matter’ of Hume’s views. Much of this can be explained in terms of Hume’s very different primary concern, which is to discredit religion and religious ethics without leaving himself open to the charge that he aims to undermine morality as such.53 Nevertheless, even when we allow for this, it remains true that Hume suggests a more ‘conservative’ picture than Williams with regard to the wider implications of his naturalistic understanding of moral responsibility. This is certainly a feature of Hume’s ethics that Williams is concerned to reject.54 51 See, for example, Hume, ‘A Dialogue’ (EM 314–​43). For further discussion of this, see Russell 52 See, in particular, Hume’s remarks at EU 8.23/​95, where he suggests that the whole free will debate ‘has been hitherto merely verbal’. 53 On this, see Russell (2018/​2021: 277–​81). 54 It is arguable that a similar gap—​in tone if not in substance—​opens up between P. F. Strawson’s views in ‘Freedom and Resentment’ and those of Williams. In contrast with Williams, Strawson [p. 206] When it comes to ‘recasting our ethical conceptions’, Williams is clear that, not only does this not leave our ordinary conceptions undisturbed, but these changes come at some cost. Proponents of ‘morality’ will be disposed to present the costs as taking the form of scepticism about ‘genuine’ or ‘true’ moral responsibility. Williams’s vindicatory genealogy aims to block and discredit that suggestion. This does not, however, entirely erase the taint of ‘pessimism’ that hangs over this effort to dispense with the apparatus of ‘morality’ and replace it with a more ‘realistic’ and ‘truthful’ account of moral responsibility. The relevant source of pessimism is based not on any kind of general scepticism about moral responsibility, but rather on the observation that moral responsibility is not immune to conditions of fate and luck. In light of this, we are invited to accept a kind of ‘pessimism of strength’ (in Nietzsche’s language) that involves recognizing that while we cannot, on one side, deny or repudiate our standing as responsible moral agents acting in the world, we also cannot deny that there are features and forces operating in the world that we cannot control but that, nevertheless, substantially shape and influence the ethical trajectories of our lives.55 While Hume’s own understanding of moral responsibility is subject to pessimistic reflections of this nature, Hume’s presentation of his ethical system tends to conceal these ‘darker’ implications of his views.56 Perhaps the most significant gap between Hume and Williams on this subject concerns the free will problem itself. Hume’s approach, particularly as presented in the sections concerned with ‘liberty and necessity’, is (famously) presented as a ‘reconciling project’ (Hume, EU 8.23/​95).57 This suggests, as most Hume commentators would agree, that Hume belongs squarely on the side of compatibilism. It is, on this reading, the compatibilist/​incompatibilist divide that Hume is primarily concerned with, as structured around the traditional free will problem. It is evident that this is not how Williams understands his own position on this subject. As Williams understands it, the division that matters here is not between compatibilist and incompatibilist but between those who are committed to ‘morality’ and those who reject it. The ‘old compatibilists’ accept the framework of the free will problem as presented by ‘the morality system’. They have, as their critics presents his naturalistic account of moral responsibility as essentially ‘descriptive’ and supportive of our ‘ordinary’ ethical conceptions. On this, see Russell (2024). 55 For a discussion of the wider relevance of this mode of ‘pessimism’ for the free will problem, see Russell (2017). Also of relevance here is Krishnan and Queloz (2023). They argue that Williams’s ‘pessimism of strength’ reveals more than just a commitment to view the world truthfully, without comforting illusions. Williams was, they suggest, a ‘shaken realist’, living in the shadow of the horrors of the Second World War and the late twentieth century. Arguably, considerations of this kind go some way to accounting for the distance that Williams perceives between his concerns and Hume’s. 56 Here too, it may be argued, the distance between Hume and Williams should not be exaggerated, since despite the (superficial) ‘optimistic’ tone of Hume’s presentation, the substance of his position is plainly in line with the more ‘pessimistic’ outlook that Williams describes. 57 The tone of the parallel discussion in the Treatise is somewhat different, emphasizing a defence of necessity against ‘the doctrine of liberty’ (Hume, T 2.3.2/​407–​12). But the difference is largely superficial, as a ‘reconciling project’ is implicit in both. [p. 207] Paul Russell suggest, failed to show ‘that our actual ethical notions are compatible with determinism’, which leaves them open to the charge that ‘they have changed the subject’ (Williams 1985/​1995: 6). This situation leaves Hume’s position straddled between the (morality-​oriented) aims of the ‘old compatibilists’ and the ‘recasting’ ambitions that Williams’s critique proposes. This is not entirely surprising, since Hume’s problem is not, primarily, with ‘the morality system’ but with religion and the (flawed) ethical systems that it suggests. Where do these observations concerning the relationship between Hume and Williams on the subject of moral responsibility leave us? There is, of course, an asymmetry in their relationship in that Hume serves to provide the foundations for much of Williams’s thinking about ethics, and this does not hold in the other direction. We can make sense of Hume’s relevance to Williams’s critique of ‘the morality system’ to the extent that he provides much of the basic (more ‘realistic’) psychological material that Williams draws from. The significance of this ‘Humean’ material is that Williams pushes it further to reveal the illusionary assumptions and aspirations that ‘morality’ relies on and aims to satisfy. Looking in the other direction, backwards from Williams to Hume, we can consider Hume’s ethics and naturalistic views about moral responsibility in terms that were not available to Hume but that serve, nevertheless, to illuminate what he has to say. More specifically, we can understand the way in which Hume’s concern to describe secularized ethics, grounded in human nature, involves rejecting many of the assumptions and prejudices of ‘morality’. Despite their different interests and concerns, and their different idioms, what Hume and Williams share is a fundamental commitment to providing a more ‘truthful’ and ‘realistic’ understanding of moral responsibility and our human ethical predicament.58 References Árdal, Páll S. 1966. Passion and Value in Hume’s Treatise. Edinburgh: Edinburgh University Press. Darwall, Stephen. 2006. The Second-​Person Standpoint: Morality, Respect and Accountability. Cambridge, MA: Harvard University Press. Dennett, Daniel. 1984. Elbow Room: The Varieties of Free Will Worth Wanting. Oxford: Clarendon Press. Duff, R. A. 2001. Punishment, Communication, and Community. Oxford: Oxford University Press. 58 I am grateful to the editors of this collection, Marcel van Ackeren and Matthieu Queloz, for the support, patience, and helpful comments and suggestions that they have provided. I should also mention, in this context, that my own thinking about both Hume’s philosophy and problems of free will and moral responsibility dates back to the early 1980s, when my PhD work at Cambridge was supervised by Bernard Williams. This was also the period when he was writing Ethics and the Limits of Philosophy. Suffice it to say that this chapter reflects my considerable debt to—​and respect for—​his philosophical contributions and influence. [p. 208] Hart, H. L. A. 1959/​1968. ‘Prolegomenon to the Principles of Punishment’. Proceedings of the Aristotelian Society 60: 1–​26. Reprinted in Hart, Punishment and Responsibility: Essays in the Philosophy of Law. Oxford: Clarendon Press, 1–​27. Hume, David. 1739–​40. A Treatise of Human Nature. Edited by D. F. Norton and M. J. Norton. Oxford: Oxford University Press (This edition includes _​An Abstract of A Treatise of Human Nature_​[1740], abbreviated as TA). References are also provided to the Selby-​Bigge/​Nidditch editions of the Treatise and Enquiries. Hume, David. 1748/​2000. An Enquiry concerning Human Understanding. Edited by T. L. Beauchamp. Oxford: Clarendon Press (abbreviated as EU). Hume, David. 1751/​1998. Enquiry concerning the Principles of Morals. Edited by T. L. Beauchamp. Oxford: Oxford University Press (abbreviated as EM). Hume, David. 1757/​2008. The Natural History of Religion. In A Dissertation on the Passions; The Natural History of Religion: A Critical Edition. Edited by T. Beauchamp. Oxford: Clarendon Press (abbreviated as NHR). Hume, David. 1777/​2021. Essays: Moral, Political, and Literary, 2 vols. Edited by T. L. Beauchamp and M. A. Box. Oxford: Clarendon Press. References are also provided to the edition by E. F. Miller. Indianapolis: Liberty Classics (abbreviated as ESY ). Krishnan, Nikhil, and Matthieu Queloz. 2023. ‘The Shaken Realist: Bernard Williams, the War, and Philosophy as Cultural Critique’. European Journal of Philosophy 31.1: 226–​47. Mackie, J. L. 1977. Ethics: Inventing Right and Wrong. Harmondsworth: Penguin. Nagel, Thomas. 1976/​2013. ‘Moral Luck’. Proceedings of the Aristotelian Society, Supplementary Volume 1: 137–​51. Reprinted in Russell and Deery (2013: 31–​42). Queloz, Matthieu. 2022. ‘A Shelter from Luck’. In Morality and Agency: Themes from Bernard Williams. Edited by A. Szigeti and M. Talbert, 184–​211. New York: Oxford University Press. Russell, Paul. 1995. Freedom and Moral Sentiment: Hume’s Way of Naturalizing Responsibility. New York: Oxford University Press. Russell, Paul. 2007. ‘Hume on Free Will’. In The Stanford Encyclopedia of Philosophy. Revised and 2020. Edited by Edward N. Zalta, . Russell, Paul. 2015/​2021. ‘Hume’s “Lengthy Digression”: Free Will in the Treatise’. In The Cambridge Companion to Hume’s Treatise. Edited by A. Butler and D. Ainslie, 230–​51. Cambridge: Cambridge University Press. Reprinted in Russell, Recasting Hume and Early Modern Philosophy: Selected Essay, 158–​84. New York: Oxford University Press. Russell, Paul. 2017. ‘Free Will Pessimism’. In Oxford Studies in Agency and Responsibility, vol. 4. Edited by David Shoemaker, 93–​120. Oxford: Oxford University Press. Reprinted in Russell, The Limits of Free Will, 243–​75. New York: Oxford University Press. Russell, Paul. 2018/​2021. ‘Hume’s Optimism and Williams’ Pessimism’. In Ethics beyond the Limits: New Essays on Bernard Williams’ Ethics and the Limits of Philosophy. Edited by S. G. Chappell and M. van Ackeren, 37–​52. London: Routledge. Reprinted in Russell, Recasting Hume and Early Modern Philosophy: Selected Essays, 265–​82. New York: Oxford University Press. Russell, Paul. 2022. ‘Free Will and the Tragic Predicament: Making Sense of Williams’. In Morality and Agency: Themes from Bernard Williams. Edited by A. Szigeti and M. Talbert, 163–​83. New York: Oxford University Press. Russell, Paul. 2024. ‘Responsibility after “Morality”: Strawson’s Naturalism and Williams’ Genealogy’. In P. F. Strawson and His Philosophical Legacy. Edited by S. Heyndels, A. Bengtson, and B. De Mesel, 234–​59. Oxford: Oxford University Press:. Schlick, Moritz. 1939/​1966. ‘When Is a Man Responsible?’ In Problems of Ethics. Translated by D. Rynin, 143–​56. New York: Prentice-​Hall. Reprinted in Free Will and Determinism. Edited by Bernard Berofsky, 54–​63. New York: Harper & Row. Strawson, P.F. 1962/​2013. ‘Freedom and Resentment’. Proceedings of the British Academy 48: 187–​ Reprinted in Russell and Deery (2013: 63–​83). Stroud, Barry. 1977. Hume. London: Routledge & Kegan Paul. [p. 209] Paul Russell Wallace, R. Jay. 1994. Responsibility and the Moral Sentiments. Cambridge, MA: Harvard University Press. Watson, Gary. 1987/​2013. ‘Responsibility and the Limits of Evil’. In Responsibility, Character and Emotions: New Essays on Moral Psychology. Edited by F. Schoeman, 256–​ 28. Cambridge: Cambridge University Press. Reprinted in Russell and Deery (2013: 84–​113). Williams, Bernard. 1971/​1973. ‘Morality and the Emotions’. In Morality and Moral Reasoning. Edited by John Casey. London: Methuen. Reprinted in Williams (1973: 207–​29). Williams, Bernard. 1973. Problems of the Self. Cambridge: Cambridge University Press. Williams, Bernard. 1976/​ Moral Luck’. Proceedings of the Aristotelian Society, Supplementary Volume 1: 115–​35. Reprinted in Williams (1981: 20–​39). Williams, Bernard. 1980/​1981. ‘Internal and External Reasons’. In Rational Action. Edited by R. Harrison, 000–​0. Cambridge: Cambridge University Press, 1980. Reprinted in Williams Williams, Bernard. 1981. Moral Luck: Philosophical Papers 1973–​1980. Cambridge: Cambridge University Press. Williams, Bernard. 1983. ‘The Uses of Philosophy: An Interview with Bernard Williams’. Centre Magazine, November/​December: 40–​9. Williams, Bernard. 1985/​2011. Ethics and the Limits of Philosophy. With a Commentary by A. W. Moore and a Foreword by J. Lear. London and New York: Routledge (Routledge Classics edition, 2011). Williams, Bernard. 1985/​1995. ‘How Free Does the Will Need to Be?’ Lindley Lecture, University of Kansas, 1985. Reprinted in Williams (1995a: 3–​21). Williams, Bernard. 1986. ‘Making Ends Meet’. Philosophical Books 27.4: 193–​208 (with Simon Blackburn). Williams, Bernard. 1989/​1995a. ‘Internal Reasons and the Obscurity of Blame’. Logos 10: 1–​11. Reprinted in Williams (1995a: 35–​45). Williams, Bernard. 1991/​1995a. ‘Making Sense of Humanity’. In The Boundaries of Humanity: Humans, Animals, Machines. Edited by J. Sheehan and M. Sosna. Berkeley: University of California Press. Reprinted in Williams (1995a: 79–​89). Williams, Bernard. 1993. Shame and Necessity. Berkeley, CA: University of California Press (abbreviated as SN). Williams, Bernard. 1993a/​1995a. ‘Nietzsche’s Minimalist Moral Psychology’. European Journal of Philosophy 1.1: 4–​14. Reprinted in Williams (1995a: 65–​76). Williams, Bernard. 1993b/​1995a. ‘Moral Luck: A Postscript’. In Moral Luck. Edited by D. Statman, Albany: State University of New York Press. Reprinted in Williams (1995a: 241–​7). Williams, Bernard. 1994. ‘Interview with Bernard Williams’. Cogito 8.1: 3–​19. Williams, Bernard. 1995a. Making Sense of Humanity and Other Philosophical Papers 1982–​1993. Cambridge: Cambridge University Press. Williams, Bernard. 1995b. ‘Ethics’. In Philosophy: A Guide Through the Subject. Edited by A. C. Grayling, 545–​82. Oxford: Oxford University Press. Williams, Bernard. 1995c. ‘Replies’. In World, Mind, and Ethics: Essays in the Ethical Philosophy of Bernard Williams. Edited by J. E. J. Altham and R. Harrison, 185–​224. Cambridge: Cambridge University Press. Williams, Bernard. 1999. ‘Seminar with Bernard Williams’. Ethical Perspectives 6.3–​4: 243–​65. Williams, Bernard. 2000. ‘Naturalism and Genealogy’. In Morality, Reflection and Ideology. Edited by E. Harcourt, 148–​161. Oxford: Oxford University Press. Williams, Bernard. 2002. Truth and Truthfulness: An Essay in Genealogy. Princeton: Princeton University Press. Williams, Bernard. 2002/​2004. ‘A Mistrustful Animal’. Interview in Harvard Review of Philosophy interview dated December 2002). Reprinted (in part) in Conversations on Ethics. Edited by Alex Voorhoeve, 195–​212. Oxford: Oxford University Press. Williams, Bernard. 2014. Essays and Reviews 1959–​2002. Foreword by Michael Wood. Princeton: Princeton University Press. [p. 210] The Predicament of Temporality Williams’s Challenges to Kant’s Practical Reason* Carla Bagnoli Williams credits Kant with having ‘given the purest, deepest, and most thorough representation of morality’ (1985: 174). This acknowledgement entails that there is an ordinary conception of morality,1 which is largely coincidental with the Kantian ethics of respect for humanity: ‘Kant’s view not only carries to the limit the notion that moral worth cannot depend on contingencies, but also emphasizes . . . the idea of respect which is owed to each man as a rational moral agent’ (Williams Williams relentlessly denounces the mortifying effects of this conception of moral worth, which rests on a conception of rational agency that is equally divorced from contingencies.2 His arguments have shaped the landscape of recent analytic ethics and made a crucial contribution to the scholarly reassessment of Kant’s ethics, giving an unprecedented impulse to neo-​K antian ethics in Anglo-​ American moral philosophy. While the debate has primarily focused on Williams’s attack on the supremacy of morality, this chapter claims that these arguments are best appreciated as being aimed at the underlying conception of practical reason. It thus chooses an unorthodox perspective from which to assess Williams’s dispute with Kant, refocusing on a dimension of contingency that has not received quite so much attention: temporality. Characters such as Elina Makropulos (Williams Anna Karenina, and Paul Gauguin (Williams 1981: 20–​39) are the best evidence that temporality is a central theme in Williams’s work. In contrast to the Kantian approach to temporality, Makropulos illustrates that mortality is the very condition for valuing life, while the tragic choices of Karenina and Gauguin raise issues regarding the modes and requisites of agential integrity. Here is the roadmap: in section 1, I explain the benefits of refocusing on temporality as a key aspect of Williams’s critique of Kant’s conception of practical reason. * Work on this chapter has been funded by the University of Modena and Reggio Emilia (FAR 2023, grant n. CUP E83C23002460001). T]‌he Kantian conception embodies, in a very pure form, something that is basic to our ideas of morality’ (Williams 1981: 36). See also Williams (1973: 235; 1993: 7, 91). 2 In his view, the philosophical disregard for emotions, individuality, and difference are all but a by-​product of a conception of reason divorced from contingency (Williams 1973: 225–​9; 1995a, b: 246, 1995c), see also Williams 1963. On the importance of emotions and affects in temporalizing one’s existence, and associated varieties of estrangement, see Moran 2022. Carla Bagnoli, The Predicament of Temporality In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. [p. 211] Carla Bagnoli In section 2, I argue that Williams’s rejection of the Kantian approach raises significant and intricate questions about diachronic rational agency, and I examine distinct arguments regarding (2.1) the structure of practical reflection and its deliverances, (2.2) the possibility of categorical desires that are at once temporal and unconditional, and (2.3) the consequent fragmentation of practical reason. In section 3, I explore different rejoinders, which build upon Kant’s claim that reasoning is generative or constructive, and which rely on the purported integration between practical reasoning and appropriate normative attitudes. In section 3.1, I illustrate Kant’s programme regarding the possibility of autonomous self-​integration via the adoption of the categorical imperative (CI), which is the constitutive standard of reason. I surmise that this constitutive norm is not limited to individual self-​ organization, but it generates the rational collective called humanity. In section I argue that Kant’s insight is that reasoning can be subversive in that it generates distinctive modes of motivation, which explain rational action. In section 3.3, I point to some practical presuppositions that explain human (individual and collective) engagement in action in spite of past failures. These arguments show that Williams’s criticisms of Kant are largely inconclusive, although they rule out some contemporary rendering of the Kantian project. 1. Thematizing the Predicament of Temporality This chapter argues that the disruptive force of Williams’s critique of Kantian ethics is best evidenced by bringing into focus the impact of temporality on rational agency and, consequently, the theory of practical reasoning. Temporality is a feature of contingency, but its role in this dispute is more important than it might first appear. For Kant, any empirical determination is contingent, and belongs to the phenomena, and since all phenomena are temporally structured, temporality must be a constitutive feature of them. Agents and actions are subjected to time in so far as they are phenomena, and temporality marks the phenomenology of practical choice. However, rational agents and rational actions cannot be seen entirely as mere phenomena: they are autonomous because and in so far as they are capable of reasoning.3 For Kant, temporality is a predicament to which embodied rational agents are exposed. It imposes several conditions that impact on rational choice: finitude, mortality, and fragility. These features of the human condition play a key role in human self-​representation and make humans vulnerable and sensitive to time in different ways. Our interests, concerns, personal relations, and fundamental projects arise, develop, and decline in ways that threaten to 3 Reasoning emancipates humans from the influences of alien causes (G 4: 440–​1, 446). Autonomy consists of the capacity to be motivated independently of needs and desires, and even in contrast to them (Allison 1990: 97; Nagel 1970). [p. 212] undermine our diachronic integrity and agential authority. Kant does not believe that this threat can be dismissed or overcome once and for all, but that humans—​ animals endowed with reason—​can and should actively engage with it, by engaging in practical reasoning. Thus, the practice of reasoning represents the best response to the predicament of temporality, and the only way in which humans can achieve autonomy over time is by engaging in the activity of rational self-​integration. According to Williams, in order to warrant the requisite autonomy, Kant consistently pursued the view that ‘the source of moral thought and action must be located outside the empirically conditioned self ’ (Williams 1973: 228). On this reading, the very notion of practical choice is predicated on the alleged atemporality of reason: the ascription of any empirical determination would make agency structured by the category of time. In so far as Williams objects against the purity of Kant’s conception of rational agency, he also rejects its atemporality and astoricity. Therefore, my change in focus from contingency to temporality does no violence to Williams’s argument. On the contrary, this strategy brings to the fore some crucial aspects of Williams’s dispute with Kant that have been underappreciated precisely because of the failure to properly thematize temporality. While the purpose of this chapter is not exegetical, the strategy of centring on temporality brings numerous benefits to Williams’s argument. First, it identifies a fundamental point of disagreement: while Kant considers temporality as a (practical and cognitive) predicament to which embodied rational agents are liable, Williams appreciates temporality as the condition upon which the meaning of life is based. Secondly, it accounts for Williams’s reluctance to accept the ‘mystery’ of rational action, which happens over time but results from the legislative activity of pure rational will (Williams 1995b: 15n). Thirdly, it highlights that Williams is sensitive to the complex ways in which temporality affects human experience, and which—​in his view—​elude the Kantian approach. Human agents are embodied, temporally situated, and subjected to temporal constraints. However, they also represent themselves as extended through time; their choices are deliberated in the present, originated in the past, and projected into the future (Williams Temporality has disaggregating effects on practical reason, but it is also the condition upon which the meaning of life rests. Williams’s ultimate point is that recognizing the temporal structure of human willing and agency requires a perspectival account of practical reasons, which is the view that practical reasons apply from specific temporal perspectives. In the face of the multifaceted experience of temporality, the model of the pure will seems both illusory and self-​ deceptive (Williams 1995b). A further advantage of this argumentative strategy is that it relates apparently different strands of Williams’s resistance to Kantian philosophy: the critique of the categoricity and unconditionality of moral obligations, and the disregard for emotions (Williams 1966/​1973: 207, 225–​9), personal attachments and fundamental personal projects (Williams 1981). According to Williams, Kant’s ‘universalistic [p. 213] Carla Bagnoli and legalistic’ conception of morality neglects these elements, which are central to the good life, and profoundly shaped by time. These different critiques run in parallel as they all track the alleged atemporality of the normative principles that are constitutive of the exercise of reason. While, for Kant, these principles apply to embodied as well as to disembodied rational agents, Williams urges that valuing human life is rooted in desires that enjoy a temporal dimension, and whose existence relates to embodiment. Williams’s critique apparently echoes a longstanding debate regarding the alleged atemporality of the Kantian paradigm of rational agency (Williams 1985: 104, 184, 197).4 Kantian rejoinders are based on an alternative conception of the role of the normative standard of reason, which organizes agency through time. 2. Williams’s Critique of the Atemporal Model of Rational Agency In Ethics and the Limits of Philosophy, Williams praises the Groundwork as an ‘extraordinary book’ and ‘the most significant work of moral philosophy after Aristotle’ (1985: 55), and yet it constitutes the main target of his criticisms.5 Ultimately, Williams denies that the faculty of reason can entail all the benefits promised by Kant. He contests universality as the norm of reason and doubts that self-​legislation is the organizing form of rational willing and rational agency. The appeal to universality is questioned in its formality, but also in its capacity to produce normative reasons for action. The problem arises as early as at the stage of the agent’s self-​representation: ‘Why should I think of myself as a legislator and—​since there is no distinction—​at the same time a citizen of a republic governed by these notional laws? . . . The argument needs to tell us what it is about rational agents that requires them to form this conception of themselves as, so to speak, abstract citizens’ (Williams 1985: 63). One way to respond to Williams’s line of questioning is to say that Kant’s normative model is confined to ‘rational agents as such’;6 but this cannot be the definitive answer if ethical theory is to address human agents. Humans are rational, but they are not ‘rational and no more’ (Williams 1985: 63). The expression ‘rational agents 4 See Hegel’s critique of Kant’s conception of reason as a ‘dimensionless activity, as a pure concept of infinitude’ which is ‘held fast in opposition to the finite’ (Hegel 1802). Allison notes Williams’s deep affinities with Hegel’s critique, but does not comment on temporality (see Allison 1990: 47–​53, 81, 191–​ Williams’s resistance to Kantian moral philosophy is in line with Murdoch’s comment that ‘The centre of this type of post-​K antian moral philosophy is the notion of the will as the creator of value’ While it is doubtful that the slogan captures Kant’s conception of the rational will, it is certainly apt to describe recent variants of Kantian ethics. 5 Analytic philosophy has predominantly considered the Groundwork as the paradigm of Kantian ethics, and neo-​K antians have taken this work as the main measure of reference. 6 ‘The normativity of reasons requires the perspective of a rational agent as such as the standpoint from which all reasons, including those grounded on what motivates an agent from his own point of view, are ultimately assessed’ (Darwall 1983: 113). [p. 214] as such’ is almost unintelligible, since ‘there is no way to be a rational agent and no Williams’s dissatisfaction with the formality and abstractness of Kant’s characterization of humanity in terms of the capacity for rational self-​determination extends to contemporary philosophers inspired by Kant who have not taken advantage of the Hegelian objection that Kant’s practical reason is a universal and empty form, temporally unhinged and ahistorical (Williams 1985: 104, 184, n16). However, this dissatisfaction concerns, basically, the methods of transcendental philosophy, which is regarded as a shattering failure (Williams Thus, a pressing question is whether and how contemporary neo-​K antians meet the challenge of proving that rational agents can represent themselves as citizens and legislators of the notional republic, in a way that ‘does not bring back the more extravagant metaphysical luggage of the noumenal self ’ (Williams 1985: 65).7 Starting with John Rawls, the main strategy to respond to the metaphysical queerness associated with Kant’s conception of practical reason is to leave metaphysics behind and to defend Kant’s moral theory at normative level entirely (Rawls 1980), but Williams quickly rules out this methodological option, on the basis that it fails to vindicate the distinctive ambition of Kant’s programme. If there is any hope of carrying on Kant’s project, it must start with Kant’s presumptions about the metaphysics of rational agency (Williams 1985: 64). Critique of ‘Reflective Freedom’ Since Kant’s metaphysics and its contemporary normative correctives are both considered to be unpromising, Williams engages with Thomas Nagel’s claim that bases practical freedom on the capacity of self-​consciousness and reflection. Nagel believes that the acknowledgement of the reality of others is tantamount to the awareness of equal moral standing, which depends on the agent’s metaphysical conception of him/​herself (Nagel 1970: 14).8 This metaphysical conception translates into an epistemic awareness of the equal value of oneself as well as others. The opposite is a form of practical solipsism. Williams’s assessment of this argument assumes that the kind of freedom entailed by the exercise of rational agency is ‘reflective freedom as a thinker’ (Williams According to Williams, the appeal to the reflective stance trades on an equivocation between factual and practical deliberation, and ultimately fails to 7 Williams’s point is reminiscent of Mackie’s objection of queerness, which is applied to Kant as much as to G. E. Moore (see Mackie 1977: 40 ff.; cf. Bagnoli 2015). 8 ‘If he acts on reasons, then he must not only be an agent but reflect on himself as an agent, and this involves his seeing himself as one agent among others’ (Williams 1985: 65, see also Williams 1995a). [p. 215] Carla Bagnoli account for rational action. Kant himself has insisted on the risk of this equivocation (cf. section 3.1 below). Williams grounds his argument on the alleged asymmetry between factual and practical deliberation. In factual deliberation, the reflective self is unified with the pre-​reflective self; they are both committed to seeking truth. Reflection imposes corrections and revisions of pre-​reflective beliefs, but this is a process informed by the constitutive aim of truth, and the selection of beliefs in the face of evidence is a straightforward matter of rationality. In practical deliberation, instead, the reflective self radically differs from the pre-​ reflective self. First, the former is not an agent at all, while the latter is a particular agent faced with a particular deliberative predicament. The deliberator belongs in the context of choice, suffers the pressure of time, and is divided between different clashing desires, interests, and concerns, all of which demand to be satisfied. Deliberation aims to rank and order these demands so that the agent can act on the best reasons. Supposedly, reflection assists in this practical task by requiring the agent to detach from all his/​her current desires and to suspend their normative and motivational pressure. Williams believes that this marks a second crucial difference between the pre-​reflective and the reflective self, in that the reflective agent acquires totally different reasons for action. Furthermore, Williams argues, the reasons acquired via reflective detachment aspire to be practical and action guiding, but it is doubtful that they actually are, as they are not anchored in the agent’s current deliberative set. This is a third element that undermines the practical significance of reflective detachment. The reasons that drive the pre-​reflective self are grounded in his/​her deliberative set, while the ones that are reached by mere reflection have no such grounding and thus no motivational support. The latter claim is integral to the entire argument and stands in stark contrast to the Kantian view (cf. section 3.2 below).9 Finally, Williams remarks that practical deliberation is first-​personal in a way that factual deliberation is not. It is precisely this first-​personal feature that warrants the practical significance of reasons for action: ‘Practical deliberation is in every case first-​personal, and the first person is not derivative or naturally replaced by anyone’ (Williams 1985: 66). On the surface, this can be seen as a point of agreement with Kantians.10 However, for Williams, the first-​personal mode is also and necessarily agent-​relative: it is rooted in particular interests, concerns, desires, beliefs, and outlooks, and therefore—​claims Williams—​is anchored in the agent’s present deliberative set. Such perspectival elements belong in the contents of the 9 This is hardly Williams’s finding: Kant himself has formulated the problem of how a purely speculative principle can motivate. 10 As rational deliberators ‘our relation to our actions and choices is essentially authorial: from it, we view them as our own’ (Korsgaard 1996a: 378). Williams brings together the claim that deliberation is first-​personal, in the sense that it is authoritative in the first person, and the claim that it is personal or agent-​relative. [p. 216] agent’s reasons for action, which play a pivotal role in explaining how such contents can guide the agent in action. On the basis of these considerations, Williams claims a decisive theoretical advantage over the Kantian approach, as his proposal is better suited to explaining the practical significance of reasons. If this argument succeeds, it is a definitive blow to the Kantian approach, since Kant claims the transcendental method is the only one capable of identifying reasons that are objective and authoritative in the first person, as the unique way of vindicating the objectivity and normative authority of moral obligations (G 4: 441–​4).11 It should be noted, however, that the comparative merits of the two models should be assessed considering also the evolution of Williams’s reflection on the nature of practical reasons. While he holds that all the reasons ascriptions true of an agent are made true by the agent’s current deliberative set, he offers different specifications of what he means by ‘deliberative set’ (cf. Williams 1979/​1981, 1989/​ b, 2001). Williams’s settled position is that: ‘A has a reason to ϕ only if there is a sound deliberative route from A’s subjective motivational set . . . to A’s ϕ-​ing’ This formulation captures the relevant connection between an agent’s reason to ϕ and a ‘motive which will be served or furthered by his ϕ-​ing’, which is crucial to Williams’s understanding of internalism about reasons (1979/​1981: 101), and to his critique of Kant (Williams 1985). However, it also adds a complication that renders his internalism less straightforward than in its first stark proposition. This is because the qualification of soundness introduces a considerable ambiguity regarding the content of the subjective deliberative set of reasons. Presumably, the subjective motivational set is expanded to include all those reasons that could potentially explain something that this particular agent does in the circumstances, were she to deliberate soundly from her current ‘S’ (subjective motivational set). Interestingly, that equivalence need not be something known to an agent. This is to say that the counterfactual is true of the agent, now, but not its truth-​makers given that they would be made true by some process of deliberation in which the agent has (ex hypothesi) not yet engaged. What warrants the relation that explains action is, however, the present motive, and thus the issue of temporality is of first importance.12 On the other hand, the qualification of soundness softens the disagreement between Williams’s and the Kantian perspective on the nature of practical reasons. Indeed, in recent debates, Williams’s argument based on the distinction between practical and speculative deliberation has met a curious fate. It has become something of an asset to a prominent variety of Kantian ethics championed by Christine M. Korsgaard. Korsgaard repeatedly uses this argument in an attempt 11 For Williams, authority derives from conjoined categoricalness and self-​address (Williams 12 I would like to thank a referee for pressing me to clarify this point. [p. 217] Carla Bagnoli to distinguish constructivist and realist conceptions of Kantian ethics (Korsgaard b: ch. 4; 2002).13 Correspondingly, Williams’s critique has generated a divide between different ways of understanding and carrying on Kant’s legacy.14 Furthermore, Kantians have elaborated further on the importance of partial perspectives for an objectivist account of rational action.15 In an absolute perspective, in which one is deprived of all perspectival mental states and attitudes, particular interests and desires, there is no drive to agency. Humans may have a hard time constructing reasons sub specie aeternitatis, in so far as this endeavour requires absolute detachment from anything they care about, and which is for them alienating and ultimately meaningless. Nevertheless, there is a deeper problem: the complete success of reflective detachment involves the denial of any subjective trait and, with it, the suppression of any drive to action, which makes this device inapt for making sense of rational action. The Temporality of Categorical Desires A second argument against Kant’s morality concerns the temporality of categorical desires. Kant argues that moral obligation is a unique form of practical necessity by which humans respond to the predicament of temporality, which is a threat of desegregation. By contrast, Williams argues that categoricalness is not distinctive to moral obligations and, more importantly, that it is inextricably linked to temporality. First, temporality is not a plight, but rather the condition that makes sense of the way human life is infused with meaning and value, by way of the dynamics of desire. Desires propel the agent into the future and are the building block of a strategy aimed at dismantling Kant’s theory of practical reason. Secondly, the temporality of desires explains the phenomenology of value and choice, in ways that evade the Kantian framework: that is, by displacing the centrality of moral obligations. These claims are established via the power of example.16 The first proposition is based on Makropulos’s case, which imposes a reflection on the tedium of immortality. For the sake of argument, I focus on the point that immortality is intolerable because it is meaningless, and it is meaningless because immortality necessarily muffles and eventually extinguishes the propelling force of desire.17 Williams identifies categorical desires as the source of meaning and 13 Korsgaard believes that the epistemological model is misguided when applied to ethics (see Korsgaard 2009: 302–​26; 2008a: 234, 30–​1, 55–​7, 67–​8). For a different assessment of Kant’s conception of practical knowledge, cf. Engstrom (2009, 2013); Bagnoli (2013b). 14 The very distinction between realism and Kantian constructivism has been questioned on different grounds (see Bagnoli 2022). 15 Nagel (1986) is perceptive of this problem, and replicates an issue raised by Kant. 16 For a thematization of the use of examples in Kant, see O’Neill (1986). Immortality, or a state without death, would be meaningless, I shall suggest; so, in a sense, death gives the meaning to life’ (Williams 1973: 82). [p. 218] the drives of future agency, which ‘propel forward into longer life’ (1973: 86–​7, What makes desires categorical is not that their objects are special (e.g. moral interests), but rather that they play a grounding role: ‘some kinds of desires are unconditional or categorical in a distinctive sense, they are fully categorical, and one’s existence itself wanted as something necessary to them’ (1973: 86). The desirability of life supplants the transcendence of reason: ‘The question of life being desirable is certainly transcendental in the most modest sense, in that it gets by far its best answer in never being asked at all’ (1973: 87). The undesirability of immortality is not merely contingent, but necessarily related to the dynamics of desire.18 Rather than the appeal to a transcendent (and allegedly atemporal) reason, it is a human psychological mechanism that explains how categorical desires lose force. Boredom is an effect of the absence of change, which is to say that the categoricalness of desires is bound up with temporality.19 Thus, Williams’s notion of categoricalness does not serve the purpose of responding to the predicament of temporality by allowing for agential authority over time. Kant uses categoricalness to make reasons invulnerable to time: moral obligations are inescapable because and in so far as they are rational requirements. By contrast, Williams relates the categoricalness of desires to their temporality, and shows that moral obligations can be divisive and alienating rather than providing the unifying ground of human identity.20 The effect of his argument is to denounce not so much the extraordinary demandingness of morality as its deeply alienating effects, which depend on its being disconnected from the agent’s real desires, concerns, and projects. These effects, along with Williams’s second claim regarding the role of temporality in practical reasoning, are brought to the fore by a broad repertoire of difficult choices. Gauguin and Karenina are required to deliberate while driven by the propelling force of categorical desires that diverge from moral obligation (Williams These agents experience a conflict between reasons grounded in their current personal projects, which they regard at the time as the foundational source of meaning of their lives, and reasons deriving from past commitments and obligations to others. This conflict is divisive because moral obligations In EM’s case, her boredom and distance from life both kill desire and consist in the death of it’ (Williams 1973: 91). 19 Likewise, Millgram (2004) argues that (1) all final ends are bound to become boring after a time, and (2) this is not regrettable. Correspondingly, he considers the capacity for boredom as being among the top-​level components of rationality. With regard to the inescapability of moral obligations, see Queloz (2022: 195–​6). 20 Cf. Korsgaard (1996b: 102). The limiting case of this might be that the promised life held out some hope just to that desire mentioned before, that future desires of mine will be born and satisfied; but if that were the only categorical desire that carried me forward into it, at least this seems demanded, that any image I have of those future desires should make it comprehensible to me how in terms of my character they could be my desires’ (Williams 1973: 91–​2). [p. 219] Carla Bagnoli command with unconditional authority, but they seem unreasonable in the deliberative stance. Gauguin’s and Karenina’s categorical desires have arisen over time and, possibly, will decline over time. However, while they are in force, they infuse their lives with categorical meaning and thus undermine the rational authority of moral obligations towards their families.22 Note that neither Gauguin nor Karenina has any expectation that their desires should diminish the authority and importance of moral obligations; thus, they feel bound by categorical forces that are pulling them in different directions. Apparently, the main culprit is the Kantian claim that moral obligations are the only kind of practical necessity, categorical imperatives, which hold true for all rational agents at all times because their authority is grounded on reason.23 However, this is only a consequence of a broader claim that highlights the fragmentation of practical reason, the different sources of which exert different forms of authority across time. At the same time, Williams unveils a weakness specific of the Kantian project. If moral obligations lack rational authority, then their authority must reside in social power and external enforcement by way of incentives and sanctions—​something at odds with Kant’s claim about the autonomy of reason. This is how the critique of the sovereignty of moral obligations relates to and culminates with the critique of morality as a peculiar institution centred on the ‘system of blame’ (Williams 1985: 177).24 However, the driving point is that by advocating for the plural sources of practical necessity, Williams opens the possibility of radical conflicts within the domain of reasons. 22 The significance of this essay is generally appreciated as a contribution to the issue of moral luck, but the temporal aspect is crucial to the phenomenology of choice: ‘Momentous decisions such as those of Gauguin and Anna Karenina . . . are especially dramatic examples of this simple truth, cases in which an agent’s choices set in motion causal processes whose final shape cannot be known at the time when they are made, where those processes in turn determine the standpoint from which the agent will later look back on them’ (Wallace 2003: 137). Kant understood practical necessity in terms of a reason for action that was not conditional on any desire at all, and he thought that there could be such a thing because he believed that the reasons of morality were based on reason alone. That is why he identified practical necessity as being uniquely moral necessity, and why, for him, unconditioned possibility and unconditioned necessity ultimately coincide, so that he could be led to say that the only truly free acts were those done for the sake of duty’ (Williams 1995b: 17). ‘Practical necessity is in no way peculiar to ethics. . . . The fundamental point is that a conclusion of practical necessity is the same sort of conclusion whether it is grounded on ethical reasons or not. . . . [P]‌ractical necessity, even when it is grounded on ethical reasons, does not necessarily signal an obligation’ (Williams 1985: 188). Williams aligns with Murdoch on this point. ‘If I attend properly I will have no choices and this is the ultimate condition to be aimed at . . . The ideal situation . . . is to be represented as a kind of “necessity” ’ (Murdoch 1997: 331). See also Frankfurt (1999). The institution of blame is best seen as involving a fiction, by which we treat the agent as one for whom the relevant ethical reasons are reasons. . . . This fiction has various functions. One is that if we treat the agent as someone who gives weight to ethical reasons, this may help to make him into such a person. . . . The fiction of the deliberative community is one of the positive achievements of the morality system’ (Williams 1985: 193). See Queloz (2022); cf. Bagnoli (2011) on Kant’s argument that sanctions and emotions take away from the obligatoriness of moral obligation. [p. 220] The Fragmentation of Practical Reason and Temporally Anchored Deliberation Williams’s novel conception of categoricalness provides the materials for a third argument that targets Kant’s conception of rational deliberation. While Kant emphasizes the lawful and unifying activity of reasoning, Williams’s view is that practical reason is fragmented, and its motivational force derives from categorical desires that are effective in the present. The resulting view is that rational deliberation is anchored in the present deliberative set. In his view, then, reasons constructed via the appeal to the reflective stance fail to account for rational action precisely because and in so far as they are temporally disconnected. His whole argument rests on the assumption that in order for reasons to be action-​guiding they must be present or rooted in the agent’s current deliberative set. This assumption explains the nature of the conflict experienced by Karenina and Gauguin. These are best described as intrapersonal diachronic conflicts of values, due to the fact that the agent acquired different categorical desires over time, and thus developed radically different views of what is, for them, a life worth living.25 How should intertemporal conflicts be assessed? In the limiting cases, such as that of Karenina, it is no longer clear why past reasons (to stand by her husband and son) are rationally overriding, given her unconditional love for Vronsky. Williams’s view is that the ‘correct perspective on one’s life is from now’ (1981: 13).26 This affirmation establishes a temporal bias to the present which is built into the very idea of intentional action. If reasons have a hold over the agent only when they are anchored in the present deliberative set, then categorical desires are not really categorical in any interesting sense: there is nothing beyond the fact that for some time, they infuse life with meaning, and for that period of time they provide the basis for action. These cases represent a peculiar challenge to Kant’s conception of morality because they are not morally indifferent but experience the burden of moral obligations as external constraints that threaten to undermine the meaning of their life.27 On Williams’s construal of the predicament, in this case, moral obligations (e.g. to stand by the family) lose not only their motivating force but also their supreme authority because they have no normative grip on the agent. These agents are fragmented across time: their diachronic integrity—​rather than their morality—​is at stake. At least in its practical function, reason fails to ensure the unity of reasons. The principles of reason may still have a synchronic function, and provide agential 25 From within Williams’s theory of reasons, analogous changes may be due to self-​correction: that is, the engagement with a sound deliberative route that corrects previous inconsistencies, cf. section 2.1 on the qualification of ‘soundness’. 26 The contrary view treats bias to the present as a form of imprudence (see Korsgaard 2014: 7). 27 In general, it is questionable whether Kant took the threat of moral scepticism as seriously and saw it in the same way as contemporary philosophers do (see Timmermann 2007, 129–​30; Stern 2013). [p. 221] Carla Bagnoli unity at any one time, but they do not have a unifying function over time. Most importantly, synchronic unity is predicated on the contingency that past deliverances of reasons are coherent with current motives. Practical reasons are as sensitive to time as desires are. It follows that practical reason defines a heterogeneous and fragmented domain, in which no single paramount kind of interest has the utmost importance and thus deliberative priority across time. The predicament of temporality is the predicament of having even a single practical reason being unified throughout one’s life. The cost of this approach, which privileges the present deliberative stance, is the loss of diachronic rational authority. Nagel objects that Williams’s approach fails to make sense of action at a distance, across time and across people. Williams’s point is that this is a cost that should be borne, as it promises an explanatory advantage.28 By referring to reasons that accord with the present ranking we are in a better position to explain the relationship that rational agents hold with themselves and their actions, because and in so far as this ranking relates to the present agent’s motivational set. The option of searching for absolute reasons is more costly as it forgoes the explanation of acting on reasons. 3. The Kantian Project of Rational Self-​Integration In the following sections, I present some rejoinders to Williams’s arguments. They are intended to establish that the previous critiques of Kant are misplaced, although they may still have some bearing on some recent neo-​K antian theories of reflective detachment. In fact, Kant does not propose reflective detachment as such, and he is adamant that practical and epistemic deliberation differ.29 In fact, he focuses on the same concerns as Williams (cf. Reflexionen; Allison 1990: 195). To gain an adequate appreciation of this difference, Kant holds, we need to understand the specific role played by the norm of universality in speculative and practical cognition, respectively. Similarly, Kant’s view is that reasoning is generative of a distinctive moral incentive, known as respect.30 In section 3.1, I illustrate the role of the CI as a mode of autonomous organization, rather than an exercise of detachment from the influences of desires. I surmise that this constitutive norm 28 Williams acknowledges this cost in rejecting Rawls’s model of the agent as a ‘trustee of one’s own future self ’ (Williams 1981: 49). 29 Kant does acknowledge a close kinship between practical and epistemic spontaneity (see Allison ch. 2). However, the practical function of reason is distinctive, and it can be appreciated instead in the ‘direction of existential dependence’ (Engstrom 2009: 119). In the case of theoretical cognition, the object of knowledge affects the mind, and thus it is ‘given from elsewhere’ (2009: 119). Conversely, objects of practical cognition (or objects of the will) are produced by the very act of cognition, and thus they do not exist prior to and independently of the act of cognition. 30 I use the term ‘incentive’ as a subjective determining ground of a will, which does not yet conform to the objective law (C2 5: 72, 74). [p. 222] is not limited to individual self-​organization, but generates the rational collective called ‘humanity’. In section 3.2, I argue that Kant’s insight is that reasoning can be subversive in that it generates distinctive modes of motivation, which explain rational action. In section 3.3, I point to some practical presuppositions that explain human (individual and collective) engagement in action despite past failures. The CI as the Form of Autonomous Self-​Organization A common thread among neo-​K antian approaches to practical reason is the appeal to ‘reflective detachment’ as a source of rational authority.31 In a way, this is not surprising because Kant focuses on cases in which moral motivation arises from the act of discarding desires and demoting self-​interest (G 4: 440). However, it is misleading to characterize these reflective modes in terms of detachment as the purpose of engaging in them is not merely to remove oneself from current motives, but to organize one’s conflicted mind and thus constitute it in action autonomously. Practical reflection is aimed at action, and action requires the elimination of any conflict between motives. However, this necessity is not only pragmatic, since rational agents aspire to integrity, and not only express, since rational agents also constitute themselves in action.32 Arguably, practical reflection is intended to organize one’s agency in time and across time, not only as individuals (uti singuli), but as eligible members of the kingdom of ends. These complex functions are all fulfilled by conforming to the CI, which is not merely a deliberative test, but the very form of reflection. It is the constitutive norm that governs the very activity of practical thinking, which basically consists of the agent’s activity of self-​constitution, a form of mental representation that is both efficacious and increasingly self-​transparent. Persons constitute themselves as persons per se through rational activity that is public (Engstrom 2013: 148; see Bagnoli 2013b: 171; 2017). If the main point of the CI is self-​constitution, then it is reductive and misleading to understand it as a form of synchronic individual deliberation. In fact, this norm attends both the diachronic and collective dimensions of human agency. According to Kant, human beings ‘necessarily represent’ themselves as rational ends in themselves, in so far as they hold that rational nature is an end in itself, and they represent themselves as rational (G 4: 429). This subjective principle of 31 The notion of reflective detachment is absent in Nagel (1970), while it is the main polemical target in Williams (1985). It is also questionable that this is the central device at work in the process of objectivity in Nagel (1986). However, reflective endorsement is a key mechanism at work in Korsgaard b) as ‘our capacity to turn our attention on to our own mental activities is also a capacity to distance ourselves from them, and to call them into question’ (1996b: 92–​3, 129, 229–​31; cf. Korsgaard Williams 1996; Nagel 1996). It remains questionable whether this mechanism can adequately serve as the source of agential authority without a subsequent stage of reflective endorsement. 32 Unity of agency is a practical need and ‘it is not based on a metaphysical theory, nor on a unity of which you are conscious’ (Korsgaard 1996a: 369, 370–​3, 378; 1996b: chs 3–​4). [p. 223] Carla Bagnoli human action commits them to treat all rational agents the same, but also to consider irrelevant the particular temporal instantiations of rational agency, in so far as the mere temporal location is morally irrelevant. Thus, the CI constrains deliberation in two ways: across time and across persons (in dealing with others and with ourselves at future times).33 Both diachronic and collective human agency require the work of reasoning as we cannot expect shared rationality to arise naturally: that is, without the exercise of rational will. However, the constraining effects of the CI do not have any alienating consequences. Importantly, incentives are incorporated into the subjective maxims of action.34 The normative power of the CI consists of transforming the subjective maxim into a law that can be accepted by a totality of rational wills. This is the device by which self-​legislation is also and at the same time co-​legislation in the temporally extended and socialized sense which represents the structure of the kingdom of ends, and thus always includes dealings with all rational agents and across time. Secondly, this understanding of the CI shows that some objections against the allegedly individualistic features of Kant’s humanity are misplaced. Following in the footsteps of Hegelian critics of Kant, Williams remarks that the emphasis on the ‘purely moral’ conceals the fact that ethical life lies outside the individual However, the main function of the CI is not to set aside the urgency of present desires and allow for purity of the will, but to select motives that can form the stable basis of shared rational agency, and around which the notion of humanity is built. Humans engage in practical reasoning to determine what is rational to do independently of the pressure of current desires because they represent themselves as finite but extended over time. They care about their future, and the future of what they care about. The CI warrants a sort of rational action that is not determined by its historical root or by the impellent desires and needs of the moment. Importantly, its function is not only to achieve mere consistency in time, or a form of stability that relies on the contingent stability of character, or on traditions and habits. Rather, the CI attains a distinctive form of stability, marked by autonomy. To support this claim, Kant’s conception of practical reason presupposes both a consistent relationship with and an abstraction from time. Human agents act in time, and necessarily so. Thus, action is situated in a temporal succession. On the other hand, to act freely is to act without any determining link with previous motivations. For an action to be free, the past may incline to but not determine action. Rather, free action is originated by the activity of agency. Importantly, this origination is not unruly, but it is governed by norms whose authority does not 33 See Korsgaard (2021: 186). 34 As Allison notices, Williams’s objection of alienation gains traction proportionally to discounting the ‘incorporation thesis’ (see Allison 1990: 192). According to Kant, we must ‘incorporate’ our incentives into our maxims if we are to act on them. [p. 224] depend on temporal constraints. Such norms are at once constitutive of the exercise of rational agency which happens in time and is not conditioned by temporality. However, action must also be explicable in terms of its relation to future and past states, but the relation between an agent and his/​her action does not belong in time just because it is a normative relation of authority.35 While the work of practical reason is not in itself subject to temporal relations, its operation is a matter of setting aside essentially temporal items and establishing connections through time—​so that practical reason ultimately has an effect in time, and its main function is to allow agents to achieve integrity over time. The Subversive Effect of Reasoning Williams presses the case that the predicament of temporality arises because agents are temporal subjects, and their reasons appear to be temporal items which, like their needs and desires, are destined to vanish. Kant’s main rejoinder is based on the categorical imperative, which captures the requisites of rational action that do not depend on arbitrary or external factors. However, the predicament of temporality puts pressure on Kantian theory to demonstrate how practical reasons that are not sensitive to time can be motivating. This concern is not allayed by the previous explanation about the constitutive function of the CI. Williams instils doubts that universal and temporally unanchored reasons may be of any practical guidance: he points to cases in which the CI has alienating effects, rather than working towards self-​integration. Kant acknowledges the gap created by reflection but opposes Williams’s view that rational deliberation must be temporally anchored in order to be action-​ guiding, which would lead to a form of imprudence.36 However, Kant’s practical reflection differs from reflective detachment in so far as it produces a moral incentive, which reorients the will away from self-​love, and thus furthers attention to others. Kant’s argument is that practical reasoning generates a distinctive moral incentive, named respect, which carries the subjective awareness of the moral law (G 4: 400, 401; C2 5: 71–​89).37 The first important function of respect is precisely to provide the sort of first-​personal authority that Williams found lacking in Kant’s account of moral obligations. 35 See Kant’s discussion of the third antinomy (A536/​B564, A445/​B473; cf. Wood 1984). On the interpretation of Allison (1990: 25–​8), the apparent oddity is resolved by reference to two criteria: the explicability of the unity of experience as an ordered flow (A446/​B474) and the activity of rational agency, which is not subjected to the order of time (A551–​2/​B579–​80). The compatibility of such criteria depends on the success of transcendental idealism. In almost anyone’s account of imprudence, present desire uses its unchallenged incumbency to silence representatives of the future self ’ (Korsgaard 2014: 7). 37 I leave aside here the complications associated with this claim but see Bagnoli (2021). [p. 225] Carla Bagnoli Secondly, respect marks a sort of moral motivation distinctive to the finite rational agent (C2 5: 79, 82), whose will does not naturally conform to the objective law (C2 5: 72, 74), and for whom reasoning proves necessary.38 The normative role of respect is subversive, in that it can contrast desires and inclinations that exist prior to engaging in reasoning and set them aside as lacking normative authority. Respect is the moral feeling that subverts the ranking of maxims, and allows the human agent to act on reasons, even in the presence of contrary motives. Thus, respect establishes the possibility of moral action irrespective of the current pressure of desires. This is the counterpart of the claim that establishes the normative power of reasoning in order to counteract past patterns, habits, and desires. Thirdly, in deliberating about future action as much as in deliberating with others, agents must presume that reasoning has this subversive effect, and does not merely confirm past patterns. This normative power establishes the practical freedom from one’s past, a crucial dimension of autonomous (individual and collective) agency. Respect is tantamount to the subjective experience of autonomy, and it is built into one’s self-​representation as an autonomous agent. This is the straightforward answer to Williams’s issue about being a citizen of the notional republic: such a description is operative in humans because it belongs in the first-​ personal conception of oneself as an autonomous agent. Due to this difference, Kant’s view is not vulnerable to Williams’s objection in the same way as the neo-​ Kantian conception of reflective detachment is. Of course, Williams dismisses Kant’s theory of respect as the sole moral incentive, because of its dubious underpinnings.39 Admittedly, Kant’s argument is partly based on the distinction between moral and pathological feelings, which does sound somewhat obsolete; however, this is not a sufficient ground for rejecting Kant’s view that practical reasoning produces novel incentives. Williams’s objection is circular precisely because it denies the motivating and generative function of reasoning.40 The claim that reasoning is not generative entails that all rational actions are motivated by a pre-​existing desire, which resides in the agent’s deliberative set (Nagel 1970: 28–​9). On Williams’s view, intertemporal and interpersonal distance can be bridged only by the contingent persistence of desires, which are the only drive to future action and shared agency. By contrast, Kant sees the activity of reasoning as being explanatorily prior to incentives.41 Even though motivation presumes the presence of desires, desires are 38 For analogous reasons, it is risky to have other motives to cooperate with the law (C2 5: 72, 75); the main risk is that their interference may destroy the moral worth of action (C2 5: 93, 96–​7). 39 Williams opposes Kant’s distinction between pathological and moral feelings (1973: 260). The ‘mystery’ of the inescapability of moral obligation is exposed as involving a sort of misrepresentation and illusion (1985: 177, 191, 196), while claiming paramount importance (1985: 182) and deliberative priority (1985: 183; 1995b; cf. Bagnoli 2015). No purely rational process can require a man to move from I-​desire to non-​I desires’ (Williams 41 Nagel argues that desires can be explained by practical reason rather than vice versa; they therefore do not constitute its condition of possibility. The structure of (practical) reason explains motivation by [p. 226] not always present as the original drives of action: some desires are generated by reasoning. Action is explained by reason, in so far as it is brought under the control of normative principles. Practical reasoning generates efficacious reasons because respect works at the same time as the incentive to morality and as a normative constraint on the contents of reasons. The appeal to a moral incentive does not establish that reasoning always produces a decisive moral motive for action. In animals endowed with rationality, it does not, and to presume that this is always the case would be contrary to the facts. However, it is also incorrect to assume that reasoning never motivates anyone to do the right thing, in the absence of prior motives. The Kantian contention is that there is a structural connection between motivation and reason, which is constitutive of humanity (Nagel 1970: 14, 31–​2). To this extent, humans are not fully free to be amoral or unsusceptible to moral demands, because the latter are rooted in or constitutive of their distinctive form of agency. This is the modest sense in which moral obligations are inescapable. This sense of inescapability is not incoherent with the painful experience of the predicament of temporality. The recognition of the inescapability of moral obligations may be quite burdensome—​as the cases of Gauguin and Karenina illustrate—​but it is not alienating, in the proper sense, not because these agents harbour a deep moral motivation or are altruistically oriented, but because they are sensitive to motivations generated by reasoning. This generative aspect of reasoning is perhaps the most misunderstood feature of Kant’s theory of moral obligation. Affective Orientation towards the Future By and large, Williams’s critique of Kant calls attention to the role of desires, emotions, and sensibility in a theory of practical reasoning, which is shown to require an appropriate ‘ethicized psychology’.42 However, his condemnation of Kant’s moral psychology is not fully justified, and his uncharitable attitude can partly be explained by his focus on the Groundwork. Kant’s mature moral psychology prominently includes future-​oriented practical attitudes which directly respond to the concerns about interpersonal and diachronic rational agency. Humans act under the idea of freedom: that is, they must assume that they are able and free to determine rational ends for themselves, and that they can exercise some deliberative control over their inclinations, desires, and instincts, having at desire (Nagel 1970: 33–​4). Due to my constructivist understanding of Kant’s conception of practical reason, I prefer to talk of the priority of reasoning over desire. 42 To ethicize psychology is ‘to provide a psychology that gets its significance from ethical categories’ (Williams 1993: 43), and to define ‘the functions of the mind, especially with regard to action . . . at the most basic level in terms of categories that get their significance from ethics’ (1993: 160). [p. 227] Carla Bagnoli least some capacity to resist their pressure (G 4: 448). These are necessary presuppositions for practical reason, and they inform deliberation for the future. If humans were convinced that their past would determine them, then how could they engage in action? Given the conflicted minds of animals endowed with reason, there is no expectation that the moral incentive will always prevail. In fact, the predicament of temporality exposes humans to the recurrent experience of failure. Kant is well aware of the psychological drawbacks of being a temporal and fragile subject. This awareness is particularly acute in his reflections on history, in which failures are perceived as not merely recurrent, but pervasive and cumulative. To counteract their despairing effects, Kant brings to the fore the practical attitude of rational hope. The interesting feature of his conception is that it differs from outcome-​oriented hope, and from any optimistic expectations or wishful thinking. Its credentials are wholly practical rather than speculative, and its function is not merely to strengthen the human will that has been weakened by past failures. Rather, hope invests oneself and others with the responsibility to try and resist, thereby providing novel grounds for action. Hope entitles rational agents to enjoin action, even when the history of failures is so bruising (TP 8: 309). This is another way in which Kant envisions individual and collective agency as being organized over time, by virtue of engaging in practical attitudes (PP 8: 368). In contrast to theories that rely on the conservative pressure of desires, and that anchor deliberation in the present, the Kantian approach points to practical attitudes as modes of authorizing action at a distance, across time and across persons. In contrast to theories that rely on character traits or traditions, it allows for modes of self-​integration that make room for normative change and radical departures from past experiences. 4. Conclusion I have argued that by refocusing on the predicament of temporality we can gain a better insight into the nature and depth of Williams’s critique of Kant. Williams celebrates mortality and contingency. Kant’s ambition is to provide normative tools to manage the predicament of temporality, suggesting that humanity is bound together not only by interdependency, desire, and habits that are easily placed into patterns, but also by rational attitudes and norms that design rational action anew. This dispute reveals a radical disagreement regarding the capacity of reasoning to provide the normative and motivational structure of agency: while Williams relies on the force of current desires, Kant identifies the potentially subversive force of reasoning as a generator of motives.43 I have argued that Williams 43 Nagel underlines that this is not a psychological explanation of moral motivation; rather, it is what explains one of the core elements of motivation. [p. 228] does not offer any decisive argument to dismiss the Kantian account of practical rationality. In fact, he seems to lack the resources to explain how rational agents can exercise their rational authority over time and across persons.44 In spite of the traction of Williams’s arguments, Kant’s thesis remains a lively research programme. However, its success is predicated on articulating the integration of reason and normative attitudes coherently with empirical psychology. To this extent, Williams’s objections have set the normative standard for any plausible Kantian theory of practical rationality.45 References Allison, Henry. 1990. Kant’s Theory of Freedom. Cambridge: Cambridge University Press. Bagnoli, Carla. manuscript. ‘Kant’s Empowering Conception of Hybrid Agency’. Bagnoli, Carla. 2011. ‘Emotions and the Categorical Authority of Moral Reasons’. In Morality and the Emotions. Edited by Carla Bagnoli, 62–​81. Oxford: Oxford University Press. Bagnoli, Carla, ed. 2013a. Constructivism in Ethics. Cambridge: Cambridge University Press. Bagnoli, Carla. 2013b. ‘Constructivism About Practical Knowledge’. In Constructivism in Ethics. Edited by Carla Bagnoli, 153–​82. Cambridge: Cambridge University Press. Bagnoli, C. 2015. ‘Moral Objectivity: a Kantian Illusion?’ Journal of Value Inquiry 49: 31–​45. Bagnoli, Carla. 2017. ‘Kant in Metaethics: The Paradox of Autonomy, Solved by Publicity’. In The Palgrave Kant Handbook/​Palgrave Handbooks in German Idealism. Edited by Matthew Altman, 355–​77. London: Palgrave Macmillan. Bagnoli, C. 2021. ‘Respect and the Dynamics of Finitude’. In Respect. Edited by Richard Dean and Oliver Sensen, 121–​39. Oxford: Oxford University Press. Bagnoli, C. 2022. Ethical Constructivism. Cambridge: Cambridge University Press. Darwall, Stephen L. 1983. Impartial Reason. Ithaca, NY: Cornell University Press. Engstrom, Stephen. 2009. The Form of Practical Knowledge. Cambridge: Cambridge University Press. Engstrom, Stephen. 2013. ‘Constructivism and Practical Knowledge’. In Constructivism in Ethics. Edited by Carla Bagnoli, 133–​52. Cambridge: Cambridge University Press. Frankfurt, H. 1999. ‘On the Usefulness of Final Ends’. In Frankfurt, Necessity, Volition, and Love, Cambridge and New York: Cambridge University Press. Hegel, Georg Wilhelm Friedrich. 1802. ‘Faith and Knowledge (Glaube und Wissen)’. Translated by William Desmond. In The Hegel Reader. Edited by Stephen Houlgate, 39–​58. Oxford: Blackwell, 1998. Kant, Immanuel. Citations are to Kants gesammelte Schriften. Ausgabe der Preussischen Akademie der Wissenschaften. Berlin: Georg Reimer, subsequently Walter de Gruyter, 44 It could still be objected that this view relies on a form of motivational dualism, which for Williams marks a powerful and deeply rooted misconception of the good life (Williams 1985: 196); but there are ways of arguing for this claim coherently with non-​reductive naturalism, see Bagnoli manuscript. 45 Work on this publication was completed during my tenure as a Visiting Fellow of All Souls College at the University of Oxford, in 2023–​4. Abridged versions have been presented at the conference What is the Point of Morality? at the University of Konstanz, and as the Paton Lecture 2025 at the University of St. Andrews. I would like to thank the editors, Marcel van Ackeren and Mattheu Queloz, and Roger Crisp, Stephen Fischer, Lorenzo Greco, Edward Harcourt, Elijah Millgram, Jacob Rosenthal, Jens Timmermann, and the referees, for their constructive comments on a previous draft. This publication is part of the project PID2022-​139226NB-​I00, funded by MCIN/​AEI/​10.13039/​501100011033/​and by FEDER—​A way of doing Europe, EU. [p. 229] Carla Bagnoli Translations are based on the Cambridge Edition of the Works of Immanuel Kant. Edited by Paul Guyer and Allen Wood. Cambridge: Cambridge University Press, 1992–​. Korsgaard, Christine M. 1996a. ‘Personal Identity and the Unity of Agency, a Kantian Response to Parfit’. In Korsgaard, Creating the Kingdom of Ends, 363–​97. Cambridge: Cambridge University Press. Korsgaard, Christine M., et al. 1996b. The Sources of Normativity. Cambridge: Cambridge University Press. Korsgaard, Christine M. 2008. The Constitution of Agency. Oxford: Oxford University Press. Korsgaard, Christine M. 2009. Self-​Constitution: Action, Identity and Integrity. Oxford: Oxford University Press. Korsgaard, Christine M. 2014. ‘The Normative Conception of Agency’. In Rational and Social Agency: Essays on the Philosophy of Michael Bratman. Edited by Manuel Vargas and Gideon Yaffe, 201–​26. New York: Oxford University Press. Korsgaard, Christine M. 2021. ‘The Value of Humanity’. In Respect. Edited by Richard Dean and Oliver Sensen, 171–​91. Oxford: Oxford University Press. Mackie, John L. 1977. Ethics: Inventing Right and Wrong. Harmondsworth: Penguin Books. Millgram, Elijah. 2004. ‘On Being Bored Out of Your Mind’. Proceedings of the Aristotelian Society Moran, Richard. 2022. ‘The Philosophical Retreat to the Here and Now: Notes on Living in Time’. Philosophy 97.4: 413–​33. Murdoch, Iris. 1997. Existentialists and Mystics: Writings on Philosophy and Literature. Edited by Peter J. Conradi. London: Chatto & Windus. Nagel, Thomas. 1970. The Possibility of Altruism. Oxford: Clarendon Press. Nagel, Thomas. 1986. The View from Nowhere. New York: Oxford University Press. Nagel, Thomas. 1996. ‘Universality and the Reflective Self ’. In The Sources of Normativity. Edited by Christine M. Korsgaard et al., 200–​10. Cambridge: Cambridge University Press. Queloz, Matthieu. 2022. ‘A Shelter from Luck: The Morality System Reconstructed’. In Morality and Agency: Themes from Bernard Williams. Edited by Árpád Szigeti and Michael Talbert, Oxford: Oxford University Press. Rawls, John. 1980. ‘Kantian Constructivism in Moral Theory’. In Collected Papers. Edited by S. Freeman, 303–​58. Cambridge, MA: Harvard University Press. Stern, Robert. 2013. ‘Moral Skepticism, Constructivism and the Value of Humanity’. In Constructivism in Ethics. Edited by Carla Bagnoli, 22–​40. Cambridge: Cambridge University Press. Timmermann, Jens. 2007. Kant’s Groundwork of the Metaphysics of Morals: A Commentary. Cambridge: Cambridge University Press. Wallace, R. Jay. 2003. The View from Here. Oxford: Oxford University Press. Williams, Bernard. 1963. ‘Freedom and the Will’. In Freedom and the Will. Edited by David Pears, New York: Palgrave Macmillan. Williams, Bernard. 1966/​1973. ‘Morality and the Emotions’. In Problems of the Self: Philosophical Papers 1956–​1972. Cambridge: Cambridge University Press, 1973. Williams, Bernard. 1973. Problems of the Self. Cambridge: Cambridge University Press. Williams, Bernard. 1979. ‘Internal and External Reasons’. Reprinted in Williams, Moral Luck, Cambridge: Cambridge University Press, 1981. Williams, Bernard. 1981. Moral Luck. Cambridge: Cambridge University Press. Williams, Bernard. 1985. Ethics and the Limits of Philosophy. London: Fontana. Williams, Bernard. 1989/​1995b. ‘Internal Reasons and the Obscurity of Blame’. Reprinted in Making Sense of Humanity, 35–​45. Cambridge: Cambridge University Press. Williams, Bernard. 1993. Shame and Necessity. Berkeley: University of California Press. Williams, Bernard. 1995a. ‘How Free Does the Will Need to Be?’ In Williams, Making Sense of Humanity, 3–​21. Cambridge: Cambridge University Press. Williams, Bernard. 1995b. Making Sense of Humanity. Cambridge: Cambridge University Press. [p. 230] Williams, Bernard. 1995c. ‘Replies’. In World, Mind and Ethics, Essays on the Ethical Philosophy of Bernard Williams. Edited by J. E. J. Altham and Ross Harrison, 185–​224. Cambridge: Cambridge University Press. Williams, Bernard. 1996. ‘History, Morality, and the Test of Reflection’. In Christine M. Korsgaard et al., The Sources of Normativity, 210–​18. Cambridge: Cambridge University Press. Williams, Bernard. 2001. ‘Postscript: Some Further Notes on Internal and External Reasons’. In Varieties of Practical Reasoning. Edited by Elijah Millgram, 90–​7. Cambridge, MA: The MIT Press. Williams, Bernard. 2002. Truth and Truthfulness: An Essay in Genealogy. Princeton: Princeton University Press. Wood, Allen W. 1984. ‘Kant’s Compatibilism’. In Self and Nature in Kant's Philosophy. Edited by Allen W. Wood, 73–​101. Ithaca, NY: Cornell University Press. [p. 234] Genealogy Williams, Hume, and Nietzsche P. J. E. Kail 1. Introduction Prior to the publication of Bernard Williams’s Truth and Truthfulness: An Essay on Genealogy (2002), it is fair to say that philosophers in the Anglo-​American tradition considered genealogies and State of Nature theories (SNTs) as different kinds of account. That is not to say that this assumption indicated some deep reflection on the differences between SNTs and genealogies: instead, it reflected the fact that the former are staples of political philosophy curricula while the notion of a philosophical genealogy inhabited the altogether murkier philosophical waters where Nietzsche and Foucault swam. Roughly, the differences came out as follows. SNTs are explicitly fictional accounts of how life would be for human beings who have certain needs and interests but lack some practice or concept. One can then show how that practice or concept meets those needs and interests and, in so doing, vindicate that practice. Such accounts offer explanations of the target practice or concept in the sense that we can see how the relevant social practice might have emerged from the hypothesized situation, but for the account to vindicate the practice it is inessential that it really had such origins. The picture of genealogy, prior to Williams at any rate, has Nietzsche and Foucault at its centre and appears to differ from SNTs in a number of ways. First, genealogy appears to insist strongly on ‘real history’ and should involve that which ‘can be documented, which can be actually confirmed and has actually existed’ (GM Preface 6). Second, whereas SNTs typically vindicate, genealogy was thought to be an essentially critical exercise. Third, in the case of SNT the connection between the needs and interests of those in the State of Nature, and the relevant normative practice is relatively transparent: given the needs and interests, one can see that there are practical reasons in favour of the relevant normative practice. In the case of Nietzsche and Foucault, however, this connection seems, how shall we say, somewhat murkier, involving power relations, ressentiment, shameful origins, or the cunning of the priests. For Williams and his followers, both SNTs and what I have referred to as genealogy fall under the umbrella of genealogy. Williams doesn’t explain why he lumps them in together and perhaps he doesn’t need so to do: after all, in the P. J. E. Kail, Genealogy In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. DOI: 10.1093/​9780191966361.003.0012 [p. 235] P. J. E. Kail philosophical context, ‘genealogy’ is a term of art and, as Hugh Mellor was fond of saying, artists can employ terms how they wish. Thus, Williams characterizes genealogy as ‘a narrative that tries to explain a cultural phenomenon by describing a way in which it came about, or could have come about, or might be imagined to have come about’ (2002: 20), which evidently comprehends both SNTs and traditional genealogies by including ‘real histories’ as well as fictional ones. This characterization, however, leaves it open as to how we are to understand the ‘way in which it [the target phenomenon] came about’, but some illumination is to be had a few pages later where the notion of a function is introduced. Williams illustrates this notion in the company of Edward Craig’s account of knowledge (Craig Craig eschews the traditional accounts of knowledge that seek to analyse knowledge into conceptually more basic elements and instead offers an illumination of the concept by asking what practical needs the concept meets. He starts by making some abstract but plausible assumptions about the needs and interests human beings might have when they lack the concept of knowledge and then builds a different concept, that of a ‘reliable informant’, that meets those needs. Human beings need true beliefs and, while they have ‘on board’ sources such as the senses, they need to avail themselves of other people when determining whether p. But upon whom to rely? We need reliable informants and that in turn creates a need to identify them as such. Roughly speaking, the features that pick out reliable informants are discussed by Craig in relation to the traditional accounts of the properties that were thought to turn true belief into knowledge. Crucially, however, the variations are considered in the context of increasingly complicated practical situations which condition the difficulties involved in identifying a reliable informant. The complexities of the concept then emerge from the complexities of various practical situations, and so are understood functionally, even within the midst of complexity. It emerges as instrumental to the end of acquiring true belief, and in showing how we can show why the notion of knowledge has the shape it does. For Williams, this general approach can be extended to other concepts. One starts with a concept and imagines what interests and needs it might serve, and then considers further practical complexities to illuminate the concept at hand. Such accounts are typically fictional or imaginary, but, nevertheless, explanatory, if only potentially so. How so? Williams discusses this by referring to Nozick’s distinction between ‘fact-​defective’ and ‘law-​defective’ potential explanations—​ explanations where some antecedent condition is false and those involving the postulation of a false law. Where such imaginary explanations are fact-​detective they can be illuminating even when the presumed antecedent condition—​the ‘State of Nature’—​is not only false but impossible. Thus, Nozick’s account involves a State of Nature where people have ‘only economic motivations and moral ideas of individual right’ (1974: 32), but, as Williams notes, we ‘know . . . that there could not have been a pre-​political condition with just those properties’ (2002: 31). [p. 236] Nevertheless, the (potential) explanation is useful because it can display functional relations between those starting points. [The] relation between the derived, more complex, reason and the simpler ‘more primitive’ reasons or motivations is rational, in the sense that in the imagined circumstances people with the simpler motivations would welcome, and, if they could so, aim for, a state of affairs in which the more complex reasons would operate. (2002: 34) The ‘if they could do so’ is important here. An imaginary genealogy is explanatory because it represents as functional a concept, reason, motivation, or other aspect of human thought or behaviour, where that item was not previously seen as functional; the explanation of the function is unmysterious, because in particular it does not appeal to intentions or deliberations or (in this respect) already purposive thought; and the motivations are ones that are agreed to exist anyway. (2002: 34) Famously, Williams proceeds to give an imaginary functional account of our valuing truth, of our imbuing it with intrinsic value. This, however, is not my concern. An aim more modest is to show that it is worth keeping apart accounts that aim to give genuine accounts of the emergence of some phenomenon from the fictional accounts, and reserving the term ‘genealogy’ for the former. I shall slightly revise this aim, but it is worth pointing out that Williams’s followers seem to recognize this, albeit implicitly, but elect to see this difference as a matter of accounts occupying different places on a spectrum. Matthieu Queloz, for example, writes of the accounts offered by Hobbes, Locke, and Nozick that ‘these are genealogical insofar as they are developmental narratives, but it . . . is doubtful that these narratives have any serious explanatory ambitions. They are more plausibly read as justificatory arguments in genealogical guise’ (2021: 5). At the other end of the spectrum are the documented histories of Foucault. One might query whether there is really a spectrum rather than just two different kinds of things. Queloz claims that there are ‘hybrid’ genealogies that are a mixture of ‘real history’ and functional fictions. Significantly, however, Queloz mentions only three examples of such hybrid genealogies that are not themselves influenced by Williams’s own conception of genealogy. One of these, Edward Craig’s Knowledge and the State of Nature, was not conceived by Craig when originally formulated as a genealogy, and Craig’s own view of its relation to Williams’s project is somewhat ambivalent (Craig 2007). The other two exemplars that Williams draws upon are Hume’s account of justice, and some of Nietzsche’s writings. Since the subject of this volume is Williams, I shall focus on an even more modest aim than I mentioned above and discuss the few brief remarks he makes in connection [p. 237] P. J. E. Kail with Hume, Nietzsche, and genealogy. Williams mentions these two because he sees them as fellow travellers in his project. But although these accounts are genealogical, there is little to say about them as engaged in a fictional-​cum-​f unctional project that is distinctive of Williams’s view. In brief, I find no fictionality in either Hume or Nietzsche, and, in the case of Nietzsche, little to suggest that the functional story is at the centre of his own genealogy. 2. Hume, Fiction, and the State of Nature If we take genealogy to be an account of origins then we can read the entirety of Hume’s A Treatise of Human Nature as an exercise in genealogy. Throughout that work he seeks the origins of ideas and beliefs, a project conditioned by the search for the simple impressions from which simple ideas are derived. Sometimes the results of such investigations can be surprising or destabilizing. Thus, for example, when he determines that our idea of causal power is derived from some internal impression rather than an experience of genuine causal powers, he recognizes that his conclusion will be seen as ‘extravagant and ridiculous’ (T 1.3.14.26). More dramatic is his treatment of the belief in the external body in ‘Of scepticism with regard to the senses’ (T 1.4.2), where he begins by stating that his topic is not whether ‘there be body or not’ but ‘what causes induce us to believe in the existence of body’ (T 1.4.2.1). However, once these causes are examined Hume declares himself of ‘quite a contrary sentiment’, and is ‘inclin’d to repose no faith at all in my senses, or rather imagination, than place in it such an implicit confidence’ (T 1.4.2.56). Somehow, knowledge of those causes destabilizes the belief. By contrast, knowledge of the principles governing the ‘rise and origin’ of our moral sensibility means that those principles ‘acquire a new force’ since we find ‘nothing but what is great and good in [their] rise and origin’ (T 3.3.6.3). Nothing in the Treatise suggests that Hume takes the accounts of origins he offers to be fictional. Instead, they reflect aspects of Hume’s general theory of human nature and its explanatory aspirations. However, during his discussion of the emergence of the artificial virtue of justice Hume says something that Williams seizes upon to support the claim that Hume’s explanation of justice is a fictional one. Hume writes that the ‘suppos’d state of nature . . . [should be allowed] as a mere philosophical fiction, which never had, and never cou’d have any reality’ and that the ‘very first state and situation [of humanity] may justly be esteemed social’ (T 3.2.2.14). Williams sees this as indicating that Hume takes his own account of justice to be ‘fact-​defective’ in Nozick’s sense (2002: 29). The state of affairs from which the story starts is one in which people are self-​ interested and have a capacity for limited sympathy, but have no motives of justice and, correspondingly, no concept of property. . . . The state of affairs from which [p. 238] this process starts, and hence the process itself, Hume recognizes to have been impossible. (2002: 29) The first part of the paragraph is correct in outline, but the final sentence is not. For when Hume talks of a ‘mere philosophical fiction’ that is the supposed State of Nature, he is not referring to his own starting point but to a Hobbesian one that he has considered, and rejected, earlier.1 Since my concern is not with Hume’s account per se but only its alleged fictional character, I shall focus solely on this point and not outline Hume’s own explanatory genealogy. ‘Of all the animals, with which this globe is peopled,’ Hume writes, ‘there is none towards whom nature seems, at first sight, to have exercis’d more cruelty that towards man’ (T 3.2.2.1). Unlike other animals, humans have a host of needs which are not readily met in nature. What is requisite to meeting these needs is ‘society’ or social cooperation. Society helps the human being become ‘in every respect more satisfy’d and happy’ than they would be in a ‘savage and solitary condition’. When ‘every individual person labours apart, and only for himself, his force is too small to execute any considerable work’. Society allows not only for an increase in force but also for a division of labour, and its concomitant benefits, together with increased security (T 3.2.2.2). In this ‘savage and solitary’ condition, however, the fact that society is advantageous is insufficient to explain its emergence. What is further required is that humans ‘be sensible of its advantages; and ’tis impossible, in their wild uncultivated state, that by study and reflection alone, they shou’d ever be able to attain such knowledge’ (T 3.2.2.4). The emergence of society from a mere set of solitary humans is thus impossible. However, ‘the natural appetite betwixt the sexes’ initially brings people together, which is typically followed by ‘a new tye’ which is ‘their concern for common off-​spring’, which itself is ‘also a principle of union betwixt the parents and off-​spring, and forms a more numerous society’. This sociability, albeit a narrow family-​centred one, operates ‘on the tender minds of the children, [and makes] them sensible of the advantages, which they reap from a society’ (T Thus the, albeit limited, social nature of the family remedies the epistemic knowledge requirement of the benefits of cooperation confronting the solitary human. A second point he makes in this connection concerns the sentiments and behaviours within the family, which include amiable sentiments, contrasting these with the ‘delight’ that ‘certain philosophers’ take in representing human beings as egregiously selfish, a view ‘as wide of nature as any accounts of the monsters, which we meet with in fables and romances’ (T 3.2.2.5). What is significant here is that Hume is contrasting a family, and hence a social starting point for his account of justice, with a solitary and selfish one. The latter 1 I have benefited from discussions with Alessio Vacarri on this point. [p. 239] P. J. E. Kail he rejects as explanatorily impossible and psychologically false. Thus the ‘very first state and situation [of humanity] may justly be esteemed social’ (T 3.2.2.14). Nevertheless, this social state is certainly prior to the invention of justice and property, these notions emerge from inter-​group cooperation, and nothing in that suggests Hume’s account is a fiction. The ‘mere philosophical fiction’ Hume refers to later is not his own account but the Hobbesian State of Nature populated by the solitary and the selfish. This is evident from the context of the philosophical fiction claim. Hume reiterates his point about the family, stating that it is ‘utterly impossible’ for men to be solitary and that ‘his very first state and situation may justly be esteem’d social’. This is Hume’s starting point. Other philosophers ‘may, if they please, extend their reasoning to the suppos’d state of nature’ (T 3.2.2.14; original emphasis), one that is described as full of war and violence (T 3.2.2.14). This is not Hume’s starting point, but quite obviously Hobbes’s, one that Hume earlier rejects as offering any chance of a genuine explanation. Certainly, Hobbes’s conception of the political state can be seen to serve the interests of those in the State of Nature, and so in that sense states of nature are revealing, but Hume thinks that they are not explanatory. 3. Nietzsche and Fiction Hume, then, doesn’t think that his account is fictional. On the other hand it is not something that is conventional history, if by that we mean one that adverts to documentary or other forms of evidence. Instead it is what Dugald Stewart called a ‘conjectural’ or ‘theoretical’ history, a coinage made with explicit reference to Hume’s own Natural History of Religion. When, Stewart says, ‘we are unable to ascertain how men have actually conducted themselves upon particular occasions, [we consider] in what manner they are likely to have proceeded, from the principles of their nature, and the circumstances of their external situation’ (1929: 32). Conjectural histories are conceived to be inferences to best explanation.2 We cannot know what the detailed history of the emergence of some given phenomenon was, but we can make assumptions about the general psychology and interests of human beings and the kind of environment in which those humans were placed so that the conjunction of the psychology and the environment explain the emergence of that phenomenon. Such accounts are conjectural, theoretical, or speculative, but they are not fictional: they are conceived as inferences to the best explanation, and not as accounts that are explicitly conceived to be false. This brings us to Nietzsche’s On the Genealogy of Morality, which is the obvious reference point for any discussion of philosophical genealogy. Nietzsche 2 For discussion, see Evnine (1993) and Santos Castro (2017). [p. 240] himself seems to emphasize that his own genealogy is a ‘real history of morality’ (GM Preface 7), and I shall try to sketch here quite what he means by this claim. First, however, let us take note of Williams’s own, and somewhat sparse, remarks about Nietzsche and genealogy in Truth and Truthfulness. Williams states that the term ‘genealogy’ is connected with Nietzsche and that he intends ‘the association to be taken seriously’ (2002: 13). Much of what Williams says about Nietzsche here consists in well-​placed remarks about misapprehensions over Nietzsche’s views of truth, but he also makes a few brief remarks on Nietzsche and genealogy. He views his own conception of genealogy as a ‘descendant’ (2002: 18), but claims it is only one of many such descendants. When it comes to Nietzsche’s particular genealogy, Williams makes the following remarks. First, since Nietzsche is trying to explain ‘the whole of morality’, and morality in this sense ‘demands to be understood as self-​sufficient’, the ‘resistance to a functional style of explanation, here, is very deep’ Secondly, the ingredients that enter into the explanation are not merely non-​moral but include the ‘supposed enemies of morality’, emotions like hatred and resentment (2002: 37). Thirdly, the explanation is necessarily unconscious. Fourthly, this aspect leads to a puzzle. Nietzsche, Williams maintains, is trying to explain a social phenomenon and this leaves a question about what role is played by this type of ‘individual psychological reaction’. Finally, Williams claims that Nietzsche’s ‘genealogy is not meant to be entirely fictional’ (2002: 37). I will not address all these points. But Williams’s last remark returns us to the question of ‘real history’. While Williams notes that Nietzsche’s ‘genealogy is not meant to be entirely fictional’, he adduces no evidence at all that any of it is meant to be fictional. He is not alone in this flat assertion. More recently, Bernard Reginster, in his attempt to offer a more detailed reading of Nietzsche’s genealogy as functional in character, writes that Nietzsche’s ‘actual genealogies thus combine fictional narrative with some historical documentation’ (2021: 35). We will return to the historical documentation point, but the ‘thus’ in Reginster’s claim is only justified by an assertation he makes a sentence earlier: namely, that his genealogies involve ‘vague, conjectural, and even fictional allusions to human “prehistory” ’ We might agree that what Nietzsche offers is conjectural, and dispute the degree of vagueness, but Reginster offers nothing to support the claim that fiction is involved: indeed, he slides from asserting that there are ‘fictional allusions’ in one sentence to a claim that there is nothing but fictional narrative with some historical documentation in the next sentence, completely eliding the ‘conjectural’.3 One reason for thinking that Nietzsche’s explanations are fictional is precisely their conjectural character. Thus Queloz writes that the problem is we have a ‘hard 3 Reginster also adds a footnote to the sentence about Nietzsche’s genealogy being ‘vague, conjectural, and even containing fictional allusions’ that this is a view held by other commentators, but none of those commentators—​including the present author—​asserts that Nietzsche’s account contains fiction. [p. 241] P. J. E. Kail time accounting for what can only appear, in a trained philologist, as a poor effort at writing history’ (2021: 101), and the conjectural character of the contents of Nietzsche’s genealogy certainly seems at odds with his Preface claim that his ‘grey’ genealogy is one that must involve ‘that which can be documented, which can be really ascertained, which has really existed’ (GM Preface 7). However, we need to pause a little and ask quite what documentary evidence could be expected, given what it is that Nietzsche is trying to explain. In the first essay the explanandum is the emergence of a particular family of values, ‘slave morality’ against and in reaction to a ‘master’ morality, and, in particular, the emergence of Judeo-​Christian morality in contrast to Roman aristocratic values (GM I: 16). Now, what is documentable is the existence of these two broad kinds of morality, and the supplanting of the one by the other. Nietzsche refers to various different human communities who exemplify master morality, taking their existence as illustrative for the central project of explaining how the slave morality dominating the western world emerges from a prior Greco-​Roman ‘master’ morality. This is one of the points which accounts of the ‘English psychologists’, whom Nietzsche nevertheless calls ‘genealogists’, fail to appreciate properly. They fail to see this shift in the history of morality because they are lacking in ‘historical spirit’. ‘Good’ didn’t originate in utility but instead in the noble and the powerful (GM I: 2), and the morality the English psychologists seek to explain is rather a reaction to the noble and the powerful. Admittedly, Nietzsche doesn’t do much to provide historical documentation for that shift, but his readers would not need much by way of convincing that there was such a shift from the non-​Christian to the Christian evaluative outlook. Nietzsche’s explanation of the shift, however, is such that it is difficult to know quite what documentary evidence there could be for it. Central to the explanation is the postulation of the psychological state of ressentiment and an unconscious biasing of psychological mechanisms that issues in evaluative beliefs which downgrade what the masters take to be positively valuable and a corresponding positive gloss on what otherwise is taken to be base and lowly about the slaves. There is a good deal of plausibility to the psychology involved in the explanation, but it is difficult to imagine what documentary evidence there could be for its operation. It is instead a psychological postulate, together with some other assumptions, that explains what can be documented: namely, the shift in values.4 In the case of the second essay, the absence of documentary evidence is again not surprising, given what Nietzsche is trying to explain. Underlying all the phenomena with which Nietzsche is concerned is ‘the internalizing of man: thus first grows in man that which he later call his “soul” ’ (GM II: 16). Bad conscience is the 4 Williams’s concern about the relation between the social and the psychological needs to be addressed. A full answer is beyond the scope of this chapter, but, in outline, we need to factor in the priestly interpretation of existence, an interpretation which appeals to the dispositions of the psychology of ressentiment and, in giving it focus and meaning, promotes and sustains its manifestation. [p. 242] ‘true womb of ideal and imaginary events’ (GM II: 18), and constitutes a ‘forceful separation from his animal past’ (GM II: 16). Nietzsche is trying to explain the emergence of evaluative self-​consciousness which is coeval with bad conscience, and which is a presupposition of other phenomena that are the subject of the second essay, such as guilt. Necessarily, such phenomena pre-​exist any documentary evidence. Again, it’s a matter of postulating a ‘generic psychology, not properly localized to times, places, or individuals’ (Janaway 2007: 11).5 That is there is no documentary evidence for Nietzsche’s explanatory hypotheses should not, therefore, be seen as a surprise, but, more importantly, it does not itself imply the account is a fiction. There are changes in the history of humanity, changes for which no particular times and places can be localized, and Nietzsche offers explanatory hypotheses that are conjectural but nevertheless conceived to be true. It is worth noting in this connection that Darwin’s account of natural selection, generally considered to be an inference to the best explanation, afforded no direct evidence. Thus, in a letter of 1863 to George Bentham, Darwin writes: In fact the belief in natural selection must be grounded entirely on general considerations. (1) on its being a vera causa, from the struggle for existence; & the certain geological fact that some species do somehow change (2) from the analogy of change under domestication by man’s section. (3) & chiefly from this view connecting under an intelligible view a host of facts.6 Causes are posited which have an analogy with current phenomena and which explain a range of facts. Commenting to J. D. Hooker on some remarks on his theory made by Frederick Hutton, Darwin writes that ‘he is one of the very few who sees that the changes of the species cannot be directly proved & that the doctrine must sink or swim according as it groups & explains phenomena’.7 Again, to Hooker, he writes: ‘I have always looked at this doctrine of Nat. selection as an hypothesis, which if it explained several large classes of facts would deserve to be ranked as a theory deserving acceptance.’8 Admittedly, what Nietzsche frequently calls his ‘hypothesis’ is highly conjectural, speculative, and lacking in detail, but it explains more than the competing hypotheses Nietzsche mentions: namely, the religious and metaphysical accounts 5 Pace Reginster (2021: 36), this non-​specificity provides no reason to think that the account is fictional. 6 Darwin Correspondence Project, ‘Letter no. 4176’, accessed 26 October 2022, . The notion of a ‘vera causa’ comes from Newton, who states that ‘no more causes of natural things should be admitted than are both true and sufficient to explain the phenomena’. For discussion of this in the context of Darwin, see Pence (2018). 7 Darwin Correspondence Project, ‘Letter no. 3098’, accessed 26 October 2022, . 8 Darwin Correspondence Project, ‘Letter no. 2696’, accessed 26 October 2022, . [p. 243] P. J. E. Kail which seek the origins of good and evil ‘behind the world’ (GM Preface 3), and the other naturalistic explanations of the English psychologists. What also should not be forgotten is Nietzsche’s note to GM 1.17, where he proposes a series of academic essay contests to promote ‘moral-​historical studies’, indicating quite clearly that his project is empirical rather than fictional. 4. Nietzsche and Function Williams himself suggests that Nietzsche’s genealogy is a functional account, though he doesn’t offer details of how that is to be fleshed out. However, Reginster attempts so to do. Part of his motivation for so doing we have already seen and rejected: he thinks that to a very great extent we must read what Nietzsche has to say as fictional. If it is fictional, then a functional interpretation of genealogy is required. But let us put that to one side. The leading idea for Reginster is that we seek ‘to identify the practical problem, to which morality can be seen to constitute a response’ (2021: 41), telling then a fictional story about how morality meets that problem. This further affords a way to understand how Nietzsche’s genealogy can constitute a critique of morality. By identifying the problem moral beliefs were ‘designed’ to solve we can address a question about the ‘value’ of moral values. Do they adequately solve the problem? Are they beneficial or harmful? (2021: 41) Reginster then points us to GM Preface 3, where this question seems to be at the fore. As well as asking for the origins of good and evil, we need to ask, says Nietzsche, what value those values themselves possess. Thus: Have they inhibited or furthered human flourishing until now? Are they a sign of distress, of impoverishment, of the degeneration of life? Or, conversely, do they betray the fullness, the power, the will of life, its confidence, its future? (GM Preface 3) We give a functional account of the ends of some particular moral phenomenon and ask whether it serves that function and thereby address this question. Hence we can take Williams and Nietzsche to have approaches that are more similar than it may seem, and, furthermore, the notion of functionality can help us to understand the critical dimension of Nietzsche’s own genealogy. There are, however, some real difficulties in reading Nietzsche this way. In classical State of Nature accounts their fictionality is not an issue since it is quite clear how they have a normative function. Roughly, we suppose that we share the motivations and interests of the supposed inhabitants of the relevant State of Nature, and [p. 244] see how a particular practice accords with those motivations and interests. Since we share the motivations and interests of those inhabitants, and those motivations provide them with a reason to adopt the practice, it also provides us with a reason, which itself vindicates our practice. Matters are far less straightforward for the idea that some fictional account can show us that certain values do not solve the problem they were ‘designed’ to solve. Suppose we identify a problem that some value is supposed to solve, and then declare it is inadequate. If, however, it is not true that the value is ‘designed’ to solve that problem, then its failure so to do is simply irrelevant. So we need to know what that purported function of the value is. If we know that in advance, and know the value fails to meet its aim, then the genealogical excursion into its emergence seems irrelevant to any critical aim. If the corkscrew doesn’t work, it doesn’t work. I don’t need to know anything about its origin. But, of course, the point of offering a genealogy is to give a functional account of a value when, precisely, its function isn’t evident, as Reginster notes (e.g. 2021: 35). We might then suppose that humans have some need (and not intend that supposition as a fiction) and tell a fictional story about how some practice is designed to meet that need, just as we might hold truly that humans constructed some building to provide shelter, but tell a fictional story about how they constructed it. However, on Reginster’s view, the fictional stories are supposed to illuminate values and, at the same time, show that they do not solve the problem that they are ‘designed’ to solve, and this leads to a question about why we should find such accounts intelligible. In classical vindicatory states of nature, the very intelligibility of the fictional accounts offered rests precisely on the fact that the practices do solve the relevant problems. Reginster is aware of something like this issue, though not quite this precise one. He notes first what we shall call ‘the problem of dysfunctionality’ (2021: 42). If the supposed best explanation of moral judgements is functional usefulness, how could they turn out to be harmful? He then notes a number of different ways in which this can be accommodated, such as (1) some judgements may be claimed to be useful to a certain set of persons, but harmful to others, (2) such functionality is lost in a change of circumstances, and (3) such functionality is self-​undermining. Secondly, he notes that some value might have multiple functionality. Here Reginster has in mind Nietzsche’s famous discussion of punishment, where a certain practice is subject to different interpretations or ‘meanings’ of its function. Punishment can be understood to be retributive or deterrent. So how can we say that a value has only one function? In these cases, the ‘historical’ character of Nietzsche’s genealogy comes into play. We tell the story of how the value came into being by appeal to its solving some problem, but then a further story about how certain changes engender a loss of utility, or a change of function, as well as recording matters such as a claim that a value is useful only to a certain set of persons. [p. 245] P. J. E. Kail All this is possible, but it still doesn’t address the redundancy issue. If the question is ‘does this value perform the function it is designed to do?’, then we can move to that question without an excursion into genealogy. The response that the functionality might not be apparent, and so a fictional genealogy is required, would be to the point only if we could be sure that the value did have that function. If, however, the function is itself obscure, then we are open to competing fictional accounts that yield a completely different account. It is not too difficult to think, for example, that the practical need that slave morality addresses is that human beings seek love and that promoting slave morality increases the amount of love in the world. How are we to decide which account is better in a way that doesn’t point us to questions and considerations regarding which one is more likely to be the true account of the value? If I am not convinced that values have a supposed function (or any function at all), I am going to be quite unmoved by someone who concocts a story that supposedly shows that they do, and equally entitled to concoct a different story that reveals that the value does fulfil its function. I shall close by indulging myself a little and making some remarks on an alternative view of the conjectural history of the genealogy and its role in the revaluation of values, partly because it suggests an entirely different view of what is going on in the GM and partly because Reginster motivates his account by attempting, unsuccessfully, to show that it is mistaken. Not only that, but it also seems to be something Williams himself thinks is operative in Nietzsche’s genealogy. The ‘slave revolt’ in morality is a matter of the slaves turning the table of values on the masters, and coming to conceive of what was previously schlecht or base as morally good and conceiving what was good in the masters as evil. It is a ‘revenge’, born of seething ressentiment, but an imaginary or conceptual one. The slaves come to believe what we previously conceptualized as good to be its inverse, and what was bad, good. But this revaluation cannot be the result of a mere decision that is the result of pragmatic advantage. We could represent it that way: it would be a good idea for slaves to come to believe themselves to be morally superior to the masters because doing so palliates the gnawing ressentiment, but the actual acquisition of that belief cannot be modelled that way. The core values of the ‘slave revolt’ are sourced in mechanisms of wishful thinking and self-​deception. And, as Williams says, the process of the creation of these new beliefs must be unconscious, since no-​one could arrive at that result while acknowledging this route to it. For the same reason, people who are identified with the result, the outlook of morality, must be resistant to this explanation of it, and if they come to accept the explanation, their outlook will have to change. (2002: 37) This affords a different way of reading the critical aspect of the GM. The leading idea here is that values and other commitments of slave morality emerge in beliefs, awareness of the aetiology of which provides grounds for suspicion. If we come to [p. 246] be aware that they have epistemically shameful origins, then we have reasons to seek justification for them. This feeds into the project of the revaluation of values. Nietzsche thinks a key problem is that many have ‘taken the value of these “values” as given, as a fact, as beyond all calling-​into-​question; until now one has not had the even the slightest doubt or hesitation in ranking the “good” as of higher value than the “evil” ’ (GM Preface 6). If such values are taken to be self-​evident, their revaluation cannot get going since these values trump any competing values. However, this ‘givenness’ is destabilized by the knowledge that the originating source of those values is an epistemically unreliable mechanism. Hence Williams’s remark that if the explanation of them is accepted by those attached to such value, ‘their outlook will have to change’. Awareness of the source of such values dislodges their alleged givenness, depriving them of their privileged position and then allowing for their potential revaluation. This is not the same as a critique of those values: it is, however, a way of destabilizing those values so that they may be subject to one. Thus, as Nietzsche writes in the Will to Power: The inquiry into the origin of our evaluations . . . in no way coincides with a critique of them . . . even though the insight into some pudenda origo certainly brings with it a feeling of a diminution in value of the thing that originated thus and prepares the way to a critical mood and attitude to it. (WP 254) Obviously, much more is needed to present a full case for this reading, and here is not the place so to do.9 Nevertheless, it is clearly in line with what Williams suggests the critical role of Nietzsche’s genealogy might be. Reginster, as I mentioned, marshals a number of objections to this reading, none of which is remotely compelling. First, Reginster asserts that ‘it ignores his [Nietzsche’s] emphatic claim that the epistemic standing of moral beliefs—​their truth or their justification—​has no bearing on their “value” ’ (2021: 17). Nietzsche certainly believes that, but that doesn’t take away from the force of the destabilization of the genealogy for those who think that it does, and that is whom his genealogy is directed towards: it is aimed at his fellow naturalistic philosophers, ‘we knowers’ (GM Preface 1), who are in for the final lesson that their unconditional will to truth is itself the last gasp of the ascetic ideal.10 They are not going to be remotely persuaded of the point of a revaluation if they assume both the unconditional will to truth and the truth of core moral beliefs. Second, Reginster (2021) claims that there would be nothing new in Nietzsche’s approach since previous philosophers have queried the rational grounds, but also notes that Nietzsche takes all such disputes to revolve around what he, Nietzsche, calls ‘the common faith in the prevalent morality’ (BGE 186). All that destabilization does is ‘call for a fresh reexamination of their grounds 9 See Kail (2011) for a full case. 10 For a superb discussion of this, see Gemes (unpublished). [p. 247] P. J. E. Kail [for moral beliefs]’ (2021: 18). Reginster thinks this is insufficient to do any critical work. Part of the problem is that Reginster himself seems to expect more from genealogy than Nietzsche does. Thus, he writes that such a view of the genealogy of morals is presented only ‘as a necessary preparation for their actual critique’ original emphasis). Indeed it is, but that is not a foible of this reading: it is what Nietzsche himself says genealogy is supposed to do. As we noted above, he writes that the ‘inquiry into the origin of our evaluations . . . is in no way identical with a critique of them’, and in Ecce Homo he describes the GM as ‘a psychologist’s three crucial preparatory works for a revaluation of all values’ (EH ‘Why I write such good books’). So while Reginster thinks that genealogy itself must be a critique, neither Nietzsche nor this reading thinks it is. Be that is it may, there is a further issue: Reginster interprets ‘the common faith in the prevalent morality’ as a view that we are convinced that there are good reasons for our moral beliefs, even if we are presently unable to produce them. No doubt that is one aspect of it, but it is not all. At the centre are the contents of the beliefs—​the values—​which constitute the space in which moral thought is engaged. Such a space presents itself as what morality is—​as Nietzsche puts it, it says ‘I am morality itself and nothing else is moral!’ (BGE 202). The ‘good faith’ (gutens Glauben) in common morality is that what is moral—​and what is not—​is simply a ‘given’. So it is not merely, as Reginster describes, that the destabilized reader might simply cling to some faith that there are some reasons for those beliefs even though they can’t find them at present (a situation I evidently find more alarming than Reginster). It is rather that the very identity of what is taken to be distinctively moral—​the selfless slave values—​is itself now open to criticism since it cannot be taken to be, somehow, constitutive of morality without that claim itself requiring some further justification. Of course, as Nietzsche himself knows well, and as Williams notes, someone highly invested in morality ‘must be resistant to this explanation of it’: but it is far harder to resist this form of destabilization than the spinning of some idle fiction.11 References Craig, Edward. 1990. Knowledge and the State of Nature. Oxford: Clarendon Press. Craig, Edward. 2007. ‘Genealogies and the State of Nature’. In Bernard Williams. Edited by Alan Thomas, 181–​200. Cambridge: Cambridge University Press. Evnine, Simon. 1993. ‘Hume, Conjectural History, and the Uniformity of Human Nature’. Journal of the History of Philosophy 31.4: 589–​606. Gemes, Ken. Unpublished. ‘Strangers to Ourselves: Nietzsche on the Will to Truth, the Scientific Spirit, Free Will, and Genuine Selfhood’. 11 Thanks to Chris Fowles for very helpful comments on an earlier version of this chapter. [p. 248] Hume, David. 2000. A Treatise of Human Nature. Edited by David Fate Norton and Mary J. Norton. Oxford: Oxford University Press. References are to book, part, section, and paragraph number. Janaway, Chris. 2007. Beyond Selflessness: Reading Nietzsche’s Genealogy. Oxford: Oxford University Press. Kail, P. J. E. 2011. ‘ “Genealogy” and the Genealogy’. In Nietzsche’s On the Genealogy of Morality: A Critical Guide. Edited by Simon May, 214–​233. Cambridge: Cambridge University Press. Nietzsche, Friedrich. 1973. The Will to Power. Translated by Walter Kaufman and R. J. Hollingdale. New York: Vintage. Nietzsche, Friedrich. 1988. On the Genealogy of Morality. Translated by Maudemarie Clark and Alan J. Swensen. Indianapolis/​Cambridge: Hackett Publishing. Nietzsche, Friedrich. 2002. Beyond Good and Evil. Translated by Rolf-​Peter Horstmann and Judith Norman. Cambridge: Cambridge University Press. Nietzsche, Friedrich. 2005. Ecce Homo. In The Anti-​Christ, Ecce Homo, Twilight of the Idols, and Other Writings. Translated Judith Norman. Cambridge: Cambridge University Press. Nozick, Robert. 1974. Anarchy, State, and Utopia. New York. Basic Books. Pence, Charles. 2018. ‘Sir John F. W. Herschel and Charles Darwin: Nineteenth Century Science and Its Methodology’. Hopos: Journal of the International Society for the History of Philosophy of Science 8.1: 108–​40. Queloz, Mattieu. 2021. The Practical Origins of Ideas: Genealogy and Conceptual Reverse Engineering. Oxford: Oxford University Press. Reginster, Bernard. 2021. The Will to Nothingness: An Essay on Nietzsche’s On the Genealogy of Morality. Oxford: Oxford University Press. Santos Castro, Juan Samuel. 2017. ‘Hume and Conjectural History’. Journal of Scottish Philosophy Stewart, Dugald. 1829. An Account of the Life of Adam Smith. In Collected Works, Vol. 7. Cambridge: Hilliard & Brown. Williams, Bernard. 2002. Truth and Truthfulness: An Essay in Genealogy. Princeton: Princeton University Press. [p. 249] Ethics, Untimeliness, and Redlichkeit On the Character of Williams’s Relationship to Nietzsche David Owen Bernard Williams’s relationship with Nietzsche’s philosophical work was long and increasingly intense. Although Williams’s explicit written engagement with Nietzsche’s philosophy takes place in the later part of his career,1 that Williams had a keen interest in Nietzsche’s salience for modern moral philosophy is made apparent as early as his 1968 lecture ‘Universities: Protest, Reform, and Revolution’, when he remarks apropos student demands for curriculum reform: To take for a moment my own subject, which notably, though by no means uniquely, raises these problems: it would be an excellent idea for a philosophy department to provide, if it could, classes on certain thinkers not widely studied in this country, such as Nietzsche. But if, as things are, there is no one in the department who really cares about Nietzsche, to whom he means very much, it will be pointless to make them do it: a course on Nietzsche cannot be just ‘laid on’. Nietzsche is to some extent a special case, but the same point applies more generally, and particularly for the sorts of subjects for which there is at present a lot of demand. (2014a: 61) Apart from his revealing choice of Nietzsche as the relevant exemplar of what it would be good to teach in widening the philosophical curriculum beyond the parochial frame of English philosophy in the post-​war period, these comments suggest that Williams has at least begun to engage with Nietzsche sufficiently to feel able to express the claims that ‘a course on Nietzsche cannot be just “laid on” ’ and that ‘Nietzsche is to some extent a special case’. Thus, although Michael Tanner is rightly reported as saying that Williams was dismissive of Nietzsche in the 1960s (O’Grady 2003),2 it seems clear that by 1968 this had at least begun to change and, at some point in the 1970s, Williams apparently contemplated 1 It is primarily concentrated in Shame and Necessity, Truth and Truthfulness, and some late essays reflecting on naturalism in ethics and on genealogy, as well as in Williams’s introduction to the Cambridge University Press edition of Nietzsche’s The Gay Science. 2 And I can confirm this report since Tanner also communicated this to me on the occasion of examining Tamsin Shaw’s PhD thesis (where Williams had been one of Shaw’s supervisors). David Owen, Ethics, Untimeliness, and Redlichkeit In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. [p. 250] writing a book on Nietzsche: ‘I remember spending a summer reading him in Italy, and writing a lot of notes. But I came to the conclusion that I didn’t know how to write a book about Nietzsche’ (Williams 1996: 15).3 Williams may later have changed his mind on this score,4 but regardless of that, his engagement with Nietzsche’s work and the issues that Nietzsche’s philosophy raises for us, reflective moderns, continues to deepen through the remainder of his life and become more central to his self-​understanding. As Williams wryly recounts of a conversation with Habermas: [Habermas] is very keen to get his opponents categorized, he wants to know which class of opponent you belong to. He and I had a long discussion and he was very worried whether I was an Aristotelian or a Wittgensteinian, and I said I don’t think I’m either Aristotelian or Wittgensteinian and he said: ‘What are you then?’ I was too modest to say: ‘I’m me’, or something like that so I said, ‘How about I’m Nietzschean?’, and that really did put the petrol on the fire. (1999: 246) Yet it is important to note—​as Williams will remark in an interview from 1994—​ that this engagement with Nietzsche has a particular character: I don’t see Nietzsche as the sort of philosopher whose views you just adopt: there are all sorts of problems with his positive views about the future, about politics, and so on. I see him in the same way that Foucault saw him, as a sort of resource. Foucault said that there isn’t any one Nietzsche: everybody gets out of him what they find most helpful. I’m also convinced, from my own experience, that you get most out of him when you’ve got part of the way there on some path of your own. I think it’s arriving at some thoughts of my own which turned out to be not dissimilar to things that Nietzsche had developed in greater depth that has greatly increased my interest in him. (1994: 9) In particular, Williams saw that an abiding concern in his own work was closely related to a thought articulated by Nietzsche: I suspect that there is one idea, or perhaps obsession, which does tie together a number of the things I’ve been interested in. It’s related to a phrase of Nietzsche’s 3 Maudemarie Clark (2015: 42n3) reports Williams as having told her that the project foundered on the problem of Nietzsche’s style (and, given how fundamental Williams takes the problem of style to be for writing ethics, this was not a minor issue), but Williams also commented that Nehamas’s book Nietzsche: Life as Literature would have helped him work through this issue. 4 Mark Philp (the series editor) had a discussion with Williams about his doing the volume on Nietzsche in the Oxford University Press ‘Founders of Social and Political Thought’ series in the 1990s. Philp reports that ‘in typically wry fashion he said that if he wrote a book on Nietzsche then he’d be happy to have it in the series’ (personal communication). He didn’t and, in a twist, I am currently writing this volume. [p. 251] David Owen about becoming what you are. One thing that has continued, in various forms, to interest me is the question of what constraints, what sorts of authority, there are over ways in which one might develop, ways in which one’s life might develop. Are those constraints somehow given internally, given by an ethical order, or given by something you already are? (1994: 4) These diverse notes and reflections by Williams concerning Nietzsche raise the question of how we should aim to engage with Williams’s relationship to Nietzsche—​and thereby also a more general issue for philosophy’s engagement with its history. In this chapter, the opening section will briefly note some of the ways in which Nietzsche appears in Williams’s thinking prior to Ethics and the Limits of Philosophy (ELP), which work can be seen as marking his increasing recognition of affinities which will come to fruition in the transition of Williams’s philosophy from a broadly Humean to a distinctively Nietzschean sensibility.5 My purpose here is twofold: first, to demonstrate that Williams’s engagement with Nietzsche is of longer standing than typically acknowledged; and, secondly, to highlight the specific areas where Williams found Nietzsche to be of value as a fellow-​traveller. In section 2, I focus on the question of the character of Williams’s ‘turn’ from Hume to Nietzsche that is marked by ELP, before considering in section 3 how we should understand the character of Williams’s relationship to Nietzsche. Sections 4–​6 develop the view that I defend by addressing, respectively, Williams’s engagement with classical Greek literature and philosophy, with issues of naturalism and genealogy, and with truth and truthfulness. 1. Nietzschean Intimations in Williams’s Early Work There are a number of passing references to Nietzsche in Williams’s work prior to ELP. Thus, for example, in his 1972 book Morality, Williams references Nietzsche in objecting to what he characterizes as the Aristotelian enterprise of trying draw out moral ends or ideals from the distinctive features of human nature (1993: 60). Another en passant use of Nietzsche to highlight a point that Williams will develop in his own way comes with the epigraph, taken from Twilight of the Idols, to the essay ‘A Critique of Utilitarianism’: If we possess our why of life we can put up with almost any how.—​Man does not strive after happiness; only the Englishman does that. (TI Maxims and Arrows, cited in Williams 1973: 77) 5 This is no doubt too broad and bald a claim, but it is a useful heuristic for my purposes. [p. 252] While it is possible that Williams is simply taking rhetorical advantage of Nietzsche’s capacity for pithy and witty put-​downs, we should notice that this aphorism speaks to a central issue in Williams’s critique. Most obviously, it draws attention to the point that people have projects and some of these projects—​what Williams here calls ‘commitments’ and will later call ‘ground projects’—​are what agents build their lives around and take to ground the value of their lives (and, indeed, of living). ‘Commitments’ are constitutive of ‘our why of life’ and, as Nietzsche stresses, given such ground projects ‘we can put up with almost any how’ (note, importantly, the qualified nature of this claim: some may be beyond our ability to bear). Happiness is not an aim, on Nietzsche’s view, and treating it as such (even in negative mode as the desire to abolish suffering) is both an intellectual and ethical error.6 This is not to say that Nietzsche anticipates the specific argument that Williams advances against utilitarianism. It is to say that Williams’s criticism that utilitarianism constructs a picture of the moral agent, which necessarily fails to acknowledge that the mature agent is identified with his actions as expressive of commitments that give meaning and value to his life, has a close affinity with the concerns that Nietzsche raises. A sign of Williams’s deepening engagement with Nietzsche comes with his review of Iris Murdoch’s book The Fire and the Sun: Why Plato Banished the Artists, in which Williams criticizes Murdoch’s reading of Plato (effectively her assimilation of Plato to Kant) and contends that, once deprived of a conception of moral order as transcendental, the view that ‘goodness lies in correct perception and freedom from error’ leads, in an ironic reversal of its original character, towards ‘cosiness’: Only connect, understand others, you cannot will evil. It is a pity—​though, granted much else of what he said, hardly a surprise—​that Nietzsche has not yet succeeded in persuading us what a hopeless thought that is. (2014b: 145) Nietzsche’s attack on modern morality as a ‘religion’ of snug cosiness (Behaglichkeit) (GS 338) is bound up with his rejection of the view that suffering is intrinsically bad7 and his diagnosis of the tendency towards nihilism in modern culture that he represents as finding its expression in the figure of ‘the last [or ultimate] man’ who has no desire to engage in, and has become incapable of committing to, ground projects (Z prologue). Williams’s deliberate use of the term ‘cosiness’ in criticism of Murdoch is surely intended to register this point. Indeed, Williams’s engagement with Nietzsche’s reflections on nihilism is further evidenced in his 1980 review of 6 Part of Nietzsche’s more general point in remarks such as this is directed against what he takes to be a characteristic feature of modern morality: namely, that it is a morality of ‘selflessness’ in the dual sense of promoting selfless actions and doing so through an impartialist conception of moral reasoning tied to an idea of impersonal objectivity. See Janaway (2007). 7 For a sophisticated discussion of this view, see Janaway (2017a & 2017b). [p. 253] David Owen Gloudsblom’s Nihilism and Culture alongside Lasch’s The Culture of Narcissism, which concludes thus: At the extremes of his prophetic disgust, Nietzsche did not foresee the precise deformations of social and personal life that Lasch holds up for castigation. But it may be that he understood a great deal about their structure. Was he right in his basic reaction—​that the fault must lie with the Socratic aspiration to moral truth and reflexive understanding? And if so, is there any way of overcoming that drive to ever greater self-​consciousness without invoking a catastrophe? These questions are more pressing than ever. (2014c: 173) What is manifest here is Williams’s clear commitment to the view that Nietzsche, as philosophical cultural critic, raises questions that are of central importance to contemporary ethical culture. Williams’s later work on ethical confidence, reflection and the loss of ethical knowledge, and vindicatory genealogy is, in part, a way of engaging with these questions. That Williams has been engaged, with what appears to have been increasing attention, with Nietzsche’s thinking in the period from 1968 to 1980 should now be clear, but this engagement will take a more explicit turn in 1981.8 Probably the central event in this shift is Williams’s reading of Alasdair MacIntyre’s After Virtue with the central opposition that it constructs for modern ethics (and ethical culture) as the choice between Aristotle or Nietzsche. While Williams regards MacIntyre’s claim ‘that this is the only choice we have’ as ‘one of MacIntyre’s most illuminating exaggerations’, he is also clear about where he stands: If, at our present juncture, we do have to choose between Aristotle and Nietzsche, it may rather be Nietzsche who is right: it has finally got out that there is something wrong with morality. (2014e: 186) Indeed, in the same year, Williams makes the following—​ quite uncharacteristically—​unqualified claim in the course a review of three books on Nietzsche: It is certain, even if not everyone has yet come to see it, that Nietzsche was the greatest moral philosopher of the past century. This was, above all, because he saw how totally problematical morality, as understood over many centuries, has become, and how complex a reaction that fact, when fully understood, requires. 8 Leiter (2022: 31n20) reports that ‘Paul Russell tells me that he thinks Williams only became seriously engaged with Nietzsche after reading Alasdair MacIntyre’s After Virtue [1981].’ I have given some reasons to be sceptical of this reported thought, although much here depends on how much one builds into the concept ‘seriously’. I would agree that the intensity of Williams’s engagement with Nietzsche increases after 1981 for reasons that will become clear as we continue. [p. 254] To help himself understand it, he resourcefully explored, in twenty years of increasingly hectic activity, our feelings about art, guilt, violence, honesty, and indeed every element of that moral consciousness which the Greeks helped to invent. (2014d: 183) It is important to recognize that Williams’s sense of why Nietzsche is so important is not independent of the fact that Nietzsche’s concerns are closely bound up with Williams’s own views. What is clear, though, is that by 1981 Williams has come to see Nietzsche as a central resource for the development of his philosophy and this will be increasingly manifest in its direction and projects for the remainder of his career. 2. From Hume to Nietzsche I once had a great admiration for Hume. Now I think that he suffered from a somewhat terminal degree of optimism. Nobody’s who’s got to 1999 can take it that seriously. (Williams 1999: 256) In the light of the previous section, we can reasonably surmise that Williams’s engagement with Nietzsche plays some role in this transformation of his attitude towards Hume. But for the purposes of this section, it will be useful to approach this issue by asking why Williams admired Hume, what it is about Hume’s optimism that leads Williams to turn away from him, and how this is related to his turn towards Nietzsche. Addressing these questions will provide a basis for understanding the salience that Nietzsche has for Williams’s philosophy that can then be developed by attending to his post-​ELP texts. The grounds for Williams’s admiration of Hume can be elucidated by noticing that Hume offers an approach to ethics that responds to what Williams takes to be the fundamental predicament of ethics that modernity raises and an alternative to the other, dominant, form of response which he will label ‘the morality system’. Thus, Williams begins ELP by drawing attention to a structural predicament faced by modern ethics. The question of the justification of ethical life can be asked either from a perspective that is ‘inside’ ethical life or from a perspective that is ‘outside’ ethical life: the former asks what reasons we have for continuing to live such a life, where the first-​personal reasons invoked may draw on the ethical dispositions that agent takes to be part of who they are; the latter asks why we should take up ethical life at all, where the third-​personal reasons invoked take those ethical dispositions as objects of evaluation. For Aristotle, the virtuous agent experiences no conflict or tension between inside and outside perspectives because, on Aristotle’s theory, there is a view of ‘a certain kind of ethical, cultural and indeed political life [p. 255] David Owen as a harmonious culmination of human potentialities, recoverable from an absolute understanding of nature’ (Williams 2006a: 59). However, we (moderns) have no reason to believe that any such natural harmony exists and: Once we lose the belief . . . a potential gap opens up between the agent’s perspective and the outside view. We understand—​and, most important, the agent can come to understand—​that the agent’s perspective is only one of many that are compatible with human nature, all open to various conflicts within themselves and with other cultural aims. With that gap opened, the claim I expressed by saying that agent’s dispositions are the ‘ultimate supports’ of ethical value takes on a more sceptical tone. It no longer sounds enough. I believe that claim is true . . . At the same time, we must admit that the Aristotelian assumptions which fitted together the agent’s perspective and the outside view have collapsed. No one has yet found a good way of doing without those assumptions. (2006a: 59) In the light of this predicament and the condition of ‘ethical vertigo’ (Sagar to which it gives rise, one response is to surrender the project of giving an account of a fully developed human life and to adopt a way of justifying ethical life in terms of ‘morality’ through an appeal to rational agency in which ‘morality’ presents itself to the rational agent as a categorical demand. On this view, there is no tension between an inside and outside perspective since the reasons to be moral are intrinsic to rational agency as such. ‘Morality’ can be seen here as involving two claims: a claim to be comprehensive with respect to the domain of ethical value and a claim to be normatively authoritative with respect to that domain. Morality’s supposed hegemony in turn renders unintelligible, and hence serves to suppress, the thought that there may be a viable ethical outlook distinct from morality, giving rise to ‘the powerful feeling that morality just is the ethical in rational form’ (Williams 1995a: 246). One reason that Hume is important and admirable for Williams is that his philosophy offers an alternative to, and a critical view of, this type of response to the predicament of ethics in modernity. Hume argues that ethics is grounded in the sentimental dispositions and sympathetic reactions of humankind. In sketching out this view, Hume provides reasons for those who are inside ethics to affirm the value of living an ethical life, and also aims to provide reasons to show that recourse to a view of ethics grounded in naturalistic moral psychology can ward off the sceptical threat posed by the gap between ‘inside’ and ‘outside’ perspectives (Russell This Humean alternative, as Greco (2007) in particular highlights, aligns with key features of Williams’s own project: the account of internal reasons and rejection of external reasons, scepticism towards the comprehensive claims of 9 Russell’s point here is, in part, a rebuttal of Sagar (2014) and the argument that Hume views the inside/​outside gap as illusory. [p. 256] ‘morality’ and a rejection of its privileging of obligation, its emphasis on the voluntary, and the purified system of blame that it deploys as well as a sceptical attitude towards the project of ‘ethical theory’ which it exemplifies. Given these commonalities, what reasons does Williams have to turn away from Hume—​and towards Nietzsche? One part of the answer to this question concerns the continuity between Hume’s philosophical approach and Nietzsche’s. The alignment just noted between Hume’s ethical thinking and that of key features of Williams’s project applies equally to Nietzsche.10 But this continuity simply heightens the salience of the question: ‘What is the difference Nietzsche makes?’ Williams’s stated reason for this shift concerns his rejection of what he takes to be Hume’s ‘terminal degree of optimism’, which Nietzsche doesn’t share. In what respect, then, is Hume’s ethical thinking inappropriately optimistic? There are two dimensions along which Williams may be taken to make this charge. As Russell remarks, on Williams’s view: Hume is insufficiently impressed by ethical diversity and the limits of human nature—​or philosophy—​to dictate any particular form of ethical life, much less adjudicate among them when they come into conflict. Related to this, Hume is too hopeful that ethical life, in its various forms, neatly and reliably integrates with human needs and interests of a broader kind. (2019: 276–​7)11 Both Nietzsche and Williams can be seen as situated within a context in which the claims to authority of religion that Hume’s philosophy aims to dismantle (Craig no longer command confidence—​God is dead—​but this condition acutely raises the question of whether ethics, in the form of the morality system or more generally, can sustain our allegiance and of whether we have reason to endorse the general orientation of philosophy towards the view that human beings and the world are designed for each other. Williams’s objection to Hume’s optimism, and the reason for his decisive turn to Nietzsche, may finally be this: Hume’s sanguine confidence that our ethical concepts will prove stable under genealogical reflection. . . . Hume thought that our sense of morals ‘must certainly acquire new force when, reflecting on itself, it approves of those principles, from whence it is derived, and finds nothing but what is great and good in its rise and origins’. This sunny picture contrasts with the darker one painted by Nietzsche, 10 Beam (1996, 2001) has drawn attention to significant commonalities between Hume’s and Nietzsche’s philosophical approaches and concerns, while Kail (2009, 2016) has provided a rich account of the parallels (and differences) between their projects of providing naturalistic genealogical explanations of human thought and practice, especially in relation to ethics. 11 There is a further question of whether Williams’s view of Hume is right, but that is not my concern here. The engagements of Greco (2007), Sagar (2014), and Russell (2019) provide insightful and contrasting views of this issue. [p. 257] David Owen who is more pessimistic about our prospects of being left with a sufficient number of ethical concepts to sustain a meaningful life once we subject them to truthful scrutiny. (Queloz 2021: 188–​9) Williams’s turn to Nietzsche lies in his sense that they share this post-​Humean scepticism and Nietzsche has developed some helpful philosophical resources and strategies for addressing it. But what kind of relationship is this? 3. The Character of Williams’s Relationship to Nietzsche For any philosopher, as for any artist, the question of their relationship to a predecessor whom they take to be significant for their work may have various possible answers. One is that they adopt the predecessor’s substantive philosophical views and revise, extend, and defend them. Another is that their work is informed and shaped by the questions and problems identified by the predecessor. A third is that the predecessor supplies them with tools for taking forward their own concerns. That Williams does not stand in the first kind of debtor relationship to Nietzsche is reasonably clear. Indeed, Leiter (2022) argues, with characteristic robustness, that Williams’s debt to Nietzsche is illusory and betrays only a superficial understanding of Nietzsche’s philosophy.12 Leiter’s is what we might call ‘the scorecard model’ for judging relationships; he proceeds by evaluating whether Williams is committed to Nietzsche’s project and views (as Leiter reconstructs them), and, at least figuratively, entering a cross in the box when he finds, as he does, that Williams’s project and views do not cohere with those of Leiter’s Nietzsche. Leiter’s starting place is the reasonable point that Nietzsche and Williams appear prima facie to be engaged in different enterprises: Nietzsche is engaged in social and cultural criticism in a way that is more akin to figures like Marx, Freud, Adorno, and, more recently, Foucault than the work of someone like Williams. Nietzsche is criticizing moral culture; Williams is merely criticizing moral philosophy (Leiter 1997, 2022). This is a fair point as far as it goes, but the distinction between the critique of moral culture and of moral philosophy is not a categorical one. Nietzsche’s criticisms of philosophical views such as Kantianism and utilitarianism are part of his critique of moral culture because these philosophical views are intellectual expressions of that moral culture. Williams’s criticisms of the philosophical views that he groups under the label of ‘the morality system’ are part of a critique of contemporary moral culture on the same basis. While it is importantly the case that Williams takes 12 Leiter identifies his task as being ‘to assess the Williams–​Nietzsche relationship, the extent to which Williams learned from Nietzsche, and the extent to which he retreated from or ignored Nietzsche’s views’ (2022: 17). This appears to construe learning from Nietzsche as adopting his views. [p. 258] ‘morality’ not to be ‘an invention of philosophers’ but ‘the outlook, or, incoherently, part of the outlook, of almost all of us’ (2006a: 194), he is explicit about the fact that his criticisms are directly aimed at the philosophical errors integral to ‘morality’ that appear most clearly when we consider, to borrow Leiter’s words, ‘ “morality” as conceived, systematized, and refined by philosophers’ (1997: 257): Many philosophical mistakes are woven into morality. . . . Its philosophical errors are only the most abstract expressions of a deeply rooted and still powerful misconception of life. (Williams 2006a: 218) Criticizing these ‘abstract expressions’ of the ‘misconception of life’ manifest in our moral culture is a way of criticizing that moral culture. Williams’s practice is no way identical to Nietzsche’s, but it is certainly continuous with one dimension of it.13 Nor, as we will see, is it straightforwardly the case that Williams is otherwise not engaged in cultural criticism. Leiter also claims as a further categorical difference that there is a sharp distinction between Nietzsche’s focus on the Extraordinary Life and his view that ‘morality’ is only a problem with respect to those few potentially great individuals whose flourishing is being made impossible by modern moral culture ‘as a matter of empirical fact’ (2022: 24; original emphasis), on the one hand, and Williams’s more egalitarian orientation, in which his objections to ‘morality’ have salience to lives that are full of ordinary projects, on the other. If Leiter’s reading of Nietzsche is right, then there is certainly an important difference here, but it is a mistake to think that it lies in Williams’s being unconcerned with the possibility of the Extraordinary Life in contemporary moral culture. On the contrary, one of Williams’s deepest concerns is with the way in which the morality system denies the radically first-​personal character of our ethical lives and, in doing so, acts to repress what Williams takes to be central to ethical life: namely, the call to become what you are (in Nietzsche’s phrase). Williams may disagree with Leiter’s Nietzsche on whom the audience of the critique of ‘morality’ is, not least because he may reasonably regard it as less determinate or fixed in advance than Leiter’s Nietzsche does, but he certainly shares Nietzsche’s view that ‘morality’ is inimical to the highest forms of human flourishing for the simple reason that it is inimical to all forms of human flourishing.14 13 As Krishnan and Queloz note, Williams reiterates a more general version of this point in his 1990 Preface to the French translation of ELP, ‘noting that “the critique of ethical theory does not speak only to ethical theory”, since ethical theory “is not itself the basic condition with which we should be concerned, but a symptom, the expression of that condition in the tissue of a certain type of philosophy” ’ 14 Williams’s view may be taken to resonate with the Emersonian dimension of Nietzsche’s thinking in which the extraordinary is seen less as a matter of projects of remarkable value than of the discipline and commitment to pursuit of ‘one’s genius’ through a ground project that expresses one’s deepest values. [p. 259] David Owen A rather different approach to thinking about Williams’s relationship to Nietzsche, which focuses less on whether Williams adopts the substance of Nietzsche’s philosophical views (or Leiter’s version of them15) and more on whether Williams learns from Nietzsche’s efforts at diagnosing and attempting to address the problem of modern ethical culture (including its philosophical expressions), is found in Krishnan and Queloz (2023). The focus here is more on the second (and, to a lesser extent, third) forms of indebtedness. The reading of ELP that Krishnan and Queloz offer is particularly interesting because it illustrates the way in which this work as a whole can be taken as enacting a form of indirect cultural criticism that is Williams’s response to what he sees as a problem for doing ethical reflection in modernity that is sharply raised by Nietzsche: namely, the problem of style.16 It is a feature of Nietzsche’s work that he takes the question of style to be intrinsic to the question of engaging in ethical reflection and criticism, not least because he comes to recognize that his project of re-​evaluation needs to bring his audience to recognize the need for such a project. Williams faces that challenge and the further one of not writing, after Auschwitz and in the face of possible nuclear war, in a way that collapses into sentimental kitsch. Developing their reading of Williams from detailed attention to the two epigraphs to ELP (both taken from writers deeply engaged with Nietzsche)—​a passage from Wallace Stevens’s poem ‘Esthétique du Mal’ and a French sentence from Albert Camus’ novel La Chute—​Krishnan and Queloz sensitively demonstrate how attending to these orientates us to the self-​consciously post-​war location of Williams’s thinking as well as to its Nietzsche-​inflected character. The epigraph drawn from Camus’ novel La Chute is this: ‘Quand on n’a pas de caractère, il faut bien se donner une mèthode’ (cited in Krishnan and Queloz 2023: 229). In the context of the novel, this is explicitly a reference to the Holocaust as an exemplar of the primacy of method and implicitly to modern moral culture as a form of life that, in valuing method over character, leads to the atrophying of capacities for ethical excellence. This epigraph thus picks out a central theme of ELP that Camus, Nietzsche, and Williams all share: namely, the priority of character to method. Both Williams and Nietzsche are committed to the primacy of virtue, even as they reject Aristotle’s account of ethics; thus Nietzsche identifies his outlook with the concept of virtù manifested in, for example, Machiavelli’s work: ‘virtue in the style of the Renaissance, virtù, 15 Notably, Clark (2015) offers a reading of Nietzsche that is more sympathetic to Williams’s views. 16 In the 1990 Introduction to the French edition of ELP, with its revealing title ‘Ethics, a Matter of Style?’, Williams remarks: ‘At certain points in this book, particularly in its attention to philosophies of the ancient world and in its critique of morality, it is possible perhaps to hear a Nietzschean tone. It was not developed with Nietzsche consciously in mind, and certainly not in any close relation to his writings’ (2021: 282, see also Babbiotti et al. 2021). One of the merits of Krishnan and Queloz’ close reading is to show how Williams can be interestingly unaware of how closely his concerns in ELP relate to themes developed by Nietzsche. [p. 260] moraline-​free virtue’ (A §2). The view that this discloses is one that affirms the radically first-​personal standpoint of ethics and is expressed by both Nietzsche and Williams. The second epigraph, drawn from section VIII of Wallace Stevens’ poem, addresses the condition of humanity after the death of god, and the relations of realism, tragedy, and affirmation. As Krishnan and Queloz note, the impress of Nietzsche on Stevens’ thinking is readily apparent in this passage (they note GS and §377), and the passage raises three Nietzschean themes that run through ELP: ‘the modern challenge of finding meaning in a world from which the phantoms have gone (disenchantment); the temptation to install new idols in place of evicted ones (re-​enchantment); and the difficulty of affirming life in full awareness of its horrors (pessimism)’ (2023: 236). The last of these is particularly pressing for Williams, not least since, in ELP, he does not have a cogent response to the question of whether we, reflective moderns, having dispelled God from our picture of the world, can sustain our affirmation of life without embracing idols. Notably, it is in relation to this topic that one of the very few direct references to Nietzsche in ELP occurs—​namely, Williams’s invocation of the idea of a ‘pessimism of strength’ that Nietzsche associates with the Greeks (and Greek tragic drama): One question we have to answer is how people, or enough people, can come to possess a practical confidence that, particularly granted both the need for reflection and its pervasive presence in our world, will come from strength and not from the weakness of self-​deception and dogmatism. (Confidence is not the same as optimism; it could rest on what Nietzsche called the pessimism of strength.) a: 190) Here, Williams’s thinking is situated squarely within the thematic frame of Nietzsche’s reflective engagement with modern moral culture. These two central themes that Williams’s epigraphs highlight for his readers in orienting them to ELP are joined together in the Postscript to that work. As Krishnan and Queloz remark: If the ideas of Christianity and its less obvious legacies no longer make sense to us now, however, ‘some extension of ancient thought, greatly modified, might be able to do so’, Williams thought. . . . He returns to this idea in the Postscript, contending that a historical story could be told to show why, in a Europe in which the pale Galilean has come and gone, ‘very old philosophies may have more to offer than moderately new ones’ . . . The challenge [though] was to reconcile disenchanted realism with what was most alive in the Enlightenment’s legacy, in particular the values of critical reflection and truthfulness, and this meant that the temptation to place new idols in place of evicted ones should be resisted. . . . In [p. 261] David Owen the Postscript to ELP, Williams wrote that the hopes expressed in the book ‘can be compressed into a belief in three things: in truth, in truthfulness, and in the meaning of an individual life’. . . . But perhaps more so even than Stevens and Nietzsche, Williams was impressed by the difficulty of combining these three things. (2023: 237) The Postscript to ELP thus sets the agenda that Williams takes up in Shame and Necessity and in Truth and Truthfulness respectively. My point in highlighting these two contrasting ways of approaching Williams’s relationship to Nietzsche is to draw attention to the fact that they are concerned with different kinds of question. Leiter is marking Williams against his Nietzsche scoresheet. Krishnan and Queloz are attending to subtle lines of inheritance and originality. One issue thus concerns what kind of enquiry we want to conduct and the value of these different enquiries. But there is also a wider philosophical question about what it is to stand in a truthful relationship to a predecessor: is it to imitate or to follow (to borrow Kant’s distinction)? Williams’s relationship to Nietzsche, like Goya’s relationship to Rembrandt or Picasso’s to Velázquez, is, I think, most fruitfully approached in the latter mode, as the remainder of this chapter will attempt to vindicate. And perhaps Williams nudges us in this direction. Apart from his use of Camus and Stevens in ELP, we might note that already at the end of his first attempt at a book, Morality, Williams misquotes17 D. H. Lawrence: ‘Find your deepest impulse, and follow that’ (1993a: 79). This misquote combines two of Lawrence’s injunctions ‘Resolve to abide by your own deepest promptings, and to sacrifice the smaller thing to the greater’ and ‘Try and find your deepest issue, in every confusion, and abide by that’, both drawn from the essay on Benjamin Franklin to which Williams refers (Lawrence 1923: 25–​6), and can be seen as Lawrence’s own gloss on Nietzsche’s idea of becoming what you are (Milton 1987), as Williams is likely to have been aware. This practice of using remarks by thinkers whose thinking has been deeply shaped by their engagement with Nietzsche’s philosophy, but who ‘follow’ rather than ‘imitate’ Nietzsche, suggests that this is how Williams wants us to understand his own relationship to Nietzsche. As he will later remark: Some of the concerns to which [Nietzsche] speaks are going to be better met—​ that is to say, met in a way in which we can better make something of them—​by some quite other styles of thought, and perhaps by some theory that comes from elsewhere; certainly not by theoretical, or again anti-​theoretical, incantations supposedly recovered from Nietzsche himself. (1995b: 66–​7) 17 I am grateful to Matthieu Queloz and Nikhil Krishnan for directing me to this point. [p. 262] 4. Being Untimely: Nietzsche, Williams, and the Greeks It is a feature of both Nietzsche and Williams that they see the ancient Greeks as inhabiting a form of ethical life that was, in certain important ways, in better shape than modern moral culture and its philosophical expressions. Neither believes that this form of ethical life is available to us or that, if per impossible it were, it would be desirable to take it up without fairly significant changes. Nietzsche refers wryly to ‘those trusting and muscular virtues for which we hold our grandfathers in honour—​but also slightly at arm’s length’ (BGE §214) and more dramatically to the ancient ethical ideal as ‘this synthesis of Unmensch and Übermensch’ (GM I §16). But each does hold, first, that it is important that we can acknowledge the value and achievements of Greek ethical life and, secondly, that, despite our distance from the Greeks, there are elements of their outlook that make better sense of our ethical lives than ‘morality’ does and from which we can learn in confronting the pathologies of modern morality. In this section, I will focus on Shame and Necessity (SN), but it should be noted that some the issues Williams raises in SN were already prefigured in a 1981 essay ‘The Legacy of Greek Philosophy’, in which Williams stresses the salience of Nietzsche’s insight concerning the Greeks: But however much [Nietzsche] or we may qualify his account of Greek tragedy and Greek thought, what he pointed to is truly there: Greek philosophy, in its sustained pursuit of rational self-​sufficiency, does turn its back on kinds of human experience and human necessity of which Greek literature offers the purest, if not richest, expression. If there are features of the ethical experience of the Greek world which can not only make sense to us now, but make better sense than many things we find nearer to hand, they are not at all to be found in its philosophy. Granted the range, the power, the imagination and inventiveness of the Greek foundation of Western philosophy, it is yet more striking that we can take seriously, as we should, Nietzsche’s remark: ‘Among the greatest characteristics of the Hellenes is their inability to turn the best into reflection.’ (2006b: 46) This final point is striking not least because it is a distinctive and problematic feature of European modernity as Williams portrays it that it potentially converts everything into reflection—​and, as he argues in ELP, that this can be destructive of ethical knowledge and of ethical confidence. This is an issue to which we will return later, but for now it is time to turn to the enterprise of SN. Nietzsche’s engagement with the Greeks is well known. However, while he certainly held that the ancient Greeks were too distant from us for any general recuperation of their form of life to be an intelligible project, he is committed to the recovery in historical understanding of Greek ethical culture as a point of contrast [p. 263] David Owen to, and critical resource for reflection on, ‘morality’ in the sense that he targets. This contrastive use of Greek ethics is one of the tactics that Nietzsche regularly mobilizes in making both general and specific points about what he takes to be the problems of ‘morality’. But there is a more particular problem that the modern hegemony of ‘morality’ poses for Nietzsche’s project of the re-​evaluation of values to get a grip on its audience: namely, the structuring of the historiography of ethics imposed by the claim of ‘morality’ to be the rational form of ethics—​‘I am morality itself, and nothing besides is morality’ (BGE §202). An important prerequisite for the contrastive use and critical value of Greek ethics is that it should be recognized as being, in some salient sense, on a par with ‘morality’ as a more or less coherent expression of human ethical life. Nietzsche adopts various strategies of disrupting the historiographic imperialism of ‘morality’, but the important point is that he recognizes he needs to do so if his use of the Greeks is to do the work that he needs it to do.18 The salience of these remarks on Nietzsche is that we can reasonably read SN as engaged in a kindred enterprise of attempting to free our understanding of Greek ethical life from a framework of ‘progressive’ interpretation that presupposes, independently and in advance of enquiry, that ‘morality’ is the rational form of ethics. Williams is explicit about this claim: If these similarities [concerning the concepts that we use in interpreting our own and other people’s feelings and actions] between our own ways of thought and those of the ancient Greeks are, in some cases, unobvious, this is not because they arise from a structure hidden in the unconscious, but because they are, for cultural and historical reasons, unacknowledged. It is an effect of our ethical situation, and of our relations to the ancient Greeks, that we should be blind to some of the ways in which we resemble them. (1993b: 2–​3) He is also explicit about the relationship of his project to Nietzsche as one in which Nietzsche constructed the problematic within which Williams envisages his own enquiry (1993b: 9–​11). SN thus has two main tasks. The first is to free us from the grip of a myth propagated by the progressivist interpretation according to which the Greeks had primitive conceptions of action, responsibility, justice, and other key ethical notions that have been displaced by more sophisticated and cogent conceptions with the rise of moral consciousness. The form of Williams’s claim here is straightforward: 1. These interpretations are misleading, historically and ethically, and are so in a systematic way. 18 Owen (2007) addresses how this concern shapes the structure of Nietzsche’s Genealogy. [p. 264] 2. This systematic misinterpretation is generated by judging the ethical ideas and experience of the Greeks in terms of the conceptions of modern morality, which it is presupposed are both clear and valid. 3. But we actually ‘have no clear idea of the substance of these conceptions, and hence no clear idea of what it is that, according to the progressivist accounts, the Greeks did not have’ (1993b: 5–​6). The second, having recovered a realistic view of the Greek ethical outlook, is to consider the ways in which it helps us to make sense of features of our ethical lives that are obscured by the morality system that drives the progressivist interpretation. There are three features to highlight in considering Williams’s relationship to Nietzsche in SN. The first is that Williams’s engagement with Greek tragedy allows him to articulate a view that he shares with Nietzsche in a wider historical and philosophical frame than he has done hitherto. This is his rejection of what may be seen as the general orientation—​or temptation—​of philosophy, and in particular moral philosophy, towards optimism (1993b: 161–​7). As Geuss notes: This optimism has several related aspects. First of all, traditional philosophers assumed that the world could be made cognitively accessible to us without remainder . . . Second, they assumed that when the world was correctly understood, it would make moral sense to us. Third, the kind of ‘moral sense’ which the world made to us would be one that would show it to have some orientation towards the satisfaction of some basic, rational human desires or interests, that is, the world was not sheerly indifferent to or perversely frustrating of human happiness. Fourth, the world is set up so that for us to accumulate knowledge and use reason as vigorously as possible will be good for us and will contribute to making us happy. Finally, it was assumed that there was a natural fit between the exercise of reason, the conditions of healthy human development, the demands of individuals for satisfaction of their needs, interests and basic desires, and human sociability. (2005: 223) Williams is not under any illusions about his ability to rebut this view wholesale, but he does take it that there are indirect methods available to unsettle its grip retail and, perhaps, support its displacement, not least since he thinks that our cultural-​historical moment underwrites an elective affinity with the pre-​Socratic Greeks: In important ways, we are, in our ethical situation, more like human beings in antiquity than any Western people have been in the meantime. More particularly, we are like those who, from the fifth century and earlier, have left us traces of a consciousness that had not yet been touched by Plato’s and Aristotle’s attempts to make our ethical relations to the world fully intelligible. (1993b: 166) [p. 265] David Owen SN is thus a conscious continuation of the challenge to ‘morality’ and the task of finding an ethical language adequate to our experience begun in ELP. The second feature is the importance of shame in our ethical lives that Nietzsche and Williams both stress. The priority of shame over guilt is directly related to the priority of character over method. If we are concerned with the development of character, of ethical virtues or excellences, then shame is pivotal because it is the engine of self-​overcoming. More generally, as Salkever nicely puts it: His use of Greek literary texts to urge the primacy of shame over guilt is in aid of the philosophic project Williams shares with Nietzsche, that of restoring the dignity and importance of the world of appearances against what both men see as the depreciation of that world by idealizing and absolutizing philosophers ancient as well as modern, by Plato as well as Kant. (Salkever 1993) Against this shared background, however, there are differences between Williams and Nietzsche at least in so far as the latter emphasizes the dynamic character of this internalized other within the context of his attention to the agonic character of Greek ethical culture. Nietzsche’s focus on admired exemplars as constituting the standpoint of the internalized other and on the overcoming of such exemplars as involving the transformation of the internalized other suggests that he regards it as a key advantage of shame that its standards develop as one’s character does (thus in the text of ‘Schopenhauer as Educator’, we see the displacement of Schopenhauer by Emerson). Williams’s use, here and elsewhere, of the language of maturity suggests he holds a less agonist view of a healthy ethical culture. While Williams holds that ‘morality’ represses the creation of extraordinary lives in a way that is disvaluable, he does not think, contra Nietzsche in the vein stressed by Leiter, that the goal of culture is creating such lives. The third feature is Williams’s use of Greek literature to make the argument that while ethical life requires a conception of responsibility and the basic elements of responsibility can be identified as ‘cause, intention, state, and response’, ‘[t]‌here is not, and never could be, just one appropriate way of adjusting these elements to one another—​as we might put it, just one correct conception of responsibility’ (1993b: 55). This argument links back to Williams’s earlier work on moral luck, agent-​regret, tragic conflicts, and his objection to the morality system that it cannot make intelligible sense of the ethical weight that the acknowledgment ‘I did it’ may play in the course of our lives. The relationship of these issues to Nietzsche’s thinking can be drawn out by reflecting briefly on Nietzsche’s understanding of freedom in terms of ‘the will to self-​responsibility’ (TI Expeditions What this means is that the agent, in acknowledging their actions as their own, as expressions of their character, affirms that they bear responsibility in relation to the outcomes to which their actions give rise (this is why the sovereign individual ‘promises like a sovereign—​seriously, seldom, slowly’, GM II §2). This [p. 266] is, simultaneously, the ground of what Williams terms ‘agent-​regret’ and what we may call ‘agent-​affirmation’, each of which may be expressed through the spontaneous avowal ‘I did it’, albeit in very distinct tonal registers. This point is critical to Nietzsche’s concern with the affirmation of life since this affirmation must be first-​personal in character; the question of affirmation is one that the daemon asks me about my life (GS 341). We can see here why Nietzsche might identify himself with the Renaissance concept of virtú, with its register of prowess in which both the concepts of virtue and of virtuosity are equally present.19 One part of Nietzsche’s objection to the morality of selflessness and the selflessness of morality is precisely that it fails to register this dimension of our ethical lives and, indeed, structures our relationship to ourselves in a way that undermines our ability to affirm our lives in first personal terms. It seems clear that Williams’s philosophy shares, and is motivated by, this concern, although Williams—​in contrast to Nietzsche—​is much more alert to the fact that we will require different conceptions of responsibility for distinct areas of our ethical lives; thus his point is not just a point about distinct ethical outlooks involving different views of responsibility, but that any cogent ethical outlook will require more than one view of responsibility. 5. Nietzsche, Naturalism, and Genealogy A central suggestion of SN is that the ethical language of pre-​Socratic Greek literature exhibits a richer and more realistic moral psychology than that of the morality system, and a part of Williams’s argument is that articulating a form of philosophical enquiry that does not succumb to the moralized outlook of optimism requires an approach to naturalizing ethics that can help us develop a psychologically realistic account of ethical life. Nietzsche’s remarks on why he prefers Thucydides to Plato point us in the right direction here, Williams holds, because ‘Thucydides’ conception of an intelligible and typically human motivation is broader and less committed to a distinctive ethical outlook than Plato’s’ (1993b: 161–​2). This thought informs Williams’s take-​up of Nietzsche as a resource for working out an approach to naturalizing ethics. What is it that Nietzsche offers here? Williams identifies two contributions that he takes to be significant for his concerns (1995b, The first concerns a general attitude. The second concerns specific explanations that Nietzsche develops. As Williams notes, there is a systematic difficulty in specifying exactly what naturalism demands in relation to ethics, since the top-​down approach of naturalizing ethics presupposes that we know what the relevant terms of natural description 19 Reginster (2015) and Owen (2018) both pick up this aspect of Nietzsche. [p. 267] David Owen are –​and we don’t (1995b: 68–​9 see also 2000: 148–​54). The general attitude that Williams finds helpful has two elements: First, to the question ‘how much should our accounts of distinctively moral activity add to our accounts of other human activity’, it replies ‘as little as possible’ . . . [Secondly,] Nietzsche’s approach is to identify an excess of moral content in psychology by appealing first to what an experienced, honest, subtle, and unoptimistic interpreter might make of human behaviour elsewhere. (1995b: 68) The first element pertains to the demands of a non-​reductive approach (Williams the second glosses the psychological realism that Nietzsche admired in Thucydides. The second thing that Williams takes Nietzsche to offer are examples of such accounts that illustrate specific mechanisms through which new concepts and values emerge. Thus, Williams points to Nietzsche’s account of the slave revolt in morality as an example of how certain general psychological features of human beings can, under a specific set of circumstances, give rise to a special and peculiar picture of human agency that is characterized by ‘a kind of double-​counting’, where the purpose of this account is to provide the conditions of articulacy for ‘a certain purified conception of blame’ (1995b: 71–​2). One way in which this naturalism can be developed is through genealogy as exemplified by Nietzsche’s Genealogy, which aims to show that a particular ethical outlook—​‘ morality’—​is unstable under genealogical explanation (Williams but the same naturalistic approach can also be developed in ways that can bolster our practical confidence through a vindicatory form of genealogy20 that aims to stabilize some feature of our ethical outlook—​and it is to this task that Williams’s final book turns. 6. On Redlichkeit: Nietzsche, Truth and Truthfulness It might justly be argued that Truth and Truthfulness (Williams 2002) demonstrates how the Nietzschean turn in Williams’s work is based in, and develops from, his Humean starting point. Methodologically, it owes something to Hume’s genealogy of justice and something to Nietzsche’s quasi-​historical mode of genealogy (as well as the more contemporary work of Edward Craig). However, its substantive focus is centrally Nietzschean in several respects. My concern here is not to scrutinize the form of Williams’s argument or assess its cogency; rather, I want 20 This view has been systematically developed by Queloz in an important series of articles (2018, and a book (2021). [p. 268] to highlight the relationship to Nietzsche’s concerns of the exercise that Williams aims to carry out. In Williams’s 2001 editorial introduction to Nietzsche’s The Gay Science, he focuses very heavily on Nietzsche’s relentless commitment to truth and truthfulness, his concerns about the relationship of the value of truth to human flourishing, and the nihilistic dynamic of truth and truthfulness he identifies in modern moral culture. It is to these issues that Truth and Truthfulness relates. How should we understand Nietzsche’s commitment to truth-​seeking? A clear statement of this, as Williams also notes, is provided in The AntiChrist: Truth has had to be fought for every step of the way, almost everything else dear to our hearts, on which our love and our trust in life depend, has had to be sacrificed to it. Greatness of soul is needed for it, the service of truth is the hardest service.—​For what does it mean to be honest in intellectual things? That one is stern towards one’s heart, that one despises ‘fine feelings’, that one makes every Yes and No a question of conscience! (AC §50) In the context of Nietzsche’s reflections on the valuing of truth and truthfulness as grounded not in utility but in morality, and his critique of morality and experimental calling into question of the value of truth, we might wonder how Nietzsche can simultaneously endorse this commitment to truthfulness—​yet it is clear from his statements and, even more, from his philosophical practice that he does endorse it. The question that Nietzsche faces is, then, whether it will be possible for the denizens of modern culture to sustain this commitment in the light of the very dynamics of modern culture that he has uncovered. These dynamics pertain to two points. First, we have no reason but a ‘moral’ reason to accept the view that truth automatically coheres with human flourishing and trumps other values. Truth can be bitter and destructive, as Oedipus discovers. ‘How much truth can you bear?’ is often a pertinent question. Second, this moralized grounding of the unconditional value of truth drives the process of its own self-​overcoming and leaves us caught up in a process of dissolution: ‘not to esteem what we know, and not to be allowed any longer to esteem the lies we should like to tell ourselves’ (WP §5; cited in Williams 2001: xix) With this process, our concern for truthfulness seems be undercutting the grounds on which its value has been secured, which is to say that the moral grounds on which we have hitherto valued truth-​seeking are not stable under genealogical explanation. Truth and Truthfulness is directly concerned to address these issues. Methodologically, ‘[Williams’s] explanation of the value of truth is an exercise in human self-​understanding in the sense that it starts out from human dispositions to value the truth, and explains these in terms of their practical value to human life’ (Queloz 2021: 158–​9). Substantively, it aims to provide an explanation of why we [p. 269] David Owen have reason to value truth as an intrinsic value that is grounded not in ‘morality’ but in an account of basic human needs initially developed in a minimal idealized State of Nature model in terms of the needs for Accuracy and Sincerity—​and, then, de-​idealized through recourse to history in order to illustrate how we have made sense of, and developed, the virtues of Accuracy and Sincerity that are constitutive of our valuing of the concern for truth as an intrinsic value. In this way, Williams aims to naturalize our valuing of the value of truth in a way that is reflectively stable. Moreover, this account of truth as an intrinsic value points to how truth is integrally connected to basic human needs without arguing that valuing truth, in some mysterious way, always promotes flourishing or trumps other values. It is not always best to know the truth, but it is always best to value truth as an intrinsic value. It is, I submit, fairly clear how Williams’s view takes up and responds to concerns that Nietzsche raises. What may be less clear is that Williams’s genealogy may be seen as an instantiation of what Nietzsche describes as ‘the youngest virtue’ (Z I: 3; cf. D §456, §556): Redlichkeit. This German word has resonances with honesty, integrity, probity, and frank speech, but Nietzsche’s identification of it as the youngest virtue points to a more specific feature that is drawn out in GS namely, a form of reflective intellectual conscience that drives us to ‘discover everything lawful and necessary in the world’ and thereby enables us to become ‘creators’ (GS §335). It is, Nietzsche tells us, the last virtue available to ‘free spirits’ (BGE §227). The significance of these remarks is to point out that it is the virtue of Redlichkeit that is manifest in the naturalization of ethics that Nietzsche undertakes and the genealogical exercises that he engages in (GS §335)—​and which, as the preceding section illustrates, guides Williams’s reflection on how genealogy can be used to undermine or support practical confidence in features of our ethical outlook. Truth and Truthfulness can thus be seen as an exercise in Redlichkeit that aims to stabilize our commitment to the values of truth and truthfulness against the scepticism that emerges from the self-​undermining of their ‘moral’ foundations. 7. Conclusion Addressing Williams’s relationship to Nietzsche raises a number of questions, foremost among them the issue of how to approach this relationship. I have indicated that there are reasons to avoid the ‘scorecard model’ (what is Williams’s Nietzsche Quotient?) and instead to focus on perhaps more subtle but certainly no less real debts that emerge from Williams’s genuine philosophical engagement with Nietzsche as an interlocutor who shares various intuitions and concerns, who has tried to get to grips with some of the problems and issues that Williams is struggling with, and whose work may provide some leads or paths to pursue and others to avoid. The question for Williams is: what resources does Nietzsche offer for Williams’s project? The question for us is: what difference does Nietzsche make to [p. 270] Williams’s philosophy? I have tried to show that the answers to both questions are that the role Nietzsche plays is significant. Michel Foucault once remarked: For myself, I prefer to utilise the writers I like. The only valid tribute to thought such as Nietzsche’s is precisely to use it, to deform it, to make it groan and protest. Would Williams demur? Hardly: I do think that, on condition that one is ruthlessly selective and entirely disinclined to regard him as an authority, one must regard Nietzsche as an indispensable contributor to any future moral philosophy that has any hope of being worth doing; and—​paradoxical though this may seem—​this is so even if that philosophy is quite opposed to Nietzsche’s own politics and addresses itself, as I think it should, to questions of how to sustain a liberal concern with social justice without the illusions fostered by much of the moral philosophy of modernity. (Williams References Babbiotti, P., Krishnan, N., & Marquier, M. 2021. ‘Commentary to B. Williams’s French introduction to ethics and the limits of philosophy’. Philosophical Inquiries 9.2: 259–​68. Beam, Craig. 1996. ‘Hume and Nietzsche: Naturalists, Ethicists, Anti-​Christians’. Hume Studies Beam, Craig. 2001. ‘Ethical Affinities: Nietzsche in the Tradition of Hume’. International Studies in Philosophy 33.3: 87–​98. Clark, Maudemarie. 2015. ‘On the Rejection of Morality: Bernard Williams’ Debt to Nietzsche’. In Clark, Nietzsche on Ethics and Politics, 41–​61. Oxford: Oxford University Press. Craig, Edward. 1987. The Mind of God and the Works of Man. Oxford: Oxford University Press. Craig, Edward. 1997. Hume on Religion. Shimla: Indian Institute for Advanced Study. Foucault, Michel. 1980. ‘Prison Talk’. In Power/​K nowledge: Selected Interviews and Other Writings, 1972-​1977. Edited by Colin Gordon, 53–​54. New York: Pantheon. Geuss, Raymond. 2005. ‘Thucydides, Nietzsche, Williams’. In Guess, Outside Ethics, 219–​33. Princeton: Princeton University Press. Greco, Lorenzo. 2007. ‘Humean reflections in the ethics of Bernard Williams’. Utilitas 19: 311–​25. Janaway, Christopher. 2007. Beyond Selflessness: Reading Nietzsche’s Genealogy. Oxford: Oxford University Press. Janaway, Christopher. 2017a. ‘Attitudes to Suffering: Parfit and Nietzsche’. Inquiry 60.1–​2: 66–​95. Janaway, Christopher. 2017b. ‘On the Very Idea of “Justifying Suffering” ’. Journal of Nietzsche Studies 48.2: 152–​70. 21 I am grateful to the editors for the invitation to contribute to this volume. Thanks to Brian Leiter for sending me a copy of his own essay on this topic, to Chris Janaway for helpful comments on an early draft, to Peter McLaughlin for alerting me to the essay by Krishnan and Queloz, and to Matthieu Queloz for detailed and helpful suggestions on the draft chapter. [p. 271] David Owen Kail, P. J. E. 2009. ‘Nietzsche and Hume: Naturalism and Explanation’. Journal of Nietzsche Studies 37: 5–​22. Kail, P. J. E. 2016. ‘Hume and Nietzsche’. In Paul Russell (ed.), The Oxford Handbook of David Hume, 755–​76. Oxford: Oxford University Press. Krishnan, Nikhil, and Matthieu Queloz. 2023. ‘The Shaken Realist: Bernard Williams, the War, and Philosophy as Cultural Critique’. European Journal of Philosophy 31.1: 226–​47. Lawrence, D. H. 1923. Studies in Classic American Literature. New York: Thomas Seltzer. Leiter, Brian. 1997. ‘Nietzsche and the Morality Critics’. Ethics 107.2: 250–​85. Leiter, Brian. 2022. ‘Williams’s debt to Nietzsche: Real or illusory?’ In Morality and Agency: Themes from Bernard Williams. Edited by A. Szigeti and M. Talbert, 17–​ 35. New York: Oxford University Press. Milton, Colin. 1987. Lawrence and Nietzsche: A Study in Influence. Aberdeen: Aberdeen University Press. O’Grady, Jane. 2003. ‘Sir Bernard Williams’. The Guardian, 13 June. Owen, David. 2007. Nietzsche’s Genealogy of Morality. Abingdon and New York: Routledge. Owen, David. 2018. ‘Nietzsche’s Antichristian Ethics: Renaissance Virtu and the Project of Re-​evaluation’. In Nietzsche and the AntiChrist. Edited by Daniel Conway, 67–​88. London: Bloomsbury. Queloz, Matthieu. 2018. ‘Williams’s Pragmatic Genealogy and Self-​Effacing Functionality’. Queloz, Matthieu. 2020. ‘From Paradigm-​Based Explanation to Pragmatic Genealogy’. MIND Queloz, Matthieu. 2021. The Practical Origins of Ideas: Genealogy as Conceptual Reverse-​ Engineering. Oxford: Oxford University Press. Queloz, Matthieu. 2022. ‘Genealogy, Evaluation, and Engineering’. The Monist 105.4: 435–​51. Reginster, Bernard. 2015. ‘Nietzsche, Proficiency, and the (New) Spirit of Capitalism’. Journal of Value Inquiry 49.3: 453–​77. Russell, Paul. 2019. ‘Hume’s Optimism and Williams’s Pessimism: From “Science of Man” to Genealogical Critique’. In Ethics beyond the Limits: New Essays on Bernard Williams’ Ethics and the Limits of Philosophy. Edited by Sophie Grace Chappell and Marcel van Ackeren, 37–​ London: Routledge. Sagar, Paul. 2014. ‘Minding the Gap: Bernard Williams and David Hume on Living an Ethical Life’. Journal of Moral Philosophy 11.5: 615–​38. Salkever, Stephen. 1993. Review of Shame and Necessity. Bryn Mawr Classical Review, 20 April, . Williams, Bernard. 1973. ‘A Critique of Utilitarianism’. In Williams, Utilitarianism: For and Against, 77–​155. Cambridge: Cambridge University Press. Williams, Bernard. 1990. ‘L’éthique, question de style?’ Translated by M.-​A . Lescourret. In Williams, L’éthique et les limites de la philosophie, v–​xix. Paris: Gallimard. Williams, Bernard. 1993a [1972]. Morality: An Introduction to Ethics. Cambridge: Cambridge University Press. Williams, Bernard. 1993b. Shame and Necessity. Berkeley: University of California Press. Williams, Bernard. 1994. ‘An Interview with Bernard Williams’. Cogito 8.1: 3–​19, . Williams, Bernard. 1995a. ‘Moral Luck: A Postscript’. In Williams, Making Sense of Humanity and Other Philosophical Papers, 1982–​1993, 241–​7. Cambridge: Cambridge University Press. Williams, Bernard. 1995b. ‘Nietzsche’s Minimalist Moral Psychology’. In Williams, Making Sense of Humanity and Other Philosophical Papers, 1982–​1993, 65–​78. Cambridge: Cambridge University Press. Williams, Bernard. 1996. ‘A Fugitive from the Pigeonhole: Interview by John Davies’. Times Higher Education Supplement 1252, 1 November. Williams, Bernard. 1999. Seminar with Bernard Williams. Ethical Perspectives 6.3–​4: 243–​65. Williams, Bernard. 2000. ‘Naturalism and Genealogy’. In Morality, Reflection, and Ideology. Edited by E. Harcourt, 148–​61. Oxford: Oxford University Press. [p. 272] Williams, Bernard. 2001. ‘Introduction’. In Nietzsche: The Gay Science. Edited by Bernard Williams, vii–​x xii. Cambridge: Cambridge University Press. Williams, Bernard. 2002. Truth and Truthfulness: An Essay in Genealogy. Princeton: Princeton University Press. Williams, Bernard. 2006a. Ethics and the Limits of Philosophy (Reissued edn). Abingdon: Routledge. Williams, Bernard. 2006b. ‘The Legacy of Greek Philosophy’. In The Sense of the Past: Essays in the History of Philosophy. Edited by Myles Burnyeat, 3–​48. Princeton: Princeton University Press. Williams, Bernard. 2014a. ‘Universities: Protest, Reform and Revolution’ (1968). In Essays and Reviews 1959–​2002. Edited by Michael Wood, 55–​69. Princeton: Princeton University Press. Williams, Bernard. 2014b. ‘The Fire and the Sun: Why Plato Banished the Artists by Iris Murdoch’. New Statesman (1977). In Essays and Reviews 1959–​2002. Edited by Michael Wood, 142–​6. Princeton: Princeton University Press. Williams, Bernard. 2014c. ‘The Culture of Narcissism by Christopher Lasch & Nihilism and Culture by Johan Goudsblom’. London Review of Books (1980). In Essays and Reviews 1959–​ Edited by Michael Wood, 169–​73. Princeton: Princeton University Press. Williams, Bernard. 2014d. ‘Nietzsche on Tragedy by M. S. Silk & J. P. Stern, Nietzsche: A Critical Life by Ronald Hayman, & Nietzsche vol. 1: The Will to Power as Art by Martin Heidegger, translated by David Farrell Krell’. London Review of Books (1981). In Essays and Reviews 1959–​ Edited by Michael Wood, 179–​184. Princeton: Princeton University Press. Williams, Bernard. 2014e. ‘After Virtue: A Study in Moral Theory, by Alasdair MacIntyre’. Sunday Times (1981). In Essays and Reviews 1959–​2002. Edited by Michael Wood, 184–​6. Princeton: Princeton University Press. Williams, Bernard. 2021. ‘Ethics, a Matter of Style? Introduction to the French edition’. Philosophical Inquiries 9.2: 269–​84. [p. 273] The Sense of the Past Williams and Collingwood on Humanistic and Scientific Knowledge Giuseppina D’Oro and James Connelly 1. Scientism, Narrow and Broad One of the important questions which has faced philosophy since the rise of modern science concerns the relation between scientific knowledge and humanistic understanding. The growth and success of natural science has given rise to a view known as ‘scientism’, a philosophical conviction in the epistemic superiority of science and its right to impose its methods onto the territory of the humanities. This in turn has caused a backlash against the alleged epistemic superiority of scientific method, a backlash aiming to expose scientific enquiry as a historical enquiry embedded in human practices, thereby purportedly showing that the reality scientific method uncovers is a historical construct, akin to the cultural norms by which human beings have led their lives in different periods of time. In this debate between proponents of the right of science to extend its methods onto the territory of the humanities, and denigrators of science who deny it is a sui generis mode of inquiry, Bernard Williams and Robin George Collingwood occupy the sensible and much neglected middle ground. They teach us that ‘scientism’ should not be understood merely as the view that the methods of science can be extended to answer the questions addressed by the humanities or as ‘a matter of putting too high a value on science in comparison with other branches of learning or culture’ (Sorell x). This is the most common use of the term, but scientism should also be understood more broadly as the attempt by any form of knowledge to impose its methods onto the subject matter of any another. On this broader understanding of ‘scientism’, historicism, understood as the claim that all knowledge, including scientific knowledge, is historical, is a topsy-​turvy form of epistemic imperialism which denies the sui generis nature of scientific knowledge. In so doing it reverses the power relations between the humanities and science, without challenging the epistemic power structures as such. As critics of scientism in both the narrow and the broader sense, Williams and Collingwood challenge the idea that it is possible Giuseppina D’Oro and James Connelly, The Sense of the Past In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. [p. 274] to understand scientific reality through the lens of the humanities (as if nature were a social/​historical construct); such attempts are just as misguided as the proposal to understand culture through the lens of scientific method. There is increasing awareness of Williams’s critical engagement with Collingwood. An account of the philosophical influence that Collingwood may have had on the work of Williams would be a worthwhile enterprise in itself, but this is not the goal of this chapter; nor is our goal simply that of highlighting the affinities and differences in their philosophical outlooks in the spirit of a ‘compare and contrast’ exercise. The aim is rather to show that Williams and Collingwood make an important contribution to the question concerning the relation between humanistic and scientific knowledge and how the concerns of science and those of the humanities bear upon our understanding of the past. They have a great deal to say that is of enduring philosophical significance both about the point and value of understanding the past from a humanistic point of view, and about what it really takes to reject scientism. Section 2 explores Williams’s and Collingwood’s views of the point and value of understanding the past historically. Both Williams and Collingwood thought that the value of engaging in historical research is misunderstood by those who fail to see its point. The point of understanding the past historically, or from a humanistic point of view, is not to hold it up as a mirror in which the concerns of the present are reflected. It is rather to understand it in its own terms, thereby uncovering what is peculiar to it. Williams’s remarks on this topic occur in the context of his engagement with the question concerning the relation in which philosophy stands to its history and how this relationship differs from that in which science stands to the history of science. Collingwood’s reflections on this topic occur in the context of his philosophy of history and how the past should be understood if it is to be understood historically. But they both agree that the categorial structures and concerns of the present are not the right filters through which to see the historical past because such structures only distort past historical reality rather than enabling us to understand it appropriately. Williams and Collingwood therefore speak with one voice against the presentist orthodoxy according to which the past can only be known from the perspective of the present. Section 3 considers Williams’s critical assessment of Collingwood’s metaphilosophy and suggests that (certain differences notwithstanding) both Williams and Collingwood stood united in condemning all forms of what we might call ‘knowledge imperialism’, irrespective of which discipline happens to assert its dominance beyond the boundaries of its domain of inquiry. To establish that Williams and Collingwood were effectively allies in the struggle against scientism (in both the narrow and the broad sense), one needs to remove a great deal of interpretative debris and to challenge two interpretative myths. The first, discussed in section 4, is that Williams’s appeal to what he calls an ‘absolute conception’ of reality requires reintroducing a form of unreconstructed realism in the [p. 275] Giuseppina D’Oro and James Connelly domain of scientific knowledge. The second (a myth to which Williams himself succumbs) is that Collingwood advocated a form of radical historicism which denied the sui generis nature of scientific knowledge. This is discussed in section 5. Once these interpretative myths are dismantled, as we conclude in section 6, Williams and Collingwood can be seen as articulating a sophisticated critique of scientism which avoids the error of replacing one ‘ism’ (as in scientism) with another (as in historicism). 2. The Point and Value of Understanding the Past Historically Both Williams and Collingwood were suspicious of ‘presentism’, the view that the past is necessarily known from the perspective of the present, and its corollary: namely, that there is not just one past but many, because the past is rewritten anew by each new generation of historians (White 1973; Ankersmit 1983; Jenkins Roth 2020). At a time when this sceptical view about the possibility of knowing the past has become rather prevalent and is indeed now widely seen as the most appropriate contemporary approach, Williams and Collingwood defend the alternative view that understanding the past in its own terms is in fact possible: hence they stand out as champions of common sense against the excesses of postmodernism. They both believed that historical understanding is just a special case of understanding others (Williams 2002: 249; Collingwood 1946: 219), and that while the task of understanding past agents may be harder, as there are more epistemic obstacles in the way (lost records, fewer documents), it is in principle possible to see the world as past agents did. The point of understanding the past historically, they tell us, is precisely to see what is peculiar to it, by understanding it in its own context, from the perspective of past agents, not from that of contemporary historians (Williams 2002: 236 ff.). Williams and Collingwood undertake to explain the nature of historical understanding by comparing the relationship in which the present stands to the past in science and in the humanities. The history of science, Williams claims, is ‘vindicatory’. A vindicatory history is one where the transition from an earlier outlook to a later one involves an improvement that can be recognized as such by both the earlier and the later outlook (Williams 2006c: 189). Since the progress of science corrects the errors of the past, past scientific theories may often have to be discarded as false and the history of science becomes irrelevant to the practice of science. This is not so in the humanities. Unlike natural science, philosophy (which Williams regards as a humanistic discipline) has no vindicatory history, notwithstanding thinkers such as Marx and Hegel who thought that such vindicatory histories could be written (Williams 2006c: 191). Unlike the case of science, therefore, philosophy’s history is relevant to the discipline itself; past philosophers still have important things to say to us; they do not simply stand corrected as [p. 276] philosophy progresses, in the way in which past scientific theories are falsified by the progress of science. Yet, Williams argues, if we are to learn something from past philosophers, we must first understand them. And if we are genuinely to understand them, Williams warns, we should not ‘treat something written by Plato, for instance, as though it had come out in Mind last month’ (2006c: 181; 2006e: 258). The philosophers of the past can be worthwhile conversational partners, but in order to converse with them one must first understand what it is they are saying, and to understand them one cannot directly transpose their claims into the argumentative context of the reader. Such crude transpositions create straw men which can all the more easily be knocked down. He quotes approvingly Collingwood’s claim, in An Autobiography (Collingwood 1939: 63), that ‘in ethics, a Greek word like dei cannot be legitimately translated by using the word “ought”, if that word carries with it the notion of what is sometimes called “moral obligation” ’ (Williams That Williams should focus on this particular example is not surprising since much of his critical engagement with modern moral philosophy requires distinguishing the meaning of ‘ought’ in the context of the Socratic question of how one ought to lead one’s life from the context in which the deontological and consequentialist questions concerning how one ought to act are raised (Williams 1985, a). But the example has a more general significance as it is meant to illustrate the historiographical principle that ‘the point of reading philosophers of the past is to find in them something different from the present’ (Williams 2006f: 344). Williams is very critical of the attitude that many analytic philosophers display towards the history of philosophy. They are said either to believe that the history of philosophy stands to philosophy as the history of science stands to science, or to treat the philosophers of the past as interlocutors in contemporary debates without taking due care to interpret their claims in context. In the first instance they deem the history of philosophy to be irrelevant to the contemporary practice of philosophy, and simply ignore it, just as practising scientists normally regard the history of science to be irrelevant to contemporary scientific practice. In the second instance they misconstrue the claims of earlier philosophers who are then treated with condescension and subjected ‘to instruction by current philosophical methods, and . . . reproved for their errors—​errors to which they have been committed, typically, by the way in which analytical philosophy interprets them’ (Williams 2006e: 258). Ignoring the historiographical principle that ‘the point of reading philosophers of the past is to find in them something different from the present’ (Williams 2006f: 344) is to fail to understand what the value of studying the history of philosophy is: namely, to gain a heightened awareness of one’s standpoint by standing back and becoming estranged from it. This is something that would not be possible without the encounter with a viewpoint that is radically different from one’s own. Those who fail to see the point of engaging in the study of the history of philosophy (i.e. to understand the voices of the past in their own [p. 277] Giuseppina D’Oro and James Connelly terms) therefore also fail to see its value (gaining a heightened awareness of one’s own assumptions). Williams rejects the account of historical understanding that is found, for example, in the philosophical hermeneutics of Hans Georg Gadamer The point of studying past texts in the history of philosophy is not to fuse the categorial horizon of the text with that of the interpreter (as Gadamer claims) but to let the text emerge in its strangeness so as to be able to confront oneself with a position that is genuinely novel. Nor does the point of studying texts from the past lie in redescribing them from the perspective of the present, as is implied by the narrativist approach that has dominated the philosophy of history since the publication of Danto’s Analytic Philosophy of History (1965). Williams likewise rejects the idea that the past that is of historical concern is a practical past (White 2014). For when the concern with a past text is a practical one, the text is selectively plundered for the insights which can retrospectively be seen as having had an impact on later philosophers (Williams 2006e). What Williams is criticizing here is the ‘presentist’ approach to the past, where the past is constructed from the perspective of the present in a way that prevents the encounter with something that is quite radically different. Writing about philosophy historically, Williams says, ‘cannot be identified with the history of influence . . . What we must do is to use the philosophical materials that we now have to hand, together with historical understanding, in order to find in, or make from, the philosophy of the past a philosophical structure that will be strange enough to help us to question our present situation and the received picture of the tradition, including those materials themselves’ (Williams A consideration that is sometimes invoked to motivate presentism is that a concern with the past in its own terms or for its own sake turns history into the study of dusty antiques with no connection or relevance to present interests. Commenting on Croce’s historicism, for example, Hayden White says: History ‘taught’ nothing at all, Croce maintained; and the only thing that the theory of history could legitimately teach was that while history gave information about the past, it could never say anything about the true nature of the present world. It could give insights into what had been vital and what had been moribund in any given past age, but it could say nothing about what was living and what was dead in the present age. (White 1973: 401) Whether or not Hayden White’s criticism of Croce was at the forefront of Williams’s mind, Williams was sensitive to the point it raises. He addresses it by drawing a distinction between the history of ideas and the history of philosophy (Williams e: 257). The history of ideas ‘naturally looks sideways to the context of a philosopher’s ideas, in order to realize what their author might be doing in making those assertions in that situation’. The history of philosophy ‘is more concerned to relate a philosopher’s conception to present problems, and is likely to look at his [p. 278] influence on the course of philosophy from his time to the present’ (2006e: 257). The distinction shows that while a genuinely historical concern with past philosophical texts must preserve an ‘ “untimely” perspective on our philosophical concerns’ (2006e: 261), this need not come at the expense of making the past irrelevant to the present, as suggested by Hayden White’s criticism of Croce. Preserving the distance between the past and the present does not entail that the past, when understood in its own terms, becomes something dead and sterile. On the contrary, understanding the past in its own terms estranges us from present categories; this ‘alienation effect’ (van Ackeren 2019) is a condition sine qua non for a critical reflection on the concepts and categories of the present. Were we not open to a different sense of the term ‘ought’, for example, we would not be able to gain the critical distance required to reflect on our own use of the term. Anachronistic presentism in relation to the study of the past may be a good regulative ideal for science, whose history is vindicatory, but not for philosophy, where the goal is not to produce Whig histories which legitimate present conceptions, but to enable us to stand back from, and gain a critical perspective on, the categories of the present. Advocates of the presentist approach sometimes claim that viewing the past from the perspective of the categories and concepts of the present is a condition of the possibility of taking a critical stance on the claims contained in past texts. Williams was critical of this sort of anachronism. He regards ‘the idea of treating philosophical writings from the past as though they were contemporary’ to be not merely false but ‘unintelligible’ in so far as it is self-​undermining (Williams If one abstracts entirely from their history—​including in this both the history of their context and the history of their influence—​one has an obvious problem of what object one is even supposed to be considering. One seems simply to be left with a set of words in some modern language . . . (2006e: 258) Williams rejects the suggestion that a commitment to anachronism is a condition of the possibility of adopting a critical angle and judging one period by the standards of another, as it is sometimes argued (see e.g. Rée 1991). Those who say that understanding a claim in its historical context entails endorsing that claim (as true) are guilty (at least in so far as they are motivated by this assumption) of what Williams elsewhere calls the ‘strangely tempting fallacy’ embedded in the fat oxen principle, according to which ‘who drives fat oxen must himself be fat’ (Williams [1972] 2021: 29). Just as the person who drives the fat oxen need not be fat, so the interpreter of past texts need not believe the claim contained in past texts (to be true) simply because she understands the context to which the claim belongs. He states that while the ‘triumphant anachronism’ (Williams 2006e: 258) of ‘the more extreme forms of analytic philosophy’ (2006e: 259) removes the distance between the past and the present, ‘there is a good deal of the history of [p. 279] Giuseppina D’Oro and James Connelly philosophy that uses analytic techniques, and yet is genuinely and non-​arbitrarily historical’ (2006e: 259). Arguably, Williams is precisely the kind of sophisticated analytic philosopher with a strong sense of the past who deploys the argumentative techniques of analytic philosophy while displaying sensitivity to concepts and categories that are unfamiliar to us. For Williams, therefore, rejecting the epistemic priority of the present for the sake of preserving that ‘untimely perspective on our philosophical concerns’ does not require renouncing the right to pass any critical judgement on the texts of the past; nor does it necessitate turning the past into an antique, whose study bears no relation to present concerns. A slightly different consideration fuelling the suspicion that a humanistic concern with the past may not allow for the development of a critical stance towards it arises because of Williams’s commitment to what he calls the ‘relativism of distance’ (2006a). There is no point, for Williams, in engaging critically with past forms of life that are no longer genuine options for us, such as the lives of the medieval Samurai or the Greek Bronze Age chief. The fact that they are no longer genuine options undercuts the motivation for a critical engagement with them because such disputes, were they to arise, would be ‘notional’, not ‘real’ (2006a). This ‘relativism of distance’, Williams claimed, is the truth in ethical relativism (1981, 2006a). The precise nature of William’s relativism of distance, whether or not the sheer fact that a form of life is no longer a live option provides good reasons for exempting it from the evaluative language of right and wrong (McDowell 1986), and whether or not the distinction between real and notional confrontations is a sustainable one (Scheffler 1987), has been the object of an extensive discussion that is beyond the scope of this chapter (see Jenkins ff. for an overview of this discussion). But be this as it may, it is unlikely that Williams’s relativism of distance was intended to be a close cousin of what normally goes under the name of cultural relativism. The latter is a form of subjectivism which claims that what is right and wrong is determined by what a culture believes to be right or wrong. On this view, other cultures cannot be criticized because what is right (and wrong) is not independent of what is believed to be right or wrong. Williams is very critical of this ‘unashamedly crass’ relativism (Williams [1972] 2021: xix), which he criticizes both for being unable to explain the fact of moral disagreement ([1972] 2021: 16 ff.) and for its hypocritical attachment to ‘a nonrelative morality of toleration’ ([1972] 2021: 21). Since it is this ‘vulgar’ ([1972] 2021: 25) relativism that rules out the possibility of a critical engagement with other cultures, Williams’s relativism of distance may not undermine a commitment to the idea that the claims of the past can be subjected to critical scrutiny, as van Ackeren (2019) and Ng (2019) have argued, but it is not the goal of this chapter to address this issue. Collingwood too approaches the question of what it means to understand the past historically by considering the relationship in which the present stands to the past in the humanities and in science. The presupposition which governs the [p. 280] study of history (history, for Collingwood is the paradigmatic example of a humanistic discipline) is that one must approach the past as if it were a different country, with different rules and regulations, different expectations as to how one ought to behave. While it is a presupposition of natural science that nature is uniform, that water freezes at 0 °C whether it be in the reign of Queen Victoria or in the Middle Ages, historians studying an ancient civilization cannot assume that the legal, epistemic, and moral norms of the present apply to the past they are investigating. These different presuppositions reflect the distinctive explanatory goals of history and the natural sciences. When historians study the past, they interpret the actions of past agents in the light of what Collingwood calls the ‘context of thought’. The historical significance of Caesar’s crossing of the Rubicon, for example, can only be understood in the context of Republican law (Collingwood The crossing is an historically significant occurrence because it challenged a legal norm which forbade a general to cross a border with his army. Its significance could not be gauged merely by observing certain bodily movements, such as the wading of a small stream by a bunch of men on horses, within a purely extensional context of explanation. It is this difference in the presuppositions of historical and scientific knowledge that Collingwood captures by claiming that science is concerned with the outside of the event, whereas history is concerned with its inside (1946: 213). What Collingwood meant to capture, by drawing the distinction between the subject matters of history and natural sciences along the lines of the inside/​outside distinction, is that historical explanations necessarily invoke an intensional context, while scientific explanations abstract from such a context. His view was that unless one knew the intensional context—​in this case, for example, what Roman law specifically proscribes—​one could not properly grasp why Caesar’s action is historically significant (Ahlskog and D’Oro 2021; D’Oro and Ahlskog 2021). Unfortunately, Collingwood’s unguarded use of the inside/​outside metaphor to distinguish the presuppositions of history and natural science, made him into an easy target for what Williams calls the ‘clinically literal minded’ (2006c: 183). Collingwood’s account of historical understanding (re-​enactment) was thus easily read as providing a methodological instruction to historians on how to access the ‘inside’ of actions, the inner unobservable aspect of thought, by rethinking the thoughts inside the minds of historical agents. This interpretation of re-​enactment as a methodological tool in turn gave rise to the question as to how historians could possibly gain access to the mind of historical agents unless Collingwood naively assumed telepathic powers that enabled the historians to empathize with the agents. It is to Williams’s credit (Williams 2006f ) that he did not succumb to this preposterous reading of Collingwood’s account of historical understanding. Williams saw with unusual clarity that the doctrine of re-​enactment, which constitutes an important pillar of Collingwood’s philosophy of history, was advanced not as an epistemological tool to acquire knowledge of hidden inner psychological processes, but rather as a conceptual claim about [p. 281] Giuseppina D’Oro and James Connelly what it means to understand the past historically.1 To understand the past historically is to locate the event in the intensional context or the thought-​context of the historical agents, not that of the historian.2 The view that informs Collingwood’s philosophy of historical understanding—​ namely, that the conceptual/​ categorial mediation which informs historical knowing is that of the agents who inhabited the past, not that of the historian—​has many affinities to Williams’s own views. Collingwood is very critical of what he calls scissors-​and-​paste historians (1946: 257 ff.). These historians approach the past through the filter of the present. If, for example, they come across a witness statement attesting to the occurrence of miracles, they ask themselves whether such a statement could be true or false. Having ruled it out as unreliable by their own epistemic standards, they discard the testimony as false. Rather than asking ‘what does the statement mean?’, they assess it for truth or falsity in the light of contemporary norms (Collingwood 1946: 260). In so doing they import the presuppositions of natural science into the investigation of the historical past. Natural scientists, as we have seen, presuppose their subject matter (nature) to be unchanging; the historical study of the past, by contrast, needs to be sensitive to variations in the norms by which historical agents lead their lives. The past, according to Collingwood, is studied both by natural scientists and by historians, but it is not studied in the same way. A palaeontologist who traces the evolution of an animal species through the study of its fossilized remains is concerned with what happened in the past; but he is not a historian in the sense in which that term is normally used: that is, to denote a humanistic concern with past reality as it was conceptualized, for example, by the Egyptian, Greek, or Roman civilizations. It is in the nature of scientific inquiry to presuppose that past scientific conceptions of reality which do not match up to the standards of present scientific ones are to be discarded as false (as Williams might put it, the history of science is ‘vindicatory’). The historical study of the past makes no such assumptions because the goal of the historian is not to evaluate whether the belief systems of past agents are true or false, as if they were competing scientific theories; it is rather to invoke such beliefs in terms of rationalizations that make the actions of these agents intelligible. When the presuppositions of natural science are illegitimately imported into the historical study of the past, they give rise to the kind of scissors-​and-​paste histories of which Collingwood is highly critical and generate precisely the same errors with which Williams charges a certain kind of analytic philosopher, such 1 The methodological reading of re-​enactment was criticized by many others, such as Skagestad and van der Dussen (1995). 2 Williams approvingly mentions Patrick Gardiner (1996) for pointing out that the re-​enactment doctrine was not prescribing a method to historians, seemingly unaware of the fact that it was the very same Gardiner (1952a, 1952b) who, earlier in life, had been a vocal proponent of the methodological interpretation, according to which re-​enactment is a method that enables historians to pierce through the bodily armour to reach what Ryle called the ‘ghost in the machine’. [p. 282] as interpreting the actions of past agents as if they were their contemporaries, only to poke fun at them for being either epistemically naive or morally perverse, just as the said analytic philosopher reads past philosophical texts as if they had been published in a recent philosophical journal and then quickly ridicules them for their implausible views. There are therefore clear affinities between the principles which inform Collingwood’s philosophy of historical understanding and the principles which inform Williams’s conception of the historiography of philosophy. They see the value of historical inquiry to be akin to the sort of learning experiences one has when travelling to other countries (studying the past opens up one’s mind to different forms of understanding and self-​understanding, just as experiencing other cultures does), and they are united in rejecting the scepticism which inevitably ensues from the presentist view which denies that the historian can suspend disbelief and approach the past through the eyes of the historical agents. But while they both thought that a humanistic perspective on the past requires seeing it from the agents’ perspective, not that of the reader, neither of them inferred from this that the beliefs of past agents, or the claims contained in past texts, cannot be criticized, that whereas science can be critical of its past the humanities cannot. Just as, for Williams, the driver of fat oxen need not be fat, so the historian, according to Collingwood, does not need to believe what past agents believed to be true. To hold that understanding past agents requires believing what they believed (to be true), Collingwood would have said, is to mistake a proposition (which expresses a belief with definite truth values) with a presupposition (to which the notion of truth and falsity do not apply): A man (or at any rate an intelligent man) need not regard himself as insulted if someone who has paid him a sum of money asks him for a receipt, or if the family of a lady whom he is about to marry proposes that a marriage settlement should be drawn up. He knows that the request or proposal is based on the assumption that he is capable, or will one day become capable, of acting dishonourably; but though he knows people assume this he does not necessarily think they believe it. He finds no difficulty in distinguishing their supposing him a rascal and their believing him one, and he does not regard the former as evidence of the latter. (Collingwood 1940: 28–​9) It is the affinities with Collingwood’s own conception of the nature of historical understanding that prevent Williams from making the gross errors of interpretation to which many readers of Collingwood’s philosophy of history have succumbed. As we shall see in the next sections, Collingwood agrees with Williams that the conception of reality with which historians are concerned is mutable, while the conception of reality with which natural scientists are concerned is unchanging: historical knowledge operates under the presupposition that the way [p. 283] Giuseppina D’Oro and James Connelly in which, for example, the ancient Egyptians conceptualized reality is different from the way in which the ancient Romans did; scientific knowledge, on the other hand, operates under the presupposition of the uniformity of nature. The claim that reality as investigated by natural science is unchanging does not imply that scientific inquiry is presuppositionless (the uniformity of nature precisely is a presupposition of natural science). Just as Williams denies that what he calls the ‘absolute conception’ of reality is a conception that requires no conceptual mediation, likewise Collingwood denies that describing nature as unchanging or uniform requires assuming that scientific inquiry is a presuppositionless kind of knowledge. 3. History and Science as sui generis Forms of Knowing Williams’s assessment of Collingwood’s metaphilosophy is much less sympathetic (Leach 2011). He cursorily dismisses An Essay on Philosophical Method (Collingwood 1933) as ‘a dull and dated book’ (Williams 2006f: 343) and writes off An Essay on Metaphysics (Collingwood 1940) as embracing ‘radical historicism’ (Williams 2006f: 358). The perfunctory treatment of An Essay on Philosophical Method and the simplistic reading of An Essay on Metaphysics as defending some sort of historical fundamentalism is somewhat surprising. One might have expected that Williams, as a philosopher deeply interested in the status of philosophy and where it sits in relation to natural science, might have dug a bit deeper into Collingwood’s metaphilosophical vision. Had he done so, he might have discovered an ally in the battle against scientism. Collingwood’s An Essay on Philosophical Method outlines the distinctive conception of the role and character of philosophical analysis that he first began developing in his Lectures on Moral Philosophy (unpublished). The task of philosophy, Collingwood claims, is ‘the distinguishing of concepts . . . coexisting in their instances’ (1933: 51). He illustrates this task by considering different senses of the term ‘good’. Moral philosophers distinguish between the ‘good’ in the hedonistic sense of what is pleasant, the good in the utilitarian/​consequentialist sense of what is expedient, and the good in the deontological sense of what is right. To each of these conceptions of the good there correspond a particular conception of how one ought to act. Drawing such distinctions, Collingwood argues, is not the same as sorting actions into separate empirical classes: the class of actions which are pleasant, the class of actions which are expedient, and the class of actions which are right. To distinguish between two concepts philosophically, therefore, is to disambiguate their meaning, to find them distinct, even in the absence of any empirical differences. As philosophers, for example, we would want to distinguish between an action which is carried out from duty and an action performed on the basis of prudential considerations, even if all actions done out of duty were also expedient. [p. 284] Philosophical distinctions are therefore purely intensional distinctions to which there may correspond no empirical differences. Unlike the hedonistic, consequentialist, and deontological approaches to moral philosophy, Collingwood refuses to regard any one conception of the good as basic. The point of disambiguating the different conceptions of the good is precisely to show that they are distinct. Rather than taking a position on either side of the consequentialist/​deontological fence, Collingwood simply claims that the task of philosophical analysis is to distinguish between these notions of the good. He does not share with much contemporary moral philosophy what Williams calls the impulse ‘to reduce every consideration to one basic kind (Williams [1985] a: 19). It is therefore all the more surprising that Williams dismisses An Essay on Philosophical Method as a stuffy old book, since much of Chapter 1 of his Ethics and the Limits of Philosophy engages in a diatribe against modern moral philosophy and its widespread tendency to prioritize one conception of ‘ought’ at the expense of the other. The rejection of the reductivist impulse to make one kind of consideration basic is, as we shall see, a leitmotif of Collingwood’s metaphilosophy. In his later work Collingwood invokes this conception of the role of philosophical analysis to defend the autonomy of humanistic in distinction from scientific explanations by disambiguating the different senses of the term ‘cause’ that are operative in different forms of inquiry. The conception of the role and character of philosophical analysis that Collingwood developed in An Essay on Philosophical Method, therefore, sows the seeds for his critical engagement with scientism in An Essay on Metaphysics, a theme also close to Williams’s heart, to which we shall now turn. To show that Collingwood’s and Williams’s metaphilosophies are closer than it is often assumed we need to cast doubt on two interpretative suggestions. The first, discussed in section 4, is that Williams’s reference to an absolute conception of reality rests on an unacknowledged commitment to pre-​K antian realism. The second, discussed in section 5, is the widely held view that Collingwood was committed to a form of radical historicism. 4. Williams, Scientism, and the Absolute Conception Scientism is normally defined narrowly as the trespassing of scientific methods onto the territory of the humanities.3 But scientism also refers, more broadly, to 3 Scientism, as discussed in this chapter, is not a first-​order belief or commitment to scientific knowledge, but a second-​order philosophical belief about the unity of science and the applicability of scientific methods to answer, for example, the questions raised in the humanities (see Sorell 1991: ch. 1). A critique of scientism, therefore, is not a critique of science, but of a particular thesis in the philosophy of science, according to which the methods of science are fit to answer all questions. The claim for the unity of science goes back to John Stuart Mill (1843) and was revived by Hempel (1942), but it is still very pervasive. It informs, for example, the claim that a fully fledged neuroscience would render humanistic explanations of actions obsolete. On this understanding of scientism, scientists need not [p. 285] Giuseppina D’Oro and James Connelly the trespassing of any form of knowledge onto the domain of inquiry of another. Williams is a critic of scientism in both the narrow and the broader sense. On the one hand, he is critical of any attempt to assimilate philosophy, which he regards as a quintessentially humanistic pursuit, to science: that is, of scientism in the narrow sense. He believes philosophy stands in a very different relation to its history than the history of science stands to science. First, as we have already seen, philosophy, unlike science, does not have a ‘vindicatory history’. Secondly, while the history of philosophy is not one of progress (in the way in which that of science is), it is not irrelevant to its practice, as supposed by certain analytic philosophers who think ‘that the history of philosophy is no more part of philosophy than the history of science is part of science’ (Williams 2006c: 189). But crucially, Williams is not just critical of scientism in the narrow and most common sense: that is, the attempt to assimilate the humanities to science. He is equally critical of the converse error—​ the attempt to assimilate science to the humanities—​and is thereby opposed to any kind of scientism whatsoever. He rejects the view ‘that natural science constitutes just another part of the human conversation, so that, leaving aside the small difference that the sciences deliver refrigerators, weapons, medicines and so on, they are in the same boat as the humanities are’ (Williams 2006c: 188). He explicitly warns against the incongruity involved in historicizing science by pointing to the paradox of stating that ‘as one historian of science has incautiously said, “the reality of quarks was the upshot of particle physicists’ practice” (the 1970s is rather late for the beginning of the universe)’ (2006c: 190). Failure to acknowledge the sui generis nature of scientific inquiry leads only to the reversal of the power relations between the humanities and science, not to the overcoming of scientism in the broad sense. Replacing one form of epistemic imperialism with another is not the way to leave scientism behind. Williams is therefore just as keen to defend the autonomy of scientific knowledge as he is to defend that of humanistic understanding. Williams’s defence of scientific knowledge as sui generis involves an appeal to what he calls the ‘absolute conception’ of reality. He discusses the idea of an ‘absolute conception’ in his book on Descartes, where he sees Descartes as trying to make sense of the distinction between a conception of the world as it is for us and ‘a conception of the world as it is independently of all observers’ (Williams The absolute conception is introduced in the context of the distinction between primary and secondary qualities, but Williams’s real concern is not with the contrast between primary properties (the scientific image) and secondary qualities (a thin conception of the manifest image that can be derived from the scientific image through the relation of supervenience). His concern is rather with the contrast between the scientific image and the thick moral descriptions of the be (and most often are not) guilty of scientism, and when they are, it is in virtue of committing to a particular philosophical view about the relationship between scientific knowledge and other forms of knowing. [p. 286] world, between a conception of reality according to which ‘there are only universal laws of motion and—​at the physical level at least—​no local kinds of things with local habits’ (Williams 1978: 254) and the conceptions of reality of ‘a Greek Bronze Age chief or mediaeval Samurai’ (Williams 1981: 140) with their varied ethical practices. Ethical descriptions, unlike scientific ones, are ‘local’ or historically parochial. The differences between the conceptions of ethical life of, say, the Greek Bronze Age chief and the medieval Samurai are unlike the differences between the ways in which secondary qualities are perceived by different people, or at different times of day and night. While it is possible, for example, to explain the differences in the perception of a coloured surface as bright by day and dark by night by invoking the angles of refraction, it is not possible likewise to explain differences in ethical practices by invoking thin moral concepts such as ‘good’ and ‘evil’ precisely because one cannot derive what kind of actions a particular form of ethical life will regard as morally praiseworthy or blameworthy in the way in which one might scientifically explain how a particular secondary quality, such as a colour, will be perceived under a different set of experiential circumstances (Williams [1985] 2006a: ch. 8). Contra Jackson (1998), thick ethical concepts for Williams are not ‘entailed’ by thin ethical concepts in the ways in which secondary qualities are entailed (through the relation of supervenience) by primary qualities. Williams’s idea of an absolute conception of reality relies on a contrast between thin naturalistic descriptions of nature and thick ethical concepts that has attracted a great deal of critical attention. The contrast has been criticized for reinstating the fact/​value distinction by confining the correspondence theory of truth to the domain of science, while rejecting it in ethics and, as a result, advancing an overly relativistic conception of the good (Cottingham 2009). Prominent among those who see Williams as being engaged in the project of reintroducing the fact/​ value distinction is Hilary Putnam, who objects to Williams’s claim that, while it is possible to criticize the Aztecs’ (scientific) belief that there exist supernatural beings who get angry if sacrifices are not performed, it is not possible to criticize the Aztecs’ form of ethical life. This, he believes is a double standard: If we can say that the Aztec belief about the gods was false, why can we not say that the practice to which it led was wrong (although, to be sure, understandable given the false factual belief )? If we are not allowed to call the practice wrong, why are we allowed to call the belief false? The so-​called absolute and the ethical are just as entangled as the ‘factual’ and the ethical. (Putnam 1990: 175) According to Putnam, Williams’s separation of science and ethics and the asymmetry between the ability to correct past epistemic beliefs on the one hand, and the inability to correct past ethical beliefs on the other, ultimately rests on a partial endorsement of the correspondence theory of truth with respect to the absolute [p. 287] Giuseppina D’Oro and James Connelly conception of the world and its rejection in relation to the local ethical conceptions of reality: [T]‌he absolute conception of the world was defined in terms of the idea that some statements describe the world with a minimum of ‘distortion’, that they describe it ‘as it is’, that they describe it ‘independently of perspective’—​and what does any of this talk mean, unless something like a correspondence theory of truth is in place? (1990: 174) Williams, however, saw no need to endorse a crude distinction between facts and values, one which requires us selectively to reintroduce the correspondence theory of truth, in order to distinguish between ethics and science. He was incensed by this criticism and returned to it several times. He anticipated Putnam’s objection in his book on Descartes: ‘Can we really distinguish between some concepts or propositions which figure in the conception of the world without observers, and others that do not? Are not all concepts ours, including those of physics?’ and replied there and then: ‘Of course: but there is no suggestion that we should try to describe a world without ourselves using any concepts, or without using concepts which we, human beings, can understand’ (Williams 1978: 244). The notion of historical transcendence at work in the absolute conception is not the same as the chimerical ‘view from nowhere’ for, as Williams says, the challenge is ‘to explain what we might mean by this contrast, not from outside our conceptions, but in terms of reflections we can conduct in human life, the only place (needless to say) in which we can conduct them’ (1991: 12–​13). He revisited and rebuffed this criticism in Philosophy as a Humanistic Discipline, where he again denied that the ‘minimalist perspective’ that science strives to provide transcends historically local perspectives by invoking the God’s-​eye point of view of pre-​K antian dogmatic metaphysics: ‘My aim in introducing the notion of an absolute conception’, Williams says, ‘was precisely to get round the point that one cannot describe the world without describing it, and to accommodate the fundamentally Kantian insight that there simply is no conception of the world which is not conceptualized in some way or another’ (2006c: 185). As A. W. Moore (2006, 2007) points out, in defending an absolute conception of reality, Williams was not advocating the idea of ‘a conception of reality without concepts’ (Moore 2006: xiv). Far from being ‘a conception without concepts’ or a conceptually unmediated grasp of reality, the absolute conception relies on a heuristic principle of the scientific study of reality, the expectation that scientific theories ‘converge upon a “final opinion . . . independent not indeed of thought in general, but of all that is arbitrary and individual in thought” ’ (Williams 1978: 244) The notion of an absolute conception of reality is introduced to defend the sui generis nature of scientific enquiry, by pointing out its distinctive regulative ideals, not to reinstate the correspondence theory of truth via the back door. Without invoking such a notion, it would not be possible to avoid [p. 288] the paradox involved in claiming that quarks, just like flared jeans, did not exist before the 1970s. As we shall see in the next section, Collingwood, like Williams, was opposed to scientism, both narrowly and broadly conceived. Williams did not, however, recognize him as an ally here because, like many others, he dismissed his later metaphysics as being committed to the kind of historical fundamentalism that turns the epistemic power relations between science and the humanities upside down without altering the structures of domination themselves. 5. Collingwood, Scientism, and Radical Historicism Collingwood is perhaps mostly known for his defence of the autonomy of historical understanding against the claim that historical inquiry is not different in kind from scientific inquiry: that is, for his criticism of scientism in its narrow and most common sense. According to Collingwood, history (which he took to be the humanistic discipline par excellence) and science have different presuppositions. When historians study the past, they do not presuppose that the epistemic norms which govern their own times extend to past agents. They therefore approach the past as if it were another country with different customs and mores. By contrast, natural scientists do not approach the past like another country. They do not assume that the laws of nature change from one period to another; on the contrary, they assume them to be the same in the Middle Ages as they are under the reign of Queen Victoria. Natural scientists presuppose that nature is uniform, that water will freeze at 0 °C at all times, because the presumption of the uniformity of nature is a condition of the possibility for carrying out the kind of inductive inferences through which they come to know nature. While the historian expects to find different habits, norms, and practices in different periods of history, the scientist expects the same laws to be operative at all times and places. This is not to say that, since the scientific conception of reality is that of an unchanging system of natural laws, we can say that nature is known in-​itself (so to speak) while history can only be known through the cultural norms of the historical agents. What it means rather is that historical reality and scientific reality are mediated in different ways, that they are known through different presuppositions, which give rise to different kinds of questions, which are answered by the adoption of different methods. Like Williams, Collingwood rejects scientism both narrowly and broadly conceived. He believes the view that ‘all reality is historical . . . to be an error’ (Collingwood 1946: 209). His argument against scientism in the narrow sense (the attempt to impose the methods of science onto the subject matter of the humanities) relies on his argument against scientism in the broad sense (the attempt by any form of knowledge to impose its methods onto the subject matter of another). The answers to questions raised by humanistic disciplines do not conflict with answers to questions in science because they are answers to different questions, [p. 289] Giuseppina D’Oro and James Connelly which rest on different presuppositions. There is therefore no competition between science and the humanities, but only a division of labour. The trespassing of one form of knowledge onto the subject matter of another occurs when one fails to see that there are different kinds of becausal answers which invoke different senses of causation. In An Essay on Metaphysics, for example, Collingwood argues for explanatory pluralism by disambiguating the different senses of the term ‘cause’, just as in his earlier metaphilosophical treatise he disambiguated the concept of ‘good’. The term ‘cause’, he claims, can be used in one sense (what he calls ‘sense I’) to denote what is normally called a reason. This is the sense that is operative in humanistic explanations where to cause means to afford a motive. It is in this sense that the term is used in headlines such as ‘Mr Baldwin’s speech causes adjournment of the house’ (Collingwood 1940: 290). In another sense (what he calls ‘sense II’), the term ‘cause’ is used to denote an event that is open to manipulation. When one says, for example, ‘the cause of books going mouldy is their being in a damp room’ the notion of cause invoked is that of a handle that can be turned either to prevent or to provoke a certain effect, the damp being the preventable cause of the moulding. This sense of causation corresponds to the sorts of explanation which are used in the practical sciences of nature. The notion of causation one invokes corresponds to the kinds of question one is seeking an answer. Just as in An Essay on Philosophical Method Collingwood showed no desire to take a side on the consequentialist/​deontological debate, so in An Essay on Metaphysics he resists the reductivist impulse to regard any one sense of causation as either explanatorily or metaphysically privileged. The upshot of the claim that no sense of causation is either explanatorily or ontologically basic is that it is an error to seek to order forms of knowledge hierarchically by adopting the layered view of the sciences which places physics at the base and then seeks to legitimize the sciences at the higher levels by invoking the notion of supervenience. Philosophical discussions of the problem of causal overdetermination and explanatory exclusion which emerge in the philosophy of mind arise because of a mistaken assumption concerning the relation between the sciences fostered by the view that physics has taken over from metaphysics the title of the science of pure being or being as such. The relation between the mind and the body, Collingwood says, is the relation between the sciences of body, or natural sciences, and the sciences of mind; that is the relation inquiry into which ought to be substituted for the make-​believe inquiry into the make-​believe problem of ‘the relation between body and mind’. (1942: 2.49/​11) The conception of philosophical analysis that Collingwood developed in An Essay on Philosophical Method as ‘the distinguishing of concepts that coincide in their instances’ informs his later argument for the methodological autonomy of the human sciences (D’Oro and Connelly 2020). Read in the light of An Essay on [p. 290] Philosophical Method, Collingwood’s later metaphilosophy supports an argument against scientism, not an argument for the historicization of all knowledge, including scientific knowledge (Connelly 1990; D’Oro 2002, 2023). Given all this, one might expect Williams to have identified Collingwood as an ally in the struggle against scientism and to have viewed An Essay on Metaphysics with a great deal of sympathy. But he did not. Williams’s assessment of An Essay on Metaphysics (Williams f: 345, 358) lies in a long tradition of reading this treatise as advocating the historicization of all knowledge (Rotenstreich 1972; Toulmin 1972; Donagan Moore 2012: 495), including scientific knowledge, and as subscribing to a form of topsy-​turvy scientism in which history becomes the epistemologically first or primary science. While this radically historicist reading does have some foundations in the text, it largely arises as a result of a failure to see the continuities between Collingwood’s earlier metaphilosophical treatise, An Essay on Philosophical Method, and An Essay on Metaphysics. As a result, An Essay on Metaphysics has often been interpreted through the lens of Collingwood’s philosophy of history. This has obscured the nature of Collingwood’s metaphilosophy, whose goal is to establish a division of labour between the humanities and the sciences by arguing that they answer different questions, questions which arise from different presuppositions. The role of philosophy, for Collingwood, is to render explicit the presuppositions on which different forms of knowledge rest, and the distinctive kinds of question they ask. Philosophical knowledge is not first-​order knowledge, such as historical or scientific knowledge. It is second-​order knowledge of the ways in which historians and natural scientists go about acquiring first-​order knowledge in their specific domains of inquiry. Strictly speaking, philosophy does not give us knowledge in the sense in which one gains knowledge by engaging in first-​order inquiries. By studying history or biology, for example, one comes to know something new, something that one did not know before. Philosophy instead deepens our understanding; it ‘brings us to know in a different way things which we already knew in some way . . .’ (Collingwood, 1933: 161). It is a conceptual enterprise whose task is to uncover the kinds of question that are asked, and answered, in different domains of inquiry. The task of history, for Collingwood, is, in the first instance, to understand the past. When a historian comes across a statement in a source, ‘the important question is not whether it is true or false, but what it means’ (Collingwood 1946: 260). For example, it would be ‘folly for the historian’ to rebuke a historical agent who refuses to cross the mountains ‘because he is frightened of the devils in them’ by admonishing him and pointing out that ‘this is sheer superstition. There are no devils at all’ (1946: 317). This judgemental attitude, one that is characteristic of scissors-​and-​paste historians who prioritize the perspective of the present, is out of place in a philosophy of historical understanding. This does not entail that the beliefs of past agents are beyond criticism: Collingwood defends a form of explanatory pluralism according to which the questions of concern to a humanistic approach to the past are different from the questions that are [p. 291] Giuseppina D’Oro and James Connelly of concern to a scientific approach to reality, not a form of historical relativism according to which the past cannot be criticized. It is because Williams (like many others) did not appreciate the continuities between An Essay on Philosophical Method and An Essay on Metaphysics that he read the latter as an argument for the historicization of knowledge, rather than as an argument against scientism. Had Williams read An Essay on Metaphysics as an attempt to defend the disciplinary boundaries between the humanities and the sciences, rather than as an argument for historical fundamentalism, he would have been much more sympathetic to its message. But once this misreading is set aside, it is clear that Collingwood anticipates important themes in Williams’s own critique of scientism. Both Williams and Collingwood see the attempt to reduce scientific knowledge to humanistic understanding to be as misguided as the attempt to reduce humanistic understanding to scientific knowledge. It is the same error, made in reverse gear. Correcting Williams’s reading of An Essay on Metaphysics would be a futile and merely pedantic exercise were it not for the fact that when read as an argument against scientism, rather than as an argument for the historicization of knowledge (including scientific knowledge), Collingwood’s metaphysics of absolute presuppositions offers a way of understanding how one can defend something akin to Williams’s absolute conception of reality without endorsing a crude version of the fact/​value distinction or selectively endorsing the correspondence theory of truth. As we have argued, Collingwood’s distinction between the scientific conception of reality (nature as an unchanging system of laws) and the historical conception of reality (the ever-​changing system of cultural norms) is not a distinction between reality in itself and reality for us, but a distinction between the different presuppositions that govern scientific and historical knowledge. The uniformity of nature is a presupposition of scientific method, not a feature of reality independently of how it is known scientifically. Collingwood’s metaphysics of absolute presuppositions therefore offers a defence of the autonomy of the human sciences that neither threatens the sui generis nature of scientific investigation (by inverting the power relations between the humanities and science) nor requires appeal to unreconstructed realism in science, and the selective reintroduction of the correspondence theory of truth and the fact/​value distinction. Like Williams, he seeks to draw the distinction between an absolute and a local conception of reality from within our own categories and concepts, without stepping outside them. Had Williams seen the deep continuities that link Collingwood’s two essays, he might have been more sympathetic to Collingwood’s metaphilosophical vision and recognized him as an ally in the attempt to navigate the straits of scientism (the attempt to capture the subject matter of the humanities through the adoption of scientific methods) and radical historicism (the attempt to reduce science to a humanistic pursuit). Williams’s work has been criticized for lifting science out of history; Collingwood’s work has been criticized for the opposite reason—​that is, for historicizing scientific knowledge. These criticisms rest on two interpretative [p. 292] myths, which we have sought to dispel. The first is that Williams’s defence of the absolute conception of reality involves the reintroduction of the fact/​value distinction and a commitment to a form of unreconstructed realism, albeit selectively in the domain of science rather than that of ethics. The second myth is that Collingwood’s defence of the autonomy of humanistic explanations involves historicizing science, thereby denying that scientific knowledge is sui generis. Dispelling these myths has enabled us to present Williams and Collingwood as being on the same page with regard to scientism. They are both critics of scientism and they both construe scientism in a broad sense, as a form of knowledge imperialism, irrespective of which form of knowledge ends up wearing the imperialist uniform. From their perspective, historicism, understood as the attempt to historicize science, is just as misguided as the arguably much more common attempt to deny that the humanities have a distinctive method and subject matter vis-​à-​vis the natural sciences. They teach us an important lesson: namely, that in philosophy, just as in politics, one must be wary of stepping into the shoes of the enemy whom one sought to depose. Paradoxically, the reasons why they have both been misunderstood, albeit in opposite ways, is that they were not recognized as seeking to undermine scientism in a broad sense (including the attempt to reduce scientific knowledge to humanistic understanding) as well as in the narrow sense (the attempt to deny the distinctive nature of humanistic understanding). 6. Conclusion: Beyond All ‘isms’ Collingwood and Williams agree that it is the mark of a humanistic discipline to be concerned with a conception of reality that is historically changing, and that a humanistic concern with the past must display sensitivity to the ways in which reality is construed by past agents. For Collingwood, the ‘past’ is an ambiguous term. Just as the concept of ‘good’ may refer to what is pleasant, what is expedient, or what is right, and the concept of ‘cause’ may refer either to a motive or to an antecedent condition, so the term ‘past’ has a different meaning depending on whether it reflects a humanistic or a scientific concern. From a humanistic perspective, the past cannot be approached anachronistically from the point of view of the present because what one is concerned with, when investigating the history of past civilizations, for example, is what was distinctive to those particular forms of life, and this specificity is lost if one treats them as if they shared the same norms as the historian investigating them. But the anachronistic presentism which, when applied to the study of the cultural past, gives rise to the scissors-​and-​paste histories Collingwood is very critical of is perfectly legitimate from a scientific perspective; a scientific concern with the past, we have seen, is governed by the presupposition that the laws of nature, unlike cultural norms, are constant, and thus by the view that nature, unlike culture, has no history. Williams highlights the [p. 293] Giuseppina D’Oro and James Connelly distinction between a humanistic and a scientific concern with the past by saying that the history of science, unlike history governed by a humanistic concern with the past, is vindicatory. This is not to say, of course, that there cannot be a humanistic history of science, one which considers the ways in which the concept of nature has changed from one form of life to another, but that this kind of humanistic history of science is different from the sort of vindicatory history of science that is legitimately anachronistic. Nor does the consideration that history of science is vindicatory, whereas humanistic history is not, entail that the past approached from a humanistic perspective is beyond critical appraisal, that to understand the action of historical agents or the claims contained in the texts of the past in their historical context requires endorsing the context which enables one to understand those claims. To assume this, as we have seen, would be to commit what Williams calls ‘the fat oxen’ fallacy. Collingwood’s and Williams’s heightened sense of the past does not lead them to assert that all knowledge is historical knowledge, for they are both committed to the view that there are legitimate contexts in which presentism is a legitimate heuristic principle. Both Collingwood and Williams seek to keep alive the distinction between the ways in which the history of nature and culture should be investigated, from a humanistic and scientific point of view respectively, without, and this is an important caveat, reintroducing a commitment to realism. Williams does so by defending the idea of an absolute conception of reality, one which draws the distinction between nature and culture without undoing Kant’s Copernican revolution; Collingwood does so by developing a metaphysics of absolute presuppositions. In so far as they seek to preserve the distinction between nature and culture as the distinctive subject matters of the humanities and science, they are, as we have argued, critics of scientism broadly construed (as the overreach of any form of knowing) rather than merely narrowly construed (as the overreach of natural science). They remind us that it takes much more to overcome scientism than the simplistic attempt to reverse the previously existing power relations between science and the humanities, and teach us to be wary of revolutionary changes that install a new dictator without altering the structures of domination. References Ahlskog, J., and D’Oro, G. 2021. ‘Beyond Narrativism: The Historical Past and Why It Can Be Known. Collingwood and British Idealism Studies 27.1: 5–​33. Ankersmit, F. R. 1983. Narrative Logic: A Semantic Analysis of the Historian’s Language. The Hague: Martinus Nijhoff. Collingwood, R. G. Unpublished. Lectures on Moral Philosophy, 1923 Bodleian Library Collingwood dep. 3; 1928? Bodleian Library Collingwood dep. 22/​4; 1929 Bodleian Library Collingwood dep. 10; 1932 Bodleian Library Collingwood dep. 10; 1933 Bodleian Library Collingwood dep. 8. [p. 294] Collingwood, R. G. 1933. An Essay on Philosophical Method. Oxford: Clarendon Press. Revised edition with an Introduction by James Connelly and Giuseppina D’Oro. Oxford: Oxford University Press, 2005. Collingwood, R. G. 1939. An Autobiography. Oxford: Oxford University Press. Revised edition, R. G. Collingwood: An Autobiography and Other Writings, with Essays on Collingwood’s Life and Work. Edited by David Boucher and Teresa Smith. Oxford: Oxford University Press, 2013. Collingwood, R. G. 1940. An Essay on Metaphysics. Oxford: Clarendon Press. Revised edition with an Introduction by Rex Martin. Oxford: Oxford University Press, 1998. Collingwood, R. G. 1942. The New Leviathan. Oxford: Clarendon Press. Revised edition, edited and introduced by David Boucher. Oxford: Clarendon Press, 1992. Collingwood, R. G. 1946. The Idea of History. Revised edition with an Introduction by Jan van der Dussen. Oxford: Oxford University Press, 1993. Connelly, James. 1990. ‘Metaphysics and Method: A Necessary Unity in the Philosophy of R. G. Collingwood’. Storia, Antropologia e Scienze del Linguaggio 5.1–​2: 33–​156 (Rome: Bulzoni editore). Cottingham, John. 2009. ‘The Good Life and the “Radical Contingency of the Ethical” ’. In Reading Bernard Williams. Edited by Daniel Callcut, 24–​42. Abingdon: Routledge. Danto, Arthur C. 1965. Analytical Philosophy of History. Cambridge: Cambridge University Press. Donagan, Alan. 1972. ‘Collingwood and Philosophical Method’. In Critical Essays on the Philosophy of R. G. Collingwood. Edited by Michael Krausz, 1–​19. Oxford: Clarendon Press. D’Oro, Giuseppina. 2002. Collingwood and the Metaphysics of Experience. London and New York: Routledge. D’Oro, Giuseppina, and Connelly, James. 2020. ‘Robin George Collingwood’. In The Stanford Encyclopedia of Philosophy (Winter 2020 Edition). Edited by Edward N. Zalta, . D’Oro, Giuseppina, and Ahlskog, Jonas. 2021. ‘Imagination and Revision’. In The Routledge Companion to History and Theory. Edited by C. M. van den Akker, 215–​32. Abingdon: Routledge. D’Oro, Giuseppina. 2023. Why Collingwood Matters: A Defence of Humanistic Understanding. London: Bloomsbury. Gadamer, Hans-​Georg. 2004. Truth and Method, second edition. Translated by Joel Weinsheimer and Donald G. Marshall. New York: Continuum. Gardiner, Patrick. 1952a. The Nature of Historical Explanation. London: Oxford University Press. Gardiner, Patrick. 1952b. ‘The Objects of Historical Knowledge’. Philosophy 27: 211–​20. Gardiner, Patrick. 1996. ‘Interpretation in History: Collingwood and Historical Understanding’. In Verstehen and Humane Understanding. Edited by Anthony O’Hear, 109–​19. Royal Institute of Philosophy Supplement 42. Cambridge: Cambridge University Press. Hempel, Carl G. 1942. ‘The Function of General Laws in History’. Journal of Philosophy 39: 35–​48. Jackson, Frank. 1998. From Metaphysics to Ethics. Oxford: Oxford University Press. Jenkins, Keith. 1999. Why History? Ethics and Postmodernity. London: Routledge. Jenkins, Mark P. 2018. Bernard Williams. London and New York: Routledge. First published by Acumen 2006. Leach, Stephen. 2011. ‘Bernard Williams on R. G. Collingwood: Distinctions and Assimilations’. Theoretical and Applied Ethics 1.3: 30–​5. McDowell, J. 1986. ‘Critical Notice of Ethics and the Limits of Philosophy’. MIND 95: 379–​86. Mill, J. S. [1843]. A System of Logic, Rationative and Inductive. In The Collected Works of John Stuart Mill. Edited by J. M. Robson, 1963–​ Toronto: University of Toronto Press, London: Routledge & Kegan Paul. Moore, A. W. 2006. ‘Introduction’. In Bernard Williams, Philosophy as a Humanistic Discipline. Edited and with an Introduction by A. W. Moore, xi–​x x. Princeton: Princeton University Press. Moore, A. W. 2007. ‘Realism and the Absolute Conception’. In Bernard Williams. Edited by Alan Thomas, 24–​46. Cambridge: Cambridge University Press. Moore, A. W. 2012. The Evolution of Modern Metaphysics. Cambridge: Cambridge University Press. [p. 295] Giuseppina D’Oro and James Connelly Ng, Geraldine. 2019. ‘The Irrelativism of Distance’. In Ethics beyond the Limits: New Essays on Bernard Williams’ Ethics and the Limits of Philosophy. Edited by Sophie Grace Chappell and Marcel van Ackeren, 188–​208. London and New York: Routledge. Putnam, Hilary. 1990. Realism with a Human Face. Cambridge, MA and London: Harvard University Press. Rée, Jonathan. 1991. ‘The Vanity of Historicism’. New Literary History 22.4: 961–​83. Rotenstreich, N. 1972. ‘Metaphysics and Historicism’. In Critical Essays on the Philosophy of R. G. Collingwood. Edited by Michael Krausz, 179–​200. Oxford: Clarendon Press. Roth, Paul A. 2020. The Philosophical Structure of Historical Explanation. Evanston, IL: Northwestern University Press. Scheffler, Samuel. 1987. ‘Morality through Thick and Thin: A Critical Notice of Ethics and the Limits of Philosophy’. Philosophical Review 96: 411–​34. Skagestad, Peter. 1975. Making Sense of History: The Philosophies of Popper and Collingwood. Oslo, Bergen, and Tromso: Universitetsforlaget. Sorell, Tom. 1991. Scientism: Philosophy and the Infatuation with Science. London and New York: Routledge. Toulmin, Stephen. 1972. ‘Conceptual Change and the Problem of Relativity’. In Critical Essays on the Philosophy of R. G. Collingwood. Edited by Michael Krausz, 201–​21. Oxford: Clarendon Press. van Ackeren, Marcel. 2019. ‘Williams (on) Doing History of Philosophy: A Case Study on Ethics and the Limits of Philosophy’. In Ethics beyond the Limits: New Essays on Bernard Williams’ Ethics and the Limits of Philosophy. Edited by Sophie Grace Chappell and Marcel van Ackeren, London and New York: Routledge. van der Dussen, Jan. 1995. ‘The Philosophical Context of Collingwood’s Re-​enactment Theory. International Studies in Philosophy 27.2: 81–​99. White, Hayden. 1973. Metahistory: The Historical Imagination in the Nineteenth-​Century Europe. Baltimore, MD: Johns Hopkins University Press. White, Hayden. 2014. The Practical Past. Evanston, Illinois: Northwestern University Press. Williams, Bernard. [1972] 2021. Morality: An Introduction to Ethics. Cambridge: Cambridge University Press. Williams, Bernard. 1978. Descartes: The Project of Pure Enquiry. Harmondsworth: Penguin. Williams, Bernard. [1981] 1999. ‘The Truth in Relativism’. In Williams, Moral Luck: Philosophical Papers, 132–​43. Cambridge: Cambridge University Press. Williams, Bernard. 1991. ‘Terrestrial Thoughts, Extraterrestrial Science’. London Review of Books, 7 February: 12–​13. Williams, Bernard. 2002. Truth and Truthfulness. Princeton: Princeton University Press. Williams, Bernard. [1985] 2006a. Ethics and the Limits of Philosophy. London and New York: Routledge Classics. First published by Fontana Press 1985. Williams, Bernard. 2006b. Philosophy as a Humanistic Discipline. Edited and with an Introduction by A. W. Moore. Princeton: Princeton University Press. Williams, Bernard. 2006c. ‘Philosophy as a Humanistic Discipline’. In Williams, Philosophy as a Humanistic Discipline. Edited and with an Introduction by A. W. Moore, 180–​99. Princeton: Princeton University Press. Williams, Bernard. 2006d. The Sense of the Past: Essays in the History of Philosophy. Princeton: Princeton University Press. Williams, Bernard. 2006e. ‘Descartes and the Historiography of Philosophy’. In The Sense of the Past: Essays in the History of Philosophy, 257–​64. Princeton: Princeton University Press. Williams, Bernard. 2006f. ‘An Essay on Collingwood’. In The Sense of the Past: Essays in the History of Philosophy, 341–​58. Princeton: Princeton University Press. [p. 296] Williams’s Debt to Wittgenstein Matthieu Queloz and Nikhil Krishnan 1. Williams and Wittgenstein: The Myth of Unwavering Hostility The main philosophical event in Bernard Williams’s formative period was what Elizabeth Anscombe described as ‘Wittgenstein’s second cut’: the appearance of Wittgenstein’s later work. ‘A philosopher makes a cut’, she wrote, ‘if he makes a difference to the way philosophy is done: philosophy after the cut cannot be the same as before’ (2011: 181).1 Yet it is apparent that Williams was not in any obvious sense a ‘Wittgensteinian’. In his personal manner, he was famously un-​ Wittgensteinian: sociable where Wittgenstein was solitary, egalitarian where Wittgenstein sought hierarchies, transparent where Wittgenstein was esoteric, active in politics where Wittgenstein was conspicuously apolitical. Williams also harboured less hostility to academic philosophy and its conventions, insisting that an interest in philosophy could ‘be driven by straightforward curiosity’ (2001a: xvi) rather than by a quest for salvation. He had, moreover, a great respect for historical scholarship and published extensively in the history of philosophy—​a striking contrast with Wittgenstein’s proud description of himself as ‘a one-​time professor of philosophy who has never read a word of Aristotle!’ (Drury 2017: 65). And while Williams’s prose could tend towards the overly compressed and epigrammatic, it never radically departed from the conventions of the academic essay: conclusions were generally supported by identifiable arguments, and positions distinguished from others in the previous academic literature.2 In none of these senses was Williams any kind of Wittgensteinian. If there is a debt here, it must consist in something more subtle. Williams’s most detailed exegetical discussion of Wittgenstein’s work is his ‘Wittgenstein and Idealism’ (1973c), delivered as part of a 1972–​3 lecture series of the Royal Institute of Philosophy entitled Understanding Wittgenstein. While Williams’s suggestion that there was a ‘pluralized’ transcendental idealism in the later Wittgenstein was greeted with interest and productively developed by some of those well acquainted with Williams,3 the lecture was more widely perceived as 1 Anscombe adopted this image from the Polish Wittgenstein scholar Bogusław Wolniewicz. 2 On the compressed quality of Williams’s style, see Nussbaum (2003) and Babbiotti (2020, 2023). 3 See Moore (2012: 268–​71; 2019), Lear (1982), and Lear’s contribution to Lear and Stroud (1984). Matthieu Queloz and Nikhil Krishnan, Williams’s Debt to Wittgenstein In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. [p. 297] Matthieu Queloz and Nikhil Krishnan a polemical attack on Wittgenstein and provoked a flurry of critical responses. In a recent response of his own, Stephen Mulhall remarks: A number of those well-​acquainted with Williams and his writings have suggested to me that his essay was never intended to have such polemical significance. I find it hard to accept that suggestion, in part because of Williams’s pretty much unwavering hostility to the work of Wittgenstein and his followers in his other writings. n4) Is it true that Williams displays ‘unwavering hostility’ to the work of Wittgenstein in his other writings? Or did Williams have more sympathy for Wittgenstein than Mulhall allows? What, in other words, is the extent of Williams’s debt to Wittgenstein? We shall be arguing that once Williams’s oeuvre is perused for more sympathetic remarks on Wittgenstein’s philosophy, it turns out that hostility is outweighed by sympathy, and the debt deeper than the differences. As the case of Nietzsche reminds us, some philosophers acknowledge their debts by reserving their most unwavering hostility for those to whom they are most indebted. But we are claiming not that Williams treated Wittgenstein as a source of provocation or a handy target, but that many aspects of his philosophy—​his style, his method, and his metaphilosophical views—​can be read as evolving dialectically out of Wittgenstein’s own.4 There are several biographical reasons for expecting Williams to have engaged Wittgenstein sympathetically throughout his career. Two such reasons pertain to his early academic associations. Williams’s teacher at Oxford, Gilbert Ryle, imbibed Wittgensteinian ideas through various channels—​first through his close friendship with Margaret MacDonald, who studied under Wittgenstein before coming to Oxford in 1937,5 and later through his own acquaintance with Wittgenstein.6 Asked about his formative influences in a 1983 interview, Williams identifies Ryle as his principal mentor, whom he credits with instilling in him a liberating wariness of philosophical ‘isms’. But Williams also names David Pears as one of his most important influences, a major figure in Wittgenstein scholarship who was a model for the Wittgensteinian character in Iris Murdoch’s Under the Net.7 4 Wittgensteinian strands in Williams’s thought are also emphasized in Queloz and Cueni (2021), Macedo Jr (forthcoming), and Fricker (2023); Owen (2001) shows how Williams-​style genealogy can be understood in terms of Wittgenstein’s notion of perspicuous representation; Glock (2006) compares and contrasts Williams’s genealogical method with Wittgenstein’s remarks on historical modes of philosophizing; Misak (2021) emphasizes how Williams spent time among the Wittgensteinians and picked up various pragmatist ideas from them. 5 See Kremer (2022). 6 See Tanney (2021: §1). 7 Pears, though only slightly older than Williams, had already begun teaching courses on Wittgensteinian themes that Williams appears to have attended: ‘In the early 1950s David Pears gave a seminar in Oxford together with Geoffrey Warnock on the subject of synthetic necessary truth. . . . This undertaking . . . expressed an attitude to philosophy and its problems which continued to shape [p. 298] The third influence Williams names is none other than Wittgenstein himself. Williams admits to sharing the general excitement surrounding Wittgenstein’s work in the 1950s: Like everyone else then . . . I was interested in the philosophy of that time. It was just when Wittgenstein’s posthumous work was being published. In fact, his Philosophical Investigations came out in 1950, I think. And a lot of this sort of Wittgensteinian literature had been circulated before that—​things that had not been published, but copies of which were in circulation informally, including the ‘Blue Book’ and the ‘Brown Book’, as they were called. Well, like everybody else, I was interested in that philosophy; I was turned on, excited about it. (1983: 41) By Williams’s own account, then, Wittgenstein was an important early influence on his philosophical formation. The fact that he misdates the publication of the Philosophical Investigations (which appeared in 1953) suggests that he was hardly immersed in Wittgenstein scholarship in the 1980s. But in a discussion of Wittgenstein’s work with A. J. Ayer in the early 1970s, Williams speaks of Wittgenstein with admiration, and recognizes Wittgenstein’s influence on himself in no uncertain terms: ‘it would be preposterous and wrong to deny that one had been influenced very much by this work’ (Chanan 1972). Stanley Cavell is another, slightly later Wittgensteinian influence on Williams. The two became friends when Williams spent a few months in Princeton in the spring of 1963. Williams sent Cavell a copy of ‘Morality and the Emotions’, his inaugural lecture at Bedford College, in 1965, and dedicated his 1978 book on Descartes to Cavell and his wife.8 Admittedly, Cavell, especially in 1963, was more in thrall to J. L. Austin’s work than to Wittgenstein’s. This highlights a challenge in characterizing Williams’s debt to Wittgenstein. Given the parallels between Wittgenstein and Austin, is there not a risk of mistaking debts to the latter for debts to the former? After all, Austin, who had, like Williams, been trained as a classicist, and was Williams’s older contemporary at Oxford, was better positioned than Wittgenstein to shape the thinking of the young Williams. Moreover, Austin had proven an effective school builder, counting not only Oxford philosophers such as J. O. Urmson and Geoffrey Warnock among his followers, but also a number of Americans, including Cavell.9 his work, in particular his writing on Wittgenstein’ (Williams 2001a: xiii). On Pears’s influence on Williams, see Krishnan (2023: 265–​6). 8 Cavell records his recollections of that friendship in Cavell (2010: 149–​50, 405–​6, 16, 99–​501). See Babbiotti (2023) for a detailed discussion of the Cavell–​Williams connection that is also informed by their respective Nachlass. 9 See Rowe (2023: 587). [p. 299] Matthieu Queloz and Nikhil Krishnan And in conversations with Cavell in 1955 and 1958, Austin had ‘singled [Williams] out for praise among the young Oxford philosophers’ (Cavell 2010: 149). However, Austin was himself far more deeply influenced by Wittgenstein than he cared to admit (Rowe 2023: 3, 142–​50); and, by the end of his life, Austin was in any case ‘pained to find that the brightest young thinkers in Oxford were becoming disenchanted with his methods and outlook, and that, in particular, he could not attract the interest of Bernard Williams’ (Rowe 2023: 587). Williams himself repeatedly denied that he had been much influenced by Austin (1995d: 220n1; 1999: 3; I was always rather careful of Austin’, he declared in an interview. ‘I kept at a certain distance from him . . . those who got close to him got involved in his style, which I did not find sympathetic’ (Williams 1983).10 At a more substantive level, they disagreed about the direction of British philosophy. ‘I never believed that the problem with British philosophy was that it was liable to metaphysical excess and needed to be cut back’, Williams explained, whereas Austin seemed to him ‘like a Treasury official who thought that the British economy needed deflating, when there were already three million unemployed’ (1999: 143).11 In keeping with his wariness of Austin, Williams consistently associates Cavell with Wittgenstein in his writings, and in particular with the demand that philosophy should listen to what it says (2006j: 207). Williams remarked of Austin that ‘his considerable, though unfinished, contribution is something that one can to some extent take or leave’ (2014a: 46). But he was unequivocal about Wittgenstein’s importance to philosophy: ‘His impact, both on the spirit of philosophy and on some particular issues, was enormous, and cannot be ignored’ (2014a: 46). Nor did Williams cease to make favourable remarks about Wittgenstein’s philosophy later in his career. In a late piece on modernist philosophers, Williams called Wittgenstein ‘the greatest of such philosophers’ (2006c: 119). And in the credo-​like lecture in which he summed up his view of philosophy as a humanistic discipline, he expressed great admiration for aspects of Wittgenstein’s work, declaring that ‘some of the deepest insights of modern philosophy, notably in the work of Wittgenstein, remain undeveloped’ (2006e: 181)—​an observation that blends praise for Wittgenstein’s legacy with regret at what philosophers had so far managed to make of it. Williams was usually careful to exempt Wittgenstein from the hostility he directed at some of Wittgenstein’s followers.12 To bring out the nature and extent of Williams’s debt to Wittgenstein, we begin by considering Wittgenstein as a stylistic influence on Williams, especially as regards ideals of clarity, precision, and depth (section 2). We then examine Williams’s methodological debt to Wittgenstein, in particular his anthropological 10 See Krishnan (2023: 264) for a discussion of what Williams saw as the numbing effect that Austin could have on younger philosophers. 11 For further discussion of the similarities and differences between Austin and Williams, see Krishnan (2023: 264) and Queloz (forthcoming). 12 See, for example, Williams (1995d: 218; 2001a: xvi; 2005e: 35, 7; 2006f: 161). [p. 300] interest in thick concepts and their point (section 3). In section 4, we turn to the fact that, in the 1990s, Williams started explicitly associating himself with a certain form of Wittgensteinianism, which he called ‘Left Wittgensteinianism’. Section 5 shows how this is not a sudden conversion, but the direct product of Williams’s longstanding critical engagement with Wittgenstein’s methodology and metaphilosophy: Williams arrives at this position by envisaging a Wittgensteinianism that thinks in concrete sociohistorical terms, embraces genuine explanation, and relinquishes its insistence on the purity of philosophy. When properly understood, moreover, this critique turns out to be continuous with Williams’s advocacy of a conception of philosophy as a humanistic discipline. In the final section, we show that Williams inherits from Wittgenstein a certain understanding of how philosophy can help us to live, particularly the therapeutic ambition to liberate us from distortions in our self-​understanding by assembling reminders. 2. Resisting Scientistic Ideals of Clarity, Precision, and Depth Williams’s style in philosophy was, to a considerable degree, what his coeval Richard Wollheim called a ‘group style’: a set of dispositions manifested in thought, speech, and writing that he shared with other philosophers, just as shared dispositions united artistic movements such as the Impressionists and the Pre-​ Raphaelites. That group style was what Williams characterized as the ‘analytical’ style: ‘What distinguishes analytical philosophy from other contemporary philosophy . . . is a certain way of going on, which involves argument, distinctions, and . . . moderately plain speech’ (1985: xvi). He added that analytical philosophy rejects ‘obscurity’, but sometimes finds ‘technicality’ necessary to achieve its ends. He himself did not care much about the ‘analytical’ label. However, he did care that his prose should be ‘what I call “clear” ’ (1985: xvi). Yet Williams’s style was never merely ‘analytical’ in this generic sense. Even a cursory examination of his prose reveals distinguishing features: an occasionally mannered elegance, a wide range of references (often to high cultural artefacts such as opera and classical texts), a wider diction than most analytic philosophers employ, and the use of the full expressive resources of the English language: idiom, metaphor, analogy, imagery, compression, allusiveness, and the deliberate mixing of seriousness with dry wit. The individuality of his style is already hinted at in the crucial qualifying phrase in his remark above. ‘What I call “clear” ’ suggests that he recognized something idiosyncratic in his understanding of clarity. Williams took it for granted that ‘if philosophy, or anything like it, is to have a point, the idea of “getting it right” must be in place, and so must clarity and precision’ (2014b: 367). Yet he insisted that ‘there is more than one kind of all these things’ (2014b: 367). And he regretted analytic philosophy’s tendency to narrowly [p. 301] Matthieu Queloz and Nikhil Krishnan model its conception of these ideals on the natural sciences. To be sure, physics could explain complex behaviour in terms of simple laws, and mathematics could offer sharp definitions and irrefutable proofs. But it did not follow that, in philosophy, clarity likewise had to consist in the reduction of complexity to simplicity; or that precision required the total elimination of vagueness through sharp definitions in terms of necessary and sufficient conditions; or that ‘getting it right’ simply meant arriving at a logically unassailable argument. A good place, then, to seek a description of what Wollheim might have called Williams’s ‘individual’ style—​analogous to the artistic dispositions that might distinguish Cézanne’s Impressionism from Monet’s—​is to ask how his conceptions of clarity, precision, and ‘getting it right’ differed from those that other analytic philosophers drew from the natural sciences. It is in relation to these ideals that Wittgenstein offered Williams a constructive alternative. First, Wittgenstein offered a kind of philosophy which, as Williams described it in an interview, left it ambiguous ‘how far it is harnessed to an argument’ It was striking, Williams said, how few of the conventional markers of argumentative structure (e.g. ‘therefore’, ‘since’, and ‘because’) there are in the Philosophical Investigations. Instead, ‘the work consists of curious sorts of conversations with himself, and epigrams, reminders’, suggesting that philosophy had ‘nothing to do with proof or argument at all’.13 Williams certainly never disdained argument, and explicitly condemned philosophers such as Richard Rorty, who emulated Wittgenstein in this respect, as allies of the very professionalization they scorned: a conversation held together only by ‘well then’ and ‘that reminds me’ and ‘come to think of it’ would not give anyone sufficient reason to listen: ‘the only people who will take part in such a conversation are those who are paid to do so’ (2014b: 367). Nevertheless, Williams saw no reason why philosophy should not combine the argumentative mode with the conversational, the explicit with the suggestive. Martha Nussbaum summarized the reaction of many readers when she described Williams’s compressed prose as ‘suggestive and revealing rather than systematic and finished, reaching for imaginative insight rather than hobbled by conventions of analysis’ (2003). One stylistic marker of this suggestiveness is Williams’s frequent use of ‘of course’ to remind readers of some shared human experience.14 Another is his pervasive use of the first-​person plural in inviting readers to confront the implications of some belief or conception. ‘My procedure’, he explains at the outset of 13 For a detailed discussion of the relation between Wittgenstein’s style, method, and philosophy, see Pichler (2023). 14 Some illustrative examples: ‘The combination—​discovery, trust, and risk—​are central to this sort of [Romantic] outlook, as of course they are to the state of being in love’ (2001b: 79). ‘Of course, no sane person could really believe that the goodness of the world just consisted in people keeping their obligations’ (1973a: 89). ‘Telemachus can be held responsible for things he did unintentionally, and so, of course, can we’ (1993: 54). [p. 302] ‘Moral Luck’, ‘will be to invite reflection about how to think and feel about some rather less usual situations, in the light of an appeal to how we—​many people—​ tend to think and feel about other more usual situations’ (1981a: 22). He refined this explanation in Shame and Necessity: in his usage, ‘ “we” operates not through a previously fixed designation, but through invitation. . . . It is not a matter of “I” telling “you” what I and others think, but of my asking you to consider to what extent you and I think some things and perhaps need to think others’ (1993: 171n7). The result was a style that ‘combined brilliant clarity with some of the properties of aphorism’ (2003), as Nussbaum put it. However, we need not accept Nussbaum’s implicit contrast between ‘clarity’ and ‘the properties of aphorism’. There are respects in which these suggestive qualities of Williams’s style themselves serve his aspiration to clarity, precision, and ‘getting it right’—​it is only that Williams has a more Wittgensteinian conception of these ideals than the mainstream of analytic philosophy. As emerges from a conversation between Williams and Bryan Magee, the philosophical approach of the Investigations resembles not the natural sciences, but works of art that try ‘to get people to see things in a certain way’; when that approach succeeds, it is not so much that we are compelled to adopt a new belief, but that we see things ‘in a way uncorrupted by the theoretical oversimplifications of philosophy’, thus ‘recovering the complexity of ordinary experience’ (Williams 1982: 118). The suggestiveness of Williams’s prose can likewise be understood as serving to help us see things aright, undistorted by philosophical theory. One might thus say that Williams adopts a Wittgensteinian conception of clarity. As Williams described the ideal of clarity at work in the Investigations: ‘the idea of clarity, here, is connected with substituting complexity for obscurity. Philosophy is allowed to be complex because life is complex’ (1982: 118). Instead of reducing complexity to simplicity, Wittgensteinian clarity replaces obscurity with complexity. Out of this Wittgensteinian conception of clarity then falls a Wittgensteinian conception of precision, whereby philosophy only seeks to eliminate vagueness as far as real complexity will allow. A conception of precision not merely as consistent with complexity and vagueness but as requiring it puts Wittgenstein and Williams at odds with the aspiration to arrive at strict definitions in terms of necessary and sufficient conditions. They both see a place for a suggestive vagueness that reflects real and irreducible complexity. Williams explicitly endorsed this Wittgensteinian understanding of precision when he praised David Pears for his ‘particular ironical taste for formulae which offer the tone or register of rigorous analysis but actually deliver a condition which is deliberately, and realistically, vague’ (2001a: xv). Such formulae exemplifying Wittgensteinian precision also pervade Williams’s own work. His well-​known account of the truth-​conditions for statements about reasons for action is an example: he took it to be a necessary condition on an agent A having reason to perform an action ϕ that ‘A could reach the conclusion that he [p. 303] Matthieu Queloz and Nikhil Krishnan should ϕ . . . by a sound deliberative route from the motivations that he has in his actual motivational set—​that is, the set of his desires, evaluations, attitudes, projects, and so on’ (1995b: 35). Responding later in that paper to the worry that this leaves us ‘with a vague concept of what an agent has a reason to do’, Williams responded that this consequence was ‘not a disadvantage of the position. It is often vague what one has a reason to do’ (1995b: 38).15 An account of reasons that tried to render the notion of ‘sound deliberation’ more sharply precise—​for instance, by rendering the link between rationality and imagination entirely determinate—​ would simply distort the concept of a reason. The added precision of such an account would come at the cost of Wittgensteinian precision—​a precision that allows certain forms of vagueness to reflect not eliminable obscurity, but ineliminable complexity. At the same time, Williams contrasted these rigorously articulated but realistically vague formulae with a superficially similar combination of characteristics he found exemplified in G. E. Moore’s ‘grinding style’, which ‘assists sometimes the appearance rather than the reality of precision, and is capable of conveying a kind of emphatic vagueness which curiously co-​exists with the marks of solicitor-​like caution’ (2014c: 76). Williams’s ideal of precision is not to combine apparent precision with actual vagueness, but to be as precise as possible while remaining as vague as the recognition of real complexity requires one to be. His Wittgensteinian conceptions of clarity and precision also led Williams to a correspondingly different understanding of what it means to ‘get it right’ in philosophy. He became impressed by the thought that getting it right required more than coming up with clever and logically unassailable arguments. This resistance to mere technical sophistication—​as opposed to the sort of sophistication that consists in having as many thoughts and feelings as one needs to make sense of the world—​seems, as a matter of biographical fact, to have come to him notably through the Wittgensteinian influence of Anscombe, who, he reported, ‘conveyed a strong sense of the seriousness of the subject, and how the subject was difficult in ways that simply being clever wasn’t going to get round’ (2009: 197). Above all, however, Williams was stylistically indebted to Wittgenstein for helping him see the importance of imaginative and expressive power to getting it right. Those in the analytic mainstream who modelled philosophy on science held that getting it right had nothing to do with style and modes of expression and imagination. And indeed, ‘the question of whether scientists have got it right or not’, Williams acknowledges, ‘is not much affected by the expressive power of their writing’ (2014b: 368). But philosophy is different. A philosopher’s contribution to the subject, especially in moral and political philosophy, is not 15 He doubles down on this in a later paper: ‘It is not an objection to the internalist account . . . that it involves vagueness and indeterminacy. This merely mirrors the truth that statements to the effect that A has a reason to do a certain thing are themselves vague and in various ways indeterminate’ (2006i: 110). [p. 304] independent from the imaginative and expressive power of their work. Ethics can be, as the title of the Preface to the French edition of Ethics and the Limits of Philosophy has it, ‘a matter of style’ (2021). To convey a certain picture of human life, to integrate it with what the philosopher cares about, and to adequately express those concerns—​all this can be part of what it means to get it right in philosophy. Getting it right can require one not just to say true things, but to say the right things in the right tone, which requires certain imaginative and expressive powers. As Williams was fond of saying, philosophers’ observations need not just to be true, but to ring true.16 It is in Wittgenstein’s work that the young Williams found a salutary example of philosophy that sought to ring true: Oxford philosophy in the fifties was very clever . . . But the philosophical tone had to be kept down, muted[,]‌English, dry, and that was a loss. That is why in a way I was drawn to the Wittgensteinian thing to some extent. Wittgenstein put much greater weight on the imaginative and the unpredictable aspects of philosophy. Williams warns that analytic philosophy’s ‘plain style’ modelled on natural science ‘can become a dead weight under the influence of the scientific model’ (2014b: 368). The philosophical outlooks that show the ‘most enthusiasm for natural sciences’, Williams tells Ayer, suffer from a tendency to be ‘brutally optimistic, unimaginative, short on . . . certain deeper perceptions about human life and values’ (Chanan Not for Williams the austere, Eddingtonian naturalism that views the world as largely empty, with a few scattered electric charges rushing about. Williams dislikes those ‘skeletal metaphysical pictures’, preferring what he describes as ‘emotionally and morally denser pictures of a form of life’, which Williams explicitly thinks of as ‘represented by somebody like Wittgenstein’ (Chanan 1972). The working picture that other twentieth-​century philosophers such as Carnap and Russell have of human life, Williams elaborates, lacks that emotional and moral density, and consequently lacks depth in its perception of human life and values, in a way that Wittgenstein’s picture of human life does not: [I]‌f you turn to Carnap or, indeed, Russell, come to that, and then compare that with Wittgenstein, just in terms of the tone about what human life is like, I think, although one may well repudiate a lot of Wittgensteinian propositions, particularly his obsession with the quasi-​religious issue of suicide and some other topics, it would be difficult to deny that there is some form of depth in Wittgenstein’s philosophy—​which there is also, for instance, obviously in the philosophy of 16 See Williams (2001b: xv; 2006j: 206). [p. 305] Matthieu Queloz and Nikhil Krishnan Nietzsche—​which is notably lacking in the philosophies of, say, Russell and Carnap. (Chanan 1972) Williams here casts Wittgenstein as offering a model of philosophical depth, which comes in part from ‘the tone about what human life is like’. It is true that Williams mentions Nietzsche in the same breath, as another model of depth. But Nietzsche was anathema in the Oxford of the 1940s and 1950s, and Wittgenstein would have loomed far larger in the intellectual scene of Williams’s formative years, both directly and through his influence on Ryle, Pears, Cavell, Anscombe, Foot, and Murdoch.17 It is therefore a plausible surmise that it was rather Wittgenstein who acted as a formative model of philosophical depth for Williams. Indeed, Williams avowed earlier in that conversation that Wittgenstein’s work had influenced him ‘partly because of the enormous imaginative power . . . there is such enormous literary power’ (Chanan 1972). Wittgenstein embodied the idea that getting it right could require one to deploy imaginative and expressive powers to convey the kind of depth and density that philosophy must have if it is to be truthful to human experience. Williams thus inherited something of Wittgenstein’s resistance to scientism—​ not, however, ‘Wittgenstein’s hatred of the cockiness of natural science’, which Williams found hard ‘to distinguish from a hatred of natural science’ (1973c: 91), but rather his resistance to stylistic ideals modelled on natural science.18 The scientism Williams resisted was what he called the ‘scientism of style’ (Williams j: 204).19 The resistance to the sort of cleverness that oversimplified the phenomena so that it could meet arbitrary standards of theoretical virtue was itself an idea well expressed in Investigations §107, where Wittgenstein speaks of how the ‘crystalline purity of logic’ is too cheaply won if it is simply a requirement placed in advance of inquiry, rather than a product of an inquiry successfully conducted. Wittgenstein’s famous motto in relation to logic and language, ‘Back to the rough ground!’, would serve equally well as an epigraph to most of Williams’s ethical writings, and an encapsulation of his style. 17 Along with Hegel, Nietzsche was thought to be connected with totalitarianism and was ‘ideologically suspect’ (Williams 1982: 117). And even after Nietzsche was well on his way to being rehabilitated in the Anglophone world, it took Williams years to warm to the self-​described ‘hermit of Sils-​Maria’. The Nietzsche scholar Michael Tanner remembers how, as late as the early 1960s, Williams was still capable of picking up Tanner’s copy of Beyond Good and Evil and wondering: ‘Why do you waste time over rubbish that Joad could have refuted?’ (O’Grady 2003) (C. E. M. Joad was a broadcasting personality and popularizer of philosophy who came to prominence with the BBC programme The Brains Trust in the 1940s). It was only later that Williams became seriously interested in Nietzsche, to the point of planning a book about him. See Owen (this volume, Chapter 12) for the most detailed account to date of Williams’s relationship to Nietzsche. 18 On Wittgenstein’s resistance to scientism and his defence of the autonomy of humanistic understanding against encroachment by the natural sciences, see Hacker (2011), who also sketches the history of the notion of humanism. 19 On Williams’s resistance to the ‘scientism of style’, see also Fricker (2023). [p. 306] 3. Thick Concepts and Their Point: A Functionalist Anthropological Method If we turn now to Williams’s methodological debt to Wittgenstein, it is striking that Williams makes a number of observations about Wittgenstein’s methodology that apply equally to himself. One example is his remark that Wittgenstein ‘wished to recall philosophy to the world’ (Williams and Montefiore 1966: 10). That is a methodological commitment he adopted himself. This comes out most clearly in his last book, which defends the value of truth against its postmodern deniers. ‘Philosophy here, on lines variously laid down by Hume, Wittgenstein, Stanley Cavell, needs to recall us to the everyday’ (2002: 10), he writes, stressing the need to remember that there are everyday truths, and that they are important. Williams also remarks of Wittgenstein that he ‘emphasises concrete practices and shared understandings as against abstract ethical theory, and indeed has no time for that sort of theorizing’ (2021: 277). Again, the same could be said—​with qualifications20—​of Williams. Like Wittgenstein, Williams takes considerations arising from the way concrete practices actually work, and from the shared understandings we bring to these practices, to carry more weight than considerations of systematicity arising from the desire for a philosophical theory, which threaten to distort those understandings.21 Above all, however, the pursuit of ‘emotionally and morally denser pictures of a form of life’ drove both Wittgenstein and Williams away from physical or biological explanations towards anthropological or ethnographic descriptions.22 This preference for anthropological or ethnographic descriptions is manifest in Williams’s emphasis on the explanatory and justificatory value of what he influentially labelled ‘thick’ ethical concepts.23 It also comes out in his repeated insistence on the importance to philosophy of what he called ‘the ethnographic stance’, whereby, like an ethnographer, one ‘understands from the inside a conceptual system in which 20 Although Williams was famously an opponent of moral theory, he was in favour of systematic theorizing in other areas of philosophy if the scientific model of theory was made appropriate by the systematic nature of the phenomena (as in the philosophy of language), or if there were strong practical demands for some degree of systematization (as in political and legal philosophy). On the practical demands for systematization that made Williams sympathetic to theory in political and legal philosophy, see Cueni and Queloz (2021). 21 See especially ­chapter 6 of Ethics and the Limits of Philosophy (1985: 126). See Diamond (2018) for a sympathetic discussion of an example of this in Williams’s work. 22 A helpful overview of Wittgenstein’s views on philosophical method is given by Glock (2017b). For detailed accounts of the respects in which Wittgenstein exemplified an ‘anthropological and ethnographical approach’, see Hacker (2013) and Brusotti (2014). For Williams’s account of ‘the ethnographic stance’ and its importance to philosophy, see Williams (1985: 157; 1986: 203–​4; 1995d: 207; 1995f: 239; h: 61; 2021: 278). He also endorses a form of what he variously calls ‘a priori anthropology’, ‘ideal anthropology’, or ‘philosophical anthropology’ in Williams (1997: 27; 2002: 10; 23 Williams would in turn have been familiar with Geertz’s (1973: 6) advocacy of ‘thick descriptions’ in anthropology, not least since Geertz borrowed the phrase from Ryle (2009a: 489; 2009b: 497). [p. 307] Matthieu Queloz and Nikhil Krishnan ethical concepts are integrally related to modes of explanation and description’ while being ‘conscious that there are alternatives to any such system’ (1986: 204).24 Relatedly, Williams also adopts what he recognizes is ‘basically a Wittgensteinian idea’ (1985: 263n7): namely, the idea that we would be unable to see how people ‘go on’ from one application of a thick ethical concept to the next ‘if we did not share the evaluative perspective in which this kind of concept has its point’ (1985: 157). Williams notes that he first encountered this idea in a graduate class convened by Philippa Foot, Iris Murdoch, and Basil Mitchell in the early summer of 1954.25 It seemed to him to remedy a signal failing of ordinary language philosophy as predominantly practised in Oxford during his student days: namely, its lack of interest in the background of contingent facts and human concerns from which a concept or distinction derives its point: [W]‌hat we tended to do was to pick up some distinction or opposition, and go very carefully into it and into the various nuances that might be attached to it, and order them, or state them, without enough reflection on what background made this set of distinctions, rather than some other, interesting or important. For Williams, Wittgenstein did not simply emphasize that justifications come to an end: that is, that ‘at various points we run into the fact that “this is the way we go on” ’ (Williams 2006e: 196); significantly, he also encouraged philosophers to ask why we go on in this way—​what the point of a given use of language is, if it has one at all.26 As Wittgenstein himself put it, a ‘use of language has normally what we might call a point. This is immensely important’ (1989: 205). One way of figuring out what, if anything, the point of something is, Wittgenstein thought, is to look at its history: ‘if you wish to give the point, you might tell the history of it’ (1989: 204)—​ although, as Wittgenstein remarked at the end of the Philosophical Investigations, ‘we can also invent fictitious natural history for our purpose’ (2009: II, §365).27 Wittgenstein did not suffer from the Panglossian presumption that nothing is pointless. He explicitly wondered whether there was a point to everything we do and emphasized that it is only in relation to certain facts about people’s interests and concerns, and the world in which they pursue them, that a use of language has a point. If people or the world were sufficiently different, that use of language would become pointless. Even as things are, not every use of language 25 Williams only mentions a seminar with Foot and Murdoch in the 1950s, but Murdoch’s teaching record shows that this class, entitled ‘Analysis in Moral Philosophy’, was the only one she co-​taught with Foot in that period (Broackes 2011: 5). 26 See also Wittgenstein (2009: §§467–​70). 27 On Wittgenstein’s scattered remarks on history and historical modes of philosophizing, see Glock who also compares and contrasts them with Williams’s views on the topic. [p. 308] necessarily has a point. As Williams observes, it is central to Wittgenstein’s critique of certain forms of philosophizing that some uses of language might be ‘alienated from every human purpose’ (2006j: 210): ‘They are, so to speak, timelessly out of place, because they fit no conceivable human purpose, except the misguided philosophical impulse which they are supposed to illustrate’ (2006j: 210).28 Williams self-​consciously follows Wittgenstein in approaching almost any puzzling conceptual practice by asking what its point or function is—​why we go on in this way. To the 29-​year-​old Williams, this methodological strategy stood out as an underappreciated continuity between the Tractatus and the Philosophical Investigations, as he observed in his 1958 review of a collection entitled The Revolution in Philosophy. Though the collection ranged over everything from Bradley, Frege, Moore, and the Vienna Circle to reflections on the state of the discipline and the role of the imagination, Williams devotes the bulk of his review to calling for further research on strands in Wittgenstein’s work that he thinks come closest to constituting the ‘essence’ (1958: 67) of the post-​war ‘revolution’ in British philosophy. What he singles out as ‘certainly lacking’ from the book is a ‘unified account of the work of Wittgenstein’ (1958: 67) that would help us to make sense of the relation between the Tractatus and the Philosophical Investigations: There are many similarities and connections between the two works, which need exploration. . . . Sometimes it almost looks as if the afterthoughts and marginal comments of the Tractatus grew in Wittgenstein’s thought until they edged out the central thesis: casually at 6.211 he remarks ‘in philosophy the question “Why do we really use that word, that proposition?” constantly leads to valuable results’, and in the Investigations we find the results of asking just this sort of question. Where we also find the results of asking just this sort of question is in the works that Williams himself went on to write. It is an unobtrusive but abiding feature of his method that he asks after the point or function of the concepts and practices he examines. What is achieved by thinking and speaking in this way? How does it help us to live? Of course, Wittgenstein was not alone in directing philosophical attention to the point of individual concepts and practices. But he was a particularly influential exponent of what has lately been termed ‘Cambridge Pragmatism’ (Misak a tradition that includes F. P. Ramsey, Hugh Mellor, Edward Craig, Simon Blackburn, and Huw Price, and also influenced post-​war Oxford.29 And it is revealing that what stood out to the young Williams as a continuity in Wittgenstein’s work was its pragmatist penchant for asking why people think and speak as they do. 28 As Williams remarks in an interview with Bryan Magee, this led Wittgenstein to harbour ‘great doubts about the existence of philosophy at all, except as a deep aberration that happens when our conceptions of ourselves go wrong’ (1982: 120). 29 On Oxford pragmatism, see Glock (2017c), Kremer (2022), and Misak (forthcoming). [p. 309] Matthieu Queloz and Nikhil Krishnan It was notably for raising this functionalist question about the concept of knowledge that Williams commended Edward Craig’s Knowledge and the State of Nature (1990), which offered a template for Williams’s own State of Nature narrative in Truth and Truthfulness. What Craig’s guiding question crucially introduces, Williams explains, is the notion of function, and that step itself does some of the work. If one sees the concept of knowledge as having a function—​in particular, a function in relation to very basic needs—​this in itself helps one to see why it has the features it has, and can discourage one from less fruitful approaches. (2002: 31–​2) But where did Craig get the idea of asking after the function of the concept of knowledge? At least part of the answer, as Craig acknowledged in his Wittgenstein Lectures in Bayreuth, was Ludwig Wittgenstein.30 Craig declared himself indebted to Wittgenstein for ‘loosening up the concept of a concept’, giving him licence to regard the concept of knowledge as an instrument serving some function, and to contemplate the possibility that its function might even be the most important thing about it (1993: 39–​40).31 In Williams’s oeuvre, however, the interest in the function or point of conceptual practices long predates Craig’s book, and is something that Williams firmly associated with Wittgenstein—​in his 1972 discussion with Ayer on ‘the point’ of religious practices, for example, and again in his 1973 lecture on Wittgenstein and idealism, where he notes that Wittgenstein sometimes relates people’s ‘practice in some broadly functional way to their interests’ (1973c: 91). Williams’s interest in the point of conceptual practices is already evident in the way he approaches the moral/​non-​moral distinction in his first book, Morality, which inquires into ‘the point of selecting certain motives for moral approbation’ (2001b: 68). This methodological interest in the point or function of conceptual practices endures throughout Williams’s later work.32 Discussing the concept of obligation, 30 This is only part of the story; a full account would have to mention Craig’s debts to Carnap and Hume as well as his inspiration by Williams’s own work on the concept of knowledge (1973b: 146; b: ch. 2), for which Williams declared himself indebted to ‘the Australian philosopher Dan Taylor, who may have been influenced in this direction by John Anderson’ (1995h: 211n4). See Queloz (2021a) for a detailed discussion of these connections. 31 Oswald Hanfling’s (1985) approach, which bears some resemblance to Craig’s, even more explicitly affiliates itself to Wittgenstein. 32 Examples abound: Williams’s assessment of the characteristic psychology of blame remains controlled by a sense of ‘the purpose of blame’ (1995a: 15)—​see Queloz (2021b) for a more detailed exegesis along these lines. In thinking about ethical theory, he asks after ‘the point of ethical theory: who needs such a theory? What for?’ (2005d: 54)—​see Cueni and Queloz (2021). His account of responsibility is guided by the conviction that ‘responsibility has a function’ (2006d: 125), and by the ‘purposes that are served by discriminating between actions in terms of the voluntary’ (1993: 67)—​see Queloz Writing on tort law, he insists that philosophers must understand why tort law has the principles it does (1995g: 492–​3)—​see Queloz (forthcoming). Reflecting on ‘liberty’ and ‘freedom’, he warns against missing ‘the point of why we want these terms in the first place’ (2005c: 79)—​see Queloz (2024). [p. 310] he seeks to ‘help us to understand the point and value of living a life in which obligations counted as ethical reasons’ (2006g: 73). In considering the morality system as a whole, he wonders about ‘the point of this conception of morality’ (1995c: 241). The paper he presented at the Fifth Kirchberg Wittgenstein Symposium in 1980 proposes to draw on ‘the function of the all-​in ought of practical deliberation’ (1981b: 120) to explain why this concept has just the features it does. Even in his brief essay on censorship, he finds room to consider when ‘the point of censorship is lost’ (2005a). And of course, Truth and Truthfulness aims to show that truthfulness ‘gets its point ultimately from the human interest, individual and collective, in gaining and sharing true information’ (2002: 126). It even implements Wittgenstein’s methodological suggestion for how to achieve this: by telling a partly fictional ‘history’ or ‘genealogy’ of truthfulness.33 At the same time, Williams strove to be sensitive to the fact that we do not simply value truthfulness instrumentally, as perhaps we value money, but regard truthfulness as something intrinsically valuable. The ultimate explanation of that fact may refer to its instrumental value, but the phenomenology of valuing truthfulness is not instrumental-​minded (2002: 92–​3). Here, as elsewhere, Williams makes a point of asking what the psychology of an agent who lives by the concept or practice in question looks like: what sorts of considerations actually figure in the deliberations of such an agent, and under what descriptions? But he does not leave it at that. He combines a realistic description of the phenomenology of a way of thinking with a more detached account of the human concerns that this way of thinking ties in with. Fusing the one-​eyed view of how something phenomenologically presents itself to us with the equally one-​eyed view of what function it performs, he arrives at a stereoscopic view of it. That is part of how he achieves a sense of depth. This combination of functionalist anthropology with ‘emotionally and morally denser pictures of a form of life’ is characteristic of Wittgenstein, who described his ambition to capture both the ‘dignity’ of rules and their usefulness without collapsing one into the other: What I have to do is as it were to describe the office of a king;—​in doing which I must never fall into the error of explaining the kingly dignity by the king’s usefulness, but I must leave neither his usefulness nor his dignity out of account. Williams might have said the same of his treatment of the intrinsic value of truth which aspires to leave neither its intrinsic value nor its instrumental value out of account while avoiding the error of spelling out its intrinsic value in terms of 33 For comparisons of Wittgenstein’s ‘remarks on the natural history of human beings’ (2009: §415) with Williams’s genealogical method, see also Owen (2001) and Glock (2006). [p. 311] Matthieu Queloz and Nikhil Krishnan its instrumental value. Like Wittgenstein, Williams insists on the need to explore functional hypotheses as to why we engage in certain decidedly non-​f unctionalist ways of thinking without reducing one to the other. These deep methodological consonances make for clear evidence not just of coincidence, but of a genuine debt of influence. 4. Williams’s Left Wittgensteinianism In the 1990s, Williams became more open about the Wittgensteinian roots of his own thinking. He explicitly associated himself with a certain form of Wittgensteinianism, which, in an echo of the traditional distinction between Right and Left Hegelians, he called ‘Left Wittgensteinianism’. Hegel had rejected—​as Wittgenstein later would—​the Kantian focus on thin and abstract concepts, and emphasized the importance of thick concepts in constituting shared customs and a communal way of life. But Hegelians soon split into a more conservative and a more radical camp, entrenching a distinction between the ‘Right Hegelian’ emphasis on the need to embrace tradition and its culmination in the Prussian state, and the ‘Left Hegelian’ emphasis on the need for a radical critique of the inherited order.34 Though this contested distinction made the factions appear more unified and self-​conscious than they in fact were, it had the virtue of registering that Hegelianism was not one thing, but could be elaborated in different directions. As Williams and David Bloor both argued—​apparently independently—​in a parallel divergence is possible in the elaboration of Wittgensteinian ideas. Wittgensteinianism is not one thing, and the fact that many interpretations of Wittgensteinian ideas have been of the ‘Right’ variety, encouraging the conservative embrace of inherited concepts, does not foreclose the possibility of a ‘Left’ Wittgensteinianism capable of making sense of the radical critique of inherited concepts.35 ‘So far as critique is concerned,’ Williams remarks in an essay first published as ‘Left-​Wing Wittgenstein, Right-​Wing Marx’ (1992),36 there seems no reason why non-​foundationalist political thought, characterized in the way that Wittgenstein’s philosophy suggests, should not take a radical turn. There could be, one might say, a Left Wittgensteinianism . . . we can follow 34 See Toews (1985) and Breckman (2019). 35 A variety of scholars have focused on the structurally conservative aspect of Wittgenstein’s thought: Pitkin (1972: 7–​8); Bloor (1983, 1997, 2000); Norris (2009); Nyíri (1976, 1982); Plotica Rorty (1983; 1989: 58–​60); Temelini (2015). Approaches that put a broadly Wittgensteinian picture of our conceptual apparatus to radically critical use include Pleasants (1999, 2002), Celikates and Jaeggi (2016). 36 It was then republished under the title ‘Pluralism, Community and Left Wittgensteinianism’ e) in the posthumous In the Beginning Was the Deed, from which we cite above. The essay was recently republished in Common Knowledge under yet another title, ‘Left-​Wing Wittgenstein’ (2019). [p. 312] Wittgenstein to the extent of not looking for a new foundationalism, but still leave room for a critique of what some of ‘us’ do in terms of our understanding of a wider ‘we’. (2005e: 37) On the issue of how Williams conceives of the grounds of radical critique, this essay on Left Wittgensteinianism is among the most illuminating in Williams’s oeuvre. Complementing his rejection of foundationalism and Rortyan ironism in ‘Philosophy as a Humanistic Discipline’, it stresses the possibility of non-​ foundationalist radical critique in political philosophy.37 Yet Williams does not appear to take Left Wittgensteinianism to be limited to political philosophy. In the Preface he writes for the French translation of Ethics and the Limits of Philosophy, which has only recently appeared in the original English (2021), Williams talks about Wittgenstein at some length, referring to him nine times and reiterating his case for the possibility of a ‘Left Wittgensteinianism’ in the context of moral philosophy. Philosophy, Williams insists in that Preface, ‘has to tell us how we can come to embrace new ethical concepts’ (2021: 278). But an account that ‘considers only the concepts that we pick up from our local community’, as most Wittgenstein-​inspired work in moral and political philosophy has tended to do, ‘will find it hard to explain the criticism and alteration of ethical practices’ (2021: 278). This ‘Right Wittgensteinianism’, as Williams calls it, encourages an ‘enthusiasm for the folk-​ways’ which amounts to ‘the continuation of Hegelian conservatism by other means’ (2021: 278).38 This time, however, Williams points out not only that there could be a Wittgensteinian analogue to Left Hegelianism, but that there should be (2021: 278): [T]‌here should also be a Wittgensteinian analogue to Left Hegelianism: this will be a view that accepts the insights about the thickness of our primary ethical understanding and its relation to social practices, but leaves room for a radical critique in the name of interests not adequately expressed in the folkways. 37 For an interpretation of Williams’s Left Wittgensteinianism in the context of political philosophy that brings out how it contrasts with Rortyian ironism and foundationalism, see Queloz and Cueni (2021). 38 As Bloor sketches the contrast, ‘Left Wittgensteinians’ offer interpretations that are ‘more historical, social, and materialist-​scientific’, treating ‘Wittgenstein’s ideas as embryonic social-​scientific theories’ (1992: 281), while ‘Right Wittgensteinians’ offer interpretations of Wittgenstein that draw on internal relations as ammunition against sociological approaches. For Bloor, the Right Wittgensteinians paradigmatically include G. P. Baker, P. M. S. Hacker, S. G. Shanker, Michael Lynch, Marie McGinn, and ‘other antisociological commentators’ (1992: 273; see also 281). The only Right Wittgensteinian Williams identifies by name is Peter Winch, whom Bloor does not mention, but who is hardly best described as ‘antisociological’. This suggests that Bloor’s and Williams’s contrasts do not exactly coincide, though Williams’s critique of Winch’s emphasis on internal relations at the expense of genuine social-​ scientific explanation echoes Bloor’s critique, and suggests that Winch is not sociological enough for Williams. [p. 313] Matthieu Queloz and Nikhil Krishnan This would be a peculiar way of introducing a French audience to his magnum opus in ethics unless Williams meant to invite his readers to draw three conclusions: that the book is more deeply immersed in Wittgensteinian ideas than he cared to make explicit to an Anglophone audience when it was first published; that there is an overlooked form of Wittgensteinianism that Williams considers viable not just in political, but also in moral philosophy; and that Ethics and the Limits of Philosophy can itself be read as advocating some form of Left Wittgensteinianism as an attractive alternative to ultimately unsuccessful foundationalist attempts to ‘justify the ethical life from the ground up’ (1985: 32). 5. Williams’s Methodological and Metaphilosophical Critique of Wittgenstein Williams’s late and seemingly sudden endorsement of Left Wittgensteinianism becomes completely unsurprising once one reconstructs how it falls out of his earlier critical engagement with Wittgenstein’s methodology and metaphilosophy. Left Wittgensteinianism, we argue, is what Williams ended up with after correcting the shortcomings he perceived in extant forms of Wittgensteinianism. And when properly understood, his critique of these shortcomings turns out to be continuous with his advocacy of a conception of philosophy as a humanistic discipline. We saw that Williams follows Wittgenstein in two important respects: in concentrating on thick concepts, and in asking after the point of our conceptual practices. But Wittgenstein remains non-​committal about what the exact boundaries of ‘our’ conceptual practices are. Is he talking about a particular subset of human beings? All human beings? All rational creatures? Where he contrasts ‘our’ conceptual practices with those of others, he is not interested in offering genuine explanations of why different forms of life differ;39 rather, the different forms of life are primarily meant to aid self-​understanding. This means that they can also be imaginary: by considering, even purely notionally, why people might think differently if they had different interests, or if certain general facts of nature were different, we can come to see the contingent dependence of the way we actually think on certain extraconceptual presuppositions: our own way of thinking will be revealed to derive its point from certain facts about us and our environment, and to be pointless without them. As Williams puts it, Wittgenstein’s imagined alternatives to our form of life are not so much ‘alternatives to us’ as ‘alternatives for us’, in that ‘the business of considering them is part of finding our way inside our own 39 Particularly in his middle period (1929–​36), Wittgenstein was intent on contrasting the giving of reasons with the citing of causes, and on that basis developed a stark dichotomy between rational justification and causal explanation. For a synthetic exposition of the passages in Wittgenstein to this effect, see Queloz (2016, 2017). [p. 314] view’ (1973c: 91). They are precisely not offered up as real alternatives calling for empirically informed explanation, but as imaginative crutches designed to elucidate our own form of life. Combined with Wittgenstein’s substantive focus on the ‘universalistic preconditions on interpretation and intelligibility’ (Williams 2006a: 358), which is to say the general conditions of the possibility of understanding and linguistic meaning, these ideas pulled Wittgenstein’s functionalist anthropological method towards a conception of philosophy as an exclusively a priori enterprise that remains indifferent to empirical information from the human sciences: it takes no interest in what exactly marks off a clearly delimited local ‘us’ from other expressions of human life, and it does not really seek to explain these differences in terms of contingent sociohistorical developments. Wittgenstein’s anthropology is a philosophical anthropology to the end. Williams, by contrast, resists this methodological confinement to a priori anthropology. He diverges from Wittgenstein in two significant respects. First, he takes issue with Wittgenstein’s tendency to understand the notion of a ‘form of life’ in an inclusive sense, encompassing anyone with whom we could intelligibly communicate.40 That inclusive interpretation of ‘form of life’ may, Williams grants, be appropriate when thinking about the conditions of the possibility of understanding and linguistic meaning.41 But it should not be carried over into philosophical reflection on ethics and politics. Williams is interested in contrasting actual groups of human beings by identifying some clearly delimited divide between ‘us’ and ‘them’. Once we think in concrete sociohistorical terms about groups of human beings and their distinctive ethical and political concepts, we will use the notion of a form of life in a contrastive sense: the ‘we’ in question will not be the inclusive ‘we’, but rather the contrastive ‘we’ denoting some actual ‘us’, here and now, as distinct from concrete others (Williams 2005e: 36; 2006a: 358). In appropriating the Wittgensteinian emphasis on concrete practices and shared understandings, Williams thus recasts it in more politicized and sociohistorically embodied terms (2002: 10). In this respect, his reinterpretation of Wittgensteinian ideas self-​consciously parallels Cavell’s: like Wittgenstein, they both recall philosophy away from high theory and philosophical scepticism and back to the rough ground of concrete practices; but, unlike Wittgenstein, they seek to understand these concrete practices in a way that ‘engage[s]‌with history or our present cultural situation’ (Williams 2006j: 210).42 Secondly, Williams also diverges from Wittgenstein in insisting on the need for real sociohistorical explanation. Williams does not rest content with 40 See Williams (1986: 204). 41 See Williams (2005e: 36). 42 On the unhistorical nature of Wittgenstein’s conception of philosophy and how it contrasts with that of R. G. Collingwood, whom Williams admired, see Blackburn (2000). [p. 315] Matthieu Queloz and Nikhil Krishnan self-​understanding achieved by contemplating ‘purely imaginary and schematic ethnographic case[s]‌’ (2006a: 357). He does not deny that there is a place for a priori anthropology in philosophy: Williams himself repeatedly advocates forms of a priori anthropology (1995e: 140; 1997: 27; 2002: 10; 2005c: 76; 2006h: 61), and he often begins his own philosophical reflections by considering what any human beings anywhere would need, or what they would be bound to develop in some form. Examples include Williams’s derivation, in Shame and Necessity, of the four elements of any conception of responsibility from ‘universal banalities’ about human beings, or his use, in Truth and Truthfulness, of a ‘State of Nature’-​fiction to reveal the most generic needs to which the virtues of truth answer. But a priori anthropology can be, for Williams, no more than a starting point on the path towards an a posteriori understanding of how the generically human was in fact extended, inflected, and elaborated into more specific forms by historical and cultural forces. Williams’s approach thus early on finds a role for a posteriori anthropology, which is to say anthropology as a social science, characterized not principally by a philosopher’s armchair reflections on human nature, but by fieldworkers applying methods of participant observation to some particular community. He also finds a role for historical, sociological, and psychological explanation. In discussing sociobiological accounts of morality, for instance, Williams warns that we cannot fully understand how a functional account of some piece of moral thought, such as a prohibition, relates to the markedly non-​f unctionalist spirit and content of this prohibition, unless we draw on the explanatory resources of the human rather than the natural sciences: This is an area in which it is certain that sociobiological theorists go too quickly. They find a functional explanation with a certain content; they find a prohibition with a certain content; they say ‘A h! There you are.’ But what we need is an explanation of how the one got into the other. (1980a: 278) Any functionalist anthropological approach that Williams can endorse will have to display greater openness to empirical findings and to genuine sociohistorical explanation. This twofold methodological divergence has its roots, fundamentally, in the different ways in which Wittgenstein and Williams conceive of philosophy itself. Williams repeatedly opposes what he sees as Wittgenstein’s conception of philosophy as a discipline that must necessarily remain pure of empirical material and leave everything as it is (2006e: 196).43 It is a conception of philosophy that 43 Williams also remarks on the ‘impurity of philosophy’ in Williams (1995e: 148). For illuminating discussions of Williams’s idea of a humanistic discipline and its impurity, see Moran (2016) and Cueni (forthcoming). For a conception of philosophical methodology which likewise combines Wittgensteinian commitments with a rejection of purity, see Glock (2017a). [p. 316] Williams dismisses as typical of early analytic philosophy, which aspired to use the ‘distinction of fact and value’, or of ‘theory and value’, to ‘segregate the philosophical from the normative, while the companion distinction of analytic and synthetic served to segregate the philosophical from the historical or social-​scientific’ b: 58). Williams doubts that there could be a ‘purely philosophical’ (1986: 207) understanding of our ethical ideas and our deliberative practices, unaffected by history, psychology, and the social sciences: ‘If there were such a thing, but it were somehow guaranteed not to upset our ethical ideas and our deliberative practices—​presumably by its being a criterion of correctness in that subject that it left everything where it was—​I do not see why we should have any reason to be interested in it’ (1986: 207). When Williams invites us to conceive of philosophy as a humanistic discipline, therefore, he envisages an enterprise that is deliberately and unabashedly impure. ‘Impure’ is not a pejorative term for Williams, as he clarified already in his 1969 essay on philosophy as a subject. On the contrary, he warns against ‘the attempt to keep philosophy too pure, of other subjects, or of particular cases, or of taking an evaluative side in vexed issues’ (1969: 153). Philosophy, on Williams’s humanistic conception, can take a legitimate interest in factual issues and the empirical features of actual groups of human beings, even if those groups are agreed to be local e: 196); it can draw on genuine explanations of those groups’ distinctive features in terms of contingent empirical material; and it can consider whether these explanations are subversive or vindicatory, even if it has to take a normative stance on first-​order moral or political issues in order to do so. This makes philosophy impure in three respects: first, through its involvement in the factual issues and results of other subjects, such as history, sociology, and psychology; secondly, through its involvement in the contingent empirical features of sociohistorically local situations; and thirdly, through its involvement in evaluation. The later chapters of Truth and Truthfulness illustrate all three forms of impurity. Deeply informed by history, sociology, and psychology, they offer an expressly vindicatory appraisal of the practice of valuing the truth for a local ‘us’: namely, the citizens of modern liberal democracies living in the shadow of Romanticism and the events of the twentieth century.44 Where Wittgenstein was content to show that some of our ideas and procedures were unhintergehbar, admitting of no justification in terms of something more basic, Williams thus goes further in wanting to distinguish between ‘different ways in which various of our ideas and procedures can seem to be such that we cannot 44 For an account of the role that the memory of the Second World War plays in Williams’s moral philosophy, see Krishnan and Queloz (2023). For an overview of the war’s influence on Oxford philosophy more generally, see Krishnan (2023). Queloz (2018) offers a detailed reconstruction of Williams’s genealogical argument in Truth and Truthfulness. [p. 317] Matthieu Queloz and Nikhil Krishnan get beyond them, that there is no conceivable alternative’ (2006e: 196). To achieve this, philosophy needs to take seriously the questions, first, of who ‘we’ are, and second, of what explains the seeming Unhintergehbarkeit of a given idea or procedure. If this means that philosophy needs to abandon its commitment to purity by pursuing its task in consultation with the human sciences, taking an interest in the peculiarities of local situations, and adopting an evaluative stance, then so be it: ‘philosophy cannot be too pure if it really wants to do what it sets out to do’ (2002: 39), Williams concludes in Truth and Truthfulness.45 ‘It is significant’, Williams goes on to note, ‘that Wittgenstein, who took as seriously as anyone could the question of what philosophy might now be, but stuck firmly to a conception of it as quite separate from other intellectual enterprises, came to the conclusion that philosophy could not offer any explanations at all’ (2002: 283n23). Williams in effect agrees with Wittgenstein that if philosophy is to be pure, it cannot offer genuine explanations; but whereas Wittgenstein was willing to embrace a vision of philosophy without explanations, Williams preferred philosophy not to be so committed to its own purity.46 Williams’s critique of Wittgenstein’s methodology and metaphilosophy is echoed in his critique of Right Wittgensteinianism, ‘an early and influential example’ (2006a: 357n37) of which he sees in Peter Winch’s The Idea of a Social Science and Its Relation to Philosophy (1958). Winch’s great virtue in Williams’s eyes is to propose a Wittgensteinianism that actually takes an interest in concrete societies; he also commends it for being attuned to the integral relation between thick concepts and the ways of describing and explaining things that make up a form of life, and for emphasizing the possibility of understanding such forms of life ‘from the inside’ without sharing them.47 But Williams still regards Winch’s Wittgensteinianism as suffering from ‘an over-​close assimilation of social to conceptual understanding’ (1980b: 64): it views social science primarily as a form of conceptual investigation, which leads it to eschew causal explanation in favour of studying the conceptual interrelations within the outlook of a certain group. As Williams sees it, the study of the conceptual interrelations is only a part of social scientific study: he rejects as ‘unacceptable’ the idea that ‘social science is, more than everything else, conceptual investigation (cf. Winch)’ (1980b: 75n1). Another problem with this ‘purely conceptual stance’ is that it results in a conservative ‘immunity from social reflexion’ (1980b: 63).48 To remedy this, Williams advocates ‘greater openness to the impurities of the social 45 Compare Williams (1969: 153; 1996a: 34). 46 Glock (2017a) echoes Williams’s critique in advocating ‘impure’ conceptual analysis on the grounds that even the conceptual issues of philosophy sometimes interact with the factual issues of science. 48 Another aspect of Winch’s work that Williams does not address, but which he well might have, is the issue of the authority that Winch claims for philosophy and a priori conceptual investigation. See Pettit (2000: 76) for a critique of Winch along these lines. [p. 318] sciences’ (1980b: 63) in order to render Wittgensteinianism responsive to critical social reflection. Above all, however, it is in the tendency to picture a form of life as a ‘fully functioning and coherent system’ (2006a: 357) that Williams locates what gives Wittgensteinianism a conservative or ‘Right’ inflection. Wittgensteinianism yields conservative conclusions due to its holism (2005e: 34). If one regards each concept and practice as performing its part within a smoothly functioning whole, any criticism will threaten to distort that delicately calibrated harmony. The result is a static picture, which simultaneously fails to account for the dynamic impetus of radical critique and discourages it. Williams’s critique of this holism is of a piece with his earlier critique of Wittgenstein’s lack of interest in the boundaries of ‘we’ and in genuine explanation. If one asks after the points or functions of conceptual practices, but never seriously investigates for whom exactly they have a point, one is limited to discerning functionality in conceptual practices from a single, unified perspective. But it is by no means given that the ‘we’ for whom the practice is pointful will be identical to the ‘we’ that engages in the practice. One could be merely a small subset of the other. Or they could be entirely distinct. The inclusive use of ‘we’ thus already leads to a holistic picture, because it suppresses the very questions that might encourage a more nuanced differentiation between functional perspectives. The Left Wittgensteinianism that Williams advocates, by contrast, asks after the points of conceptual practices without making any assumption that they will all be harmoniously functional for everyone. It asks for whom the practice is pointful, mindful of the possibility that it might serve a point for someone other than those who engage in it, and perhaps not one they want to see served. And just because it takes this possibility seriously, it takes an interest in sociohistorical explanations of why those who engage in a practice do so despite the fact that the practice does not benefit them—​Williams frequently alludes to Marxian analyses of false consciousness, for example.49 This Left Wittgensteinian picture of our practices as a tension-​ridden assembly of parts, which might or might not be functional for a clearly delimited ‘us’, is itself the product of rendering philosophy less pure by allowing it to be informed by the social sciences. ‘Once we regard the ethical life we now have as a genuinely historical and local structure,’ Williams writes, ‘we shall have less temptation to assume that it is a satisfactorily functioning whole; and we shall be more likely to recognize that some widely accepted parts of it may stand condemned in the light of perfectly plausible extrapolations of other parts’ (2005e: 136–​7).50 For example, the widely accepted practice of factory farming may stand condemned in the light of plausible extrapolations of a commitment to avoiding needless cruelty. 50 This line of argument is developed at length in Queloz (2025). [p. 319] Matthieu Queloz and Nikhil Krishnan There is thus an underlying systematic connection between Williams’s critique of Wittgenstein’s methodology and purist metaphilosophy, on the one hand, and Williams’s own advocacy of a Left Wittgensteinianism and a humanistic metaphilosophy, on the other. Williams’s Left Wittgensteinianism is, in more than one sense, impure Wittgensteinianism. That is to say, it is a humanistic Wittgensteinianism. In keeping with his longstanding aversion to ‘isms’, Williams does not go quite so far as to nail his flag to the mast of ‘Left Wittgensteinianism’. But there is every indication that he means, at the very least, to indicate the possibility of reading his own work as an example of a Left Wittgensteinianism that promises to correct what he sees as the shortcomings of extant Wittgensteinian approaches. And once we recognize how this critique is of a piece with Williams’s advocacy of a conception of philosophy as a humanistic discipline, his debt to Wittgenstein emerges as a unifying interpretative key to his work. Williams’s conception of philosophy can be read as evolving dialectically out of Wittgenstein’s own. 6. Philosophy as Therapy and Ethical Reminders We have assembled plenty of evidence to debunk the myth of Williams’s ‘unwavering hostility’ to Wittgenstein. In this final section, we show that even where Williams comes closest to being hostile to Wittgenstein, which is in connection with Wittgenstein’s views on the nature of philosophy, Williams inherits from Wittgenstein a certain conception of what philosophy can do for us—​of how philosophy can help us to live. Williams identified Wittgenstein as the originator of ‘an informal but very powerful style of reflection which sought insight into the origins of philosophical problems’ (2014a: 46). This is usually described as a ‘therapeutic’ conception of philosophy, on which ‘philosophy dissolves the conceptual confusions to which philosophical problems are alleged to owe their existence’ (Glock 1996: 294). Williams himself endorsed a vision of philosophy that had evident therapeutic elements. He wrote, for example, that philosophy, particularly when it sets out to destroy extant theories, aims ‘to liberate a reader . . . from distortions or misunderstandings involved in his or her own experience’ (1995d: 218). This therapeutic element also comes out when he responds in an interview to the charge that Ethics and the Limits of Philosophy is a purely negative contribution to philosophy: I don’t see [it] as negative, I see it hopefully as liberating. It seems to me people get themselves in situations in which they feel they have no right to have certain kinds of moral thoughts because they don’t fit in with some very impoverished theoretical picture of what constitutes moral thought. (1996b) [p. 320] In order to realize its therapeutic aims, Wittgenstein thought, philosophy had to consist notably in the ‘assembly of reminders’—​typically reminders of how we used some piece of language in a non-​philosophical context—​as a way of dissolving various putative philosophical problems. This is evident in his distinctive use of examples. As examples are conventionally used in philosophy, they play a role roughly analogous to that of experiments in the natural sciences, providing support for a hypothesis by eliciting pre-​theoretical judgements or intuitions, or undermining a hypothesis by confronting it with a counterexample. In Wittgenstein’s use, however, examples act as reminders of truths we overlooked because we were in the grip of some ‘picture’ that ‘held us captive’ (2009: §115). Wittgenstein’s influence in this respect also shows itself in Williams’s own use of examples. Two of Williams’s most famous examples are the mini-​narratives of ‘George’ and ‘Jim’, which centre on the notion of integrity, and why an agent’s displaying integrity—​by refusing to take up a job he regards as immoral (George), or refusing to kill an innocent even if doing so may prevent what he acknowledges as a worse outcome (Jim)—​may be inconsistent with his acting as the utilitarian recommends. Williams later clarified that the cases of George and Jim had never been intended as counterexamples. Instead, he described his aims in language strongly reminiscent of Wittgenstein: When I brought in integrity, it was . . . as a quality that many people prize and admire. It is in such ways that people put the notion to ethical use. My claim was that if people do put it to ethical use, they cannot accept the picture of action and of moral motivation that direct utilitarianism requires—​and here were two stories to remind them, perhaps in different ways, of that truth. (1995d: 212; emphases added) Clearly, there is a Wittgensteinian conception of philosophy in the background of this remark: a conception on which philosophy can help us to live by assembling reminders that liberate us from a picture holding us captive. And just as the pictures that Wittgenstein most fervently attacked tended to be the ones he used to be in thrall to himself, so the utilitarian picture that Williams invites us to break free of is one that had once held Williams himself captive: there was a time when he had ‘very pious utilitarian views’, he confessed in a late interview; but he came to see that ‘consequentialist reasoning could just lead you on and on in the wrong direction’ (Jeffries 2002). At the same time, the cases of George and Jim illustrate how Williams goes beyond Wittgenstein when he gives examples. The reminders assembled are not, as is usually the case in Wittgenstein, linguistic or ‘grammatical’: they are ethical reminders, and their concern with the constitutive features of moral dilemmas requires literary means more expansive than those afforded by Wittgensteinian epigrams: Williams’s narrative is replete with incidental contextual detail, [p. 321] Matthieu Queloz and Nikhil Krishnan biographical facts, and psychological speculation. George’s dilemma—​whether to accept a job at a chemical and biological weapons laboratory—​is not a timeless one, but one that belongs to the 1970s, the decade in which Williams was writing—​ the age of napalm bombings in Vietnam. The same is true of Jim, whose own impossible choice is taken against the background of the military coups across Latin America and the sustained persecution of indigenous people, facts (albeit seldom remarked ones) which provide significant historical context for that philosophical example. The deliberate allusion to real-​world political events in Williams’s examples illustrates how a Wittgensteinian style might be adapted to the nature of its subject matter; the assembly of ethical reminders calls for a different style from the assembly of grammatical ones. Williams admitted that his examples were ‘extremely schematic’ in comparison to biography, history, or realist fiction, but they at least had the virtue of being ‘nearer to psychological and social reality than the theory’, and could bring out the fact that certain ethical theories were simply ‘frivolous’ d: 217). Utilitarianism, in particular, is a clear instance of the sort of philosophy that Williams might have seen as having ingenuity without insight, or, as he himself put it, ‘combining technical complexity with simple-​mindedness’, which he took to mean ‘having too few thoughts and feelings to match the world as it really is’ (1973a: 149). Moreover, Williams’s therapeutic conception of philosophy differs from therapeutic conceptions that locate the root cause of distortions and misunderstandings within philosophy itself. Williams showed considerable impatience with such views, because they underestimated the seriousness of the problem: if philosophy was itself the cause of philosophical problems, the best thing would surely be to stop doing philosophy altogether. But to make ‘an academic philosophy out of denouncing academic philosophy’, Williams thought, was at best ‘wonderfully perverse, rather like setting up as a Kierkegaardian bishop’ (1995d: 218). If there was to be an academic philosophy inspired by the Wittgensteinian conception of philosophy as therapy, it had better be a response to problems that were not merely artefacts of philosophy itself. A more serious and less frivolous form of philosophical therapy, Williams maintained, should aim to liberate us from deception by ‘forces worth worrying about, such as our own fears and resentments, our misunderstanding of social representations, and the effects of tradition’ (1995d: 219). Far from being the disease to which it is itself the cure, philosophy is at most a symptom: ‘ethical theory is not itself the basic condition with which we should be concerned, but a symptom, the expression of that condition in the tissue of a certain type of philosophy’ (2021: 275). If the destruction of philosophical theory can play a part in liberating us from distortions in our experience, it is because ‘a powerful philosophical theory can be an effectively articulated expression of those distortions’ (1995d: 218). [p. 322] On this conception, one will still look for ways of thinking and speaking that are ‘philosophically alienated from human purposes’ (2006j: 210), as Williams puts it; but one will have to press further the question of what explains the tendency for people to be alienated in this way. Is it an eternal fact of the human condition? Or is it an especially prominent tendency in certain kinds of society, or in particular historical periods, such as, say, ‘modernity’? Williams mentions Cavell’s The Claim of Reason (1979) in this connection, which concerns what Williams summarizes as ‘ever-​present possibilities of scepticism, implicit in the human condition’ (2006j: 210). Are the possibilities literally ‘ever-​present’? Or are they, as Cavell sometimes also suggests, the feature of some more local ‘cultural situation’? But if the latter, is it sufficient to engage with the cultural situation in purely ‘metaphysical terms’ rather than in those of empirical history? Does a purely metaphysical treatment not risk transforming the Wittgensteinian project into something more objectionably Heideggerian, where the modern world is seen as ‘a fall from primary unity’, and turns what should be a matter of evidence-​based history into a ‘mythical story of severance’ (2006j: 210)? Therapeutic philosophy, for Williams, needs to operate not at the level of myth, but at the level of sober history, or of psychological and social explanation, if it is to understand the underlying causes of philosophical distortions and misunderstandings. Williams’s humanistic conception of philosophy thus owes a great deal to Wittgenstein: its stylistic ideals of clarity and precision, and of the depth and density required to get things right; its anthropological interest in thick concepts and their point; and its therapeutic ambition to liberate us from distortions in our self-​ understanding by assembling reminders. But Williams also departs from Wittgenstein in going beyond a priori anthropology, insisting that philosophy engage with concrete societies’ empirical determination by contingent historical forces. And for all that Williams’s conception of the aims of philosophy can be set out in therapeutic terms, his Left Wittgensteinian form of therapy is precisely not guaranteed to be conciliatory or quietistic, still less to bring the philosopher ‘peace’. It has more in common with what Nietzsche, Foucault, or the Frankfurt School would have called ‘critique’ than any activity that Wittgenstein would have recognized as properly philosophical. As we have shown, however, even these differences are departures from Wittgenstein’s conception of philosophy, marking out Williams’s own conception as descending, by dialectical evolution, from Wittgensteinian materials. ‘One might say that the subject we are dealing with’, Wittgenstein once remarked, ‘is one of the heirs of the subject which used to be called “philosophy” ’ (1958: 28). In light of the philosophical family resemblances we have brought out, one might equally say that Williams’s humanistic conception of philosophy is one of the heirs of the subject Wittgenstein was dealing with. Of the many inheritors of Wittgenstein’s legacy, Williams’s impure humanism may yet prove the more enduring. As in the [p. 323] Matthieu Queloz and Nikhil Krishnan Parable of the Talents, there is something to be said for not merely burying one’s inheritance, but transforming it. References Anscombe, G. E. M. 2011. ‘Wittgenstein’s “Two Cuts” in the History of Philosophy’. In From Plato to Wittgenstein: Essays by G. E. M. Anscombe. Edited by Mary Geach and Luke Gormally, 181–​ 6. Exeter: Imprint Academic. Babbiotti, Paolo. 2020. ‘Compression: Nietzsche, Williams, and the Problem of Style’. European Journal of Philosophy, . Babbiotti, Paolo. 2023. Problems of Style in Philosophy: Stanley Cavell and Bernard Williams. PhD thesis, Università degli Studi di Torino. Blackburn, Simon. 2000. ‘Reenactment as Critique of Logical Analysis: Wittgensteinian Themes in Collingwood’. In Empathy and Agency: The Problem of Understanding in the Human Sciences. Edited by Hans Herbert Kögler and Karsten R. Stueber, 270–​87. Boulder: Westview Press. Bloor, David. 1983. Wittgenstein: A Social Theory of Knowledge. New York: Columbia University Press. Bloor, David. 1992. ‘Left and Right Wittgensteinians’. In Science as Practice and Culture. Edited by Andrew Pickering, 266–​82. Chicago: University of Chicago Press. Bloor, David. 1997. Wittgenstein, Rules and Institutions. London: Routledge. Bloor, David. 2000. ‘Wittgenstein as a Conservative Thinker’. In The Sociology of Philosophical Knowledge. Edited by Martin Kusch, 1–​14. Dordrecht: Springer Netherlands. Breckman, Warren. 2019. ‘The Young Hegelians: Philosophy as Critical Praxis’. In The Cambridge History of Modern European Thought, Volume 1: The Nineteenth Century. Edited by Peter E. Gordon and Warren Breckman, 88–​110. Cambridge: Cambridge University Press. Broackes, Justin. 2011. ‘Introduction’. In Iris Murdoch, Philosopher. Edited by Justin Broackes, Oxford: Oxford University Press. Brusotti, Marco. 2014. Wittgenstein, Frazer und die ‘ethnologische Betrachtungsweise’. Berlin and Boston, MA: De Gruyter. Cavell, Stanley. 1979. The Claim of Reason: Wittgenstein, Skepticism, Morality, and Tragedy. New York: Oxford University Press. Cavell, Stanley. 2010. Little Did I Know: Excerpts from Memory. Stanford: Stanford University Press. Celikates, Robin. 2015. ‘Against Manichaeism: The Politics of Forms of Life and the Possibilities of Critique’. Raisons politiques 57.1: 81–​96. Chanan, Michael. 1972. Ep. 3: Appearance and Reality. In Logic Lane. United Kingdom: Chanan Films. Craig, Edward. 1990. Knowledge and the State of Nature: An Essay in Conceptual Synthesis. Oxford: Clarendon Press. Craig, Edward. 1993. Was wir wissen können: Pragmatische Untersuchungen zum Wissensbegriff. Wittgenstein-​ Vorlesungen der Universität Bayreuth. Edited by Wilhelm Vossenkuhl. Frankfurt: Suhrkamp. Cueni, Damian. Forthcoming. ‘The Idea of Humanistic Middle-​R ange Theory’. In Bernard Williams: From Responsibility to Law and Jurisprudence. Edited by Veronica Rodriguez-​ Blanco, Daniel Peixoto Murata, and Julieta Rabanos. Oxford: Hart. Cueni, Damian, and Matthieu Queloz. 2021. ‘Whence the Demand for Ethical Theory?’ American Philosophical Quarterly 58.2: 135–​46. Diamond, Cora. 2018. ‘Bernard Williams on the Human Prejudice’. Philosophical Investigations [p. 324] Drury, Maurice O’Connor. 2017. The Selected Writings of Maurice O’Connor Drury: On Wittgenstein, Philosophy, Religion and Psychiatry. London: Bloomsbury. Fricker, Miranda. 2023. ‘A Project of “Impure” Enquiry: Williams’ Historical Self-​Consciousness’. European Journal of Philosophy, . Geertz, Clifford. 1973. ‘Thick Description: Toward an Interpretive Theory of Culture’. In The Interpretation of Cultures: Selected Essays, 3–​30. New York: Basic Books. Glock, Hans-​Johann. 1996. A Wittgenstein Dictionary. Oxford: Blackwell. Glock, Hans-​Johann. 2006. ‘Wittgenstein and History’. In Wittgenstein: The Philosopher and His Works. Edited by A. Pichler and S. Säätelä, 277–​303. Frankfurt: Ontos. Glock, Hans-​Johann. 2017a. ‘Impure Conceptual Analysis’. In The Cambridge Companion to Philosophical Methodology. Edited by Giuseppina D’Oro and Søren Overgaard, 77–​100. Cambridge: Cambridge University Press. Glock, Hans-​ Johann. 2017b. ‘Philosophy and Philosophical Method’. In A Companion to Wittgenstein. Edited by Hans-​ Johann Glock and John Hyman, 231–​ Chichester: Wiley-​Blackwell. Glock, Hans-​Johann. 2017c. ‘Wittgenstein’s Rain in the “Philosophical Desert”: Pragmatist Ideas in Post-​war Oxford’. In The Practical Turn: Pragmatism in Britain in the Long Twentieth Century. Edited by Cheryl Misak and Huw Price, 131–​48. Oxford: Oxford University Press. Hacker, Peter Michael Stephan. 2011. ‘Wittgenstein and the Autonomy of Humanistic Understanding’. In Wittgenstein, Theory and the Arts. Edited by Richard Allen and Malcolm Turvey, 39–​74. London: Routledge. Hacker, Peter Michael Stephan. 2013. ‘Wittgenstein’s Anthropological and Ethnological Approach’. In Wittgenstein: Comparisons and Context, 111–​27. Oxford: Oxford University Press. Hanfling, Oswald. 1985. ‘A Structural Account of Knowledge’. The Monist 68.1: 40–​57. Jaeggi, Rahel. 2016. Critique of Forms of Life. Cambridge, MA: Harvard University Press. Jeffries, Stuart. 2002. ‘The Quest for Truth’. The Guardian, 30 November. Kremer, Michael. 2022. ‘Margaret MacDonald and Gilbert Ryle: A Philosophical Friendship’. British Journal for the History of Philosophy 30.2: 288–​311. Krishnan, Nikhil. 2023. A Terribly Serious Adventure: Philosophy at Oxford 1900–​60. London: Profile. Krishnan, Nikhil, and Matthieu Queloz. 2023. ‘The Shaken Realist: Bernard Williams, the War, and Philosophy as Cultural Critique’. European Journal of Philosophy 31.1: 226–​47. Lear, Jonathan. 1982. ‘Leaving the World Alone’. Journal of Philosophy 79.7: 382–​403. Lear, Jonathan, and Barry Stroud. 1984. ‘The Disappearing “We” ’. Proceedings of the Aristotelian Society, Supplementary Volume 58: 219–​58. Macedo Jr, Ronaldo Porto. Forthcoming. ‘An Unusual but Successful Philosophical Marriage: Nietzsche and Wittgenstein’. In Bernard Williams: From Responsibility to Law and Jurisprudence. Edited by Veronica Rodriguez-​Blanco and Daniel Peixoto Murata. Misak, Cheryl. 2016. Cambridge Pragmatism: From Peirce and James to Ramsey and Wittgenstein. Oxford: Oxford University Press. Misak, Cheryl. 2021. ‘Williams, Pragmatism, and the Law’. Res Publica 27.2: 155–​70. Misak, Cheryl. Forthcoming. Oxford Pragmatism. Oxford: Oxford University Press. Moore, Adrian W. 2012. The Evolution of Modern Metaphysics: Making Sense of Things. Cambridge: Cambridge University Press. Moore, Adrian W. 2019. ‘Transcendental Idealism in Wittgenstein, and Theories of Meaning’. In Language, World, and Limits: Essays in the Philosophy of Language and Metaphysics, 71–​89. Oxford: Oxford University Press. Moran, Richard. 2016. ‘Williams, History, and “the Impurity of Philosophy” ’. European Journal of Philosophy 24.2: 315–​30. Mulhall, Stephen. 2015. ‘ “Hopelessly Strange”: Bernard Williams’s Portrait of Wittgenstein as a Transcendental Idealist’. In The Transcendental Turn. Edited by Sebastian Gardner and Matthew Grist, 324–​41. Oxford: Oxford University Press. [p. 325] Matthieu Queloz and Nikhil Krishnan Norris, Andrew. 2009. ‘ “La chaîne des raisons a une fin”: Wittgenstein et Oakeshott sur le rationalisme et la pratique’. Cités: Philosophie, Politique, Histoire 38: 95–​108. Nussbaum, Martha C. 2003. ‘Tragedy and Justice: Bernard Williams Remembered’. Boston Review, 1 October, . Nyíri, J. C. 1976. ‘Wittgenstein’s New Traditionalism’. Acta Philosophica Fennica 28: 501–​12. Nyíri, J. C. 1982. ‘Wittgenstein’s Later Work in Relation to Conservatism’. In Wittgenstein and His Times. Edited by Brian McGuinness, 44–​68. Chicago: University of Chicago Press. O’Grady, Jane. 2003. ‘Sir Bernard Williams’. The Guardian, 13 June. Owen, David. 2001. ‘Wittgenstein and Genealogy’. Nordic Journal of Philosophy 2.2: 5–​29. Pettit, Philip. 2000. ‘Winch’s Double-​Edged Idea of a Social Science’. History of the Human Sciences 13.1: 63–​77. Pichler, Alois. 2023. Style, Method and Philosophy in Wittgenstein. Cambridge: Cambridge University Press. Pitkin, Hanna Fenichel. 1972. Wittgenstein and Justice: On the Significance of Ludwig Wittgenstein for Social and Political Thought. Berkeley: University of California Press. Pleasants, Nigel. 1999. Wittgenstein and the Idea of a Critical Social Theory: A Critique of Giddens, Habermas and Bhaskar. London: Routledge. Pleasants, Nigel. 2002. ‘Towards a Critical Use of Marx and Wittgenstein’. In Marx and Wittgenstein: Knowledge, Morality and Politics. Edited by Gavin Kitching and Nigel Pleasants, London: Routledge. Plotica, Luke Philip. 2015. Michael Oakeshott and the Conversation of Modern Political Thought. Albany, NY: State University of New York Press. Queloz, Matthieu. 2016. ‘Wittgenstein on the Chain of Reasons’. Wittgenstein-​Studien 7.1: Queloz, Matthieu. 2017. ‘Two Orders of Things: Wittgenstein on Reasons and Causes’. Philosophy Queloz, Matthieu. 2018. ‘Williams’s Pragmatic Genealogy and Self-​Effacing Functionality’. Queloz, Matthieu. 2021a. The Practical Origins of Ideas: Genealogy as Conceptual Reverse-​ Engineering. Oxford: Oxford University Press. Queloz, Matthieu. 2021b. ‘The Self-​Effacing Functionality of Blame’. Philosophical Studies 178.4: Queloz, Matthieu. 2022. ‘The Essential Superficiality of the Voluntary and the Moralization of Psychology’. Philosophical Studies 179.5: 1591–​620. Queloz, Matthieu. 2024. ‘The Dworkin–​Williams Debate: Liberty, Conceptual Integrity, and Tragic Conflict in Politics’. Philosophy and Phenomenological Research 109.1: 3–​29. Queloz, Matthieu. 2025. The Ethics of Conceptualization: Tailoring Thought and Language to Need. Oxford: Oxford University Press. Queloz, Matthieu. Forthcoming. ‘Law as a Test of Conceptual Strength’. In Bernard Williams on Law and Jurisprudence: From Agency and Responsibility to Methodology. Edited by Veronica Rodriguez-​Blanco, Daniel Peixoto Murata, and Julieta Rabanos. Oxford: Hart. Queloz, Matthieu, and Damian Cueni. 2021. ‘Left Wittgensteinianism’. European Journal of Philosophy 29.4: 758–​77. Rorty, Richard. 1983. ‘Postmodernist Bourgeois Liberalism’. Journal of Philosophy 80.10: 583–​9. Rorty, Richard. 1989. Contingency, Irony, and Solidarity. Cambridge: Cambridge University Press. Rowe, M. W. 2023. J. L. Austin: Philosopher and D-​Day Intelligence Officer. Oxford: Oxford University Press. Ryle, Gilbert. 2009a. ‘Thinking and Reflecting’. In Collected Papers, Volume 2: Collected Essays Abingdon: Routledge. Ryle, Gilbert. 2009b. ‘The Thinking of Thoughts: What Is ‘Le Penseur’ Doing?’ In Collected Papers, Volume 2: Collected Essays 1929–​1968, 494–​510. Abingdon: Routledge. [p. 326] Tanney, Julia. 2021. ‘Gilbert Ryle’. In The Stanford Encyclopedia of Philosophy (Summer 2021 edition). Edited by Edward N. Zalta. . Temelini, Michael. 2015. Wittgenstein and the Study of Politics. Toronto: University of Toronto Press. Toews, John. 1985. Hegelianism: The Path toward Dialectical Humanism, 1805–​1841. Cambridge: Cambridge University Press. Williams, Bernard. 1958. ‘The Revolution in Philosophy, by A. J. Ayer et al.’. Philosophy 33.124: 65–​7. Williams, Bernard. 1969. ‘Philosophy’. In General Education: A Symposium on the Teaching of Non-​specialists. Edited by Michael Yudkin, 138–​64. Harmondsworth: Penguin. Williams, Bernard. 1973a. ‘A Critique of Utilitarianism’. In Utilitarianism: For and Against, 77–​ Cambridge: Cambridge University Press. Williams, Bernard. 1973b. ‘Deciding to Believe’. In Williams, Problems of the Self, 136–​51. Cambridge: Cambridge University Press. Williams, Bernard. 1973c. ‘Wittgenstein and Idealism’. Royal Institute of Philosophy Supplements Williams, Bernard. 1980a. ‘Conclusion’. In Morality as a Biological Phenomenon: The Presuppositions of Sociobiological Research. Edited by Gunther S. Stent, 275–​85. Berkeley and Los Angeles: University of California Press. Williams, Bernard. 1980b. ‘Political Philosophy and the Analytical Tradition’. In Political Theory and Political Education. Edited by Melvin Richter, 57–​75. Princeton: Princeton University Press. Williams, Bernard. 1981a. ‘Moral Luck’. In Moral Luck, 20–​39. Cambridge: Cambridge University Press. Williams, Bernard. 1981b. ‘Ought and Moral Obligation’. In Moral Luck, 114–​23. Cambridge: Cambridge University Press. Williams, Bernard. 1982. ‘The Spell of Linguistic Philosophy: Dialogue with Bernard Williams’. In Men of Ideas: Some Creators of Contemporary Philosophy. Edited by Bryan Magee, 110–​24. Oxford: Oxford University Press. Williams, Bernard. 1983. ‘The Uses of Philosophy: An Interview with Bernard Williams’. The Center Magazine, November/​December, 40–​9. Williams, Bernard. 1985. Ethics and the Limits of Philosophy. Routledge Classics Edition. London: Routledge. Williams, Bernard. 1986. ‘Reply to Simon Blackburn’. Philosophical Books 27.4: 203–​8. Williams, Bernard. 1990. ‘Reply to the President’. Proceedings of the Aristotelian Society 90.1: Williams, Bernard. 1992. ‘Left-​Wing Wittgenstein, Right-​Wing Marx’. Common Knowledge 1.1: 33–​42. Williams, Bernard. 1993. Shame and Necessity. Berkeley: University of California Press. Williams, Bernard. 1995a. ‘How Free Does the Will Need to Be?’ In Making Sense of Humanity and Other Philosophical Papers 1982–​1993, 3–​21. Cambridge: Cambridge University Press. Williams, Bernard. 1995b. ‘Internal Reasons and the Obscurity of Blame’. In Making Sense of Humanity and Other Philosophical Papers 1982–​1993, 35–​45. Cambridge: Cambridge University Press. Williams, Bernard. 1995c. ‘Moral Luck: A Postscript’. In Making Sense of Humanity and Other Philosophical Papers 1982–​1993, 241–​7. Cambridge: Cambridge University Press. Williams, Bernard. 1995d. ‘Replies’. In World, Mind, and Ethics: Essays on the Ethical Philosophy of Bernard Williams. Edited by J. E. J. Altham and Ross Harrison, 185–​224. Cambridge: Cambridge University Press. Williams, Bernard. 1995e. ‘Saint-​ Just’s Illusion’. In Making Sense of Humanity and Other Philosophical Papers 1982–​1993, 135–​50. Cambridge: Cambridge University Press. Williams, Bernard. 1995f. ‘Truth in Ethics’. Ratio 8.3: 227–​42. [p. 327] Matthieu Queloz and Nikhil Krishnan Williams, Bernard. 1995g. ‘What Has Philosophy to Learn from Tort Law?’ In The Philosophical Foundations of Tort Law. Edited by David G. Owen, 487–​98. Oxford: Oxford University Press. Williams, Bernard. 1995h. ‘Who Needs Ethical Knowledge?’ In Making Sense of Humanity and Other Philosophical Papers 1982–​1993, 203–​12. Cambridge: Cambridge University Press. Williams, Bernard. 1996a. ‘Contemporary Philosophy: A Second Look’. In The Blackwell Companion to Philosophy. Edited by Nicholas Bunnin and Eric P. Tsui-​James, 23–​35. Oxford: Blackwell. Williams, Bernard. 1996. ‘A Fugitive from the Pigeonhole: Interview by John Davies’. Times Higher Education Supplement 1252, 1 November. Williams, Bernard. 1997. Der Wert der Wahrheit. Vienna: Passagen. Williams, Bernard. 1999. ‘Bernard Williams’. In Key Philosophers in Conversation: The Cogito Interviews. Edited by Andrew Pyle, 142–​63. London: Routledge. Williams, Bernard. 2001a. ‘Foreword: Some Philosophical Recollections’. In Wittgensteinian Themes: Essays in Honour of David Pears. Edited by David Charles and William Child, xiii–​ xviii. Oxford: Clarendon Press. Williams, Bernard. 2001b. Morality: An Introduction to Ethics. Cambridge: Cambridge University Press. Williams, Bernard. 2002. Truth and Truthfulness: An Essay in Genealogy. Princeton: Princeton University Press. Williams, Bernard. 2005a. ‘Censorship’. In In the Beginning Was the Deed: Realism and Moralism in Political Argument. Edited by Geoffrey Hawthorne, 139–​ Princeton: Princeton University Press. Williams, Bernard. 2005b. Descartes: The Project of Pure Enquiry. London: Routledge. Williams, Bernard. 2005c. ‘From Freedom to Liberty: The Construction of a Political Value’. In In the Beginning Was the Deed: Realism and Moralism in Political Argument. Edited by Geoffrey Hawthorne, 75–​96. Princeton: Princeton University Press. Williams, Bernard. 2005d. ‘The Liberalism of Fear’. In In the Beginning Was the Deed: Realism and Moralism in Political Argument. Edited by Geoffrey Hawthorne, 52–​61. Princeton: Princeton University Press. Williams, Bernard. 2005e. ‘Pluralism, Community and Left Wittgensteinianism’. In In the Beginning Was the Deed: Realism and Moralism in Political Argument. Edited by Geoffrey Hawthorne, 29–​39. Princeton: Princeton University Press. Williams, Bernard. 2005f. ‘Realism and Moralism in Political Theory’. In In the Beginning Was the Deed: Realism and Moralism in Political Argument. Edited by Geoffrey Hawthorne, 1–​17. Princeton: Princeton University Press. Williams, Bernard. 2006a. ‘An Essay on Collingwood’. In The Sense of the Past: Essays in the History of Philosophy. Edited by Myles Burnyeat, 341–​60. Princeton: Princeton University Press. Williams, Bernard. 2006b. ‘The Human Prejudice’. In Philosophy as a Humanistic Discipline. Edited by A. W. Moore, 135–​54. Princeton: Princeton University Press. Williams, Bernard. 2006c. ‘Janáček’s Modernism: Doing Less with More in Music and Philosophy’. In On Opera. Edited by Patricia Williams and Michael Tanner, 118–​20. New Haven: Yale University Press. Williams, Bernard. 2006d. ‘Moral Responsibility and Political Freedom’. In Philosophy as a Humanistic Discipline. Edited by A. W. Moore, 119–​25. Princeton: Princeton University Press. Williams, Bernard. 2006e. ‘Philosophy as a Humanistic Discipline’. In Philosophy as a Humanistic Discipline. Edited by A. W. Moore, 180–​99. Princeton: Princeton University Press. Williams, Bernard. 2006f. ‘Political Philosophy and the Analytical Tradition’. In Philosophy as a Humanistic Discipline. Edited by A. W. Moore, 155–​68. Princeton: Princeton University Press. Williams, Bernard. 2006g. ‘The Primacy of Dispositions’. In Philosophy as a Humanistic Discipline. Edited by A. W. Moore, 67–​75. Princeton: Princeton University Press. Williams, Bernard. 2006h. ‘Understanding Homer: Literature, History and Ideal Anthropology’. In The Sense of the Past: Essays in the History of Philosophy. Edited by Myles Burnyeat, 60–​70. Princeton: Princeton University Press. [p. 328] Williams, Bernard. 2006i. ‘Values, Reasons, and the Theory of Persuasion’. In Philosophy as a Humanistic Discipline. Edited by A. W. Moore, 109–​18. Princeton: Princeton University Press. Williams, Bernard. 2006j. ‘What Might Philosophy Become?’. In Philosophy as a Humanistic Discipline. Edited by A. W. Moore, 200–​14. Princeton: Princeton University Press. Williams, Bernard. 2007. ‘Truth and Truthfulness’. In What More Philosophers Think. Edited by Julian Baggini and Jeremy Stangroom, 130–​46. London: Continuum. Williams, Bernard. 2009. ‘A Mistrustful Animal’. In Conversations on Ethics. Edited by Alex Voorhoeve, 195–​214. Oxford: Oxford University Press. Williams, Bernard. 2014a. ‘The Concept of a Person, by A. J. Ayer’. In Essays and Reviews 1959–​ Edited by Michael Wood, 45–​7. Princeton: Princeton University Press. Williams, Bernard. 2014b. ‘On Hating and Despising Philosophy’. In Essays and Reviews, 1959–​ Edited by Michael Wood, 363–​70. Princeton: Princeton University Press. Williams, Bernard. 2014c. ‘Russell and Moore: The Analytical Heritage, by A. J. Ayer’. In Essays and Reviews, 1959–​2002. Edited by Michael Wood, 75–​7. Princeton: Princeton University Press. Williams, Bernard. 2019. ‘Left-​Wing Wittgenstein’. Common Knowledge 25.1–​3: 321–​31. Williams, Bernard. 2021. ‘Ethics, A Matter of Style? Introduction to the French Edition’. Philosophical Inquiries 9.2: 269–​84. Williams, Bernard, and Alan Montefiore. 1966. ‘Introduction’. In British Analytical Philosophy. Edited by Bernard Williams and Alan Montefiore, 1–​16. London: Routledge. Winch, Peter. 1958. The Idea of a Social Science and Its Relation to Philosophy. London: Routledge. Wittgenstein, Ludwig. 1958. The Blue and Brown Books: Preliminary Studies for the Philosophical Investigations. Oxford: Blackwell. Wittgenstein, Ludwig. 1978. Remarks on the Foundations of Mathematics. Edited by G. H. von Wright, R. Rhees, and G. E. M. Anscombe. 3rd edition. Oxford: Blackwell. Wittgenstein, Ludwig. 1989. Lectures on the Foundations of Mathematics, Cambridge, 1939. Edited by Cora Diamond. Chicago and London: University of Chicago Press. Wittgenstein, Ludwig. 2009. Philosophische Untersuchungen =​Philosophical Investigations. Translated by G. E. M. Anscombe, P. M. S. Hacker, and Joachim Schulte. Edited by P. M. S. Hacker and J. Schulte. Revised 4th edition. Chichester, UK: Wiley-​Blackwell. [p. 332] Why Bernard Williams Is a Bad Example for Historians of Philosophy John Marenbon Bernard Williams was an incisive, insightful, and learned historian of philosophy. But his explicit discussions of the method and aims of history of philosophy founder, ultimately because, much as he values history, he regards history of philosophy as valuable only instrumentally, for what it can do for first-​order philosophers. Moreover, his implicit conception of what belongs to the history of western philosophy was far too narrow, a reflection of his own prejudices (especially about religion) and those of his time, rather than of the broad-​minded, imaginative intelligence that characterizes most of his thinking. The first two sections here examine in turn Williams’s earlier and later views about methodology in history of philosophy, and the final section considers his implicit gappy conception of what constitutes western philosophy. 1. Williams on History of Philosophy: The Earlier View Williams begins his first and only full-​length monograph on a philosopher of the past, Descartes: The Project of Pure Enquiry, published in 1978, roughly in the middle of his career, with a very brief, assertive methodological Preface (1978: 9–​ The study that follows, he states, is one ‘in the history of philosophy rather than in the history of ideas’, and ‘the history of ideas is history before it is philosophy, while with the history of philosophy it is the other way round’. ‘For the history of ideas,’ Williams explains, ‘the question about a work what does it mean? is centrally the question what did it mean?’ By contrast, although ‘history of philosophy . . . has to constitute its object, the work in genuinely historical terms’, there is a ‘cut-​off point’, where the objective of authenticity, pursued by the historians of ideas, is replaced by that of ‘articulating philosophical ideas’. 1 In the 2005 reprint, the Preface is on pp. xiii–​xv. Many of the references in what follows are to essays by Williams collected into three books: Williams 2006a, 2006b, and 2014a. Each of these pieces has been assigned its own date and letter (e.g. Williams 2006f ). The citation in the References includes in square brackets the original date of publication and, except for his occasional writings in periodicals, details of the original edition. John Marenbon, Why Bernard Williams Is a Bad Example for Historians of Philosophy In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. DOI: 10.1093/​9780191966361.003.0016 [p. 333] John Marenbon It might seem that Williams should be allowed to stipulate what he means by ‘history of philosophy’ and ‘history of ideas’, even if his use of these labels turns out to be unusual. But in his way of making this distinction, he has already introduced the potential for some confusion. What does it mean for a book or an essay to be philosophy before it is history? The question can be answered with regard to procedure, or with regard to accomplishment. In the procedural answer, certain ways of writing—​for instance, setting out highly abstract arguments, analysing general concepts—​will be identified as philosophical (the choice will probably be based on the procedures philosophers use now), and a piece of work will be identified as primarily philosophical if it consists mainly of this sort of writing. For the answer in terms of accomplishment, a work will be primarily philosophical only if it contributes to philosophy, and since it is impossible to contribute to what is past, it must make a contribution—​adding, for instance, some new position, argument, refutation, or analysis—​to some area of today’s philosophy. Consider in this light what might be meant by ‘history of philosophy’. Most areas of human activity have a history, which, as studied by scholars, is called ‘history of x’: for example, ‘history of farming’, ‘history of warfare’, ‘history of the novel’. Different though farming, warfare, and novels are from each other, histories of each—​that is to say, what the historians produce (as is designated here by ‘history of x’, by contrast with ‘the history of x’, which is the subject matter of history of x)—​are similar in kind: they are writings, aiming to recount and consider what happened in the past in these particular areas of life. History of x, that is to say, is history-​like, but it need not be like x. In some cases, indeed, it cannot be at all like x: history of farming is not done with combine harvesters, nor history of warfare with battleships and tanks. In other cases, history of x will be more x-​like. Simply and uncontroversially, where the x itself consists of written texts, as in the case of novels, its history will be of the same kind in this respect. It would not, however, be unreasonable, although it might be controversial, to go much further and demand that a historian of the novel should have, and demonstrate in their writing, at least some of the talents of a real novelist. Someone without these talents, it might be held, is not really an historian, but an antiquarian or a chronicler, giving the external facts about when novels were written, how they are divided up into chapters, what sort of subjects they treat. By contrast, genuine historians understand each novel from the inside and can see and explain how it works, because they could write a novel themselves—​not that, of course, in writing a history of the novel, they are actually writing a novel. The accomplishment at which they aim is historical truth, but their procedures embrace the fiction-​making skills of the novelists they are discussing. The analogy with history of philosophy is all too obvious. A brand of history of philosophy, which dates back to Russell’s study of Leibniz (1900), was widely practised in the late twentieth century, when Williams wrote the Project, and it [p. 334] is still popular today, though not unchallenged, in analytic departments—​call it ‘classic analytic history of philosophy’. Its exponents are proud of their credentials as philosophers, who are able to get inside the arguments of the texts they study, translating them into terms understood today, testing their logical coherence, probing them with counter-​arguments, unafraid to say where they are unclear, inadequate, or plain wrong. Nonetheless, they present their work as history. They tend to scorn non-​analytic approaches, which stick too close, they say, to the language and structures of the original texts, and so fail to make them properly comprehensible to readers today. Is Williams in his Preface simply endorsing classic analytic history of philosophy, while giving non-​analytic approaches their due by recognizing them as a different kind of pursuit, with different aims, ‘history of ideas’? There are two good reasons to think so. First, he does indeed call what he is doing ‘history of philosophy’ and so there is an expectation that like any history of x, the task it accomplishes is historical, though using philosophical procedures. Secondly, what Williams actually does in most of the Project seems to be classic analytic history of philosophy, as a look at other writers on the same subject confirms. The Project of Pure Enquiry appeared in the same year as two other books on Descartes’ philosophy by analytically minded authors, Margaret Dauler Wilson’s Descartes (Wilson 1978) and Edwin Curley’s Descartes Against the Sceptics (Curley A decade earlier, another analytical treatment of Descartes, by Anthony Kenny ([1968] 1993) had been published. The interpretations by the various authors offered are markedly different, but their approach is very similar. All the writers allow the plan of the Meditations to do a good deal of the structuring of their presentation, but all draw widely on Descartes’s Responses, his correspondence, and other works to illuminate his position. None of the interpreters is afraid to use the language and techniques of late-​twentieth-​century analytical philosophy in explaining Descartes’ arguments, and all of them take it as part of their job to assess these arguments, sometimes finding ways of rescuing them, sometimes declaring them irremediably flawed. In short, they adopt the classic analytical approach. Williams’s approach in the Project is in most respects indistinguishable from that of these colleagues of his. If anything, for the greater part, Williams sticks more closely to the order of exposition in the Meditations, and he is equally or more assiduous in bringing other comments by Descartes to bear on his interpretation. Yet, despite these two reasons, and despite the oddness of the position in which he leaves himself, where history of philosophy is not history at all, Williams does not finally ally himself with the classic analytic historians. The difference emerges in practice in the Project, although only on a few occasions there, where Williams is willing to leave Descartes behind completely. For example, Williams claims that Descartes takes the search for truth as a search for certainty, and that in doing so he is motivated by ‘a quite natural conception of enquiry’. He goes on: [p. 335] John Marenbon To gain some insight into that motivation, it will be helpful to leave Descartes’s own line of argument for a while, and examine in our own terms a very basic model of the search for truth. (1978: 37) The ten following pages discuss this basic model in contemporary terms. These non-​Cartesian interludes are very rare, but there is also a difference in overall aim between Williams and the classic analysts. The classic analysts seem to see their aim as to provide an interpretation of Descartes, and no more. For Williams, however, the ‘absolute conception’ of reality, which he argues Descartes is putting forward as the guiding principle of his search for truth, is a central idea in his own wider thinking, not because it has the same place for him as for Descartes, but because it will allow him to articulate the distinctiveness of humanistic disciplines, including philosophy, which should not try to seek this absolute conception. These differences are magnified in Williams’s explicit remarks in his Preface. Articulating the philosophical ideas he wishes to propose requires him, Williams explains, to go beyond the ‘cut-​off point’ where authenticity is the main aim; the ideas are not claimed as being in any direct, straightforward sense, Descartes’. What Descartes meant is ‘essentially ambiguous, incomplete, imperfectly determined by the author’s and his contemporaries’ understanding’. In its place, Williams offers a ‘rational reconstruction . . . where the rationality of the reconstruction is essentially and undisguisedly conceived in a contemporary style’ (1978: 10). The term ‘rational reconstruction’ (rationale Nachkonstruktion) had been introduced half a century earlier by Carnap to denote a type of ‘contextual definition or logical abstraction’ (cf. Beaney 2013) central to the practice of analytic philosophy itself. But the label came to be used for the way many analytic philosophers approached philosophers of the past, so that, writing in 1984, a few years after The Project of Pure Enquiry, Richard Rorty (1984) identified rational reconstruction as one of the genres of writing about the history of philosophy. For Rorty, the rational reconstructor uses past philosophers’ writings to engage in conversation with them, not as they really were, but as if they had been educated in today’s philosophy. Williams himself explains ‘rational reconstruction’ by means of a comparison: the method is like that Stravinsky uses in Pulcinella, where the melodies are Pergolesi’s, but the harmony his own; except that, whereas what counts as melody was determined largely by Pergolesi himself, the philosophical melodies Williams harmonizes by rationally reconstructing them are ‘determined, in some part, by subsequent philosophical experience’. We do not judge Pulcinella harshly because it is inauthentic (nor, even less, because, in fact, it is mostly not by Pergolesi, as Stravinsky believed it to have been). Stravinsky aimed to make the eighteenth-​century pieces he took his own. Similarly, Williams wants to do his own philosophizing. He allows himself, as he says, to be guided by what seems to him, in the late twentieth century, Descartes’ most interesting philosophical concerns, but he is in the business of formulating his own [p. 336] arguments, thereby making ‘the new work . . . broadly of the same genre as the original’ (1978: xiv). Williams therefore escapes the central complaint that classic analytical historians of philosophy must face: that their free-​ranging exploration of the arguments in the texts from the past, of their consequences and their failings, goes beyond what is needed for an understanding of the material and sometimes even leads away from it. Since Williams is aiming, not just to use philosophical procedures, like the classic analytic historians, but at a philosophical accomplishment, he must be judged according to what he has in fact accomplished philosophically in the Project—​and that is a good deal. Yet problems remain. Grant that The Project of Pure Enquiry makes an important, distinctive contribution to late-​ twentieth-​ century philosophy. Does it need to do so by spending much of its length tracking Descartes’ Meditations closely, by fitting (despite the view reached in the Preface) for 80 per cent of the time into the mould of classic analytic history of philosophy? Even if it is itself an achievement, is it one to be emulated? Williams does not stop to give any other examples of rational reconstruction. Rorty gives three: Strawson’s The Bounds of Sense, Jonathan Bennett on Locke, Berkeley, Hume and (!) A. J. Ayer’s Language, Truth and Logic. For Rorty, though, rational reconstruction is just one of four genres of writing history of philosophy, three of which he accepts. For Williams, whose concept of rational reconstruction is certainly no broader than Rorty’s, it is what constitutes history of philosophy (as opposed to history of ideas). History of philosophy seems destined to be sparsely populated, since, despite the success of the Project, it seems doubtful that there will be many other instances of what a philosopher now wants to say coming so serendipitously together with some of the concerns (seen through contemporary spectacles) of a philosopher from the past. 2. Williams on History of Philosophy: The Later View In the last decade or so before his premature death in 2003, Bernard Williams’s approach to philosophy changed. He distanced himself from many aspects of analytic philosophy, and very sharply from the scientism of some of its exponents (Williams 2006e). He commented cuttingly on its customary methods and limited goals, and its style ‘which seeks precision by total mind control’ (2006e: 183) He took a strong interest in Nietzsche (Williams 2006c, 2006f, 2006g, 2006h) and began to think intensely about the idea of genealogy (cf. Williams 2002), and about the relationship between philosophy and history (cf. Williams 1993). These developments had an impact on his views about history of philosophy. In 1994, Williams published his only article-​length treatment of the methods and aims of history of philosophy (Williams [1994] 2006d). Officially, it is about Descartes and the historiography of philosophy, but its scope is much more general. Many of Williams’s points reflect his self-​conscious distancing from analytic [p. 337] John Marenbon philosophy. He attacks an extreme form of analytic history of philosophy (‘we should approach the works of Plato as though they had appeared in last month’s issue of Mind’), But, he believes, the extreme forms of the approach ‘now belong to the demonology of the subject, and are rarely found in inhabited places in the daylight’ (2006d: 258–​9).2 In the light of this change in approach, it seems at first surprising that Williams begins this article by referring back to his 1978 contrast between history of philosophy and history of ideas and, apparently, repeating it. But what he goes on to argue shows that, although he uses nearly the same terms as before, they end up by distinguishing different activities. Williams’s central point, repeated in other places (e.g. 2006e: 181), is that history of philosophy should make the familiar seem strange, and the strange familiar. It should do so because Williams sets up as the aim of history of philosophy ‘to deploy ideas of the past in order to understand our own’, and this process is ‘one way’ of doing this, he says—​though this is the only way he gives (2006d: 259). In order to produce this effect of contrast (cf. Van Ackeren 2014: 311–​12: ‘Kontrastmittel’), historians of philosophy, while providing philosophy, must not treat their object ‘as though it were merely contemporary’, because in that way they ‘lose the point of historical distance’ and end up repeating our own preconceptions and giving us ‘our philosophy’ (2006d: 263, 261; original emphasis). Williams, therefore, would now regard what he had taken in 1978 to be the mark of the genuine historian of philosophy—​the practice of rational reconstruction, in which the link with a past thinker, such as Descartes, is no more than what appear now to be his ‘most interesting philosophical concerns’—​as losing the whole point of the enterprise. This might seem like a shift from demanding that history of philosophy be philosophy in accomplishment to the looser requirement that it should be philosophy in procedure, but the change is in fact more complex. By Williams’s later account, so long as history of philosophy is procedurally philosophy, it no longer needs to make direct accomplishments in contemporary philosophy because, through the effect of contrast, it makes accomplishments in contemporary philosophizing automatically, though indirectly. Williams’s criticisms of historians of philosophy who neglect the historical dimension apply not just to his own earlier approach, but in part, at least, to the classic analytical historians writing at the same time, who were not driven by Williams’s desire to be contributing directly to contemporary philosophy, but were arguably, when judged in the light of this new recognition of the distinctiveness of the past, over-​eager to interpret and correct, according to their own, 2 Matthieu Queloz suggests, I think rightly, that the binary division of Williams’s thinking about philosophy of the past into an earlier and a later view should be more nuanced, since the 1994 essay on Descartes and the historiography of philosophy, the main source used here for the later view, antedates the interest in genealogy that characterizes his last writings. [p. 338] late-​twentieth-​century criteria. Williams points explicitly to the ‘obtrusive cockiness, the condescension with which earlier writers are treated to instruction by current philosophical methods, and are reproved for their errors—​errors to which they have been committed, typically, by the way in which analytical philosophy interprets them’ (2006d: 258). From Williams’s new perspective, then, the central challenge that history of philosophy faces is to remain philosophy and yet be historical too. He complains—​ talking about Descartes, but the same point could be made about any past philosopher—​that ‘as things are, the history of ideas and the history of philosophy are likely to yield, respectively, one of two types of understanding, one purely historical, the other largely anachronistic’ (2006d: 260). The obvious solution, suggested by the very terms in which Williams puts the problem, seems to be to put the talents of the historians of philosophy together with those of the historians of ideas, to see the division of labour as an obstacle to the best understanding of philosophy of the past. But Williams explicitly rejects any such move. ‘It would be a mistake’, he says, ‘. . . to suppose that the distinction [between history of ideas and history of philosophy] is simply baseless, or that the best possible historical approach to a philosopher would consist in an ideal fusion of these two activities’ (Williams Williams gives two reasons for this view. The first is that ‘the best possible history of ideas is likely to show that the philosophy did not in fact mean in contemporary terms what subsequent philosophy has most made of it’ (2006d: 257). Why does Williams think this is an objection? One of the points he goes on to make is that by following through a past thinker’s influence on the course of philosophy up to the present, historians of philosophy will lose what they most need to discover—​ the strangeness of past thinking. As he summarizes it: ‘[The history of philosophy] cannot be identified with the history of influence, the progressive exploitation of original writing in one or more philosophical traditions; this . . . destroys strangeness, by following a path which necessarily lands us at precisely the place we are at’ d: 264). Those following what Williams describes as the best possible ‘history of ideas’, who go back beyond the history of influence that has led to how we now understand a certain notion (e.g. Cartesian dualism as discussed by contemporary philosophers of mind) to the idea as it was originally meant (Descartes’ own dualism), have found the way to avoid this pitfall. To achieve Williams’s goal, historians of philosophy need to follow this approach, not distinguish themselves from it. Williams’s second reason is that ‘the kinds of sensibility needed for the two activities are bound to yield partly incompatible products’ (2006d: 257). He gives the analogy of Impressionist painting which, by concentrating on the surface effects of light, does less than other styles of painting to explore structure. The analogy suggests a stronger claim than the literal statement: not, implausibly, that no painter could combine sensibilities so as to be able paint some canvases that show [p. 339] John Marenbon up wonderfully the effects of light, and others that successfully render structure, but rather, perhaps correctly, that no single canvas can achieve both these results to a high degree because rendering each requires a different arrangement of paint on the canvas. But, put in this way, it is immediately clear that the analogy does not hold. A painting is seen all at once, but a study in the history of philosophy is read sentence by sentence, and one sentence might convey contextual information, the next enter into analysis of an argument. In fact, in the process of research and composition, the historical and philosophical sensibilities need to be even more intimately interfused. The case is like that of historians of the novel; if they are not actually novelists themselves, they need to be so at least potentially, if they are to do more than describe externals. ‘What does this phrase mean?’ the historian of philosophy wonders. The structure of the argument suggests an answer (which would not be evident to someone who was not seeking out the philosophical point of the argument). But perhaps that answer seems impossible, given the normal meanings at the time of the words, or the wider intellectual context. Maybe at this point the historian finds a different, but also philosophically plausible reading; or they revise their understanding of the words or the intellectual context and, in doing so, find it necessary to revise the interpretation of the previous passage and then, perhaps, revise yet again how this phrase is to be taken. Just as, in the same gaze at a real landscape but in different instants, we take in the colours and the underlying structure, so the historian of philosophy moves almost imperceptibly between historical and philosophical considerations, each of them affecting the other. If philosophical and historical sensibilities were incompatible, as Williams claims, the writings of past thinkers would be left utterly beyond our comprehension. It is true that he does not rule out, but rather requires, some degree of combination: the activities of history of ideas and history of philosophy ‘cannot be totally separated from one another, and each needs to some extent the skills of the other’ d: 257). But the ‘totally’ and ‘to some extent’ here are a long way from doing justice to the idea that interpreting old philosophical texts, which is always the starting point and often the only business of an historian of philosophy, requires both sensibilities, both sets of skills, inextricably fused. There are at least some writers on the history of philosophy who live up to these requirements, Williams himself at times among them, and certainly there is no reason at all why the two sensibilities should not coexist in the same scholars. Williams’s unsustainable point about sensibilities does, however, have in its background a widely evident truth about academic institutions and academic careers. There is a strong pressure on anyone in such an institution to align themselves with the ideals of a particular type of department, to be a historian, or to be a philosopher. Historians are usually not at home with philosophical argument, or, even if they are, judge that their readers would be bored or baffled by it. As a result, they do not reach, or they jump over, or they rush through at breakneck speed and so [p. 340] mishandle the one indispensable stage in history of philosophy: interpreting the texts from the past as philosophy. Such historians of philosophy can provide materials for history of philosophy; they are also well placed to interpret the wider influence of philosophical ideas, because what is important here are the positions and slogans taken from philosophical works by a mainly non-​philosophical public. But they cannot do history of philosophy. That is left for philosophers. But in philosophy departments there is a Fundamental Assumption about the history of philosophy: that it is there because of its value to real, first-​order philosophers. Most historians in philosophy departments accept the Fundamental Assumption, or at least pretend to accept it—​not surprisingly, since they are very much in a minority, and many of their colleagues need convincing of their value. Analytic philosophy retains a great deal of its original indifference to history. In his last twenty years, Williams questioned much about the methods and aims of analytic philosophy, especially in its relation to history. Yet he never questioned the Fundamental Assumption. Indeed, he seemed not even to notice that he had made it. He simply proceeded as if it were a matter of course that the value of history of philosophy was in what it could contribute to real philosophy. In his early period he looked for a direct contribution to philosophical accomplishment. In his later period, he found a way, through the effect of contrast, in which history of philosophy could always contribute to the work of philosophy today, but it was still a matter of the historians justifying their existence through the instrumental value of their work.3 Once unmasked, the Fundamental Assumption is seen to be entirely unsupported except by the arrogance of first-​order philosophers. There are two academic pursuits, first-​order philosophy and history of philosophy, which can be justified as practices for themselves in the complex, vague, and problematic ways that academic pursuits can (or some will insist, wrongly, cannot) be justified. As well as their intrinsic value, the pursuits have instrumental value, for other pursuits and for each other. First-​order philosophy has great instrumental value for history of philosophy (which cannot be done without it), since it provides some of the necessary intellectual tools to understand the texts and a framework within which to explain them and, ultimately, where and how they do not fit it. Some first-​ order philosophers have held that history of philosophy has little instrumental 3 Williams does not merely ignore the intrinsic value of history of philosophy. He also unduly restricts its instrumental value for real philosophy by locating it only in the effect of contrast—​a point well argued in Van Ackeren (2014: 318–​20). Queloz (2017: 151) not only elaborates the idea of the effect of contrast, but indicates other instrumental uses for history of philosophy within the framework of Williams’s way of thinking: as a ‘test of reflective stability’ and as ‘a way of developing our reflective attitude’ towards our and others’ outlooks. I have suggested elsewhere (Marenbon 2011: 72) that a very important instrumental use of history of philosophy for philosophy is in trying to understand the nature of philosophy, which is not a natural kind but a particular human practice. Williams’s distinctions, as developed by Queloz, help to refine this point. Physics, too, is a particular human practice, but because its history is one of discovery, there is no need to look much at the details of its history to discover what sort of pursuit it is, because what matters is what it is now and will become. [p. 341] John Marenbon value for them, rather as history of science has limited instrumental value for science. Williams thought little of them (cf. Williams 2014f: 405). His clear understanding that philosophy does not progress in the manner of the natural sciences, that its history is non-​vindicatory—​one of change, and sometimes progress, but not discovery—​opens up especially many ways in which history of philosophy can contribute to it (cf. Queloz 2017). There is no reason, however, why its high instrumental value for philosophy should be taken to imply that history of philosophy lacks its own intrinsic value. Historians of science have been given the freedom to develop their discipline as they want, and to be judged according to its own criteria, just because scientists do not find it of much value for their own work. It is a strange irony that, because of the high instrumental value Williams and many other philosophers find in history of philosophy, it should not be allowed to develop as a discipline in itself, like history of science, and realize its full intrinsic value too. 3. Williams and the Contents of History of Philosophy When history of philosophy is considered for its instrumental value, the criteria suggested for it tend to be exclusively methodological. Either its content is left open—​it does not matter what bits of philosophy’s past are treated or not—​or it is determined by the methodology: for instance, for the early Williams, texts that contribute directly to present discussions; for the later Williams, texts that make the familiar strange. Once history of philosophy is granted its intrinsic value, its content too will be seen to need criteria (cf. Marenbon 2018: 43–​6). One is breadth in what is regarded as philosophy. The other is comprehensiveness of scope historically and geographically. These criteria of breadth and comprehensiveness apply most obviously to the field as a whole or in a wide area: history of philosophy should be judged to be going badly, in (say) Anglophone countries, if these criteria are not met by the combined production of those in the field. The criteria clearly do not apply to each piece of research in the area, nor in many cases to the whole of a scholar’s work. Most articles and books are specialized and so neither broad nor comprehensive, and such writing constitutes the life work of many fine historians of philosophy. Still, even in these cases a limited application of these criteria is appropriate. For instance, a Nietzsche specialist who excluded Zarathustra entirely from their discussions, regarding it as literature and not philosophy, would fail in breadth. Breadth would not inevitably demand that the specialist write a book or a chapter or even a paragraph about Zarathustra, but that at least its existence was recognized in thinking about Nietzsche’s philosophy. Similarly, a Frege specialist who lacked an active awareness of the German-​language influences on him would fail in comprehensiveness. The work of Williams, who writes about philosophers from [p. 342] many periods and places, and who makes general declarations about philosophy and history and philosophy and its history, can reasonably be held up against both these criteria: however, they must be used, of course, not to measure his actual writing (any individual’s productions would fall short), but the breadth of his intellectual horizons, what he admits within, and what he unthinkingly excludes from, the purview of philosophy. Williams meets the criterion of breadth. Although he preferred to use the word ‘philosophy’ rather restrictively, in a number of works but especially in Shame and Necessity, which he called a ‘philosophical description of historical reality’, Williams made the borders between what he called philosophy and other non-​ philosophical reflections deliberately vague and fluid, setting Homer, the tragedians, the historian Thucydides, Plato, and Aristotle side by side (Williams 1993: 4; cf. 12–​15 and passim). The problem for Williams is the criterion of comprehensiveness. There can be no complaint about how far back in the western tradition his view extends, since it runs to the pre-​philosophical precursors of Greek philosophy (as just indicated) and then to its earliest exponents, before the time of Socrates, who receive a considered treatment in his survey of the ‘Legacy of Greek Philosophy’ (Williams i: 12–​22). But Williams’s search for origins is geographically limited. At much the same time as Heraclitus and then Parmenides were thinking, Confucius was establishing one of the schools of Chinese philosophy, and the ultimate origins of Indian philosophy are usually traced equally far back into the past. When Williams j: 148) attributed to Plato the invention of ‘the subject of philosophy as we know it’, he was careful to clarify his point in the next paragraph by specifying ‘Western’ philosophy. Williams nowhere, however, shows any interest in non-​ western philosophy. It fell outside his mental horizon—​as it did and does, indeed, of that of almost all historians of philosophy in the West. But for a thinker who has discussed cultural relativism more acutely than almost anyone (especially in Williams 1974–​5), it is, at the least, an opportunity missed. Williams’s presentation of Greek philosophy in ‘The Legacy’ is strangely truncated. Plato himself is given an even fuller consideration than his predecessors, there is a little on Aristotle, some discussion of ancient scepticism, and nothing more. Williams bluntly declares that he will not be touching on philosophy after Aristotle (and indeed he hardly touches on Aristotle, a thinker he cordially hated), except for the Sceptics, although he acknowledges that the Stoics and Epicureans were influential. There is not even a mention of the longest-​lasting and by far the most influential of the ancient schools, the Platonists, from Plotinus in the third century to Olympiodorus in the sixth. To the gap left by this very incomplete treatment of antiquity, Williams added another: the Middle Ages. It is a period which, he admitted, lacked charm for him d: 354). It is true that he once showed, in an aside (2014e: 367), a remarkable grasp of a leading characteristic of its best-​known strand: its professionalization. [p. 343] John Marenbon When the Routledge Encyclopaedia of Philosophy first appeared in 1998, Bernard Williams was the speaker at the launch. He was clearly impressed by Norman Kretzmann’s achievement in assembling intellectually exciting articles on a range of medieval thinkers and he was open-​minded enough to admit that there might be much more philosophically to the period than he had realized. In his published work, however, Williams complacently embraced the exclusion of the Middle Ages from history of philosophy, for himself at least. Reviewing Charles Taylor’s Sources of the Self, he writes: [Taylor’s] history starts with Plato and ends with Derrida, but large tracts of time are left out: in particular, thinkers and artists of later antiquity and the Middle Ages fail to appear altogether, except for Augustine. This gap is standard for secular courses on the history of Western philosophy, and comes naturally to those of us who, however much we may be rightly told about the medieval origins of Renaissance and early modern ideas, still determinedly think of the Middle Ages as resembling a historical parenthesis. (2014c: 302) Williams believes, however, that, unlike himself, Taylor should not have omitted medieval philosophy, because the story he wants to tell is a ‘catholic’ one. The comment is revealing because it shows one of the reasons why Williams and so many others of his time, and even now, alas, consider the Middle Ages, especially where philosophy is concerned, as a ‘parenthesis’: during this period (as in late antiquity, also parenthesized by Williams) philosophy in the western tradition was closely and explicitly associated with revealed religion (Christianity, in Williams’s conception, although in fact the association with Judaism and Islam was equally important). Whereas Williams spent endless, subtle intellectual effort in considering the relationship between philosophy and one of the two areas of human cognitive experience it borders and sometimes overlaps, natural science, he usually turned away from thinking about its relationship to the other—​religion. He took seriously the idea of a religion as a practice meeting a human need (cf. Williams 2014b), but he seemed unable to take seriously the idea that the claims of an Abrahamic religion—​even the most basic claim, the existence of an omnipotent, beneficent God—​might be true. It was for him an exceptional failure of intellectual imagination, like that of a religious devotee unable to conceive the intellectual possibility that God does not exist. In part at least as a result of it, Williams was left with an improbable map of the history of western philosophy: two islands, a tiny one in early antiquity, and a large one, stretching from the mid-​seventeenth century, bigger but still dwarfed by the vast, philosophy-​less ocean lying between them. There is one reason Williams himself should have recognized why it matters that he failed the criterion of comprehensiveness. By doing so, he created a large obstacle for himself, and others who might wish to follow him, in fulfilling his own [p. 344] (later) professed aspiration, to make the familiar strange and the strange familiar. The effect of contrast Williams saw as the means to this end will be found especially in texts that treat ideas and arguments similar to those discussed now, but in contexts, and for purposes, that are totally unlike any current ones. The tradition from Plotinus to Suárez, by way of Avicenna and al-​Ghazali, Maimonides and Crescas, as well as Scotus and Ockham, is a treasury for such material.4 Williams’s change of view about the aims of history of philosophy did not, as it should have done, inspire a rethinking of which work from the past should belong to it. Williams professed to seek what was strange, but the contents he gave to the philosophical past were chosen because they would be familiar. The most important reason for concern that Williams fails the test of comprehensiveness is one he himself probably would not have acknowledged, because he stuck to the Fundamental Assumption and valued history of philosophy only for what it could bring to philosophers. As a discipline in itself, history of philosophy strives to be real history, an attempt (necessarily never completely successful) to understand the past as it was (cf. Marenbon 2018). Williams claims, in both his earlier and later views, that what he is doing is history of philosophy (and what others, who do it in a different way, are writing is mere history of ideas). Yet the limited scope he implies for this history produces a glaring falsification, both of what has taken place, as a whole, and of the causal links within and between the material included—​a betrayal of the central intellectual virtue Williams defended so powerfully: truthfulness. References Beaney, Michael. 2013. ‘Analytic Philosophy and History of Philosophy: The Development of the Idea of Rational Reconstruction’. In The Historical Turn in Analytic Philosophy. Edited by Erich Reck, 231–​60. London: Palgrave Macmillan. Curley, Edwin. 1978. Descartes Against the Skeptics. Boston, MA: Harvard University Press. Kenny, Anthony [1968] 1993. Descartes: A Study of His Philosophy, Bristol: Thoemmes Press. Marenbon, John. 2011. ‘Why Study Medieval Philosophy?’. In Warum noch Philosophie? Historische, systematische und gesellschaftliche Positionen. Edited by Marcel Van Ackeren, Theo Kobusch, and Jörn Müller, 65–​78. Berlin and Boston, MA: De Gruyter. Marenbon, John. 2018. ‘Why We Need a Real History of Philosophy’. In Philosophy and the Historical Perspective. Edited by Marcel Van Ackeren and Lee Klein, 36–​50. Perler, Dominik. 2018. ‘The Alienation Effect in the History of Philosophy’. In Philosophy and the Historical Perspective. Edited by Marcel Van Ackeren and Lee Klein, 140–​54. Queloz, Matthieu. 2017. ‘Does Philosophy Have a Vindicatory History? Bernard Williams on the History of Philosophy’. Studia Philosophica 76: 137–​51. 4 It is no surprise that the most careful working out of an approach along Williams’s lines (Perler far more detailed than Williams’s general indications, though without any direct reference to Williams himself—​uses William of Ockham, a fourteenth-​century philosopher, as its main example. [p. 345] John Marenbon Rorty, Richard. 1984. ‘The Historiography of Philosophy: Four Genres’. In Philosophy in History: Essays on the Historiography of Philosophy. Edited by Richard Rorty, J. B. Schneewind, and Quentin Skinner, 49–​75. Cambridge: Cambridge University Press. Russell, Bertrand. 1900. A Critical Exposition of the Philosophy of Leibniz: With an Appendix of Leading Passages. Cambridge: Cambridge University Press. Van Ackeren, Marcel. 2014. ‘Was bedeutet der aktuellen Philosophie ihre Geschichte? Positionen –​Probleme –​Pragmatismus’, Zeitschrift für philosophische Forschung 68, 305-​327. Van Ackeren, Marcel, ed. 2018. Philosophy and the Historical Perspective: A New Debate on an Old Topic. Oxford: Oxford University Press. (Proceedings of the British Academy 214). Williams, Bernard. 1974–​5. ‘The Truth in Relativism’. The Proceedings of the Aristotelian Society Reprinted in Williams, Moral Luck, Philosophical Papers 1983–​1980, 132–​43. Cambridge: Cambridge University Press, 1981. Williams, Bernard. 1978. Descartes: The Project of Pure Enquiry. Hassocks: Harvester Press. Williams, Bernard. 1993. Shame and Necessity. Berkeley and Los Angeles: University of California Press (Sather Classical Lectures 57). Williams, Bernard. 2002. Truth and Truthfulness: An Essay in Genealogy. Princeton: Princeton University Press. Williams, Bernard. 2006a. The Sense of the Past: Essays in the History of Philosophy. Edited by Myles Burnyeat. Princeton: Princeton University Press. Williams, Bernard. 2006b. Philosophy as a Humanistic Discipline. Edited by A. W. Moore. Princeton: Princeton University Press. Williams, Bernard. 2006c. ‘ “There Are Many Kinds of Eyes” ’. In Williams (2006a: 325–​30). Williams, Bernard. [1994] 2006d. ‘Descartes and the Historiography of Philosophy’. In Reason, Will and Sensation: Studies in Descartes’s Metaphysics. Edited by John Cottingham, 19–​27. Oxford: Oxford University Press, 1994. Reprinted in Williams (2006a: 257–​64). Williams, Bernard. [2000] 2006e. ‘Philosophy as a Humanistic Discipline’. Philosophy 75.4: 477–​ Reprinted in Williams (2006b: 180–​99). Williams, Bernard. [1993] 2006f. ‘Nietzsche’s Minimalist Moral Psychology’. European Journal of Philosophy 1: 4–​14. Reprinted in Williams (2006a: 299–​310). Williams, Bernard. [2001] 2006g. ‘Introduction’ to Friedrich Nietzsche, The Gay Science. Edited by Bernard Williams, vii–​x xii. Translated by Josefine Nauckhoff. Cambridge: Cambridge University Press. Reprinted in Williams (2006a: 311–​24). Williams, Bernard. [2004] 2006h. ‘Unerträgliches Leiden’. In Zum Glück. Edited by Susan Neiman and Matthias Kroß, 95–​102. Berlin: Akademie Verlag. Translated as ‘Unbearable Suffering’ in Williams (2006a: 331–​7). Williams, Bernard. [1981] 2006i. ‘The Legacy of Greek Philosophy’. In The Legacy of Greece: A New Appraisal. Edited by Moses Finley, 202–​55. Oxford: Oxford University Press. Reprinted in Williams (2006a: 3–​48). Williams, Bernard. [1988] 2006j. Plato: The Invention of Philosophy. London: Phoenix. Reprinted in Williams (2006a: 148–​86). Williams, Bernard. 2014a. Essays and Reviews, 1959–​2002. Princeton and Oxford: Princeton University Press. Williams Bernard. [1983] 2014b. Review of The Miracle of Theism: Arguments for and against the Existence of God, by J. L. Mackie. In Williams (2014a: 197–​200). Williams [1990] 2014c. Review of Sources of the Self: The Making of the Modern Identity, by Charles Taylor. In Williams (2014a: 301–​11). Williams, Bernard. [1995] 2014d. Review of The Limits of Interpretation, Interpretation and Overinterpretation, Six Walks in the Fictional Woods, Apocalypse Postponed, Misreadings, How to Travel with a Salmon & Other Essays, by Umberto Eco. In Williams (2014: 352–​63). Williams, Bernard [1996] 2014e. ‘On Hating and Despising Philosophy’. In Williams Williams, Bernard [2002] 2014f. ‘Why Philosophy Needs History’. In Williams (2014a: 405–​12). Wilson, Margaret Dauler. 1978. Descartes. Routledge: London and New York. [p. 346] The Iniquity of Oblivion Ralph Wedgwood Throughout his whole career, Bernard Williams was intensely concerned with the history of philosophy. Not only did he write seminal works on several philosophers of the past—​most notably, on Plato, Aristotle, and Descartes—​but he also reflected deeply on the methodology of his approach to these past philosophers. In particular, starting in his 1978 book on Descartes, he firmly distinguished the kind of inquiry into these past philosophers that he was most interested in, which he labelled the history of philosophy, from another closely related inquiry, which he called the history of ideas. What is the point of the history of philosophy, of the kind that Williams was most interested in? It is only in the last ten years of his life that Williams explicitly set out to answer this question. We find one answer expressed in 1994 and also in his posthumously published essay on Collingwood. In addition, there is also a second, rather different answer to the question, which may be suggested by several works that Williams completed in the last three years of his life (2000–​3). This raises the issue of which of these two answers provides a deeper insight into this field of study. In section 1 of this chapter, I shall explain and defend Williams’s distinction between the history of philosophy and the history of ideas. In section 2, I shall explain two answers that might be given to the question about the point of studying the history of philosophy; and in section 3, I shall argue that these two answers in fact lead to two very different approaches to the subject. Finally, in sections 4 and 5, I shall attempt to evaluate these two answers. In section 4, I shall argue that the first answer is importantly true, and the approach to the history of philosophy to which it leads is an indispensable part of the discipline of philosophy. But in section 5, I shall seek to raise sceptical doubts about the second answer, and about how important or illuminating the approach to the history of philosophy to which it leads is likely to be. 1. History of Philosophy versus History of Ideas Williams first introduced his distinction between the history of philosophy and the history of ideas in the Preface to his book Descartes: The Project of Pure Enquiry The crucial distinction that he marks with these labels is this: Ralph Wedgwood, The Iniquity of Oblivion In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. [p. 347] Ralph Wedgwood [T]‌he history of ideas is history before it is philosophy, while with the history of philosophy it is the other way round. In any worthwhile work of either sort, both concerns are likely to be represented, but there is a genuine distinction. Writing in Cambridge in the 1970s, Williams is likely to have thought of his colleague Quentin Skinner—​who was working on his book The Foundations of Modern Political Thought (1978) around the same time—​as a paradigmatic exponent of the history of ideas. Today, this kind of study is often referred to as intellectual history. Williams (1978: 9) gives the following brief characterization of the history of ideas: For the history of ideas, the question about a work what does it mean? is centrally the question what did it mean?, and the pursuit of that question moves horizontally in time from the work, as well as backwards, to establish the expectations, conventions, familiarities, in terms of which the author could have succeeded in conveying a meaning. This enterprise itself cannot be uncorrupted by hindsight. . . . Yet what we are moved to, as historians of ideas, is an historical enquiry, and the genre of the resulting work is unequivocally history. He repeats this characterization more briefly in his article ‘Descartes and the Historiography of Philosophy’ (1994: 257): The history of ideas, as I intended the distinction, naturally looks sideways to the context of a philosopher’s ideas, in order to realize what their author might be doing in making those assertions in that situation. We may assume, I think, that in this characterization Williams is presupposing a mainstream conception of history, according to which the historian’s goal is to use the historical evidence that has survived from the past to construct an accurate and intelligible account of what happened in the past. Very often, such a historical account involves a narrative of past events. At all events, the goal is to construct a historical account that gives an accurate description of events that in fact happened in the past, and also helps to makes it intelligible why these events happened as they did. Among the events that happened in the past are the thoughts that philosophers had, and these philosophers’ acts of writing and engaging in discussions, along with the meanings that they conveyed in performing those acts. A historian of ideas who studies a group of philosophers would aim to give an accurate and intelligible account—​often in the form of a narrative—​of these events in the intellectual lives of these philosophers. The primary evidence that the historian of ideas [p. 348] would have to use in developing such an account consists of the texts that have come down to us. To use these texts to figure out what philosophers thought and what meanings these philosophers conveyed, the historian of ideas needs, as Williams says, to ‘establish the expectations, conventions, [and] familiarities’ of the relevant intellectual community of the time. In general, to help to make it intelligible why these philosophers had these thoughts and conveyed these meanings, we need to see these philosophers as responding to their historical context—​where this context consists of these philosophers’ contemporaries and predecessors, at least to the extent that these philosophers were aware of these contemporaries and predecessors. This is presumably why Williams says that the history of ideas ‘looks sideways to the context’ and ‘moves horizontally in time . . ., as well as backwards’ from the texts that constitute the primary evidence. By contrast, Williams (1978: 9–​10) characterizes the ‘history of philosophy’ in the following terms: The history of philosophy of course has to constitute its object, the work, in genuinely historical terms, yet there is a cut-​off point, where authenticity is replaced as the objective by the aim of articulating philosophical ideas. The ‘horizontal’ search for what Descartes meant will, if it is properly done, yield an object essentially ambiguous, incomplete, imperfectly determined by the author’s and his contemporaries’ understanding, for that is what the work—​at least if it is now of any autonomous interest at all—​cannot fail to have been. The present study, while I hope that it is not unaware that this is so, prefers the direction of rational reconstruction of Descartes’s thought, where the rationality of the construction is essentially and undisguisedly conceived in a contemporary style. . . . [T]‌he new work is broadly of the same genre as the original. As he puts it in the later article (1994: 257): [T]‌he product of the history of philosophy, being in the first place philosophy, admits more systematic regimentation of the thought under discussion. The main point here is that the history of philosophy has the goal of ‘articulating philosophical ideas’ or offering a ‘rational reconstruction’ or ‘systematic regimentation’ of the thought of past philosophers. (We shall return to the question of what such a ‘rational reconstruction’ amounts to below.) This seems to be what explains why the history of philosophy ‘is broadly of the same genre as the original’, or, in other words, why it is philosophy ‘before’ it is history. [p. 349] Ralph Wedgwood A subsidiary point that Williams (1978: 10) makes is that the sense of what is interesting in the ideas of the past that guides the historian of philosophy is itself unequivocally a product of the historian of philosophy’s own time: This study is meant to consist, to a considerable extent, of philosophical argument, the direction of it shaped by what I take to be, now, the most interesting philosophical concerns of Descartes. The argument is in twentieth-​century terms; the judgement of interest is a twentieth-​century judgement; it is absolutely certain that a work which was primarily historical would represent Descartes’s concerns in a different way. In the later article, Williams (1994: 257) makes a similar point about how the history of philosophy is often concerned to relate the ideas of the past to the problems that are of most interest in the present: The history of philosophy . . . is more concerned to relate a philosopher’s conception to present problems, and is likely to look at his influence on the course of philosophy from his time to the present. In a number of ways, this characterization of the history of philosophy is puzzling. Isn’t the historian of philosophy’s goal in working on (say) Descartes just to figure out what Descartes thought? How could such an investigation belong to any genre other than history? In fact, I believe that a similar distinction can be drawn in other fields besides philosophy. In principle, for example, a similar distinction can be drawn with the scholarly study of literature. One kind of study of literature is pursued strictly as just another branch of history.1 This strictly historical study of literature could be called, in a certain sense, literary history. The goal of this sort of literary history, like any other kind of history, is just to give an accurate and intelligible account of some of the events of the past—​specifically, events in which literature was created, published, performed, read, discussed, and so on. This is parallel to the way in which what Williams called ‘the history of ideas’ aims to give an accurate and intelligible account of events in the intellectual lives of thinkers. Just as the history of ideas treats philosophical texts as pieces of historical evidence for constructing an accurate and intelligible account of past events in the world’s intellectual life, this kind of literary history uses literary texts as evidence for constructing an accurate and intelligible account of past events in the world’s literary life. Crucially, however, to treat a literary text as a mere piece of historical evidence is not to respond to it in a way that reflects its distinctive character as a work of 1 For an example of this kind of literary history, see Miller (2001). [p. 350] literature. Not all historical evidence consists of texts. Besides texts, historical evidence can also take the form of eyewitness reports, films, and photographs, as well as forensic data and all the evidence of archaeology. Historians qua historians make use of all of these kinds of historical evidence in fundamentally the same way. Qua historians, they have no reason to treat literary texts as special; for historians, literary texts are just more pieces of historical evidence. This is not the primary way in which literary texts are designed to be treated, as mere historical evidence, of the kind that we can get equally well from photographs or archaeological data. The comedies of the Roman playwright Terence were not composed for the purpose of providing historians with historical evidence for understanding events that occurred during the Roman Republic. They were composed in order to evoke a distinctively dramatic and comedic response from their audience. In general, literary works are designed to evoke a distinctive sort of response—​although the exact sort of response that the work is designed to evoke varies from one literary genre to another. It is for this reason that there is room for a different kind of literary scholarship, which is not merely a branch of history. What is of interest here is not literary theory, which is more focused on developing general theories than on studying particular literary works.2 What is of interest for our purposes is the kind of literary scholarship that studies particular works, such as the Odes of Horace, or Goethe’s Faust—​just as Williams’s work in the history of philosophy involved a close study of particular works such as Plato’s Republic and Descartes’ Meditations. I shall refer to the sort of study of particular literary works that is of interest here as literary criticism.3 Literary criticism of this sort does not primarily aim, like history, at producing an accurate and intelligible account of past events; instead, its goal is to respond to literary works with something like the distinctive kind of response that those works were originally designed to evoke from their audiences. Admittedly, most literary works were composed for a broader audience than just literary critics. So the literary critic’s response is undeniably special, to the extent that it takes the form of a publicly accessible work of scholarship. Nonetheless, the goal is to respond to literary works with an interpretation that has a fundamental kinship with the kind of response that those works were originally designed to evoke. Such scholarship can certainly use history, as a means to an end; a knowledge of the work’s original historical context can inform and enrich the critic’s interpretation. However, to put it crudely, the literary critic who studies Terence will typically want to laugh at the jokes, while the historian who uses Terence as evidence for understanding Roman literary culture in the second century bce has no need, qua historian, to find these comedies funny at all. 2 For an influential introduction to literary theory, see Eagleton (1996). 3 Another parallel to this distinction between literary history and literary criticism might be the distinction between music history and music analysis. For examples of two works on music that seem to fall on either side of this distinction, see Magee (2001) and Clark (2011). [p. 351] Ralph Wedgwood In a similar way, the goal of ‘the history of philosophy’, as Williams understands it, is to make a distinctively philosophical response to philosophical work. Here, however, we find a crucial difference between literary and philosophical works. It is not usually expected that the audience of a literary work itself consists of authors or other literary creators. With philosophical works, by contrast, it is expected that the audience of the philosophical work itself consists of philosophers—​in the broad sense that includes amateur philosophers and students of philosophy, as well as professional philosophers. In this way, the response that a philosophical work typically seeks to evoke from its audience consists of philosophical thinking, of essentially the same kind as the thinking that is conveyed in the work itself. Thus, Williams’s goal in writing on the works of Plato and Descartes was to respond in the very way that those works call for, with the kind of philosophical thinking that Plato and Descartes aimed, in creating their works, to provoke in their readers. This is why he described his own book on Descartes’ work as ‘of broadly the same genre as the original’.4 If we succeed in understanding a philosophical work, and in responding to it with the very kind of philosophical thinking that it was originally designed to evoke, there is a sense in which we are having a kind of dialogue with philosophers of the past. Almost every surviving work of philosophy is in effect a communication from its author to its readers.5 In general, to respond to a communication with the sort of response that the communication was itself designed to evoke in its audience is a kind of dialogue. Thus, on my interpretation, the goal of the sort of history of philosophy that Williams was interested in is to create this kind of dialogue between contemporary readers and the relevant philosophers of the past. Some readers might doubt this interpretation of Williams. At a number of points, he can be read as criticizing the view that we can realistically expect any kind of dialogue with the philosophers of the past. For example, Williams (2006a: 344) says in his essay on Collingwood:6 4 Strictly speaking, Williams’s point that studies in the history of philosophy belong to the same genre as the original works that they discuss needs some qualification. Notably, some of Williams’s own work sets out to explore the ethical or philosophical ideas that are conveyed in texts that belong to non-​ philosophical genres. Specifically, in several works—​especially in Shame and Necessity (1993)—​he sets out to articulate the ethical ideas that he finds in Homer, the three Attic tragedians, and Thucydides. We can read Williams as assuming that philosophical ideas can appear in non-​philosophical works—​and so the history of philosophy can aim to articulate and study these philosophical ideas wherever they appear. 5 The exceptions to this are the private notes that some philosophers have left behind (such as Kant’s Reflexionen). Strictly speaking, interpreting and responding to such private notes is not the kind of ‘dialogue’ that I am describing here; this is because those private notes were not essentially communicative—​they were neither originally addressed to any readers nor designed to evoke any response in readers. 6 Williams (2000: 181) also discusses this remark of Grice’s in ‘Philosophy as a Humanistic Discipline’. [p. 352] Peter Strawson, talking about Paul Grice, quoted with approval a remark of Grice’s, which he claimed all Oxford philosophers would agree with, that we ‘should treat great but dead philosophers as we treat great and living philosophers, as having something to say to us’. Collingwood would have agreed with this, so long as it is not assumed that what the dead have to say to us is the same sort of thing as the living have to say to us. He would not have agreed, that is to say, with Ryle’s frequent injunction to treat something written by Plato, for instance, as though it had come out in Mind last month. Collingwood insisted, correctly, that the questions being answered by Plato and Hobbes, for instance, were not the same, and that you literally could not understand them unless you understood this. When read carefully, however, this passage does not deny that dead philosophers like Plato and Hobbes have something to say to us. This passage just insists that what they have to say to us is, in many instances, different from what any of our contemporaries have to say to us. Since they have something to say to us, it is possible for us to pay attention to what they said, and to respond appropriately: that is, with the kind of response that they, in saying what they said, aimed to evoke from their audience. In other words, it is possible in my sense to have a dialogue with these dead philosophers. In his sketch of the introduction that Williams (2006a: ix–​x) had planned for The Sense of the Past, A. W. Moore suggests that Williams conceived of his contribution in the following way: This contribution was not, as philosophers in the analytic tradition used to think, to indicate voices of yore which could be heard as participating in contemporary debates: precisely not. It was to indicate voices of yore which could not be heard as participating in contemporary debates, and which thereby called into question whatever assumptions made contemporary debates possible. The sense in which these voices from the past cannot ‘be heard as participating in contemporary debates’ is that many of the questions that they are aiming to answer, and many of the assumptions that they are presupposing, differ from those that are familiar in contemporary debates. But this does not prevent us from understanding what these philosophers were saying, nor from responding appropriately with the kind of philosophical thinking that those philosophers aimed to inspire in their readers. Moreover, as I shall argue in section 4 below, there is good chance that our contemporary debates may be changed by this kind of dialogue with the philosophers of the past. In this sense, through our understanding and responding to them, philosophers from the past can still contribute to our contemporary debates. Admittedly, if we achieve this sort of dialogue—​say, with Hume, or with Plotinus—​it will be a one-​sided dialogue, in which we can listen to and respond to [p. 353] Ralph Wedgwood what they say to us, but they cannot hear us or respond to our responses to them. It is exactly this kind of dialogue that the essays in this volume are attempting to have with Williams himself—​a one-​sided dialogue, since he too has sadly now been dead for more than twenty years. Williams (1998: 150–​2) himself sympathized with the concerns about the limitations of such one-​sided dialogues that are expressed in Plato’s Phaedrus (275c–​277a). Nonetheless, despite these concerns, both Plato and Williams were prolific writers. In writing their works, they presumably hoped to be read, not only by readers with whom they could have a mutual two-​sided dialogue, but also by readers whom they themselves could never respond to. Thus, both Plato and Williams seem to have seen some value in this kind of one-​sided dialogue with readers whom they would never know. My hypothesis, then, is that a central goal of the kind of history of philosophy that Williams was concerned with is to facilitate this kind of philosophical dialogue between contemporary readers and philosophers of the past.7 If this is the goal of this kind of history of philosophy, we can understand a number of the features that Williams ascribes to it. First, this idea that the goal of the history of philosophy is to facilitate a philosophical dialogue with philosophers of the past can also explain why the history of philosophy needs to make use, to a significant extent, of the historical methods of the history of ideas. It is also impossible to have a serious dialogue with anyone unless one makes a serious effort to understand them and the meanings that they are actually seeking to convey. Secondly, we can understand the importance of giving a ‘rational reconstruction’ of these philosophers’ ideas. Whenever anyone reads any philosophical text—​ whether it is centuries old or written only a year ago—​if the reader engages in a serious dialogue with that text, the reader has to produce a ‘reconstruction’ in their own mind of the philosophical ideas that are conveyed by that text. Crucially, this reconstruction needs to include the fine logical structure of the text’s arguments—​ where sometimes, if the text’s meaning is indeterminate, it is necessary for the 7 Marcel van Ackeren (2018a: 70–​4) criticizes what he calls the ‘dialogue model’ of our relationship with philosophers of the past, equating this model with the attempt to use the philosophers of the past as a ‘quarry’ from which useful insights can be ‘mined’. First, he suggests that most of those who advocate this model for the history of philosophy assume a ‘transhistorical identity of [philosophical] problems’ a: 70). As I make clear in section 4 below, this assumption is no part of what I have in mind by speaking of our having a dialogue with dead philosophers. Secondly, he insists that, in the strict sense of the word, in any genuine dialogue both parties respond reciprocally to each other; since philosophers of the past cannot respond to us, ‘A ll we can do is read their texts and interpret them’ (2018a: 71). So, he concludes, although we can use the word ‘dialogue’ in a looser sense, all that this use of the term means is just our reading and interpreting old texts. However, this overlooks the conception of a ‘dialogue’ that I am employing here: on my conception, a dialogue involves not only attending to and understanding a communication from someone else, but also responding to the communication with precisely the sort of response that that communication was itself designed to evoke. This goes beyond simply giving an ‘interpretation’ of evidence from the past: a palaeontologist might give an ‘interpretation’ of the fossil data from millions of years ago, but this is not in my sense a ‘dialogue’ with the fossil data—​since, unlike philosophical texts, fossils were not intentionally created with the aim of evoking a response from an audience. [p. 354] reader to consider a more precise logical structure than any that is unambiguously given by this indeterminate meaning. I propose that this is how we should interpret the kind of ‘rational reconstruction’ or ‘systematic regimentation’ that Williams refers to. Finally, if we engage in a serious dialogue with any interlocutor, while we will certainly seek to understand what strikes our interlocutor as interesting and important, we will also try to learn from them, especially about what seems most interesting and important to us. This is why Williams (1978: 10) focuses on what he takes ‘to be, now, the most interesting philosophical concerns of Descartes’, emphasizing that ‘the judgement of interest is a twentieth-​century judgement’. Thus, if we are to have a serious philosophical dialogue with (say) Plato, we will have to ‘relate [Plato’s] conception to present problems’: we can hardly have a serious dialogue with anyone if we do not relate what they are saying to issues that we care about. We can probably also understand Williams’s (1994: 257) suggestion that the history of philosophy is likely to explore Plato’s ‘influence on the course of philosophy from his time to the present’ in the following way: if we are to have a serious dialogue with Plato, we need to figure out what Plato means for us—​and one aspect of what he means for us is the influence that he has exerted over the tradition of philosophy that we have inherited.8 In general, the history of philosophy, of the kind that Williams pursued in his historical work, is certainly not a new field. It is broadly continuous with the long tradition of philosophical commentaries on the works of earlier philosophers. A particularly notable case consists of the long tradition of commentaries on Aristotle.9 These commentators include Greeks and Romans of late antiquity such as Alexander of Aphrodisias, Themistius, Olympiodorus, Simplicius of Cilicia, and Boethius; it also includes early medieval Arabic scholars such as al-​ Farabi, Avicenna, and Averroes, and later medieval philosophers writing in Latin such as Thomas Aquinas, Duns Scotus, and William of Ockham. There is another 8 This point sheds light on the relationship between Williams’s distinction and the debate that Daniel Garber (2013) had with Jonathan Bennett about the methodology of the history of philosophy. According to Garber (2013: 353), historians of philosophy should aim for a ‘disinterested historical reconstruction’ of the views and arguments of past philosophers—​where this historical reconstruction crucially needs to rely on a grasp of these past philosophers’ historical context. Indeed, in his view, the very meaning of the problems that philosophers seek to address is determined by this context. In consequence, as Garber (2005: 138) puts it, ‘it was not the same challenge for them as it is for us’. For this reason, it is a mistake to seek, as Williams (1994: 257) does, to ‘relate a [past] philosopher’s conception to present problems’. Thus, in Williams’s terminology, Garber’s work is intermediate between the history of philosophy and the history of ideas. Like the history of philosophy, it aims at ‘rational reconstruction’, in terms that are comprehensible to contemporary readers; but unlike Williams’s version of the history of philosophy, it resists the inclination to be inspired by these historical texts to think independently, as a contemporary philosopher, about the present problems that are most closely related to the texts. 9 The original texts of the Greek commentators on Aristotle are available in Diels et al. (1882–​1909). For discussion of commentaries in Arabic, see Druart (2020), Ben Ahmed and Pasnau (2021), and Gutas (2016). For discussion of the commentaries in Latin, see Marenbon (2021) on Boethius and Marenbon (2023) on the high Middle Ages. [p. 355] Ralph Wedgwood notable case in China, with the immensely influential commentaries that the neo-​ Confucian philosopher Zhu Xi wrote in the Song Dynasty on the Four Great Books of the Confucian tradition.10 In writing their commentaries, these scholars were undoubtedly engaging in philosophy, aiming at ‘articulating’ the ‘philosophical ideas’ of these older texts by providing a ‘rational reconstruction’ of them. In all these cases, these ancient and medieval commentators sought a rational reconstruction of these philosophical ideas because they viewed the ideas as having a unique kind of authority. The neo-​Confucians ascribed a unique authority to the Four Great Books of ancient Confucianism; and the ancient and medieval commentators on Aristotle believed that Aristotle represented the pinnacle of what natural reason could be expected to achieve. These scholars wrote their commentaries because they assumed that, when correctly interpreted, the works that they were commenting on would reveal the truth about philosophy. From a contemporary historical perspective, this assumption seems entirely indefensible. For example, we now know that Aristotle’s logic—​which was regarded by western philosophers as the one true logic for the whole period from late antiquity until the mid-​nineteenth century—​is in fact just a distorted fragment of the first-​order predicate calculus. We also know that Aristotle’s natural science is riddled with errors, from the spontaneous generation of insect larvae to his geocentric conception of the universe.11 In general, the insights of contemporary historians make it inconceivable that any philosopher of the past is an authority of that kind. It is clear that Williams himself did not approach the history of philosophy in anything like that spirit. Indeed, some of his most striking contributions in this field were on the works of philosophers with whom he himself had profound disagreements, such as Plato and Descartes. This raises the awkward question: what is the point of studying the history of philosophy? Some recent writers—​for example, Hanno Sauer (2022)–​have argued that the history of philosophy in fact has no point.12 According to this argument, the philosophers of the past are so unreliable that we cannot expect to learn anything worthwhile from them. Just as contemporary students of medicine do not read the works of Hippocrates or Galen, and contemporary students of astronomy do not read the works of Ptolemy, so there is no reason for contemporary students of philosophy to waste their time on the works of Plato, Aristotle, or Descartes. This is the question to which I shall turn for the rest of this chapter. What is the point of studying the history of philosophy? What is the point of seeking a philosophical dialogue with long-​dead philosophers who lived centuries ago? We shall 10 On Zhu Xi, see Thompson (2021). 11 For discussions of Aristotle’s contributions in logical theory and natural science, see Lear (1980) and Althoff (2018), respectively. 12 Sorell (2005: 44) quotes from correspondence with Gilbert Harman, in which Harman states: ‘I also think . . . that a study of the history of philosophy tends not to be useful to students of philosophy.’ [p. 356] explore two quite different answers to this question, both of which are suggested by Williams’s work. 2. What Is the Point of the History of Philosophy? In the previous section, I gave an interpretation of Williams’s contrast between the ‘history of ideas’ and the ‘history of philosophy’, and raised the question about what point there could be in the history of philosophy (on this interpretation of what it amounts to). One way to answer this question would be by arguing that this sort of history of philosophy is in some way an important component of the discipline of philosophy. Before taking up the question of whether the history of philosophy is an important component of the discipline of philosophy, however, I should like to suggest that, in fact, this sort of history of philosophy is also valuable as a contribution to history—​which I shall assume here to be an intrinsically valuable study in its own right.13 Even though the ultimate goal of this sort of history of philosophy, according to my interpretation, is to put contemporary readers into a dialogue with philosophers of the past, this may in fact be one of the best ways to discover certain truths about the past. Specifically, historians of ideas, with their wide-​ranging explorations of archival materials and of the historical context in which past philosophers were writing, may not always be best placed to uncover the fine-​grained logical structure of these past philosophers’ thought. It may sometimes be that the best way to give contemporary readers a clear and accurate account of what these past philosophers actually thought is to aim for a ‘rational reconstruction’ or a ‘systematic regimentation’ of their thought. Some historical truths are best revealed by a close analysis of a small selection of carefully chosen texts, rather than by a grand synthesis of a huge array of historical evidence.14 In what follows, however, the purely historical value of the sort of history of philosophy that Williams was interested in will be set aside. Instead, we shall focus on the value of the history of philosophy for the rest of philosophy. Williams’s first detailed discussion of the point of the history of philosophy is in his 1994 article ‘Descartes and the Historiography of Philosophy’ (1994: 258–​9): The point of any history, one might suppose, is to achieve some distance from the present, which can help one to understand the present. . . . 13 The idea that understanding the history of thought is intrinsically valuable is powerfully defended by Marenbon (2018). 14 I owe this point to Harvey Lederman. [p. 357] Ralph Wedgwood To justify its existence, [the history of philosophy] must maintain a historical distance from the present, and it must do this in terms that sustain its identity as philosophy. It is just to this extent that it can indeed be useful, because it is just to this extent that it can help us to deploy ideas of the past in order to understand our own. We can adapt to the history of philosophy a remark that Nietzsche made about classical philology: ‘I cannot imagine what [its] meaning would be in our own age, if it is not to be untimely—​that is, to act against the age, and by so doing to have an effect on the age, and, let us hope, to the benefit of a future age.’15 One way in which the history of philosophy can help to serve this purpose is the basic and familiar one of making the familiar seem strange, and conversely, but it needs to learn how better to do this. Williams (1994: 260) claims that this role—​‘reviving a sense of strangeness or questionability about our own philosophical assumptions’—​cannot be played so effectively by the history of ideas: [A]‌pplied to Descartes, . . . [t]he history of ideas quite properly invites us to learn about late scholastic influences and the syllabus at La Flèche, or introduces us to problems that were encountered in developing an adequate mechanics of inertia. . . . [This] activit[y], the history of ideas, certainly has nothing wrong with it, but, in itself, it does not yield much philosophy that can help us in reviving a sense of strangeness or questionability about our own philosophical assumptions. It may be, simply and quietly, what it seeks to be, about the past. However, if it is done properly, the history of philosophy can play this role What was called in the original distinction ‘the history of philosophy’ is essential to any activity that is going to give a philosophical point to writing historically about philosophy. That point is going principally to be found in the possibility of the past philosophy’s being untimely, and helping to make strange what is familiar in our own assumptions. As Williams emphasizes in his conclusion to this article, to do this effectively the history of philosophy often also requires us to question the ‘received picture’ of our philosophical tradition (1994: 264): What we must do is to use the philosophical materials that we now have to hand, together with historical understanding, in order to find in, or make from, the 15 The quotation is from Nietzsche (1874: 88). [p. 358] philosophy of the past a philosophical structure that will be strange enough to help us to question our present situation and the received picture of the tradition, including those materials themselves. Williams (2006a: 344) makes the same point in his essay on Collingwood: ‘the point of reading philosophers of the past is to find in them something different from the present—​and that is not just a historical but a philosophical discovery’. While these remarks are suggestive, they raise at least as many questions as they answer. What is the point of making our familiar philosophical assumptions seem ‘strange’ or ‘questionable’, or of making philosophical assumptions that initially seemed strange become more ‘familiar’ to us? Students of medicine do not find it useful to make their assumptions about the circulation of blood seem ‘strange’ by familiarizing themselves with Galen’s idea that human blood ebbs and flows like the tides. Why should it be useful for philosophy to make our assumptions seem ‘strange’? I shall attempt to answer this question in section 4 below. Before doing that, however, I shall turn to a second and rather different collection of views that Williams developed late in his life about why philosophers need to study the history of thought. Some of these views appear in his book Truth and Truthfulness (2002a), but they receive their clearest exposition in two essays that were written around the same time, ‘Philosophy as a Humanistic Discipline’ and ‘Why Philosophy Needs History’ (2002b). The central idea that Williams (2002b: 405–​6) defends in these essays can be summed up simply as follows: a central goal of philosophy is to help us to understand ourselves; to understand ourselves, we need to understand our concepts; and to understand our concepts, we need to understand the history of our concepts. So, philosophy needs to incorporate a historical study of our concepts. When Williams (2002b: 406) insists that a central goal of philosophy is to ‘understand ourselves’, whom does he mean by ‘us’? He gives a clear blunt answer to this question: Who ‘we’ are, who else is part of ‘us’, may very well be disputed, above all in ethical and political cases. But reflection must start with us in the narrowest sense—​the people who are asking the question and the people to whom we are talking—​and it starts from now. The concepts that give rise to the question are ours. It seems, then, that the word ‘we’ here refers to Williams and his readers. What ‘we’ need to understand, he says, is the ‘story behind these concepts: how people have come to think like this’. As he puts it, focusing particularly on our concepts of accuracy and sincerity [p. 359] Ralph Wedgwood These questions arise from our present ideas of such qualities or ideals as honesty, truthfulness, sincerity and realism. They are appropriate to philosophy, in that they involve a recognition that we do not adequately understand ourselves. It is obvious that our ways of conceiving these qualities have not been everybody’s, and that there is a historical story to be told about the way they came to be ours. Can we understand these concepts and so face the problems that they generate for us without understanding something of that story? How is it, for example, that we have a special ideal of personal authenticity? I think that philosophy can get a real hold on its task only with the help of history; or, rather, as Nietzsche put it, philosophising in such a case must itself be historical. This historical account of how certain ‘ways of conceiving’ these ‘qualities or ideals’ ‘came to be ours’ forms the second half of a larger project, which Williams a: ch. 3; 2002b: 409) calls a ‘genealogy’ of our concepts. At least in the case of some of our concepts, the first half of this project does not consist of ‘real history’ based on particular bits of historical evidence from the past: instead, it consists of a more abstract argument about how, in some form or other, concepts of this kind would have had to develop in any human society. In Truth and Truthfulness, Williams (2002a: ch. 3) develops this first component of his genealogy by means of the philosophical thought experiment of the ‘State of Nature’. But the whole genealogical project also includes a second component which Williams (2002a: 149) claims to consist of ‘real history’: for example, Williams (2002a: 151–​5) argues that Thucydides, writing in the late fifth century bce, helped to create a new way of understanding historical truth. How can this sort of ‘genealogy’ of our concepts help ‘us’ to ‘understand ourselves’? Williams seems to think that there are three main possibilities here. One ‘possibility’ for a genealogical account of how we came to have a certain value is ‘that the value . . . may understand itself and present itself and claim authority for itself in terms which the genealogical story can undermine’ (Williams b: 410). As Williams observes, Nietzsche thought that this is the case with the ‘morality’ that his historical account depicted as emerging from the ‘ressentiment’ of the ‘slave revolt in morality’.16 However, there is also, Williams thinks, a second possibility. In some cases, these conceptions might be ‘vindicated’ by this genealogical account. In particular, he thinks that some of our ethical concepts of the values of truth and truthfulness will be in a way vindicated by this account.17 16 See Nietzsche (1887: Essay 2). Williams (2006a: Essay 23) discusses this aspect of Nietzsche’s thought in his posthumously published essay, ‘Unbearable Suffering’; for the general idea of an undermining genealogy, see Williams (2000: 190–​4). 17 Glock (2008: 880) argues that, in fact, the vindicatory result of Williams’s genealogy of truth and truthfulness is entirely borne by the part of this genealogy that does not consist of ‘real history’. Even if Glock is right about this, Williams clearly wants to allow that, in some cases, the genuinely historical component of the genealogy may also contribute to the vindication. [p. 360] A third possibility that Williams considers, it seems, is that our present conceptions might be neither vindicated nor undermined by this historical account, but the account might nonetheless reveal these conceptions to be ones that we simply cannot give up—​they are conceptions that form part of who we are, so that living without these conceptions is simply not a live option for us.18 In this way, towards the end of his life, Williams became convinced of the need for philosophy to engage seriously with history, in order to understand the history of our concepts, and thereby to evaluate those concepts, by either undermining or vindicating them, or at least by revealing them to be ones that we simply cannot give up. If Williams is right that philosophy needs to engage with history in this way, it might be suggested that one potential benefit of studying the history of philosophy is that it can help philosophy to engage with history in the required way. In fact, Williams himself never explicitly endorsed this suggestion. But the suggestion seems to me worth exploring. For this reason, in the rest of this chapter, I shall explore the following two different answers to the question of how the history of philosophy is important to philosophy. According to the first answer, the history of philosophy is important to philosophy because it can ‘make the familiar seem strange’; according to the second answer, the history of philosophy is important because it can form part of a ‘genealogy’ of our concepts. 3. Two Different Approaches to the History of Philosophy As I shall argue in this section, these two answers to our question about the point of studying the history of philosophy lead to very different approaches towards the subject. If our goal is to find philosophical ideas that initially seem strange to us, and to make them seem less strange and more intelligible, we should presumably look far beyond the canon of western philosophical classics. For example, we should certainly look at non-​western philosophy. In this spirit, it would be valuable to study the works of ancient Chinese philosophers: these would include the ancient Confucian tradition (as expounded in such texts as the Analects and Mengzi), as well as the Mohist and Daoist traditions (including Zhuangzi). It would also be valuable to investigate the neo-​Confucian thinkers of the Song dynasty (such as Zhu Xi) and the Ming dynasty (such as Wang Yangming). It would similarly be important to explore the rich traditions of Indian philosophy, including the various schools of Hindu philosophy (such as the Nyāya and Vaiśeṣika schools) and of 18 That is, these conceptions might be, as Williams (2000: 195) puts it, unhintergehbar. [p. 361] Ralph Wedgwood Indian Buddhist philosophy (such as the Mādhyamaka and Dignāga-​Dharmakīrti traditions). When studying western philosophy, it would be particularly valuable to investigate thinkers whose works have long been ignored. Instead of studying nothing but Plato, Aristotle, Descartes, Kant, and Hume all the time, we could look at the three Hellenistic schools—​the Epicureans, Stoics, and Sceptics—​whose works were lost and largely ignored between the fourth and the fifteenth centuries; in the early modern period, we could study the lesser-​known philosophers, such as Samuel Pufendorf, Ralph Cudworth, and Richard Price, as well as the women philosophers whose work was rarely taken seriously until recently, such as Anne Conway, Margaret Cavendish, Émilie du Châtelet, and Mary Wollstonecraft. By contrast, if our goal is to understand our concepts, we should only look at the parts of history that can plausibly be claimed to have played a significant role in leading to our now having these concepts. Assuming that ‘we’ here refers to Williams and the readers whom he had in mind, ‘we’ in this sense are all westerners. So, non-​western philosophy played little or no role in leading to ‘our’ concepts. Equally, long-​forgotten philosophical works presumably also played a negligible part in this historical story. Moreover, it is not clear that much of this historical story need consist of the history of philosophy.19 Many other developments—​for example, in public political discourse, or in scientific, religious, legal, and political thought—​are at least as important as the work of philosophers, if not more so. Philosophers, after all, have always been a small group, and for most of human history they have had limited influence on the culture in which they lived. If our approach to the history of philosophy is guided by the first answer to the question about what the point of the subject is, we should go on a quest to rescue strange and interesting philosophical ideas from the risk of being forgotten. Moreover, for reasons that I shall explain in the next section, we need to focus most intensely on the work of philosophers—​as it were, our colleagues from centuries past. By contrast, if our approach is guided by the second answer, ideas that have long been forgotten are unlikely to have played a large role in the development of ‘our concepts’; if our goal is to understand this development, we may as well leave these ideas to remain forgotten—​regardless of their intrinsic merits. We also have little reason to focus our attention on works of philosophy: a more general investigation of the history of human thought is what is principally required. In this way, these two different answers to our question about the point of the history of philosophy lead to very different approaches to the subject. One response, of course, would just be to say that we should let a thousand flowers bloom. Some scholarly investigations should take the first approach, and some others, the 19 This is a point that Williams himself recognized; see Williams (2002b: 405–​6; 2006a: x). [p. 362] second. In the rest of this chapter, however, I shall take a more invidious line. In the next section, I shall argue that the kind of history of philosophy that follows the first of these approaches is of central importance to our field. However, in section 5, I shall raise some sceptical doubts about the kind of history of philosophy that follows the second of these two approaches. 4. Salvaging Philosophical Ideas from Unmerited Oblivion In this section, I shall offer a defence of the first answer to our question about the point of the history of philosophy—​and of the approach to the history of philosophy to which this answer leads. The central claim of my defence of this first approach to the history of philosophy is that we philosophers, as a community, are prone to forget philosophical ideas that it would be better for us to remember. In other words, one central justification for this approach to the history of philosophy is that it enables us to salvage these philosophical ideas from being undeservedly forgotten. It is crucial to this justification of the first approach to the history of philosophy that not everything that is forgotten deserves to be forgotten. As the seventeenth-​ century English medic and polymath Thomas Browne (1658: 75–​6) memorably complained: But the iniquity of oblivion blindly scattereth her poppy, and deals with the memory of men without distinction to merit of perpetuity . . . Who knows whether the best of men be known? or whether there be not more remarkable persons forgot, than any that stand remembered in the known account of time? As I shall argue here, one central reason why this first approach to the history of philosophy is so valuable is that it rescues important philosophical ideas from the iniquity of oblivion. We might wonder: what has this to do with Williams’s talk of ‘making the familiar seem strange and the strange seem familiar’? It is clear how salvaging philosophical ideas from the iniquity of oblivion can make ‘the strange seem familiar’. Forgotten ideas will often seem strange at first, but after they have been salvaged, they will come to seem familiar. But how will this approach make ‘the familiar seem strange’? The point seems to be that these forgotten ideas are often, in a sense, alternatives to the ideas that were familiar before we were reminded of them: after we have been reminded of these alternatives, the previously familiar ideas may come to seem more doubtful—​and it may also seem strange that such doubtful ideas were for a while treated as unquestionable. In fact, history contains numerous examples of philosophers who were convinced that they benefited from older philosophical ideas that were salvaged from [p. 363] Ralph Wedgwood oblivion. For example, in the twelfth century, the Latin-​using western half of the Christian world rediscovered the works of Aristotle, after some five hundred years when only a handful of his works were known (namely, those that had been translated into Latin by Boethius). This rediscovery was of fundamental importance for the development of medieval scholasticism, as is clear from the authority that Aquinas and later scholastic philosophers ascribed to Aristotle.20 Later, especially in the fifteenth century, the West rediscovered a large number of other ancient works. These included Lucretius’ On the Nature of Things, the works of Sextus Empiricus, and most of the works of Plato, after a period of about eight hundred years when only parts of Plato’s Timaeus were known (thanks to the Latin translations of Cicero and Calcidius). These rediscoveries of long-​forgotten ancient philosophical works had a profound influence on the development of philosophy in the Renaissance and early modern period.21 To give one more example, it has recently been established that, when Frege revolutionized the study of logic in 1879, he was profoundly influenced by the evidence that has come down to us about ancient Stoic logic—​which for centuries had been completely eclipsed by Aristotelian logic, ever since the time when ancient Stoicism faded away amid the rise of neo-​Platonism in the third century ce.22 In all these cases, one of the principal effects of the rediscovery of previously forgotten philosophical ideas was either to enable philosophers to conceive of philosophical concepts, problems, and theoretical options that they had not previously been aware of, or at least to encourage them to take these concepts, problems, and theoretical options more seriously than they had previously done. For example, the rediscovery of Epicureanism in Lucretius’ On the Nature of Things in the fifteenth century encouraged philosophers to take a host of theoretical options more seriously than they had previously done—​including an atomistic conception of matter, a non-​teleological mechanistic conception of nature, and a hedonistic conception of ethics. As is made clear by the examples of Aristotle’s influence on medieval scholasticism and of Stoic logic’s influence on Frege, the rediscovery of long-​forgotten philosophical ideas often has its effect through the later generation’s engaging with the fine detail of the earlier ideas. One way in which philosophical ideas can come to seem promising is when it becomes clear that these ideas can be worked out in detail. If it is reasonable to doubt whether there is any way of working out a philosophical idea in detail, it seems equally reasonable to doubt whether we need to take it very seriously. For a long-​forgotten idea to have this sort of fruitful impact, it is usually not enough if all that is remembered is a vague general idea: usually, it 20 For an authoritative survey of medieval philosophy, see Marenbon (2023). 21 For a famous discussion of the impact of the rediscovery of Lucretius, see Greenblatt (2011). 22 For this point about how Frege drew on the logic of the ancient Stoics, see Bobzien (2021). [p. 364] is crucial for later philosophers to be reminded of some of the fine detail of these earlier philosophers’ work. As noted above, in most of these cases, these rediscovered philosophical ideas are alternatives to the ideas that had previously been taken seriously. But why is it beneficial to consider alternatives to the currently dominant ideas and theories? In the natural sciences, the alternatives to the currently accepted theories are theories that have been discarded. These theories are generally regarded as superseded—​ such as the phlogiston theory in chemistry, or Aristotle’s theory of the spontaneous generation of insect larvae in biology. Natural scientists would not normally benefit from devoting more attention to these superseded theories. Why is the situation different in philosophy? The difference, I propose, lies in the fact that, in the natural sciences, it is typically known which of these alternative theories are true—​while philosophers typically lack such knowledge. This is not to say that philosophers have no knowledge of philosophical truths. However, such philosophical knowledge rarely if ever takes the form of knowing which illuminating philosophical theory is true. At best, philosophers can know that some such philosophical theories are false. For example, it seems to me that philosophers now know that knowledge cannot be defined as justified true belief.23 Philosophers often also have knowledge about the logical relationships between philosophical theories. For example, it seems to me that philosophers know many propositions of the form ‘Given assumptions P, Q, and R, theory S is true only if theory T is also true.’ In general, however, philosophers only know the answers to relatively small philosophical questions. The situation is different with the big questions: What is knowledge, and how is it possible? What is it to be a good person, or to lead a good life, or to act well? Do we have a fundamental nature or essence, and if so, what is it? If these questions are couched at a sufficiently general and abstract level, then my historical judgement is that they are perennial: they have gripped reflective human beings, in many different cultures, at least since the time of Heraclitus, Confucius, and the early Upanishads, in the sixth century bce. Strikingly, these big perennial questions are as wide open today as they have always been. The reason for this seems to be that knowing the answer to any of these big philosophical questions would require a tremendously difficult investigation. Ideally, all possible philosophical theories would have to be compared, on the basis of how well they fit with everything that we believe, and how well they explain everything that such philosophical theories are called upon to explain. To know the answer to any of the big philosophical questions, one would have to have identified the philosophical theory that is in fact the best of all these alternatives, and to believe it precisely because it is the best theory in this way.24 23 The discovery is due to Gettier (1963). 24 For this view of philosophical methodology, see Wedgwood (2014). [p. 365] Ralph Wedgwood On this view, the appropriate method for answering philosophical questions is deeply holistic. Part of the reason for this is that every part of philosophy seems inseparably connected to every other. Philosophers have occasionally claimed that some branches of philosophy have a kind of ‘independence’ or ‘autonomy’, in the sense that they can be fruitfully studied without taking any account of the results of other fields of study.25 But on closer inspection such claims seem dubious. For example, it is hard to study ethics without having to confront the idea that ‘ “ought” implies “can” ’; but any attempt at an adequate evaluation of that idea needs to look far beyond ethics—​to the areas of the philosophy of language that study modal terms like ‘ought’ and ‘can’, to the areas of the philosophy of action that study the notion of ‘abilities’, and to the areas of metaphysics that study the nature of possibilities. Interconnections of these kinds are ubiquitous in philosophy. This holistic conception makes philosophy methodologically different from many other fields. In many fields, hyper-​specialization is an effective way to make progress. Such hyper-​specialized researchers can ignore everything that is not directly relevant to the particular question that they are studying. For example, biochemists can often come to know truths about the chemistry that is characteristic of certain biological processes without worrying about the theory of relativity or quantum mechanics. However, if my proposal about philosophical methodology is correct, such hyper-​specialized research will never enable us to know the answers to any of the big questions of philosophy. Thus, if this view of philosophical methodology is correct, knowing the answers to these big questions would be such a difficult achievement that we may never acquire such knowledge. If we will never know the answers to the big questions of philosophy, will we never make progress in philosophy? In my opinion, it is reasonable to hope that philosophy will make progress—​not towards knowledge of the answers to these big questions, but towards a better understanding of these questions.26 What exactly do I mean here by talking of a ‘philosophical understanding’ of these questions? I propose to interpret this sort of understanding of a philosophical question as a kind of precursor to achieving knowledge of how to answer the question. Specifically, I propose, the degree to which philosophers understand a question is also a holistic matter, corresponding to the degree to which philosophers see the connections that this question has with other questions, and the 25 For a notable example, see John Rawls’s (1974) idea of the ‘independence of moral theory’. 26 For a forceful statement of the judgement that philosophers have failed to produce a ‘general agreed body of philosophical knowledge’, see Hacker (2009: 130). Nonetheless, like me, Hacker (2009: 135–​40, argues that philosophy is a contribution to understanding rather than knowledge, and that there has been progress in philosophy. The main point where Hacker and I differ is that I hold that we have knowledge of the ‘small questions’ of philosophy, and that philosophical understanding is constituted, in part, by a host of such small pieces of philosophical knowledge. [p. 366] pros and cons of all the various alternative systems of interconnected answers to these questions. This conception of the goal of philosophical inquiry as consisting in this kind of holistic understanding is undoubtedly controversial.27 However, if this conception is correct, then the achievement of philosophical understanding requires awareness of as many philosophical problems and questions as possible, and of as many theoretical options for answering those questions as possible. To keep things simple, we may concentrate solely on our awareness of philosophical questions here: in this context, problems can just be identified with questions, and awareness of a theoretical option is, in effect, awareness of some question of the form, ‘Is Theory T correct?’ Above, I claimed that the biggest questions of philosophy are perennial, and have been the focus of philosophers’ attention for more than 2,500 years. However, even if this claim is true, there are countless other smaller questions that philosophers need to pay attention to—​if, as I have suggested, philosophical understanding consists in seeing all the myriad connections between these different questions. These smaller questions are not perennial—​far from it. As we have seen, one of the points that Williams correctly stresses is that different philosophers at different times have focused on different questions. As the centuries pass, the questions that gripped earlier philosophers eventually slip out of sight and are forgotten. However, if my conception of philosophical understanding is correct, progress in philosophy requires awareness of a growing number of philosophical questions (including questions about the interrelations of questions). In other words, philosophical progress requires introducing new questions while continuing to be aware of old questions as well. Thus, forgetting philosophical questions is directly destructive of philosophical understanding, and so directly inimical to philosophical progress. For this reason, I propose, work on the history of philosophy of the sort that I have described—​in effect, work that aims to preserve and salvage philosophical questions from the iniquity of oblivion—​is an utterly central part of our discipline.28 27 It is not clear if Williams would have endorsed this conception. As we have seen, he implies that one central goal of philosophy is to ‘understand ourselves’ (2002b: 406). But this hardly seems like a unique feature of philosophy, since many fields—​including history, psychology, and all the social sciences—​all seem more directly relevant to ‘understanding ourselves’ than research on (say) the philosophy of physics. But it is arguable that a holistic understanding of the questions of philosophy could form at least part of such an understanding of ourselves. 28 Lin (2013) argues that the fact that philosophy makes progress so slowly is essential to explaining the value of the history of philosophy. I am broadly in agreement with this claim. However, what I add to Lin’s argument is a more detailed methodology that explains the slowness and the distinctive character of philosophical progress—​specifically, my distinction between big and small philosophical questions, my holistic conception of philosophical method, and my idea that we should aim for philosophical understanding instead of just knowledge of the answers to philosophical questions. [p. 367] Ralph Wedgwood Another way to make this point is to reflect on the crucial role of discussion and dialogue in philosophy. Philosophers need to consider many different questions, and many different questions about these questions and their interrelations, and listening and responding to other philosophers is practically indispensable for this purpose. The kind of dialogue that philosophers need is inclusive—​it should include as many different voices and perspectives as possible. Voices from the past form a large class that contemporary philosophers would be foolish to ignore. Just as it would be unwise to ignore the voices of your contemporaries who focus on different questions and presuppose different assumptions from you, so it would be equally unwise to ignore the voices and perspectives of your predecessors from past centuries. Admittedly, merely having a dialogue with actual philosophers (past or present) is not always enough for making progress towards philosophical understanding. It is often also important for philosophers to imagine philosophical ideas that no actual philosopher has yet articulated. There are huge areas of theoretical space that remain underexplored. Indeed, in principle, some particularly imaginative philosophers might even be able to make significant progress while ignoring the history of philosophy, and simply conjuring up a vibrant virtual debate with imaginary philosophers. In practice, however, it seems unlikely that many of us possess such extraordinary powers of imagination. For most of us, it is practically indispensable to engage in an inclusive dialogue of the kind that I have described; and at least for many philosophers, it will be highly valuable for this inclusive debate to include the voices of philosophers from the past. 5. The Genealogy of Concepts: Some Sceptical Doubts In this section, I shall turn briefly to Williams’s ‘genealogical’ project. As he describes it, the goal of this genealogical project is to understand our ‘concepts’; and, to understand our concepts, we need to understand their history (although, as I explained above, studying the ‘real history’ of our concepts is only the second half of Williams’s genealogical project—​while the first half of the genealogical project is a more abstract investigation of the role that those concepts will inevitably play in any human society). It is undeniable that it is important for us to understand our concepts. But why should we assume that understanding the history of how we came to possess these concepts, or the history of our use of these concepts, is a particularly important element of understanding these concepts themselves? If our goal is to understand the concepts that we have now, surely it is primarily to psychology, and not to history, that we should look. The general theory of concepts is a central part of the discipline of cognitive science. More specific concepts—​including all the broadly ethical concepts that Williams was particularly interested in—​can be studied in social psychology or moral psychology. For the concepts that play a crucial role in [p. 368] our political life, branches of social science such as political sociology seem likely to be particularly important. Indeed, on the face of it, the assumption that understanding our concepts centrally involves understanding the history of our concepts seems to commit the genetic fallacy.29 The facts about something’s origins in the past do not determine the character and properties that it has now. When the ancestors of birds originally developed feathers, their function was probably just to keep those animals warm; it was only later that feathers acquired the function of enabling the animals to fly. The etymology of the word ‘wrong’ lies in a word that originally meant twisted (‘wrong’ is cognate with the verb forms ‘wring’ and ‘wrung’); but it has long since lost that meaning, and has a completely different meaning in contemporary English. To be fair, Williams (2002b: 406) recognizes this concern. He agrees that it is not particularly helpful for understanding the contemporary concept of an atom to consider how the term ‘atom’ was used by Democritus in the fifth century bce. He responds to this concern just by asserting that things are otherwise with other concepts, such as the concept of freedom (2002b: 407); and in other work, Williams develops a genealogical account of our concept of freedom. However, that account does not make it clear what exactly we gain from studying the history of the concept; it seems that it is really a kind of political sociology that is most helpful for understanding our current concept of freedom. As I explained in section 2 above, Williams’s genealogical project does not just study the history of a concept in a purely disinterested way. On the contrary, it is fundamentally concerned with evaluating the suggestions that the history of a concept may either ‘undermine’ or ‘vindicate’ the concept. Unfortunately, it is obscure what these suggestions amount to. In fact, at this point, Williams often switches from talking about the history of a ‘concept’ to talking about the history of a ‘value’ or an ‘assumption’. However, on most of the ways in which philosophers have used this terminology, ‘concepts’ are importantly different from ‘values’ and ‘assumptions’. An assumption has a propositional content; and so, vindicating an assumption presumably involves demonstrating that the assumption’s propositional content is true—​while undermining an assumption presumably involves raising doubts about this propositional content. By contrast, a concept is not usually taken to have propositional content in the same way, and so it is not clear what vindicating or undermining a concept could amount to. At all events, undermining a concept need not involve showing that there need be any rational defect in possessing and using the concept. When Oscar Wilde said at his trial, ‘ “Blasphemous” is not a word of mine’, he could not plausibly deny that he possessed the concept of blasphemy.30 It is clear from everything that we know 29 Many other readers of Williams have raised the same concern; see, for example, Glock (2008: 878–​ 9) and Nagel (2009: 134). 30 For the transcript of Wilde’s first trial, see Holland (2003). [p. 369] Ralph Wedgwood about Wilde that it was precisely because he possessed the concept of blasphemy that he refrained from using the word. Moreover, in rejecting use of the word for these reasons, he was actually using this concept of blasphemy. So, although Wilde seems to have rejected the assumption that the concept provides a useful way of classifying things, he was not committed to that assumption simply by possessing and using the concept of blasphemy. At all events, while it is undeniably important to investigate whether a concept—​like that of blasphemy—​is a good or useful concept for us to use, it is unclear how the history of the concept is relevant to this investigation. In the case of blasphemy, it seems more relevant to pursue investigations in theology or sociology—​say, by inquiring whether it is possible for God to be insulted by human utterances, or whether the primary function of the concept in contemporary society is just to repress religious dissent—​than to investigate its history. When it comes to beliefs or assumptions, the idea that a historical account of why we hold a belief can undermine or vindicate that belief is familiar to contemporary epistemologists. However, it is also intensely controversial in exactly which cases a historical account of our holding a belief has this undermining effect.31 Even if the historical explanation of why we hold a belief may show that we do not hold the belief for good reasons, it does not follow that we do not have any good reasons for holding the belief—​and it certainly does not follow that the belief is not true. Similarly, even if the belief had pernicious effects in the past, it does not follow that our holding the belief now will have pernicious effects in future. In short, while it is interesting to explore the hypothesis that historical explanations of our beliefs or concepts can have this sort of undermining or vindicatory effect, the hypothesis seems too controversial for it to play anything more than a marginal role in our study of the history of philosophy. For this reason, Williams’s ideas about how ‘philosophy needs history’ seem like a shaky basis for a defence of the history of philosophy. As I noted in section 3 above, these ideas anyway do not support the conclusion that there is anything particularly important about the history of philosophy, as opposed to a more general history of human thought. But it is also unclear whether such a study of the history of human thought can really make the important contribution that Williams imagines to the quest to ‘understand our concepts’. Concepts are psychological phenomena, and prima facie it seems that the various branches of psychology are where we should look in order to understand them. For all these reasons, then, the first of the two Williams-​inspired answers that we have been considering to the question of what point there is in studying the history of philosophy seems to me a much sounder basis for a defence of the history of philosophy. 31 For example, consider the debates over Street’s (2006) ‘Darwinian dilemma for realist theories of value’. [p. 370] 6. Conclusion In the previous section, I tried to raise some doubts about Williams’s idea of the importance for philosophy of a ‘genealogy’ of our concepts. But the main goal of this chapter was to defend and elaborate on some of Williams’s other ideas about the distinctive value for philosophy of studying the history of the subject. In my terminology, it is a practically indispensable component of the quest for philosophical understanding that we should seek to salvage the worthwhile philosophical ideas of the past from the iniquity of oblivion.32 References Althoff, Jochen. 2018. ‘Aristotle, the Inventor of Natural Science’. In Oxford Handbook of Science and Medicine in the Classical World. Edited by Paul T. Keyser and John Scarborough, 235–​56. Oxford: Oxford University Press. Ben-​ A hmed, Fouad, and Robert Pasnau. 2021. ‘Ibn Rushd [Averroes]’. In The Stanford Encyclopedia of Philosophy (Fall 2021 Edition). Edited by Edward N. Zalta, . Bobzien, Susanne. 2021. ‘Frege Plagiarized the Stoics’. In Themes in Plato, Aristotle, and Hellenistic Philosophy, Keeling Lectures 2011–​18. Edited by Fiona Leigh, 149–​206. Chicago: University of Chicago Press. Browne, Thomas. 1658. Hydriotaphia: Urn-​Burial, or, a Discourse of the Sepulchral Urns Lately Found in Norfolk. London: Henry Brome. Clark, Suzannah. 2011. Analyzing Schubert. Cambridge: Cambridge University Press. Diels, Hermann, et al. 1882–​1907. Commentaria in Aristotelem Graeca, 23 vols. Berlin: Reimer. Druart, Therese-​Anne. 2020. ‘al-​Farabi’. In The Stanford Encyclopedia of Philosophy (Winter 2021 Edition). Edited by Edward N. Zalta, . Duncan, Stewart, and LoLordo, Antonia, eds. 2013. Debates in Modern Philosophy: Essential Readings and Contemporary Responses. New York: Routledge. Eagleton, Terry. 1996. Literary Theory: An Introduction. Second edition. Minneapolis: Wiley-​Blackwell. Garber, Daniel. 2005. ‘What’s Philosophical about the History of Philosophy?’, in Sorell and Rogers (2005: 129–​46). Garber, Daniel. 2013. ‘Does Philosophy Have a Future? Some Reflections on Bennett and Doing Philosophy Historically’. In Duncan and LoLordo (2013: 347–​62). Gettier, Edmund. 1963. ‘Is Justified True Belief Knowledge?’ Analysis 23.6: 121–​3. Glock, Hans-​Johann. 2008. ‘Analytic Philosophy and History: A Mismatch?’ MIND 117: 885–​97. Greenblatt, Stephen. 2011. The Swerve: How the World Became Modern. New York: W. W. Norton. Gutas, Dimitri. ‘Ibn Sina [Avicenna]’. In The Stanford Encyclopedia of Philosophy (Fall 2016 Edition). Edited by Edward N. Zalta, . Holland, Merlin. 2003. The Real Trial of Oscar Wilde. The first uncensored transcript of the trial of Oscar Wilde versus John Douglas, Marquess of Queensberry, 1895. New York: Fourth Estate. Lear, Jonathan. 1980. Aristotle and Logical Theory. Cambridge: Cambridge University Press. Lin, Martin. 2013. ‘Philosophy and Its History’. In Duncan and LoLordo (2013: 363–​73). 32 I am grateful to the editors of this volume, Marcel van Ackeren and Matthieu Queloz, for extremely helpful comments on earlier drafts of this chapter. [p. 371] Ralph Wedgwood Miller, Lucasta. 2001. The Bronte Myth. London: Jonathan Cape. Magee, Bryan. 2001. Wagner and Philosophy. Harmondsworth: Penguin. Marenbon, John. 2018. ‘Why We Need a Real History of Philosophy’. In van Ackeren (2018b: Marenbon, John. 2021. ‘Anicius Manlius Severinus Boethius’. In The Stanford Encyclopedia of Philosophy (Winter 2021 Edition). Edited by Edward N. Zalta, . Marenbon, John. 2023. ‘Medieval Philosophy’. In The Stanford Encyclopedia of Philosophy (Spring 2023 Edition). Edited by Edward N. Zalta and Uri Nodelman, . Nagel, Thomas. 2009. ‘Williams: The Value of Truth’. In Nagel, Secular Philosophy and the Religious Temperament: Essays 2002–​2008, 131–​8. Oxford: Oxford University Press. Nietzsche, Friedrich. 1874. Vom Nutzen und Nachteil der Historie für das Leben. Leipzig: E. W. Fritzsch. Cited by the page numbers in Unmodern Observations. Edited by William Arrowsmith. New Haven: Yale University Press, 1990. Nietzsche, Friedrich. 1887. Zur Genealogie der Moral: Eine Streitschrift. Leipzig: E. W. Fritzsch. Hacker, P. M. S. 2009. ‘A Contribution, Not to Human Knowledge, but to Human Understanding’. In Conceptions of Philosophy. Edited by Antony O’Hear, 129–​53. Cambridge: Cambridge University Press. Rawls, John. 1974. ‘The Independence of Moral Theory’. Proceedings and Addresses of the American Philosophical Association 48: 5–​22. Sauer, Hanno. 2022. ‘The End of History’. Inquiry, . Skinner, Quentin. 1978. The Foundations of Modern Political Thought. Cambridge: Cambridge University Press. Sorell, Tom. 2005. ‘On Saying No to History of Philosophy’. In Sorell and Rogers (2005: 43–​59). Sorell, Tom, and Rogers, G. A. J., eds. 2005. Analytic Philosophy and History of Philosophy. Oxford: Clarendon Press. Street, Sharon. 2006. ‘A Darwinian Dilemma for Realist Theories of Value’. Philosophical Studies Thompson, Kirill. 2021. ‘Zhu Xi’. In The Stanford Encyclopedia of Philosophy (Summer 2021 Edition). Edited by Edward N. Zalta, . van Ackeren, Marcel. 2018a. ‘On Interpreting Historical Texts and Contributing to Current Philosophy’. In van Ackeren (2018b: 69–​87). van Ackeren, Marcel, ed. 2018b. Philosophy and the Historical Perspective. Oxford: Oxford University Press. Wedgwood, Ralph. 2014. ‘Moral Disagreement among Philosophers’. In Challenges to Moral and Religious Belief. Edited by Michael Bergmann and Patrick Kain, 23–​39. Oxford: Oxford University Press. Williams, Bernard. 1978. Descartes: The Project of Pure Enquiry. Harmondsworth: Penguin. Williams, Bernard. 1993. Shame and Necessity. Berkeley: University of California Press. Williams, Bernard. 1994. ‘Descartes and the Historiography of Philosophy’. In Reason, Will, and Sensation: Studies in Descartes’s Metaphysics. Edited by John Cottingham, 19–​27. Oxford: Oxford University Press. Cited by the page numbers in Williams (2006a). Williams, Bernard. 1998. Plato: The Invention of Philosophy. London: Phoenix. Cited by the page numbers in Williams (2006a). Williams, Bernard. 2000. ‘Philosophy as a Humanistic Discipline’. Third Annual Royal Institute of Philosophy Lecture. Philosophy 75: 477–​96. Cited by the page numbers in Williams (2006b). Williams, Bernard. 2001. ‘From Freedom to Liberty: The Construction of a Political Value’. Philosophy and Public Affairs 30.1: 3–​26. Williams, Bernard. 2002a. Truth and Truthfulness: An Essay in Genealogy. Princeton: Princeton University Press. [p. 372] Williams, Bernard. 2002b. ‘Why Philosophy Needs History’. London Review of Books, 17 October. Cited by the page numbers in Williams (2014). Williams, Bernard. 2006a. The Sense of the Past: Essays in the History of Philosophy. Edited by Myles Burnyeat. Princeton: Princeton University Press. Williams, Bernard. 2006b. Philosophy as a Humanistic Discipline. Edited by A. W. Moore. Princeton: Princeton University Press. Williams, Bernard. 2014. Essays and Reviews 1959–​2002. Edited by Michael Wood. Princeton: Princeton University Press. [p. 373] Williams, Berlin, and the Vindication Problem Garrett Cullity One of the consistent themes of Williams’s writings on the history of philosophy is Anglophone moral philosophy’s myopic approach to its own history.1 We know that ethical attitudes and practice have changed over time, influenced by all of the currents bearing on the shape of cultural history; but then we tend to overlook the ways in which this is true of the philosophy of ethics also. The questions we now ask are not simply the questions of philosophical ethics as such: the ethical thought and practice that philosophers interrogate and to which they seek to contribute is a historically and culturally local phenomenon, conditioned by the influences of its own time and place. We see the continuity that makes western ethical thought into a unified tradition; but we fail to appreciate the ways in which the questions we now ask, and the resources and forms of argument we appeal to in answering them, are historically contingent.2 Our blindness to this is problematic in several ways. One of Williams’s complaints is shared with all serious historians of philosophy: we misread past philosophers when we approach them as though they were engaged in the same discussions as us, occupying positions in our contemporary debates. He is fond of citing Collingwood’s quip: translating the Greek word dei as ‘ought’ and then criticizing the Greeks’ theory of obligation is like translating triêrês as ‘steamship’ and then criticizing their knowledge of maritime engineering (Williams 2006d: 345; f: 181).3 Another complaint is that we can too easily adopt a complacently progressivist narrative of ethical history, appealing to our own values to argue that it displays a steady march of advancing ethical understanding (Williams 1993: 5). More generally, this narrowness of historical vision means that moral philosophy fails to understand its own subject: it fails to appreciate what ethical reflection is, missing its contingent and historically inflected character. 1 See especially the essays collected in Williams (2006a, 2006b, 2007a). 2 On this point, see Williams (2006e: 44; 2006f: §6; 2007c: 22; 2007d: 36; 2002: 21; 1995d: 135; ch. 10). 3 See also Williams (2006c: 257–​8) on ‘triumphant anachronism’. Garrett Cullity, Williams, Berlin, and the Vindication Problem In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. [p. 374] When any field of philosophical endeavour operates with a deficient historical self-​understanding, it can be compromised in those ways. This is not just a form of remediable ignorance but a fundamental philosophical error: when philosophy is blind to its own history, it misdescribes the subject it claims to address (Moran However, in ethics the problem runs deeper. Moral philosophy aims to participate in normative ethical thought, not just to observe it. And normative ethical thought, in insisting on the distinction between how I do respond to the world and how I really should respond, contains what we can call an aspiration to objectivity.4 But once we take seriously the realization that not only our ethical outlook itself, but also our mode of philosophizing about it, are a collection of attitudes, patterns of thought, and dispositions of judgement that are the inherited product of historical influences, we invite an unsettling question. If our ethical thought can only ever amount to an elaboration of one historically conditioned ethical outlook among others, what could vindicate the claims to correctness that it aspires to? How could we be warranted in thinking that our ethical judgements could ever attain the objectivity they aspire to—​that they could succeed in identifying the actions we should perform and the attitudes we should have, rather than merely embodying the attitudes we do happen to have? The first half of this chapter examines Williams’s contribution to posing this problem—​the vindication problem, I shall call it—​and his own strategy of response to it, and raises some questions about the limitations of an argument of that form. The second half then proposes a different but related strategy, drawing on some themes from one of Williams’s intellectual mentors, Isaiah Berlin. Berlin emphasizes that the history of our attempts at self-​expression and self-​understanding displays a plurality of goods around which fulfilling human lives may be structured, but also forces us to recognize that we must choose between them: we must embrace some values in competition with others, and all of the available alternatives involve some loss.5 Within your own life, this gives you the task of fashioning that life in a way that makes sense as a response to the range of circumstance-​dependent goods that happen to be presented to you. I shall argue that we collectively face a corresponding task, when we ask how it makes sense to continue the ethical project within which we happen to find ourselves, situated at our own point in its history. My contention will be that this gives us a way of meeting Williams’s challenge: it shows how ethical thought can be stably self-​aware of its own contingency, while retaining the idea that there really are better and worse answers to the ethical questions through which we shape our lives together. 4 Williams does see this distinction as part of the content of ethical thought; but he tends to reserve the name ‘objectivity’ for something that eludes ethics: ‘an objective foundation for ethical life’ 5 See Berlin (2013a); also Stocker (1990). [p. 375] Garrett Cullit y 1. The Vindication Problem Before we come to the history of philosophy, cultural anthropology already gives us ample evidence that ethical thought and practice have varied widely among human communities.6 Our liberal morality of individual rights is a modern western phenomenon, contrasting with other traditions of ethical thought that conceptualize personal virtues as contributions to a common good; and within the western tradition the further back we travel in time, the greater the differences are between our current ethical ideas and the patterns of thought that shaped past societies. We find no egalitarian concern to identify rules of conduct that are justifiable to, and hence binding on, everyone; but instead an emphasis on the forms of excellence that are available to the occupants of socially recognized roles within a hierarchical order, with an accompanying emphasis on the evaluation of oneself as fine or despicable rather than one’s actions as permitted or forbidden, and on the shameful rather than the blameworthy as the dominant form of ethical evaluation.7 The history of philosophy, according to Williams, adds three important things to this now-​familiar picture. First, it emphasizes the depth of the differences. In reading Plato, Aristotle, Hume, Kant, or Nietzsche, with their various attempts to analyse ethical thought into its most essential elements and to anchor these in their most fundamental sources, we can come to appreciate the distinctiveness and peculiarity of the ethical ideas we have inherited, and the trajectory that has led us here. Granted, there are important continuities. Philosophical ethics, as ethical thought’s reflection on itself, is recognizably one activity, and some of the ideas we have inherited can be traced back to its beginnings (Williams 1985: ch. ch. 1; 2006e). But what attention to this history shows us is differences in ethical thought that go beyond giving different answers to ethical questions: more starkly, they involve different questions, different conceptions of what is necessary in order to answer them adequately, and different repertoires of concepts in terms of which the questions themselves and the possible answers to them are understood. Secondly, Williams insists that the history of philosophy does not just reveal the peculiarity of our ethical culture: it can help us to appreciate its deficiencies, by laying bare the presuppositions embedded within it, and inviting us to question them (2007d: 36). He highlights the distinctive features of the ‘morality system’, with its legalistic emphasis on the permissible and the prohibited; its structuring of ethical thought around a concern with obligation; the prominence it gives to the practices of blaming and attributing responsibility; and as he sees it, the reliance of this system of ideas on a conception of the pure volitional self, standing above 6 For overviews, see Fassin (2012), Zigon (2008), and Fassin and Lézé (2014). 7 See, for example, Westermarck (1912), Benedict (1934), Ladd (1957), Shun and Wong (2004), Wong (2009), and Schneewind (1998). [p. 376] the desiderative and affective dispositions we have, as a pure characterless locus of decision that issues directions for action and serves as the bearer of responsibility, the target of blame, and the subject of obligations (1985: ch. 10; 1993: 94–​5, 159; b: 70–​3). This conception of ourselves is the one we have inherited; it structures the moral relationships in which we stand to each other; but it is incoherent. The third lesson that Williams draws from the history of moral philosophy is that it illustrates the failure of a persistent philosophical aspiration. This is the aspiration to find some culturally untainted, absolute standpoint that supplies the materials from which the one true ethical outlook can be constructed, thereby securing the credentials for its claim to superiority over all others. There is no such Archimedean standpoint (Williams 1985: ch. 2; 1995c: 170). Philosophy has kept searching for an extra-​cultural foundation from which to derive a universally authoritative set of ethical questions and answers—​a foundation in the purposes, dispositions, or capacities that are held to constitute human nature as such, or in the presuppositions of rationality, free agency, or mutual comprehensibility, or in the universal requirements of stable communal life—​but it has always failed. There is no such route that takes us from what it is to engage in ethical thought as such to the set of resources we have and need to deal with the particular ethical predicament we face, situated at our own juncture in history. This presents a deep problem for the philosophy of ethics: the vindication problem. We do need ‘to think seriously about a decent life in the modern world’ (Williams 1995d: 148). And serious ethical thought and practice aim to identify and live up to standards for conduct that embody not just what is thought to be good, but what really is good. They carry a commitment to standards with respect to which we can be better or worse, independently of whether we think we are: when we take ethical questions seriously, we are trying to get it right, ethically. But how can the conviction that there are correct and incorrect answers to ethical questions be vindicated, once we accept the historical contingency of our own ethical outlook?8 The point is not just that we now condemn some practices that have been widely accepted in the past: slavery, infanticide, the use of extreme cruelty as exemplary punishment, the subjugation of women, the persecution of homosexual relationships, paedophilia. The differences run deeper than that. Ethical outlooks differ in the questions they ask, the addressees of those questions, the concepts whose application we seek to determine, the forms of argument deployed in ethical thought, the repertoire of available attitudes through which human relationships are structured, and the conception of the self that bears these relationships.9 How, then, 8 See Williams (2006f: 193), where he calls this ‘a problem of historicist weariness and alienation’; see also Williams (1985: 171; 2002: 21). 9 See Schneewind (2003). [p. 377] Garrett Cullit y can any ethical outlook vindicate its own claim to be getting ethics right, once it recognizes its own historically contingent particularity? In framing the discussion that follows I will accept, with Williams, that two complacent responses to the vindication problem should be set aside. The first is a complacent objectivism that dismisses this problem as adding nothing to the task of ethical thought in the face of disagreement. We do not have to extend our horizons beyond our own immediate context to find deep and persistent disagreement in ethics, about the treatment of non-​human animals, the limits of acceptable paternalism, the just exercise of governmental power, and so on. But the persistence of disagreement should not reduce us to scepticism about whether there are answers to ethical questions. It simply imposes on us a discursive requirement to take the opposing views seriously, to seek to understand the source of our disagreements and, having identified the most fundamental considerations to which we attach different degrees of importance, to ask ourselves whether, on careful reflection, this uncovers something we have overlooked. And while this sort of thinking is no guarantee that we will arrive at the correct conclusion, the unavailability of some neutral perspective of adjudication should not lead us to doubt that there are correct answers to ethical questions. Any set of ethical views—​those of me as an individual, or those of a culture—​has its causal influences in the circumstances that have shaped it; but all this means is that we need to conduct our ethical thought carefully, with a suitable awareness of the possibility of distorting influences on our judgement—​not that we need to call into doubt the possibility of answering ethical questions correctly. But if Williams is right, that is too quick.10 It is a mistake to think that the questions we ask about obligation, responsibility, freedom, blameworthiness, permissibility, rights, and respect for autonomy are the questions of ethics. It is not simply that others in the past would not have agreed with our answers to the questions we ask—​they would not have asked these questions.11 The challenge is not one that can be met by assuring ourselves that we are pursuing our methods of ethical enquiry conscientiously. The challenge is to find something that could vindicate any confidence that our mode of pursuing ethical enquiry is the correct mode. But if complacent objectivism is too quick, then the same is true of its dismissive counterpart, complacent relativism. This happily accepts that there is no Archimedean standpoint from which to justify the commitments of any ethical outlook. But the lesson it draws is that ethical justification is always justification from within an outlook—​justification by its standards. Since the idea that there 10 Compare Williams (2006f: 191), where he asks ‘whether a tune as thin as this is worth whistling at all’, and the rejection of ‘Right Wittgensteinianism’ in Williams (2007d: 34). 11 Williams does not deny that past cultures asked questions about obligation, responsibility, freedom, and so on; but they did not always ask our questions. On obligation, see Williams (1985: 185–​ on the continuities and differences between ancient Greek conceptions of responsibility and our own, see Williams (1993: ch. 3); on freedom, see Williams (1995d: 135–​6). [p. 378] could be such a thing as justification as such is an illusion, there is nothing defective in an ethical outlook that does not claim to provide it. The only kind of correctness I can ever claim is correctness-​relative-​to-​standard-​S. I need not be reluctant to accept that standard S is no better than any other alternative standard, if there is no further standpoint from which such comparisons can be made. However, I take it that complacent relativism cannot dissolve the vindication problem either: it is a revisionary and not a vindicatory proposal.12 It is not a way of representing the content of serious ethical thought as it currently engages us. When we ask ourselves which practices of punishment are just, which conditions of employment are exploitative, and which forms of redress for disadvantage are right, we are asking for more than just the assurance that there is some standard, neither better nor worse than other alternatives, that rules a practice in or out. Seeking to insulate ethical thought from Williams’s demand for vindication by relativizing its content does not work because it changes the subject. If the problem of vindicating ethical thought as it is was insoluble, we might ultimately abandon that thought and replace it with a relativized alternative. But this would not be a way of making sense of ethical thought as it is.13 If those responses are too quick, where can we look instead for a response to the vindication problem? Williams has his own proposal, which we can examine next. Having seen its limitations, we will then look in a different direction. 2. Williams’s Genealogical Solution One of the ways in which the history of philosophy can advance our understanding is by ‘making the familiar seem strange’ (Williams 2006c: 259).14 Williams’s attack on the ‘morality system’ is an exercise of this kind. As he describes it, this is a complex of ideas built around a conception—​inherited ultimately from Plato, but focused and sharpened by Kant—​of the agential self as a pure characterless will. Purified of the psychological idiosyncrasies that distinguish us as individuals, this autonomous decision-​making self is the bearer of responsibility, the addressee of obligation, the target for guilt and blame, and the claimant protected by the individual rights that form the basis of a modern liberal morality. Williams argues from several directions that this picture is incoherent: a central complaint is that it cannot stably accommodate an awareness of its own historical contingency. In presenting itself as a transparently introspectable depiction of what human agency essentially is, it has no room for the recognition that this idea, pervasive though it has 12 For the different kind of relativism Williams wants to preserve, see Williams (1981c, 2007b). For discussion, see Fricker (2010). 13 See Williams (2007c: 24; 1985: ch. 9) and compare MacIntyre (1985: ch. 6). 14 For discussion, see van Ackeren (2018). [p. 379] Garrett Cullit y become, is a historically local phenomenon. But it is, and this is deeply problematic. This picture of the volitional self would only be right if it were compulsory—​if it were the only way in which humans could recognize themselves as agents—​but it is not.15 So the morality system should be disowned because it is ‘unstable under reflection’: it cannot tolerate the acceptance that it is a ‘peculiar institution’ (Williams ch. 10).16 Another way to press this objection is via a genealogical critique of the kind mounted by Nietzsche ([1887] 1996).17 A genealogy is not itself an exercise in historiography, but rather an imagined aetiological account of how our current practices could have arisen in response to the needs they serve. Here, the aim is to illustrate a point that is psychological, not historical: to make vivid a picture of the psychological function of moral ideas by casting this as a narrative. But again, the critique functions by exposing morality to a view of itself that it cannot tolerate. Morality ‘demands to be understood as self-​sufficient’ (Williams as being independent of any psychological needs, let alone needs of the unflattering kinds that Nietzsche ties it to, with sources in resentment and impotent hatred. A genealogical critique of this kind, Williams argues, should be deeply unsettling. The response to it cannot be simply to make some adjustments to our metaphysical doctrines about the self and carry on with moral practice as before: the errors run too deep for that. But nor can it be to shrug off ‘the morality system’ and replace it with, say, the ethical outlook of the Homeric Greeks: our relationships to each other and our ethical self-​understanding are shaped by this unsatisfactory system, in ways that none of us can easily think our way out of. The ongoing intellectual project this leaves us with is to fashion out of the materials we have to hand a set of ideas that can serve as the more coherent successors to the ones we have However, the thought Williams explores in his final monograph, Truth and Truthfulness, is that this has a positive counterpart—​a vindicatory genealogy ch. 2; 2006f: §4). There is a kind of genealogical explanation of our ethical practices that can serve not to undermine but to vindicate them in the face of—​indeed, because of—​their historically contingent character. He denies that this can be achieved for our ethical outlook as a whole. Instead, the project must be carried out painstakingly, one element at a time. He offers to do this for the virtues of truthfulness—​of accuracy (the endeavour to acquire true beliefs) and sincerity (the commitment to revealing what you believe in what you say). 16 In using the latter phrase, Williams makes a Nietzschean allusion to its use as a name for the institution of American slavery. 17 For Williams’s discussion, see (2002: ch. 2, §5). [p. 380] A genealogy of truthfulness, Williams argues, can provide a vindication of this kind if it can do two things: first, show how the practice of regarding truthfulness as intrinsically valuable serves needs which that practice can itself endorse, and then beyond this explain the relationship of truthfulness to other intrinsic goods. By doing this, we can show that it meets two conditions that Williams proposes as jointly sufficient for something to have intrinsic value: ‘first, it is necessary (or nearly necessary) for basic human purposes and needs that human beings should treat it as an intrinsic good; and, second, they can coherently treat it as an intrinsic good’ (2002: 92). Truthfulness meets the first condition through serving a culturally universal need for the pooling of information and thus for sustaining relations of trust with each other. The culturally specific variations around this basic core—​variations in ideals of sincerity, intellectual integrity, and authenticity—​are the results of changing, culturally localized conceptions of privacy, rivalry, and cooperation, which bring with them different understandings of the access to information that we can properly demand of each other. And the genealogical account shows how the second condition is satisfied by showing that truthfulness ‘has an inner structure in terms of which it can be related to other goods’—​goods such as freedom from manipulation and domination by others (2002: 91–​3). What this argument seeks to achieve is a vindication of the value of truthfulness, but not the sort of justification that would seek to derive its importance from its relationship to another more basic value. The vindicatory challenge is not to add a deeper set of justificational elements to our evaluative outlook, or to trace its justificational foundations to something outside itself. Instead, it is to assure us that this outlook can be stable under reflection—​that it can withstand the daylight of understanding its own relationship to the needs it serves. However, I think there are two kinds of reason for being unsatisfied with a genealogical argument of this form as an answer to the vindication problem described above. One worry is that the first of Williams’s conditions tends if anything to make an evaluative practice unstable under reflection. It tells us that we find ourselves with the belief that truthfulness is intrinsically valuable because of the instrumental value of that belief in serving basic human purposes and needs. But this invites the debunking thought that our convictions about intrinsic value have been implanted by culture or nature as useful illusions. We may think that there is ultimately a good reply to that challenge—​that our commitment to the intrinsic value of truthfulness can indeed be stably combined with the thought that the commitment has this instrumental explanation. But even if we reach that conclusion, the first part of Williams’s genealogy is better characterized as a challenge to the stability of our evaluative practice, rather than a contribution to establishing that stability.18 18 Compare McGinn (2003); and for a response to McGinn, see Queloz (2021: §7.4). [p. 381] Garrett Cullit y The second, larger, worry is that if we do allow that meeting Williams’s two conditions establishes a kind of stability under reflection, it is a stability that is too easily secured to amount to a solution to the vindication problem. Consider revenge: something that has been intrinsically valued throughout human ethical cultures more often than not—​‘the sweetest of all pleasures’, according to ancient Greek proverb.19 Is it so clear that this evaluative practice cannot meet Williams’s two conditions? The prevalence and longevity of the cultures of honour that contain this practice speak to the existence of basic needs that it serves—​individual needs for emotional catharsis; social needs for deterring mistreatment. And within those cultures, revenge belongs to a tightly interlocking network of values of self-​respect, retributive justice, and honour—​providing the materials to meet Williams’s second condition.20 Cannot this evaluative view, too, have the kind of stability under reflection that is delivered by a Williams-​style vindicatory genealogy? To this worry, a two-​pronged reply could be made on Williams’s behalf. First, if an honour culture met the two conditions, that would not somehow validate it as an alternative to our own. The vindicatory question for us is whether our own values are stable under genealogical reflection: if they are, then they are available as reasons for rejecting evaluative outlooks that are incompatible with them. An honour culture is bad, our values assure us:21 if our confidence in our own values survives reflection on their contingency, then we should continue to say this, whether or not some other outlook is stable under reflection too. And, secondly, to insist on some conception of ethical objectivity that is better than this—​that secures the superiority of our outlook over other internally stable alternatives—​ just is to hanker after an Archimedean point of the kind that philosophy’s failures have shown to be unavailable. The lesson we must take from the history of philosophy is that we have to settle for less than that, and we have to learn not to be disturbed by this. This kind of reply seems to me to get uncomfortably close to the complacent objectivism we wanted to supersede: our values are ours, we have no reason to abandon them, and they tell us to reject the alternatives. But whether or not that is unduly complacent, I will now argue that a better and more substantial response is available to the vindication problem. 19 Thucydides, The Peloponnesian War, vii 68.1. 20 See, for example, Boehm (1984). 21 At least, that kind of honour culture is bad. Something Williams approvingly points to in explaining how truthfulness has historically been given a structure that relates it to other goods is the connection between truthfulness, honour, and nobility (see Williams 2002: 115–​17). [p. 382] 3. A Berlinian Vindication This alternative is anchored in the value-​ pluralism championed by Berlin a).22 The human condition gives us a variety of incompatible values towards which flourishing human lives can be structured, but forces us to choose between them. This is true on a personal level: one must choose between a life structured around lifelong attachments to family, permanent friendships, and local community, and a globetrotting life of constant novelty, stimulus, and curiosity-​feeding variety: both are good, but no one can live them both. The same, Berlin maintains, is true on the political level. The great political values of equality, liberty, security, community, opportunity, welfare, and respect compete with each other for our social allegiance; we cannot satisfy them all completely. This is in one way a tragic thesis, since however we resolve these personal and political choices, there is a loss. But it is also an irenic thesis. It is a mistake to think that there is one supreme recipe for the well-​lived life, and one correct form of political order structured towards its realization. The mature recognition of this point supports a liberalism that acknowledges and celebrates the different forms that well-​directed human aspiration can take, and makes room for them all. For Berlin, this is a lesson that political history teaches both negatively and positively: negatively, by reminding us of the misery inflicted by the totalizing imposition of one master-​ideal; but also positively, by inviting us to exercise the kind of imaginative fantasia urged by Vico, and to appreciate the human fulfilments to be found in cultures we cannot ourselves inhabit. To this Berlinian idea, we can add a further, companion point. Within the life of an individual person, there are various different qualities we can refer to as ‘integrity’. One is a kind of moral fortitude: the kind displayed in standing up for what is right when there are strong incentives not to. But another is instead a kind of constancy: it is the quality of structuring one’s life by sustaining one’s allegiances to particular values. This is what would be missing in the life of a person who, confronted with the Berlinian plurality of goods around which a life can be structured, flitted temporarily between each of them, trying to sample them all within the compass of a single life. A life of this unstable kind would lack a sort of unity or coherence it makes sense to aspire to—​in trying to attach itself to as many goods as possible, it would fail to attain any of them fully. To say this is not to be committed to the mistaken view that your past irrevocably fixes the path your future life should take. You might realize that your past decisions were mistakes—​that the values towards which you have directed yourself are illusory, or that the efforts you are making to live by them are unavailing. But it suggests that there is a kind of 22 For Williams’s own description of Berlin’s value-​pluralism, see Williams (1981b: 71; and his Introduction to Berlin’s Concepts and Categories (2014). [p. 383] Garrett Cullit y value—​‘integrity-​as-​constancy’, we can call it—​that one’s life instantiates when the allegiances it sustains make sense as a cohesive way of orienting oneself towards genuine values.23 This thought can also be extended to the political level. Faced with a Berlinian plurality of goods between which we must now choose, the choices we confront are situated within a history shaped by our previous decisions and the goods to which those decisions have been responsive. The decisions we now make will constitute a further stage in an ongoing political project which has been shaped around allegiances to particular ideals. The past decisions that have formed our society as a compassionate, or a freedom-​respecting, or an excellence-​supporting one can now provide us with reasons to continue on that path, recognizing the pre-​eminence of some social goods in our national identity. Our previous choices have helped to determine the goods around which our national story has been shaped, just as an individual’s past choice can shape their own personal story, and affect the ways in which it makes sense to continue that story. This point about integrity-​as-​constancy in the political sphere comes with a qualification that parallels the one we noted for the lives of individuals. It does not commit us to a reflex conservatism. If our past decisions were incorrect, they cannot become correct just by repeating them. However, in cases of forced choice between genuine goods, the fact of having responded to those choices in the past by structuring our society around our respect for one of those goods can now provide us with further reasons that bear on our future choices: reasons to remain true to the values that have become central to our national identity.24 One way to hear talk of ‘our values’ in political debate is as a reference to reasons of this kind. This gives us a pair of ideas—​Berlinian pluralism and the value of integrity-​ as-​constancy in sustaining our allegiances—​and two applications of them, to the value-​commitments of individuals and of polities. I now want to suggest that Williams’s vindication problem can be addressed with a third, higher-​level, application of the same pair of ideas: an application to ethical cultures themselves. The application has two stages. The first stage is the recognition that, if Williams is right about the historical particularity of our ethical culture, it invites a Berlinian treatment: we can recognize our own ethical culture as one valuable way of structuring ethical thought, alongside other human possibilities. The ideas, concepts, arguments, and justificational foundations that structure our ethical thinking are not those of all ethical thought as such. This is our ethical project, with its own history, outside of which we cannot step. It is one of a range of alternatives that history displays to us. In this case, those alternatives do not present themselves as 23 For further discussion of integrity-​as-​constancy as a narrative virtue of individual lives, see Cullity (2021). 24 On the question to what extent an ongoing political community presupposes a national identity, see Habermas (1996). [p. 384] an array of possibilities between which we must choose. But we do not need to be able to show that our own inherited ethical project is superior to other alternatives by some independent measure—​any more than I need to be able to show that the goods around which my life are structured are objectively superior to other forms of human fulfilment, or that our national political story is better than any other—​ in order for a commitment to the continuation of this project to deserve our allegiance, and for us to have reasons of constancy to remain true to it. So the first stage of the application is to take Williams’s conception of a plurality of particular ethical cultures, with no external perspective of arbitration from which to declare one of them superior, and to place this within a Berlinian frame. Just as there are alternative ways to structure a political culture around rival goods, not all of which are co-​realizable, history shows us alternative ways of configuring an ethical culture through which different potentials for excellence in our recognition of and responsiveness to each other are realizable. Our own ethical culture, with its deep emphasis on moral equality and the accountability of all to all, embodies one valuable form of ethical life. But in Williams’s Nietzschean view, this also cuts us off from other, freer forms of ethical thought that we can recognize in pre-​modern cultures. Our central notions of obligation, personal responsibility, guilt, and blame carry with them an interiority, a self-​invigilation and a legalistic concern with the forbidden and the permitted that are, in comparison with other ethical cultures, both oppressive (in burdening us all with an obligation-​bound, omniresponsible self ) and complacent (in emphasizing the avoidance of wrongdoing over the aspiration to personal excellence). Berlin’s message applies: every concretely realized value-​commitment, in forgoing the ways of life with which it is incompatible, carries with it losses as well as gains. The second stage then moves beyond Berlin’s pluralism to emphasize the constraints that this imposes on us, inhabiting the ethical culture that we do. There may be a plurality of valuable forms that ethical culture can take, and it may be possible for us through the exercise of historical and cross-​cultural imagination to appreciate the forms of human fulfilment that they contain, but they are not accessible to you and me. We are subject to a psychological constraint: it is not attitudinally possible for us, either severally or together, to escape our own ethical formation and the constituent patterns of ethical thought and feeling through which our relationships to each other are structured (Williams 1981c: 132–​5, 138–​40). But independently of this there is another, evaluative, constraint that conditions our ethical thought and practice. How it is good for us to go forward from here, with the history we have, is constrained by the value of integrity-​as-​constancy. The ethical project in which we are embedded is only one among other human possibilities, but as our ethical project its continuation is something of which we are the custodians, carrying forward an ethical culture that is valuable, if not uniquely so. If this is right, it encourages a Berlinian response to the vindication problem. The ethical questions we ask may not be the ethical questions that human beings [p. 385] Garrett Cullit y must ask as such; but they can still be the right questions for us to ask, situated as we are. And if these questions have good answers, the content of those answers can indeed constitute the right way for us to live. We saw above that Berlinian pluralism does not commit us to a complacent conservatism either in personal life or in politics. The same is true of its application to ethical cultures. If we take Williams seriously, after all, we have to allow for the possibility that our ethical ideas are not in good order—​that it is not just prevailing ethical opinions that stand in need of reform but deep features of the structure of contemporary ethical thought. A Berlinian response to the vindication problem cannot guarantee that our ethical thought is without flaws and even (as Williams claims) incoherences. But it supports the thought that our ethical task is one of reform rather than abandonment. Our ethical situation is like the political predicament of those who have grown up in a society structured by flawed institutions that are a focus for valuable if imperfect forms of human community, or the personal predicament of a person whose life is structured by a commitment to relationships that are mutually enriching but flawed. This, I want to suggest, gives us a credible response to Williams’s vindication problem. It does not answer all of the worries one might have about the objectivity of ethics. But it does offer a way of reconciling the commitment that serious ethical thought has to finding correct answers to the questions we ask with a recognition of the historical contingency of those questions and answers. It allows us to retain the conviction that the ethical project in which we are engaged is itself valuable, that there are better and worse ways for us to continue that project, and that we therefore really should do the things that our ethical reflection tells us to do. 4. Ethical Confidence In proposing this Berlinian response to Williams’s vindication problem, I might seem simply to have missed his point. His problem, after all, concerns the culturally local character of contemporary liberal orthodoxies, which complacently claim superiority over other outlooks by criticizing them as illiberal.25 But when Berlin asserts that there is a plurality of mutually irreconcilable goods and that liberalism is the best political response to this, he is making substantive ethical claims from within the outlook whose authority we are being challenged to vindicate.26 His is a substantive liberalism, not a merely political liberalism of the Rawlsian kind: he thinks illiberal political orders should be rejected as ethically mistaken.27 25 See Williams (2006f: 190; 2007b: 67; 2007c: 23–​4). For discussion of Williams on this point, see Nagel (2021). 26 For a spirited criticism of the Berlinian combination of pluralism and liberalism, see Gray (1998). 27 Not that Rawlsian political liberalism escapes the vindication problem: its guiding aim of finding the political accommodation that is reasonable in sustaining a community of free and equal persons remains a distinctively modern ethical commitment. See Rawls (1993). [p. 386] So why isn’t the strategy of argument described above, in appealing to Berlin’s liberal pluralism—​along with the virtue of integrity-​as-​constancy—​just obtusely helping itself to ideas internal to the local ethical perspective that stands in need of vindication? However, this objection backfires: it is the objection that fails to appreciate Williams’s problem, not the Berlinian response. To object that the Berlinian vindication relies on materials internal to our local ethical outlook is to complain that it delivers no justification of our ethical commitments from some external foundation. But that is unattainable, according to Williams. His problem is not: how can we justify our ethical outlook by deriving its content from some external, transhistorical, and universally authoritative foundation? And it is not: how can we defend our value-​commitments against global evaluative scepticism? It is instead: how can our ethical commitments be stably combined with a reflective acceptance of their own historical contingency and the unavailability of any Archimedean grounding? To be sure, it is not an impressive reply to this challenge to maintain that our liberal outlook is superior to others because they are illiberal. But the Berlinian response is not doing that. It offers a way of meeting Williams’s vindication problem precisely because it does not rely on asserting the objective superiority of own outlook over others. It allows for a plurality of valuable ethical cultures, and explains how our commitment to finding the correct answers to the ethical questions we ask can be stable under a reflective awareness of the historical contingency of the outlook within which we are asking those questions. However, while the Berlinian strategy is not trying to establish the superiority of our outlook over all others, it does require that there are genuine values around which human lives can be structured—​that there is indeed a plurality of values, and not just of thoughts about value—​and that these include the values recognized by our own ethical culture.28 So it does still require that we can have a basis for confidence in the value-​judgements that our ethical outlook contains—​and also in our judgements about the value of our own ethical outlook itself. One might still wonder: where could that come from? It can come from several different directions. There are various ways of exercising our epistemic powers that can potentially contribute to confidence in the value-​judgements on which the Berlinian strategy relies. Some are activities within philosophy. For example, there is the kind of justification-​identifying activity that normative ethical theorists engage in, which aims to explain the ways in which one reason or value can derive from another. This kind of theorizing can potentially help to support our confidence that the multitude of considerations that make up our ethical outlook do constitute an evaluative and normative perspective that is There is a world of objective values’ (Berlin 2013d: 11). [p. 387] Garrett Cullit y worth living by if two things are true: first, there is some systematic account of the ways in which the complex surface detail of morality can be traced to a relatively small number of derivational sources; and, secondly, those sources are of transparent and compelling importance. Williams challenges both of those claims: as he sees it, the only ways we have found of giving a systematic account of the content of morality fail in the first respect, because their attempts to systematize fail to accommodate the neatness-​resistant texture of our ethical lives, and they also fail in the second respect, because the principles those theories posit as foundational to morality fail to capture what really matters to us in interpersonal relationships ch. 6). However, some of us think that it is too early to draw that pessimistic conclusion about the prospects for ethical theory.29 And if a theory could indeed meet those two conditions, it could then support our confidence in the value of our ethical outlook by showing us how it makes sense as an expression of our most fundamental values, and articulating those values in a way that makes their importance evident. Within philosophy, there are at least two other kinds of activity that can help to ground confidence in our judgements about those deeper values. One comes from work in the epistemology of value-​judgement—​work that helps to spell out the sources of distortion to which evaluative attitudes are susceptible, and (conversely) the conditions under which we are warranted in trusting our own impressions about what has importance. But in addition, there is a further kind of enquiry that philosophy can conduct. When Elizabeth Anderson reasons that intrinsic goods ‘are rationally valued, apart from the value of any other particular thing. Intrinsic values thus mark the point where justification comes to an end’ (1996: 541), what she says is true if ‘justification’ refers to the derivation of one value from another. But there remain further ‘why?’ questions we can ask about what is derivationally foundational. Once we have identified our derivationally foundational values or reasons—​those from which other values or reasons derive, but which are not themselves derived from other values or reasons—​philosophy does not run out of questions to ask, or answers to give. There are ways of explaining why something is non-​derivatively good or bad.30 Challenged by a Stoic or Puritan to explain why pleasure is good and suffering is bad, we need not try to derive the value of pleasure and suffering from the value of something else. We can instead give an account of the internal characteristics of pleasure and suffering themselves, in virtue of which they matter in the way they do. Or we can say (with Calvin!) that an outlook that denies the goodness of pleasure and badness of suffering is ‘inhuman’.31 A remark 29 For my own attempt to pursue this theoretical project, see Cullity (2018). 30 Compare Williams (2002: 92); and, for discussion, Queloz (2021: 169–​70). Away, then, with that inhuman philosophy which, while conceding only a necessary use of creatures, not only malignantly deprives us of the lawful fruit of God’s beneficence but cannot be practised unless it robs a man of all his senses and degrades him to a block’ (Calvin, Institutes of the Christian Religion III: X: 3). [p. 388] like this should not be read as proposing to derive the goodness of pleasure from the goodness of something else, humanity; but as inviting us to reflect on what it would be like to live in a way that did not treat pleasure as intrinsically good and suffering as intrinsically bad, and on the ways in which attempting to live such a life would be bad. Philosophical explanations of these kinds can be complemented by the vindicatory resources that are available to us outside philosophy. We most fully assure ourselves that something has intrinsic value by describing the lives that are oriented towards it, lives that reject it, and lives that misunderstand or mispursue it, with the kind of detail and clarity that makes its attractiveness fully compelling. For this, we need the help of literary fiction, biography, and history—​the kind of demonstration that consists in showing us the goodness or badness of a set of evaluative attitudes, not inferring or analysing it.32 If Berlin is right, another resource is also available to us. An idea he draws from Vico is that the capacity of fantasia—​of sympathetic cultural and historical imagination—​can be developed to allow us to appreciate the value of modes of thought and ways of life that are remote from our own.33 We can see other ways of living as humanly intelligible—​as alternative ways of shaping the human potentials for fulfilment—​while recognizing them as incompatible with our own. But to the extent that we can do this, it can also help to support our confidence in our judgements about the value of our own ethical culture, in two ways. By giving us confidence in our negative judgements about the past—​the stunting of lives by injustice and cruelty—​it can help us to see what led our forebears down the paths of ethical development that have given us the legacy we have inherited. But also, if the training of our historical imagination can give us confidence in our positive judgements about the forms of excellence embodied in distant ethical cultures, this can also help to support the credibility of the positive judgements we make about our own. It can help to assure us that when we hold a mirror up to our own ethical culture, describe it, analyse it philosophically, compare it to others, and find central parts of it to be compelling, we are not just reflecting back our own prejudices. Finally, there is also the possibility of appealing to a Williams-​style genealogy of our ethical views. Above, I raised doubts about whether the kind of stability under reflection that this could secure is sufficient to solve the vindication problem: it cannot itself get us all the way to the conclusion that our ethical enquiry directs us towards the correct way to live. For that, we need a Berlinian vindication: an argument that, since our ethical outlook is one valuable form of ethical culture and our historical relationship to it gives us reasons to sustain it, it can indeed tell us how 32 On this point, compare the distinction between ‘deictic’ and ‘apodeictic’ moral discourse in Garfield (2000: 203–​4). [p. 389] Garrett Cullit y to live well. A Berlinian vindication of this kind does rely on substantive evaluative claims—​the claim that we inhabit one valuable ethical tradition, and that the sustaining of allegiances through integrity-​as-​constancy is good. By helping to support our confidence in those claims, a Williams-​style genealogy could contribute towards a Berlinian solution to the vindication problem, even it if cannot solve that problem itself. 5. Conclusion According to Williams, two important lessons that the history of moral philosophy teaches are, first, the historical contingency of ethical reflection, and secondly, the unavailability of a transhistorical, culturally uncontaminated perspective from which to establish one ethical outlook as the uniquely correct one. When we take this seriously, it threatens to destabilize our commitment to treating ethical questions as mattering in the way that we do—​as being a search for the ways of relating to each other that really are right. I think Williams is drawing our attention to something deep and important—​ one of the fundamental challenges posed by our contemporary intellectual predicament. How can we sustain our conviction in the seriousness we attribute to ethical thought and practice as we inhabit it, from the inside, while also seeing that thought as itself one local phenomenon within a broader historical perspective? My suggestion has been that the problem this poses—​the vindication problem—​ can be tamed by learning from the political thought of Isaiah Berlin. Berlin points out that, confronted with a range of competing political ideals, not all of which are co-​realizable, our decision to structure our own lives around one set of values is compatible with the recognition of others, and with the denial that there is any supreme perspective from which to attempt the optimal resolution of the forced choices we face. I want to add to this: it can also make sense for us to think that this is our political project, and that this itself constrains how we should properly proceed from here, in taking forward the valuable project in which we are embedded. I have been arguing that a parallel set of ideas is available to us in responding to Williams’s larger problem. We can accept the contingency of the framework of ethical ideas within which we are embedded, without letting go of the thought—​a thought which that framework requires of us—​that there really are right and wrong ways for us to live.34 34 This chapter has been much improved by the editorial comments and advice of Marcel van Ackeren and Matthieu Queloz. [p. 390] References Anderson, Elizabeth. 1996. ‘Reasons, Attitudes and Values’. Ethics 106: 538–​54. Benedict, Ruth. 1934. Patterns of Culture. New York: Penguin. Berlin, Isaiah. 1976. ‘Vico and the Ideal of Enlightenment’. Social Research 43: 640–​53. Berlin, Isaiah. 2013a. The Crooked Timber of Humanity: Chapters in the History of Ideas. Edited by Henry Hardy. Princeton: Princeton University Press. Berlin, Isaiah. 2013b. ‘Giambattista Vico and Cultural History’. In Berlin (2013a: 51–​72). Berlin, Isaiah. 2013c. ‘The Philosophical Ideas of Giambattista Vico’. In Berlin, Three Critics of the Enlightenment, 26–​150. Princeton: Princeton University Press, 2013. Berlin, Isaiah. 2013d. ‘The Pursuit of the Ideal’. In Berlin (2013a: 1–​20). Berlin, Isaiah. 2014. Concepts and Categories. Princeton: Princeton University Press. Boehm, Christopher. 1984. Blood Revenge: The Anthropology of Feuding in Montenegro and Other Tribal Societies. Lawrence: University Press of Kansas. Calvin, John. [1536] 1960. Institutes of the Christian Religion. Translated by Ford Lewis Battles. Louisville: Westminster John Knox Press. Cullity, Garrett. 2018. Concern, Respect, and Cooperation. Oxford: Oxford University Press. Cullity, Garrett. 2021. ‘Narrative Virtues and Second-​Order Reasons’. In Virtue, Narrative, and the Self: Explorations of Character in Philosophy of Mind and Action. Edited by Joseph Ulatowski and Liezl Van Zyl, 35–​53. New York: Routledge. Fassin, Didier, ed. 2012. A Companion to Moral Anthropology. Oxford: Wiley-​Blackwell. Fassin, Didier, and Samuel Lézé, eds. 2014. Moral Anthropology: A Critical Reader. London: Routledge. Fricker, Miranda. 2010. ‘The Relativism of Blame and Williams’s Relativism of Distance’. Proceedings of the Aristotelian Society, Supplementary Volume 84: 151–​77. Garfield, Jay. 2000. ‘Particularity and Principle: The Structure of Moral Knowledge’. In Moral Particularism. Edited by Brad Hooker and Margaret Olivia Little, 178–​204. Oxford: Clarendon Press. Gray, John. 1998. ‘Where Pluralists and Liberals Part Company’. International Journal of Philosophical Studies 6: 17–​36. Habermas, Jürgen. 1996. ‘Citizenship and National Identity’. Translated by William Rehg. In Habermas, Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy, 491–​516. Cambridge, MA: The MIT Press. Ladd, John. 1957. The Structure of a Moral Code. Cambridge, MA: Harvard University Press. MacIntyre, Alasdair. 1985. After Virtu. Second edition. London: Duckworth. McGinn, Colin. 2003. ‘Isn’t It the Truth?’ New York Review of Books 50.6: 70–​3. Moran, Richard. 2016. ‘Williams, History, and “the Impurity of Philosophy” ’. European Journal of Philosophy 24: 315–​30. Nagel, Thomas. 2021. ‘Moral Reality and Moral Progress’. In Reason, Justification, and Contractualism: Themes from Scanlon. Edited by Markus Stepanians and Michael Frauchiger, Boston, MA: de Gruyter. Nietzsche, Friedrich. [1887] 1996. On the Genealogy of Moral. Translated by Douglas Smith. Oxford: Oxford University Press. Queloz, Matthieu. 2021. The Practical Origins of Ideas: Genealogy as Conceptual Reverse-​ Engineering. Oxford: Oxford University Press. Rawls, John. 1993. Political Liberalism. New York: Columbia University Press. Schneewind, J. B. 1998. The Invention of Autonomy: A History of Modern Moral Philosophy. Cambridge: Cambridge University Press. Schneewind, J. B. 2003. ‘No Discipline, No History: The Case of Moral Philosophy’. In Moral Philosophy from Montaigne to Kant. Edited by J. B. Schneewind, 107–​26. Cambridge: Cambridge University Press. Shun, Kwong-​loi, and David B. Wong, eds. 2004. Confucian Ethics: A Comparative Study of Self, Autonomy, and Community. New York: Cambridge University Press. Stocker, Michael. 1990. Plural and Conflicting Values. Oxford: Oxford University Press. [p. 391] Garrett Cullit y Thucydides. 1989–​. The Peloponnesian War. Edited by Jeffrey S. Rusten. Cambridge: Cambridge University Press. van Ackeren, Marcel. 2018. ‘Williams (on) Doing History of Philosophy: A Case Study on Ethics and the Limits of Philosophy’. In Ethics Beyond the Limits: New Essays on Bernard Williams’ Ethics and the Limits of Philosophy. Edited by Sophie Grace Chappell and Marcel van Ackeren, New York: Routledge. Westermarck, Edward. 1912. The Origin and Development of the Moral Ideas. London: Macmillan. Williams, Bernard. 1981a. Moral Luck. Cambridge: Cambridge University Press. Williams, Bernard. 1981b. ‘Conflicts of Values’. In Williams (1981a: 71–​82). Williams, Bernard. 1981c. ‘The Truth in Relativism’. In Williams (1981a: 132–​43). Williams, Bernard. 1985. Ethics and the Limits of Philosophy. London: Collins. Williams, Bernard. 1993. Shame and Necessity. Berkeley: University of California Press. Williams, Bernard. 1995a. Making Sense of Humanity. Cambridge: Cambridge University Press. Williams, Bernard. 1995b. ‘Nietzsche’s Minimalist Moral Psychology’. In Williams (1995a: Williams, Bernard. 1995c. ‘The Point of View of the Universe: Sidgwick and the Ambitions of Ethics’. In Williams (1995a: 153–​71). Williams, Bernard. 1995d. ‘Saint-​Just’s Illusion’. In Williams (1995a: 135–​50). Williams, Bernard. 2002. Truth and Truthfulness: An Essay in Genealogy. Princeton: Princeton University Press. Williams, Bernard. 2006a. Philosophy as a Humanistic Discipline. Edited by A. W. Moore. Princeton: Princeton University Press. Williams, Bernard. 2006b. The Sense of the Past: Essays in the History of Philosophy. Edited by Myles Burnyeat. Princeton: Princeton University Press. Williams, Bernard. 2006c. ‘Descartes and the Historiography of Philosophy’. In Williams (2006b: Williams, Bernard. 2006d. ‘An Essay on Collingwood’. In Williams (2006b: 341–​58). Williams, Bernard. 2006e. ‘The Legacy of Greek Philosophy’. In Williams (2006b: 3–​48). Williams, Bernard. 2006f. ‘Philosophy as a Humanistic Discipline’. In Williams (2006a: 180–​99). Williams, Bernard. 2007a. In the Beginning Was the Deed. Edited by Geoffrey Hawthorn. Princeton: Princeton University Press, 2007. Williams, Bernard. 2007b. ‘Human Rights and Relativism’. In Williams (2007a: 62–​74). Williams, Bernard. 2007c. ‘In the Beginning Was the Deed’. In Williams (2007a: 18–​28). Williams, Bernard. 2007d. ‘Pluralism, Community and Left Wittgensteinianism’. In Williams Wong, David B. 2009. Natural Moralities: A Defense of Pluralistic Relativism. New York: Oxford University Press. Zigon, Jarrett. 2008. Morality: An Anthropological Perspective. Oxford: Berg. [p. 394] Serpents in the Genealogical Garden of Eden Why Williams’s Genealogy Is Excessively Historicist and Insufficiently Historical Hans-​Johann Glock Je weniger die Leute davon wissen, wie Würste und Gesetze gemacht werden, desto besser schlafen sie. The less people know about how sausages and laws are made, the sounder they sleep. Otto Fürst von Bismarck 1. Introduction: Historicism and Genealogy Since the 1980s there has been a growing chorus from within analytic philosophy deploring its historiophobia, the alleged tendency of most analytic philosophers to ignore or despise the past (see e.g. Rorty, Schneewind, and Skinner 1984; van Ackeren 2018). Bernard Williams was a distinguished voice in that chorus. In his later writings, he persistently urged philosophy not just to pay attention to its own history, but to adopt a more historical and genetic perspective in general (Williams b). One can distinguish three positions within this new historicist trend. According to intrinsic historicism, proper philosophy is ipso facto historical as regards both its aims and methods. According to instrumental historicism, studying the past is necessary, yet only as a means to achieving ends which themselves are not historical in nature. And according to weak historicism, a study of the past is useful to the pursuit of substantive philosophy, without being indispensable. Elsewhere I have argued at length for the following verdict concerning these positions (Glock b: ch. 4; 2008a; 2018). First, intrinsic historicism is simply wrong; secondly, the case for instrumental historicism remains unproven at best; thirdly, pragmatic historicism is well supported and substantially correct. While there are drawbacks Hans-​Johann Glock, Serpents in the Genealogical Garden of Eden In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. [p. 395] Hans-Johann Glock to lavishing attention on the past (as pointed out by thinkers as diverse as Kant, Schopenhauer, Nietzsche, and Wittgenstein), these are by and large outweighed by the benefits. This dialectic constellation is typical in philosophy, and even more so in metaphilosophy. The strongest arguments are those in favour of the most modest positions. In previous contributions to this debate I have also resisted Williams’s case for instrumental historicism, which revolves around the idea of genealogy.1 This chapter is devoted to a distinct though related topic. The main issue is not so much whether genealogy provides arguments demonstrating the indispensability of historiography (whether history of philosophy or history of ideas) to substantive philosophizing (but see section 7). It is rather the scope and limits of the contribution that genealogical methods can make in philosophy and beyond. At the same time, this chapter is both an exercise in and a defence of the method of conceptual analysis, which most current genealogists regard as obsolete or at best as a stepping stone towards a higher intellectual plane. It must be kept in mind, however, that the kind of analysis propagated here is not the Aunt Sally that some avant-​garde metaphilosophers like to take pot shots at. It is connective rather than reductive, avoids pretences to being infallible, is pragmatist in looking at the overall place of concepts in our practices, and is impure in allowing for a close and dynamic interaction between philosophical analysis and the special sciences—​natural and social.2 In fact, in some respects it is more open not just to natural science but also to real historiography than Williams’s genealogy, or so I argue in sections 13–​ Regarding several of these modifications of allegedly classic conceptual analysis, my impure version is reasonably close to the ‘pragmatic genealogy’ which Queloz (2021) has presented in a clear and forceful manner. I have nevertheless refrained from engaging with this type of genealogy here. And not just for reasons of space. Having strayed slightly from the title of this collection by concentrating on metaphilosophical issues about genealogy rather than on history of philosophy, I decided to keep the protagonist firmly in view. 2. An Aside on Methodological Pluralism Whatever the merits of genealogy turn out to be, it should be clear that other aspirations and methods—​logical and conceptual analysis, methodological reflection, critical thinking, inference to the best explanation, pursing reflective equilibria, etc.—​have a role to play in proper philosophizing. Furthermore, those roles are not 1 For a more positive assessment, see van Ackeren (2019); and for a response to my stance on historicism, see Renz (2018). 2 See Glock (2017). For a defence of such analysis against the animadversions of conceptual engineers—​many of them shared by genealogists—​see Glock (2025). [p. 396] just subordinate to currently fashionable enterprises like genealogy, experimental philosophy, and conceptual engineering. Such a pluralistic outlook is in line with Williams: The starting point of philosophy is that we do not understand ourselves well enough. . . . Philosophy’s methods of helping us to understand ourselves involve reflecting on the concepts we use, the modes in which we think about these various things [nature, ethics, politics]; and it sometimes proposes better ways of doing this. (2002b: 7) In this passage he assigns a single goal to philosophy, namely self-​understanding; yet in the same breath he speaks of a plurality of its methods. And elsewhere he emphatically embraces the view that philosophy is ‘very various’ (2000: 478–​9; see also below, section 13). Methodological pluralism is at odds with methodological monists of various ilks. At present, the most vocal monistic faction consists of those single-​minded conceptual engineers who promote their own pet project as a new ‘first philosophy’, or even as providing the foundation of all other intellectual efforts.3 Such claims are misguided, and not just regarding conceptual engineering. Whatever else philosophy as an academic discipline is or should be, it is surely a supremely fundamental type of enquiry, with a completely unlimited scope. Any topic, enquiry, and form of discourse can engender philosophical questions. Moreover, at any point it is not just answers that are subject to critical scrutiny; it is also the methods for supplying them, as well as the questions these methods mean to address, as well as the methods for determining which questions should be posed and how, etc. This universally critical stance on even the most basic issues militates against the hope of starting out from a single, non-​optional, inexorable, and indubitable method. Instead, philosophy has to look at problems from highly diverse fields, and needs to tackle these from very different angles. For this reason, philosophy has to be able to purse a variety of aims, to take recourse freely to a multitude of methods, and to draw on a wide range of disciplines and intellectual tools. 3. Preview In consequence, my question here is not whether all philosophizing should be based on or at least include genealogy and related historical modes of thought. 3 Thus, Cappelen writes: ‘[C]‌onceptual engineering should be considered one of the central topics of philosophy, or perhaps even the central topic of philosophy’ (2018: 1). Even more emphatically, Eklund insists that ‘Philosophy should rather be thought of as conceptual engineering’ (2015: 364). And Cappelen (2022) opines that conceptual engineering is ‘foundational to all inquiry, and in particular to all the sciences’. For a critique of such pretensions, see Glock (2025). [p. 397] Hans-Johann Glock The aim is to assess the ambitious claims that have been made by Williams and his followers on behalf of the merits and scope of these methods. I shall start by characterizing what I call his ‘analytic genealogy’ through contrasting it with related yet distinct positions (section 4). Next, I provide a thumbnail sketch of the methodological parts of Williams’s most sustained contribution to this enterprise, Truth and Truthfulness (section 5), in particular of the way in which it combines ‘imaginary genealogy’—​possible explanations of developments from a postulated State of Nature—​with ‘historical genealogy’. Section 6 points out that Williams’s genealogy aims not just to explain the emergence of certain concepts and practices, and to assess their validity, but also to explain their content—​what the relevant expressions mean and what the practices revolving around them amount to. It is at this point that my essay touches on the aforementioned question of whether philosophy needs history (section 7). Williams’s argument for holding that a philosophical clarification of our conceptual framework requires looking at the history of concepts invokes analytic genealogy. This stance faces the challenge of committing a ‘genetic fallacy’. However, I argue in section 8, one needs to distinguish a fallacy committed by inferring the truth-​value or warrant of a statement from its origin—​a ‘genetic fallacy’ proper—​from a fallacy committed in deriving the meaning of an expression—​determining the concept it expresses—​from its origin; what I call a ‘genealogical fallacy’. Section 9 then criticizes Williams’s defence against both of these charges. Concerning the genealogical fallacy, he refuses to acknowledge that considerations looking at the current role and function of concepts can provide all the elucidation we need for philosophical purposes. Indeed, his own genealogical tale ultimately rests on functional considerations concerning a postulated State of Nature, even though he officially regards functional accounts as inadequate or insufficient. Section 10 looks at different theories of functions from the philosophy of biology and at how they might apply to genealogical accounts of conceptual frameworks. On that basis, section 11 rejects Williams’s objections to ‘presentist’ functionalism, and section 12 considers a critique of etiological functionalism that Williams develops out of considerations by Brandom. That critique targets attempts to explain the emergence of practices and concepts through functional considerations within theories of biological and/​or cultural evolution. Its crux is that the emergence of human language and culture totally transforms any of the needs that etiological functionalism invokes in explaining the phylogenesis of our current practices and concepts in biological and cultural evolution. But that contention rests on a ‘transformational model’ of mental abilities and intelligent practices that is untenable. Section 13 critically scrutinizes the way in which Truth and Truthfulness combines imaginary and historical genealogy, contrasting it unfavourably with the procedure of evolutionary anthropology dismissed by Williams and other genealogists. The general trajectory of this chapter is to put the exciting project of analytic genealogy à la Williams in perspective. It argues that non-​genealogical philosophical [p. 398] methods, on the one hand, and non-​philosophical theories about human nature and the evolutionary genesis of practices, on the other, are needed to supplement and control genealogy. Regarding the first point, it balances the promises of genealogy against the abiding merits of logical and conceptual analysis. Regarding the second point, it contends that recent advances in evolutionary theory and biological anthropology hold the promise of replacing fictions concerning the origins of some of the practices, values, and virtues discussed in genealogies by scientific hypotheses that are testable at least in principle. Section 14 ends on a more positive note. Provided that genealogical considerations concerning the content, validity, and emergence of concepts and practices are treated as heuristic devices that interact with other types of philosophical analysis and anthropological research, they have a valuable contribution to make in philosophy and beyond. 4. Analytic Genealogy Williams has been the driving force behind a distinctive movement that deserves the label ‘analytic genealogy’. That genre, as understood here, is part of a wider and somewhat older trend: namely, analytic (neo-​)Nietzscheanism (e.g. Danto 1965; Leiter 2002). But whereas many analytic Nietzscheans are enthused by their idol’s amoralism, naturalism, or alleged relativism, analytic genealogists are primarily concerned with pursuing his genealogical method. At the same time, they do so in a mode that differs from the way in which Foucault and his followers took forward Nietzsche’s genealogy. For one thing, they harbour ambitions that are almost exclusively philosophical rather than partly historiographical or sociological.4 For another, they pursue these philosophical aims in a way that attempts to be as clear in exposition and rigorous in argumentation as the topic permits.5 For a third, they invoke not just historiography but also fictional scenarios imagined to have obtained in the past. Williams’s turn to genealogy was prompted in part by Edward Craig’s Knowledge and the State of Nature of 1990 (see Queloz 2021: 4–​8). Nevertheless, Craig is not an analytic genealogist as here understood. He is influenced by Hume rather than Nietzsche. Furthermore, unlike Williams, he does not combine reflections on a State of Nature with claims about actual historical developments. In fact, he 4 This is not to deny that Foucauldian genealogy was intended to and can have philosophical implications, notably by destabilizing adherence to certain practices and concepts through pointing out their contingency (see Queloz 2021: 6n12). But it is in the main historical-​cum-​sociological rather than engaging in direct philosophical argument concerning normative issues. With apologies to the Tractatus, one might say that its ethical implications are ‘shown’ rather than stated. 5 For a characteristically dense yet trenchant defence of such analytic ideals, see Williams (1985: vi). [p. 399] Hans-Johann Glock pursues a functional analysis revealing ‘the core of the concept as it is to be found now’ (Craig 2007: 191; see also section 11 below). Something similar holds of a relative of analytic genealogy that is definitely not among its actual roots. De Mesel (2024) reveals intriguing similarities between Peter Strawson’s seminal defence of moral responsibility by reference to reactive attitudes, on the one hand, and contemporary genealogy, on the other. What De Mesel does not mention is that in early methodological writings Strawson even gave a label to this enterprise: namely, ‘explanatory metaphysics’ (see Glock 2012). It explains ‘not just how our concepts and types of discourse operate, but why it is that we have such concepts and types of discourse as we do; and what alternatives there might be’. This enterprise investigates the ‘natural foundations’ of our ‘conceptual apparatus in the way things happen in the world, and in our own natures’. And it considers counterfactual conditionals about how our conceptual scheme might change given different empirical conditions (Strawson 1956: 108; Unlike Strawson’s more famous project of descriptive metaphysics, explanatory metaphysics does not rest content with describing ‘how our concepts and types of discourse operate’. At the same time, like Craig’s State of Nature approach, it ‘is not a historical inquiry’. But by contrast to Craig, Strawson envisages an inquiry into the natural foundations of our conceptual apparatus that does not specifically target the latter’s functions but includes all of the framework conditions that render that apparatus practicable. In this respect it is more in line with Wittgenstein’s appeal to ‘natural history’, which seeks to alert us to the possibility of alternative conceptual schemes given different counterfactual circumstances.6 To sum up, what distinguishes Williams’s analytic genealogy is the combination of Nietzsche’s original project with philosophical ambitions, arguments in an analytic style, and the combination of historical claims with State of Nature speculations. 5. Truth and Truthfulness Williams’s most substantial contribution to analytic genealogy is Truth and Truthfulness, which appeared in 2002, shortly before his death. In it, Williams takes issue with what he calls the ‘deniers’, postmodern thinkers who cast aspersions on the possibility and/​or the relevance of objective truth. His prime targets here are the happy-​go-​lucky neo-​pragmatism of Richard Rorty and the lugubrious deconstructivism of Jacques Derrida. Against them, Williams insists that any recognizably human society must accept both truth and truthfulness as values, and 6 That there are connections between pragmatic genealogy, Strawson, and Wittgenstein is noted by Queloz (2021: 19). But he does not draw a connection to Strawson’s ‘Freedom and Resentment’, and, like De Mesel, he does not mention Strawson’s label ‘explanatory metaphysics’. [p. 400] sincerity and accuracy as corresponding virtues. At the same time, Williams follows Nietzsche in accusing rationalist attempts to defend these values of tacitly relying on disreputable metaphysical ideas such as theism, Platonism, and Kantian transcendentalism. This is where the idea of genealogy comes in. Williams disputes interpretations that present Nietzsche as the fountainhead of denialism—​rightly, in my view (see also Bouveresse 2021). Contrary to postmodern interpreters, and occasional lapses notwithstanding, Nietzsche is not the spiritus rector of denialism. As Williams points out: while Nietzsche was convinced that a defence of the ideas of truth and truthfulness has to be naturalistic, he also hoped to provide such a defence through what Williams calls a ‘vindicatory’ genealogy (2002a: 36). Williams likewise attempts to provide such a vindicatory genealogy. According to him, a genealogy is a ‘narrative that tries to explain a cultural phenomenon by describing a way in which it came about, or could have come about, or might be imagined to have come about’, given different circumstances a: 20). The inclusion of the last two disjuncts distances his genealogy from both Nietzsche and Foucault. Williams distinguishes a type of genealogy that purports to be ‘historically true’ from ‘imaginary genealogy’ (2002a: 20–​2, 31–​ His own genealogy combines the two. The first component, historical or, more precisely, historiographical genealogy, raises no special methodological problems over and beyond those familiar to historians of ideas. What is contentious is its philosophical relevance (see note 4 above). But as regards imaginary genealogy, the question inevitably arises, ‘how a fictional narrative can explain anything’ (2002a: 21). That question is even more acute in view of the fact that explanation is factive. X explains Y only if both X and Y are real entities (objects, events) or actual facts. Williams’s answer starts out from the idea of a ‘potential explanation’ introduced by Robert Nozick (1974: 7–​9), an explanation that ‘would be the correct explanation if everything in it were true and operated’. A potential explanation must not be ‘law-​defective’: that is, based on mistaken claims about the laws governing the area at issue. But it can be fact-​defective: that is, operate with false statements of antecedent conditions. In that case, it can nonetheless show that a certain genetic process is possible, even if its antecedent conditions fail to obtain. Potential explanations thereby provide for both the modal and the counterfactual claims that are included in Williams’s wide definition of genealogy. A final ingredient of the imaginary part of Williams’s genealogy is its appeal to a ‘State of Nature’ (2002a: ch. This is a well-​worn ploy of western philosophy from Hobbes to Rawls. At the same time, it distances Williams even further from the ideals, though not necessarily the execution, of echt Nietzschean historical genealogies. A State of Nature is not localized in either space or time, it is neither here nor there, neither now nor then. It is the condition of a fictional human community that is postulated on the basis of prima facie plausible assumptions about human nature. Because of that nature, the community is confronted with certain challenges which, according to [p. 401] Hans-Johann Glock the genealogical tale, are met through the emergence of certain dispositions, practices, institutions, and ideas. Williams recognizes that Nozick’s distinction between fact-​and law-​defective explanations is difficult to apply to a State of Nature genealogy, since it is unclear whether a suitably austere initial scenario is even anthropologically possible.7 But he nonetheless sets out on his imaginary genealogy by considering ‘a small society of human beings, sharing a common language, with no elaborate technology and no form of writing’ (2002a: 41). The members of this society need to share information about their immediate environment if they are to satisfy even their most basic needs. They then begin to cultivate the dispositions that make for good contributors to the pool of information: namely, dispositions of accuracy and sincerity, the two principal components of truthfulness according to Williams. To begin with, these dispositions are valued merely instrumentally, as means to the effective sharing of information. In due course, however, they are cherished for their own sake—​as virtues. But these two virtues are still remote from what we mean today by accuracy and sincerity, both in having a more restricted scope and in being less demanding. In order to explain the emergence of fully evolved truthfulness, Williams takes recourse to historical genealogy, paying flying historiographical visits to ancient Greece, the Romantic period, and the rise of liberalism.8 6. Explanation, Justification, and Clarification In line with all analytic genealogists I know of, Williams is convinced that the actual, prehistorical genesis of notions like truth and truthfulness cannot be recovered as a matter of principle. Given this assumption (questioned in section 13 below), it makes at least prima facie good sense to appeal to imaginary genealogy. Law-​faithful though possibly fact-​defective potential explanations starting from an anthropologically plausible hypothesis of a State of Nature seem to be the most promising and perhaps even sole pathway to insights into how notions such as truth and truthfulness did in fact arise. Furthermore, according to Williams, by furnishing such an explanation, genealogical stories can also be ‘vindicatory’. They can lead us to appreciate that the values and virtues of truth and truthfulness can and should be sustained in contemporary societies (2002a: 36, 269). That Williams and analytic genealogists more generally combine explanatory with justificatory ambitions has been widely acknowledged (Queloz ch. 1 and the works cited there). What has gone largely unnoticed is that in his metaphilosophical writings Williams puts genealogy primarily in the service 7 See Williams (2002a: 32–​3). Many thanks to Matthieu Queloz for alerting me to this passage. 8 My sketch omits important parts of the book, notably reflections on the connections between truth, belief, assertion, and sincerity (Williams 2002a: chs 4–​7) that are largely independent of genealogy. [p. 402] of a third enterprise. It aims to clarify our contemporary modes of thought, and thereby in effect our conceptual framework, particularly those parts of it that give rise to philosophical puzzlements. This is part and parcel of the attempt to reach the special kind of self-​understanding that Williams drives at in the passage quoted in section 2. In this spirit, he proposes to use genealogy to establish the content of presently contested concepts, and of how they are related not just to previous concepts but also to each other. Now, such clarification logically and methodologically precedes both explanation and justification. Neither explaining the emergence of a concept nor justifying its employment makes sense unless it is at least approximately clear what concept is at issue in the first place (see Glock 2017). For this reason, I start by critically discussing Williams’s case for historicizing conceptual clarification. This case combines his general plea for doing philosophy more historically with his specific promotion of the genealogical method. 7. History and Conceptual Self-​Understanding According to Williams, even more baneful than the neglect of the history of philosophy within analytic philosophy has been the neglect of ‘the history of the concepts which philosophy is trying to understand’ (2002b: 7). His guiding idea is that articulating our framework of concepts presupposes knowledge of its history. This position makes conceptual clarification dependent not so much on the history of philosophy as on the entire history of ideas, and perhaps on history in general, depending on what forces shape our concepts. But how can it be sustained, given that the philosophical problems we currently confront have their immediate roots in the present framework? And even if we stick to the more general brief of conceptual self-​reflection, it is our concepts that we want to understand. In the essay ‘Why Philosophy Needs History’ (2002b), Williams’s answer runs as follows. In the case of scientific concepts like that of an atom, the question of whether the same or a different concept is employed in different epochs and cultures does not matter much to ‘what may puzzle us about that concept now (for much the same reason that the history of science is not part of science)’. Unfortunately, Williams does not divulge this reason. And given what we have learnt from Kuhn (1962) about scientific revolutions and the paradigm shifts underlying them, it is difficult to see why philosophical problems concerning scientific concepts should be less sensitive to conceptual variations than philosophical problems concerning non-​scientific concepts. Indeed, Williams’s example ‘atom’ is a case in point, and even more so concepts relating to space and time. Philosophical puzzles about ultimate physical constituents and about space/​time have been completely transformed through the revolution of twentieth-​century physics. Thus, ever since Einstein, no philosophical treatment of space and time could get by [p. 403] Hans-Johann Glock without considering the relations between common-​sense, classical, and relativistic concepts (see Maudlin 2015). Be that as it may, Williams argues that the question of whether the same concept is employed in different settings does matter for some philosophically contested concepts: namely, those intimately tied to human interaction and communication—​concepts like freedom, justice, truth, and sincerity. In these cases, he insists, it is imperative to appreciate that their historical variants represent ‘different interpretations’ of a ‘common core’. We may be able to understand that core through functionalist reflections on the role these concepts fulfil in satisfying the demands of human life, as in fictions of a ‘State of Nature’ which purport to explain the emergence of morality, language, or the state. ‘But the State of Nature story already implies that there must be a further, real and historically dense story to be told.’ This is why we need a genealogy, a ‘method that combines a representation of universal requirements through the fiction of a State of Nature with an account of real historical development’ (2002b: 7).9 8. Genetic Fallacies and Genealogical Fallacies This claim, like genealogical modes of thought more generally, runs up against a distinct trend within modern philosophy. Kant’s distinction between quaestio facti and quaestio iuris and the ensuing neo-​K antian distinction between genesis and validity fuelled a pervasive, though largely implicit, suspicion of the so-​called ‘genetic fallacy’, the mistake of deducing claims about the validity of a theory or the content of a concept from information about the causes for its emergence, particularly its historical origins. Thus, Frege granted that ‘the historical perspective’ has a certain justification, while insisting that one cannot divine the nature of numbers from psychological investigations into the way in which our thinking about numbers evolved (1884: Introduction). 9 Chappell and Smyth (2023) intimate another argument in favour of the historical-​ cum​genealogical approach to ethics promoted by Williams. It starts out from the latter’s contention that there can only be ‘internal’ and no ‘external’ reasons for agents to act. All statements of the form ‘A has reason to φ’ imply that A has a motive that will be served by his φing. According to them, this ‘internal reasons thesis’ demonstrates the impossibility of impersonal reasons of the kind required by what Williams denigrates as the ‘morality system’, thereby making room for genealogy. They are alive to the possibility that Kantian moral theorizing may after all be compatible with the internal reasons thesis. But the step from abandoning general moral theorizing to genealogy is equally precarious, for example because of the option of moral particularism. Finally, the internal reasons thesis itself is based on a failure to appreciate a subtle trichotomy that Williams himself helped to initiate. ‘Justifying reasons’ are reasons for an agent to φ: that is, considerations that actually favour her acting in this way. ‘Motivating reasons’ are reasons for which an agent φs: considerations which make φing attractive from her point of view. Explanatory reasons are reasons why the agent acts in a certain way: that is, reasons which explain her acting (see Alvarez 2016). The internal reasons theory holds for motivating reasons: reasons for which A φs. But to insist that there can be no reasons for A to φ that are independent of A’s desires is to beg the question against externalism. [p. 404] In line with much contemporary genealogy, Kant’s distinction intertwines two claims. One concerns the content of concepts;10 the other concerns the justification of employing them in the formulation of judgements, including judgements concerning the legitimacy of concepts. But it is one thing to invoke the genesis of a concept in order to get clear about its content, and another to invoke it in order to ascertain its legitimacy. To keep the two apart, I stipulate that a genetic fallacy concerns what Kant would call the ‘validity’ of a concept: that is, the question of whether it is legitimate or mandatory to employ a concept in a certain area (in Kant’s case, the realm of ‘possible experience’). By contrast, a genealogical fallacy concerns the question of what precisely the concept is and what distinguishes it from other concepts—​the proper analysis of the concept through clarifying the words expressing it. It is this fallacy that Frege deplores in the course of promoting his own logico-​semantic attempt to define the concept of a number. Now, the two issues are obviously closely related. And they are persistently treated in tandem not just by Williams but also by those following in his wake, since for them specifying the content of the concept is a prelude for pronouncing on its legitimacy. Most of what follows concerns the specification of concepts and the explanation of their emergence. Nevertheless, three comments are appropriate regarding both genealogical and genetic fallacies in my narrower sense. First, the charge of committing a genetic fallacy targets the attempt to derive categorical verdicts on the truth-​value or justification of a claim from premises that exclusively concern the latter’s origin. That kind of inference is fallacious, independently of whether it purports to be logical, conceptual, or empirical. The truth-​ value of a judgement (statement, proposition) depends exclusively on how things are concerning the topic of the judgement. Therefore, it cannot depend on the origins of the judgement (the causes of its being made), except perhaps for an exotic class of self-​referential judgements like ‘What I am saying just now is false.’ Secondly, its fallacious nature is often disguised by a manoeuvre popular among thinkers who set more store by genesis than is legitimate, a category that includes some though not all genealogists, virtue epistemologists, and proponents of pragmatist theories of truth. The sleight of hand consists in describing the origins in terms that bear a conceptual connection to their truth-​value and/​or epistemic standing—​‘reliable methods’, ‘trustworthy source’, ‘ideal speech situation’, ‘expert witness’, ‘agreed upon by scientists at the end of all inquiry’. In the case of genetic explanations of concepts like truth and knowledge, genealogical and genetic thinking interact to create definitions that face a dilemma: either they evince a genetic fallacy, by assuming that the origin of a claim to truth and 10 Since concepts are themselves (semantic) contents rather than symbols having contents, it would be more felicitous to speak of the meanings of concept-​expressions. I shall disregard this complication in the sequel. [p. 405] Hans-Johann Glock knowledge directly determines its validity, or they tacitly smuggle the definiendum (e.g. being true) into the allegedly defining characterization of the genesis. Thirdly, this uncompromising stance on genetic fallacies leaves open two less apodictive ways of drawing on the origins of a claim in assessing its merits. For one thing, probabilistic assessments: one can regard certain kinds of origin as rendering it more or less probable that a claim is true or warranted. This concession to genetic modes of thinking is required in order to accommodate knowledge by testimony. For another, while facts about the origins of a claim cannot provide good reasons for simply altering one’s assessment—​for example, from the claim being true to its being false—​they can provide reasons for reconsidering it. This concession is required by a proper appreciation of ‘peer disagreement’, or so I have argued elsewhere (see Glock 2016). 9. Williams on Genetic Fallacies Williams defends genealogy against the accusation of relying on a genetic fallacy. According to him, this charge ‘overlooks the possibility that the value in question may understand itself and present itself and claim authority for itself in terms which the genealogical story can undermine’. Thus, liberal conceptions of morality ‘claimed to be the expression of a spirit that was higher, purer and more closely associated with reason, as well as transcending negative passions such as resentment’, and hence a genealogy is capable of displaying them as ‘self-​deceived in this If Williams is right, one reason why history is indispensable to philosophy is that the genesis of certain concepts or beliefs is crucial to both their content and their validity. That is to say, what such concepts or beliefs amount to and whether they are legitimate will depend on the source from which they derive. Even then, however, the basic idea of a genetic fallacy still stands. All Williams has shown is this: if a practice, belief, or mode of thought defines or justifies itself in terms of a particular origin, then that origin becomes relevant to its analysis and justification. The reason is not that there is after all no distinction between genesis, on the one hand, and content or validity, on the other. Participants in the Catholic practice of ordination, for instance, defend it by reference to the idea of apostolic succession, and hence to a particular origin. And something similar may hold of liberalism. In other cases, the genesis of a practice provides a reason for or against it even if it is not actually adduced: for example, when a legal norm has not been adopted through proper procedures. Yet the investigation into either the actual or the best possible reasons for accepting a practice, concept, or norm is not per se genetic; it merely takes on a genetic aspect in specific cases. Let us turn from genetic fallacies to genealogical fallacies (in my dichotomy). Concepts like that of a sunburn or of lava are genetic in that they apply only to [p. 406] things with a certain origin. Even in these cases, however, it is not the history of the concept itself which is part of its content, but the history of its instances. By the same token, to elucidate that content, philosophers only need to note that historical dimension; unlike empirical scientists who apply such concepts, they do not need to examine the actual origin of potential candidates. Furthermore, it is the status quo alone which determines whether a given concept is genetic or whether the actual or optimal justification of a belief or practice invokes its origins. Even if liberal morality originally laid claim to superior breeding, this entails neither that its current proponents justify it in this manner, nor that this is the best possible justification. If neither of these options holds, as seems plausible, genealogy will be immaterial to the philosophical merits of liberal morality. And whether they hold does not depend on the historical origins of liberal morality. The issue is of more than academic interest. Recent controversies over postcolonialism have demonstrated that many liberal ideas and ideals arose in contexts which were anything but savoury. This historical fact is, of course, central to assessing liberalism as a political and cultural movement. If I am right, however, it does not per se discredit the epistemic and moral credentials, such as they are, of liberal ideals like universalism and autonomy. Explanations and Functions Having dealt with elucidation and justification, we now broach the explanation of concepts and discursive practices. Remember that a genealogy is a ‘narrative that tries to explain a cultural phenomenon by describing a way in which it came about, or could have come about, or might be imagined to have come about’ (Williams a: 20). By including potential and conceivable origins, Williams links up with functional accounts. The latter explain or justify a phenomenon not by tracing its actual genesis, but by pointing out that it serves a specific role in an actual or fictional practice. As Williams realizes, however, a functional explanation is not per se genetic. It is one thing to know the function of an organ; another to know its evolutionary emergence. Similarly, one can reflect on the function of our concept of knowledge (Hanfling 1985; Craig 1990), without speculating about its origins. What counts is the current role which the concept has. At this point, it is helpful to distinguish different types of functions and functionalist theories. This is, of course, an enormous topic. All I can do here is to bring to bear a classification widely employed in the philosophy of biology on the discussion of functions within genealogy.11 11 See, for example, Garson (2008). My discussion of different types of function in biology is indebted to discussions with Aneta Zuber (2025) and to discussions with its author. Queloz (2021: 221–​7) distinguishes between etiological and presentist functionalism, and his own ‘needs-​based’ account is an instance of forward-​looking functionalism. [p. 407] Hans-Johann Glock First, backward-​looking or ‘etiological’ accounts identify the function of a trait with what it has been selected to do. Roughly, the function of a trait is the set of those of its causal effects that have made a difference to the trait having been passed down the evolutionary line to the organism that currently displays the trait. Next, presentist theories define the function of a trait as the overall causal role it plays within a system at present. Finally, forward-​looking or ‘consequentialist’ theories focus on the causal effects that the trait has on the system’s future, such as its consequences for the system’s fitness or well-​being or the satisfaction of its goals. Let us now transpose this trichotomy from biology’s concern with the functions of traits in or for an organism to genealogy’s concern with the functions of dispositions, practices, concepts, and beliefs for a human community. This yields the following types of function: The backward-​looking or etiological functions of a trait (practice, concept, belief, etc.) is what it has been selected for to do. Put differently, it is the set of those of its contributions to the survival and flourishing of the community which explains the trait’s persistence. The present function is the contribution of a trait to the system now. It is determined by the trait’s current causal, epistemic, or communicative role(s) within the community. The forward-​looking or consequentialist function of a trait is the effect it has on the community’s future, notably on its capacity to sustain itself across future generations, or to satisfy its future needs and goals. Williams on Presentist Functionalism Williams does not reckon with distinctions of the kind drawn in the previous section. Nevertheless, it is fruitful to look at his writings through that lens. As indicated in section 7, ‘Why Philosophy Needs History’ treats what are in effect presentist functions as starting points. In comparison, Truth and Truthfulness strikes a more critical chord. It takes issue with both etiological and presentist explanations of our ‘dispositions of truthfulness’, albeit on different grounds. Williams contends that purely functional accounts miss the point of these dispositions, because ‘their value always and necessarily goes beyond their function’. This in turn is due to the fact that their participants are rational agents who have their own reasons for engaging in them (2002a: 34–​5). These complaints seem to be directed against presentist accounts. But such accounts can easily allow that the current function of a trait is not confined to by-​and-​large unconscious chains of mechanical causes, but includes epistemic and communicative roles that enquirers and speakers can become aware of, at least on reflection. Indeed, that is precisely the level at which functional analyses like those of knowledge by Hanfling and Craig operate. [p. 408] What is more, Williams’s observation that we are dealing with communities of rational agents does not suggest a need to turn to the past (imaginary and/​or actual). It rather suggests that a philosophical understanding of a practice and conceptual framework must include more than looking at the causes and effects of linguistic practices, notably the way in which the agents themselves would or could explain what they mean by expressions or what the practice amounts to, and how they would or could justify its pursuit if challenged. This is a much more compelling recognition of the rational status of the linguistically competent protagonists than imputing to them concepts and concerns allegedly determined by a past of which most if not all of them will be ignorant. For this reason, Williams’s observation does not entail that the functional explanation must be temporalized by looking at the genesis of either the concepts, or the practices that give them point, or the agents that sustain them. It points in the direction not of diachronic speculations but of synchronic interrogations. Revealingly, Williams’s own purportedly genealogical vindication of the virtues of truthfulness does not presuppose any history, actual or invented.12 To be sure, he considers a State of Nature involving a fictional society with primitive speakers. But the net justificatory yield of the exercise is that a practice of acquiring true beliefs and sharing them sincerely with others is advantageous to rational social creatures, since it allows them to pool information that is not directly available to any one individual. Williams strives hard to take this line of reasoning beyond a purely utilitarian defence of an instrumental value. He insists that the beneficial practice of sharing information would be unstable unless its participants regarded accuracy and sincerity as good in their own rights. To this end, he enriches the functional story by considering further aspects of the context of the practice, as well as potential threats to it. But the vindication relies purely on what it would be rational for creatures with human requirements, capacities, and limitations to do within various scenarios. It does not depend on how either the creatures or the scenarios emerged. The philosophical case of Truth and Truthfulness is at core anthropological-​cum-​epistemological rather than historical. Williams may be right to contend that certain specific discursive practices are constituted by or should be based on genetic justifications that invite historical scrutiny (see section 9 above). Yet he fails to provide a general reason why any philosophical reflection on a concept or belief should require either a historical or a fictional genealogical account of its emergence. 12 Looking beyond Truth and Truthfulness, Williams does not defend integrity, one of the virtues of truthfulness, through any kind of genealogy. Rather, in the course of his celebrated attack on utilitarianism, he argues that actions are always performed by a particular agent whose practical deliberations are first-​personal, and cannot, by pain of compromising integrity, be dictated by an abstract theoretical system (Smart and Williams 1973: 115–​18). [p. 409] Hans-Johann Glock Williams and Brandom on Etiological Functionalism In the previous section, I cast doubt on the methodological conclusions Williams draws from his critique of presentist functional accounts. That critique is also problematic in its own right. It is targeted at attempts to explain morality, or ‘living in an ethical system’—​that is, ‘under rules and values’ (2002a: 24)—​through functionalist accounts, without a genealogy relying partly on fictional stories. Such accounts, Williams informs us ‘are simply false’. In support of this contention, he cites Brandom: Linguistic practice is not for something, . . . [it is not] a means to secure some other end specifiable in advance of engaging in linguistic practice—​not adaptation to the environment, survival, reproduction, nor co-​operation—​though it may serve to promote those ends. Even if those functions explain why we came to have language, once we did have it, our transformation into discursive creatures swept all such considerations aside. For discursive practice is a mighty engine for the envisaging and engendering of new ends. (Brandom 2000: 363; his emphasis; quoted in Williams 2002a: 34–​5) Brandom’s conclusion would rule out etiological accounts of the functions of human capacities, not just in genealogy but in evolutionary anthropology as well. For it excludes the possibility of the concepts, beliefs, and practices of a linguistic community being identical with ancestral phenomena in a non-​linguistic community, and thereby the possibility of a trait at the ancestral state being selected so as to be preserved at the present state. The function of a disposition to cooperate in a hominin group bereft of language cannot explain the presence of such a disposition in Homo sapiens, for instance, since the two cannot possibly be the same. Fortunately, that line of reasoning evinces a non sequitur. From the fact that language enables the acquisition of additional novel ends and needs, it simply does not follow that the functions fulfilled by its precursor forms of interaction and communication have simply been swept aside. This would follow only on the assumption of what has been called a ‘transformational theory of the human mind’. This theory goes back to Aristotle and Aquinas, and it has been promoted by Pittsburgh philosophers like Brandom (e.g. Boyle cp. Glock 2022). Its proponents reject the prima facie plausible ‘additive theory’ according to which we share with non-​human animals and our closest evolutionary ancestors certain basic mental capacities and forms of behaviour—​such as perception, reproduction and cooperation—​notwithstanding the fact that these have been supplemented by distinctively human rational and linguistic powers and practices. That is wrong, transformationalists aver, since the simpler capacities undergo a categorial shift when combined with linguistic ones. As a result, animal perception, reproduction, and cooperation are at best analogous to human [p. 410] reproduction and cooperation, respectively. For, if the transformational view is to be trusted, it is partly constitutive of the human versions of these abilities that they are embedded in language, and partly constitutive of their animal versions that they are not. Intriguing though the transformative model is, it does not stand up to closer scrutiny. Even if there is a difference in the way in which humans and animals perceive, procreate, or cooperate, it does not follow that animals cannot do some of the same things that we do. Nor does it follow that animals only perceive etc. in an attenuated sense rather than literally speaking. Neither cognitive nor agential expressions are equivocal between human and non-​human animals. The capacities they signify may have different preconditions and implications, since they enable humans to do a lot more than animals. But this does not imply that, for example, ‘reproduction’ linguistically means something different in the two cases. ‘Reproduction’ is a univocal term, even if it covers distinct subclasses. Transformative theories are ultimately fuelled by an extreme holism, according to which any significant difference in the context of a phenomenon per se constitutes a distinct phenomenon. But such holism reduces to absurdity, at least in the case of the mental capacities that animals seem to share with us. This was brought out by Plutarch in his criticism of the radical transformationalism of the stoics. As for those who foolishly say of animals that they do not feel pleasure, nor anger, nor fear, nor do they make preparations, nor remember, but the bee only ‘as-​if ’ remembers, and the swallow ‘as-​if ’ makes preparations, and the lion is ‘as-​if ’ angry, and the deer ‘as-​if ’ afraid:—​I do not know how they will treat someone who says that they do not see nor hear either, but ‘as-​if ’ see and ‘as-​if ’ hear, and do not give voice, but ‘as-​if ’ give voice, and, in general, do not live, but ‘as-​if ’ live. For these last statements, I believe, are no more contrary to plain evidence than their own. (Plutarch, Sollertia 961 E–​F ) However, why should holistic as-​ifness about cognitive and conative capacities imply as-​ifness about phenomena like acoustic signalling and life (Plutarch’s examples), or about reproduction and cooperation (Brandom’s examples)? Because, by the holists’ own lights, it is partly constitutive—​for example, of the concept of perception—​that the information perception provides can guide activities like signalling, reproducing, and co-​operating. Similarly, it is partly constitutive of that concept that such information can be put in the service of biological functions, and hence, roughly speaking, of life. By the same token, it is partly constitutive of those behavioural and biological concepts that the activities and functions they express are capable of being guided by perception. But now, if what is partly constitutive—​ of reproducing, for example—​differs between humans and animals, then what is constituted must differ as well. The moral for the moderns is manifest: all capacities interact, in the behavioural repertoire of a subject that possesses them, not just [p. 411] Hans-Johann Glock in causal but also in conceptual ways. Therefore, if lack of rational powers barred animals from anything other than as-​if perception, it would also bar them from anything other than as-​if running, as-​if digestion, as-​if life, etc. Anyone willing to bite that bullet incurs the consequence of intellectual lead-​poisoning. The Need for Anthropology The invocation of Brandom and some formulations in Truth and Truthfulness notwithstanding, there are reasons to suspect that Williams would not have fallen victim to this toxic condition. For he subscribed to a Nietzschean kind of non-​ reductive naturalism. He regarded ‘naturalism’ as a useful label for views that try to understand ethics in ‘worldly terms without reference to God or any transcendental [sic!] authority’ (Williams 1985: 121; on Williams’s naturalism, see Chappell 2022). Furthermore, he endorsed a ‘creeping barrage’ naturalism for which the central question is: can we explain, by some appropriate and relevant criteria of explanation, the phenomenon in question in terms of ‘the rest of nature?’ (Williams 2002a: 23, 26–​7). Given the considerations rehearsed above, it cannot be ruled out on general conceptual grounds that this explanans includes non-​and pre-​linguistic dispositions and capacities. Consequently, Brandom’s contention that language ‘sweeps aside’ all the considerations concerning animal needs and concerns invoked by functionalist accounts would have to be based on factual considerations. As such, however, it stands in need of empirical confirmation through cognitive science, developmental psychology, biological anthropology, and evolutionary theory. No such confirmation is forthcoming from either Brandom or Williams. The closest the latter gets is a reference to W. H. Durham’s verdict that the most potent forces in cultural evolution are values and convictions of people that are already enculturated (Williams 2002a: 29). Unfortunately, that leaves open the crucial issue of how hyper-​cooperative cultures like those in the hominin line evolved or could have evolved in the first place. Furthermore, the undeniable differences between biological and cultural evolution do not militate against a rising amount of evidence to the effect that the mental and linguistic capacities that enable human communities and enculturation are subject to dynamic interactions between biological and cultural evolution (see e.g. Mesoudi 2011; Tomasello 2014). These capacities are anthropological universals. Because of their linguistic and rational dimensions they enable strikingly different forms of socialization and concept-​formation. But they have not been miraculously transubstantiated through the advent of language and culture. Both the capacities and the forms of behaviour they enable have been shaped by adaptive pressures of the kind invoked by functionalist accounts and dismissed by Williams. It just [p. 412] needs to be kept in mind that the adaptive pressures facing social and cooperative primates are not of the same kind as those exerted on microorganisms and plants!13 This goes to suggest that there are serious limitations to a genealogy that for non-​philosophical input relies only on historiography, let alone philosophically motivated takes on history. Empirical findings of an anthropological and evolutionary kind need to be taken seriously. Indeed, the reservations of Williams notwithstanding, these disciplines contrast favourably in many respects with his genealogy.14 One of the most contested aspects of his procedure is the way in which it links the historical and the imaginary-​cum-​fictional components. His reflections on the State of Nature are basically armchair anthropology concerning the functions of certain practices and concepts. They are not that different from those envisaged in Strawson’s explanatory metaphysics, which later came to fruition in his anthropological naturalism. But there is a gap between explaining certain practices and concepts as having emerged because of having a function in the service of fundamental human needs and concerns, and explaining the more specific contemporary manifestations of these explananda. ‘To bridge this gap, Williams leaves the fictional state of nature behind and moves into real history’, as Queloz acutely diagnoses (2021: 9). Unfortunately, such a procedure is unsound from both a philosophical and an empirical perspective. It cannot be right to start out from speculations about an atemporal State of Nature and then make it look explanatory of extant phenomena by bringing in actual history where it suits you. To be even minimally respectable as a diachronic explanation, the combination of State of Nature anthropological speculation and genuine historiography must claim or presuppose something like the following: the practices and concepts that are potentially explained by functional requirements in the imaginary State of Nature are either identical with or part of the effective causes of the starting point of the historical tale. But remember that causal explanation is factive. This requirement does not rule out potential explanations per se. They present a potential explanation that would explain Y by reference to X, if both were actual. If both are actual, then the explanation is not just a potential but an actual one, and thereby factive. Now consider the case of the explanations featuring in Williams’s genealogy: the explanans X is fictional and hence only possible and the explanandum Y is actual. Such an explanation is a potential 13 Hrdy and Burkardt (2020) advance a credible hypothesis to the effect that our unique mental and linguistic capacities emerged out of a combination of advanced causal and social cognition shared with the last common ancestor of humans and other apes, on the one hand, and pro-​social attitudes and emotions newly derived in response to the demands of cooperative breeding, on the other. 14 This verdict does not apply to Williams’s earlier critique of attempts of evolutionary epistemology chs 7–​8). It rightly pinpoints differences between biological evolution and the development of science, and its reservations about neo-​Darwinist explanations of our distinctive cognitive capacities are more measured and hence less vulnerable to criticism. [p. 413] Hans-Johann Glock one. Qua potential explanation, however, its explanatory yield is of a different kind. It concerns not a particular causal relation between X and Y, but the general law-​like connection between phenomena of type X and type Y. The reason for this is that the requirement of factivity does rule out a specific causal explanation of an actual phenomenon Y’ (the starting point of historical genealogy) by reference to a merely fictional X’ (the State of Nature or its functionally implied consequences of imaginary genealogy). This shortcoming of Williams’s overall picture is distinct from the widely deplored fast-​forwarding from ancient Greece to the eighteenth century. It is rather the fast-​tracking from an imagined State of Nature to a part of real history, across an ontological or, as I would prefer, categorial divide. Williams’s genealogy moves from a State of Nature imagined to occur earlier in a sort-​of chronological sequence to a later historical development. Evolutionary anthropology and empirically informed research on cultural evolution takes the opposite chronological route, as regards the ordo cognescendi.15 Rightly so! After all, the explananda are current phenomena. And their emergence is in the first instance traced back along recorded history, the kind of history that should inform historical genealogies à la Nietzsche and Foucault. Where that stops, cognitive archaeology, palaeontology, and phylogenetics kick in.16 And where they give out, the ‘comparative method’ of Darwinism is used, hypothesizing about evolutionary trees by looking at the closest extant relatives (whether species or human forms of life) to the explananda. Speculations like the unduly derided just-​so stories have a role to play at this juncture. Some just-​so stories are better than others, by standards set through empirical data, on the one hand, and evolutionary theory, on the other. Even the best just-​so story, however, is to be told only where these more controlled procedures cannot be applied, and mainly in a heuristic capacity. Conclusion To me it seems that, for once, it is not overly enthusiastic empirical scientists who draw precipitate conclusions on the basis of conceptually and methodologically fraught assumptions, but overly confident philosophers. In Williams’s case, this is ironical. In a programmatic essay, ‘Philosophy as a Humanistic Discipline’ (2000), he, for better or worse, deplores Wittgenstein’s attempt to keep philosophy pure from entanglement with other disciplines on the grounds that it is a toto caelo distinct intellectual enterprise. Yet his genealogy refuses to take seriously the 15 When it comes to presenting their results in a synoptic tale, as in Tomasello (2014), they work from earlier to later. But the reasoning on which that diachronic tale is based works from present to actual past. 16 For an introduction to cognitive archaeology that is sensitive to the epistemological challenges facing the conjectures of evolutionary anthropology, see Wynn and Coolidge (2022). For potential contributions of phylogenetics, see Lewens and Buskell (2023). [p. 414] implications of his own insistence that philosophical problems concerning our conceptual framework must take the origins of concepts and practices seriously. If and to the extent to which that is correct, philosophy must engage in more than diachronic fictions sprinkled with historical excursions. It must consider the actual genesis of our modes of thinking, speaking, and acting. And that actual genesis spans both biological and cultural evolution, and thereby the divide between natural and social sciences. One moral for analytic genealogists might be this. Take seriously the factual component of the explanatory part of your enterprise. Start out by determining what the explanandum is by clarifying the concepts at issue through a liberal, connective analysis which includes reference to their present function in communicative practices. But in a next step, work in a fashion similar to evolutionary theory by taking historiography seriously and linking it up with empirical and theoretical work on biological and cultural evolution. As regards explaining the actual emergence of our conceptual framework(s), the main philosophical task may well be that of a Lockean underlabourer, who supports non-​philosophical investigations through conceptual and methodological reflection. In any event, it is important to recognize the limitations of genealogy as a fail-​ safe guide to even potential explanations.17 This makes room for a more sober appreciation of it as a heuristic device—​not just for explanatory speculation but also for the clarificatory work of establishing what our concepts are and the normative work of assessing their legitimacy. Thus, the very title of Truth and Truthfulness already indicates one valuable contribution that the book no doubt makes. It alerts us to the fact that the values and virtues of respecting objective truth and of being sincere and authentic are both closely connected and potentially in conflict. Whether or not it should count as part of philosophy proper, the anthropological study of those virtues is important to understanding that conflict. At the same time, it is difficult not to be impressed by the illumination sought and often provided by contemporary genealogies. The idea has attracted even a stalwart champion of conceptual analysis like Peter Hacker. He regards most of our emotions as anthropological universals. Yet love, he maintains, is an exception; even a purely philosophical anthropology must elucidate it through charting the history of the concept (2018: Appendix). To be sure, what Hacker provides is an impressive historical genealogy, without fictional frills. But this only goes to accentuate questions like the following: Which concepts should be understood diachronically, and for what purpose? What kind of genealogy—​historical or imaginary—​is apposite? What ends, philosophical or other, can they serve? This chapter has formulated principled reasons for doubting that Williams’s version of 17 As Williams himself does on different grounds mentioned in note 7 above. [p. 415] Hans-Johann Glock analytic genealogy yields satisfactory answers to such pressing methodological questions.18 References Alvarez, Maria. 2016. ‘Reasons for Action: Justification, Motivation, Explanation’. In The Stanford Encyclopedia of Philosophy (Winter 2017 Edition). Edited by Edward N. Zalta, . Bouveresse, Jacques. 2021. Les foudres de Nietzsche et l’aveuglement des disciples. Marseille: Hors-​d’atteinte. Boyle, Matthew. 2017. ‘A Different Kind of Mind?’ In The Routledge Handbook of Philosophy and Animal Minds. Edited by Kristin Andrews and Jacob Beck, 109–​18. London: Routledge. Brandom, Robert. 2000. ‘Facts, Norms, and Normative Facts: A Reply to Habermas’. European Journal of Philosophy 8: 356–​74. Cappelen, Herman. 2018. Fixing Language: An Essay on Conceptual Engineering. Oxford: Oxford University Press. Cappelen, Herman. 2022. ‘Conceptual Engineering and Its Place in Philosophical Methodology’, , 5 July. Chappell, Sophie Grace. 2022. ‘Bernard Williams’s Liberal Naturalism’. In The Routledge Handbook of Liberal Naturalism. Edited by Mario De Caro and David Macarthur, 134–​51. London: Routledge. Chappell, Sophie Grace, and Nicholas Smyth. 2023. ‘Bernard Williams’. In The Stanford Encyclopedia of Philosophy (Summer 2023 Edition). Edited by Edward N. Zalta and Uri Nodelman, . Craig, Edward. 1990. Knowledge and the State of Nature: An Essay in Conceptual Synthesis. Oxford: Clarendon Press. Craig, Edward. 2007. ‘Genealogies and the State of Nature’. In Bernard Williams. Edited by Alan Thomas, 181–​200. Cambridge: Cambridge University Press. Danto, Arthur C. 1965. Nietzsche as Philosopher: An Original Study. New York: Columbia University Press. De Mesel, Benjamin. 2024. ‘Strawson, Skepticism, and the Natural Roots of Responsibility’. In Scepticism and Naturalism. Edited by John Hyman and Michael Thorne, 000–​0. Leiden: Brill. Eklund, Matti. 2015. ‘Intuitions, Conceptual Engineering, and Conceptual Fixed Points’. In Palgrave Handbook of Philosophical Methods. Edited by Chris Daly, 363–​85. Basingstoke: Palgrave Macmillan. Frege, Gottlob. 1884. Grundlagen der Arithmetik. Breslau: W. Koebner. Garson, Justin. 2008. ‘Function and Teleology’. In A Companion to the Philosophy of Biology. Edited by Anya Plutynski and Sahotra Sarkar. Malden, MA: Blackwell. Glock, Hans-​Johann. 2008a. ‘Analytic Philosophy and History: A Mismatch?’ MIND 117: 867–​97. Glock, Hans-​Johann. 2008b. What Is Analytic Philosophy? New York: Cambridge University Press. Glock, Hans-​Johann. 2012. ‘Strawson’s Descriptive Metaphysics’. In Categories of Being. Edited by Leila Haaparanta and Heikki Koskinnen, 391–​419. New York: Oxford University Press. Glock, Hans-​Johann. 2016. ‘Philosophy Rehinged?’ International Journal for the Study of Skepticism 6.2–​3: 274–​308. 18 This article forms part of research under the auspices of the NCCR Evolving Language (Swiss National Science Foundation Agreement #51NF40_​180888). I am also grateful to the Institute of Philosophy at KU Leuven for awarding me a Visiting Fellowship, as well as to my host Benjamin de Mesel, whose graduate class provided the opportunity for a very fruitful discussion of the first draft. Finally, I should like to thank the editors for their comments on a second draft and for their patience. [p. 416] Glock, Hans-​Johann. 2017. ‘Impure Conceptual Analysis’. In The Cambridge Companion to Philosophical Methodology. Edited by Giuseppina D’Oro and Søren Overgaard, 83–​107. Cambridge: Cambridge University Press. Glock, Hans-​Johann. 2018. ‘What Have the Historians Ever Done for Us?’ In Philosophy and the Historical Perspective. Edited by Marcel van Ackeren, 18–​35. Oxford: Oxford University Press. Glock, Hans-​Johann. 2022. ‘Norms, Reasons, and Anthropological Naturalism’. Philosophical Topics 50.1: 9–​32. Glock, Hans-​Johann. 2025. ‘Conceptual Engineering and Conceptual Analysis: Historical and Conceptual Connections’. In New Perspectives on Conceptual Engineering. Edited by Manuel Isaac and Kevin Scharp, vol. 2, 1–​24. New York: Springer. Synthese Library. . Maudlin, Tim. 2015. Philosophy of Physics: Space and Time. Princeton: Princeton University Press. Mesoudi, Alex. 2011. Cultural Evolution: How Darwinian Theory Can Explain Human Culture and Synthesize the Social Sciences. Chicago: University of Chicago Press. Nozick, Robert. 1974. Anarchy, State, and Utopia. Oxford: Blackwell. Queloz, Matthieu. 2021. The Practical Origins of Ideas: Genealogy as Conceptual Reverse-​ Engineering. Oxford: Oxford University Press. Renz, Ursula. 2018. ‘On Philosophical Problems: Some Remarks on an Old Battlefield’. In Philosophy and the Historical Perspective. Edited by Marcel van Ackeren, 213–​28. Oxford: Oxford University Press. Rorty, Richard, Jerome B. Schneewind, and Quentin Skinner, eds. 1984. Philosophy in History. Cambridge: Cambridge University Press. Strawson, Peter. 1956. ‘Construction and Analysis’. In The Revolution in Philosophy. Edited by Alfred J. Ayer, 97–​110. London: Macmillan. Strawson, Peter. 1962. ‘Analysis, Science and Metaphysics’. In The Linguistic Turn. Edited by Richard Rorty, 312–​20. Chicago: University of Chicago Press. Tomasello, Michael. 2014. A Natural History of Human Thinking. Cambridge, MA: Harvard University Press. van Ackeren, Marcel, ed. 2018. Philosophy and the Historical Perspective. Oxford: Oxford University Press. van Ackeren, Marcel. 2019. ‘Williams (on) Doing History of Philosophy: A Case Study on Ethics and the Limits of Philosophy’. In Ethics beyond the Limits: New Essays on Bernard Williams’ Ethics and the Limits of Philosophy. Edited by Sophie Grace Chappell and Marcel van Ackeren, London: Routledge. Williams, Bernard. 1985. Ethics and the Limits of Philosophy. London: Fontana. Williams, Bernard. 1995. Making Sense of Humanity. Cambridge: Cambridge University Press. Williams, Bernard. 2000. ‘Philosophy as a Humanistic Discipline’. Philosophy 75.294: 477–​96. Williams, Bernard. 2002a. Truth and Truthfulness. Princeton, NJ: Princeton University Press. Williams, Bernard. 2002b. ‘Why Philosophy Needs History’. London Review of Books, October: 7–​9. [p. 417] Hans-Johann Glock Williams, Bernard, and J. J. C. Smart. 1973. Utilitarianism: For and Against. Cambridge: Cambridge University Press. Wynn, Thomas, and Frederick L. Coolidge. 2022. An Introduction to Evolutionary Cognitive Archaeology. New York and Abingdon: Routledge. Zuber, Aneta. 2025. ‘Forward‑Looking Concept Functions and the Function/​ Accident Distinction’. Erkenntnis 90.6. . [p. 418] Internal Reasons and Historical Thinking Geraldine Ng In this rich and diverse collection focusing on Bernard Williams’s philosophical work as it relates to historical inquiry, my contribution adds an unexpected twist. Williams’s engagement with the history of philosophy is explicit, as is his avowed ‘historicist turn’ in his later work (2009: 198). This chapter draws attention to an additional role that history, broadly conceived, plays early on in his ethical philosophy that has been overlooked. I argue that in the central conception of Williams’s ethical philosophy, reasons internalism, there is an implicit appeal to historical thinking. The chapter has three main aims. The first is to explain, by clarifying how Williams’s ethical philosophy is committed to acknowledging the contingency and complexity of ethical life, the sense in which it can be read as historicist. In the manner of R. J. Collingwood, Williams regards the task of appraising both present and past agents to involve historical understanding. Secondly, I argue that, taking Williams’s method of historical thinking seriously, reasons internalism yields a normative proposition, what I term the reflective disposition proposition. I go on to argue that Williams’s notion of a ‘thick’ ethical concept is a precursor of the ideas that are advanced in the chapter, and that Williams’s strategy for defending ethical knowledge in relation to thick concepts is of a piece with a historicist account of the normative force of internal reasons. Last, I suggest that this reconsideration of the possibilities of reasons internalism affords a robust reproach to the common objection that Williams’s ethical philosophy is merely negative. Williams’s ethical philosophy is not merely negative because his philosophical method is not merely analytic. In light of his earlier, implicit ‘historicist turn’, Williams’s ethical philosophy is historicist and positive, or so I will argue. Williams’s interest in a combination of philosophy and history took various shapes. With this contribution to the collection, we can see how the development of his thought from normative ethical philosophy to the history of philosophy and the genealogical method was not merely a change of concerns. Rather, Williams’s historical sensibility is operative in his philosophy throughout, even in its earlier, more conventional analytic applications. Ng Geraldine, Internal Reasons and Historical Thinking In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. [p. 419] Geraldine Ng 1. A Positive Moral Philosophy There are various ways in which philosophical thought can integrate historical practice into its work. First, there is the history of philosophy. Williams’s contribution to this debate is the idea that the history of philosophy could be ‘done philosophically’, such that it yielded philosophy first and history second. Next is the genealogical thesis, which holds that the reference of concepts or terms typically varies by historical context. The same concept or term can pick out very different referents at different historical moments. Accordingly, it is a mistake to assume a concept has the same meaning across historical periods. The application of historical understanding to the genealogical method animates Williams’s last work, Truth and Truthfulness. I propose that, in addition, Williams draws on historical thinking to develop his ethical outlook. On the one hand, his sceptical line of reasoning about most modern moral philosophy is expressed in the traditional, ahistorical mode of analytic philosophy. Where he offers a more positive account of ethical life, on the other hand, his method combines philosophical analysis and historical explanation. Analytical moral philosophy proceeds, largely, with ahistorical assumptions. It can be ahistorical by failing to acknowledge the importance of the historicity of concepts. It can be ahistorical, additionally, by failing to account for the contingencies of ethical life. The chapter will focus on Williams’s attempt to address ahistoricity in this second sense, in relation to contingency. An adequate account of the contingency of ethical life requires elaborating agency in light of the many dimensions of what Williams calls ‘the actual’ (1999b [1981]: 30; hereafter ML). It is important to clarify that, in his earlier sceptical ethical discussion, Williams is not concerned with the historicity of concepts.1 Rather, Williams is complaining that modern moral philosophy fails to fully grasp the contingency and complexity of ethical life. By aiming to give a simplified account of ethical life, it struggles to achieve its own explanatory goals: ‘If there is such a thing as the truth about the subject matter of ethics—​the truth, we might say, about the ethical—​why is there any expectation that it should be simple?’ (2007 [1985]: 17; hereafter ELP). Williams appeals to history both in relation to his critique of modern moral theorizing and in his positive, complicating explanation of moral practice. My point of clarification is about the distinctive part history plays in Williams’s not simple account of moral experience. In Ethics and the Limits of Philosophy, Williams observes, ‘How truthfulness to an existing self or society is to be combined with reflection, self-​understanding, and criticism is a question that philosophy, itself, cannot answer’ (ELP: 200). An agent’s (intentional) actions will reflect complex phenomena such as ‘dispositions 1 I am grateful to Colin Koopman for this point. See Koopman (2010). [p. 420] of evaluation, patterns of emotional reaction, personal loyalties, and various projects, as they be abstractly called, embodying commitments of the agent’ (2001: 81; emphasis suppressed). To be true to such an existing self, philosophy needs history to explicate moral practice. Williams concludes in Truth and Truthfulness that ‘philosophy, in order to do its business, must move into history’ (2002: 173). This thought, I propose, is unannounced but manifest in his earlier ethical writing. The interpretation of Williams’s ethics advanced here will demonstrate the way in which his moral philosophy moves into history that has been unexplored. Moreover, this interpretation of reasons internalism has the benefit of both offering the best way of understanding its distinctive kind of normativity and answering the objection that Williams’s ethical view is merely sceptical. 2. Historical Thinking and Normative Ethics Foremost in Williams’s ethical work, it appears, is the aim of debunking normative ethical theorizing. But the objective is not simply this. The arguments against ethical theory are designed (among other things) to encourage us to look in directions away from theory. ‘Philosophy should not try to produce ethical theory,’ Williams states, ‘but this does not mean that philosophy cannot offer any critique of ethical beliefs and ideas’ (ELP: 17). Williams’s position is to something which itself essentially retains an element of normativity. That element is concealed, qualified, and overlaid with other things. What is unclear, as John McDowell and others have commented, is exactly where Williams locates the ground of normativity.2 There is no doubting that Williams takes seriously the normativity of ethical reasons.3 He insists that ‘no account of “A has a reason to x” can be adequate unless it has normative force’ (1999c hereafter WME).4 The chapter attempts to identify a vindicatory explanation of normativity in Williams’s positive account of ethical life to accompany the debunking account of normative theorizing that we have already. There are two very different ways in which one might approach the normativity of ethical claims: under the assumption of the systematicity and universality of morality, and under the assumption that the actions of socially, historically situated agents in ethical life are responses to norms of internal reasons. The chapter offers 2 See McDowell (1999 [1995]). 3 For a discussion of Williams’s non-​objective ethical normativity, see Fricker (2022). 4 Also, in Chapter 3 of Making Sense of Humanity, ‘It is important that even on the internalist view a statement of the form “A has a reason to [x]‌” has normative force. Unless a claim to the effect that an agent has a reason to [x] can go beyond what that agent is already motivated to do—​that is, to go beyond his already being motivated to [x]—​then certainly the term will have too narrow a definition’ (1998b hereafter MSH). [p. 421] Geraldine Ng a sustained attempt to delineate the subject matter of internal reasons and to show why this subject matter cannot be captured by analytic philosophy alone but requires historical thinking. By ‘historical thinking’ I do not mean concern with the temporally remote past alone. Here, Collingwood is of help.5 He writes, ‘It is by historical thinking that we rethink and so rediscover the thoughts of Hammurabi or Solon’. Collingwood continues, ‘It is in the same way that we discover the thought of a friend who writes a letter, or a friend who crosses the street’ (1994: 219). I propose that a method of historical thinking unfolds from the workings of the internal reasons thesis. This should not be completely surprising. Williams’s use of historical thinking is evident in his thoughts on what he terms the ‘relativism of distance’ and his discussion of how we should understand the values of past societies.6 As well, he employs historical thinking in his notion of thick ethical concepts. In the manner of Collingwood, Williams takes advantage of the methods of historical thinking in appraising the actions and reasons of present agents as well as past societies. 3. The Limits of Reasons Internalism I identify three related problems for reasons internalism. First, the common objection to Williams’s internalist view of practical reasons is that it is too ‘psychologistic’, as McDowell puts it (1999 [1995]: 77). More generally, McDowell rejects the argumentative structure of Williams’s critique of external reasons. He objects: ‘We do not need to choose between conceiving practical reason psychologistically and conceiving it as an autonomous source of motivational energy’. There is room, McDowell insists, for an alternative account of reasons that is external but not necessarily ‘apsychologistic’ (1999 [1995]: 82). Next, objects McDowell, ethical reasons on Williams’s view are reasons ‘only for those who have motivations to which ethical considerations speak, or can be made to speak’ (1999 [1995]: 68). To speak to this worry, I want to point out an important connection between Williams’s scepticism about external reasons and his scepticism about the scope of ethics. The scope of ethics is a broader concern that Williams has with ethical theorizing. Williams asks: ‘What will the professor’s justification do, when they break down the door, smash his spectacles, take him away?’ (ELP: 23). In effect, he questions the basic aspiration of ethics or moral philosophy to speak to everyone. It is mistaken, believes Williams, to expect to discover a theory of ethics that might somehow move the intransigent. Williams 5 Williams held Collingwood in high regard. See (2006a). 6 See, in particular, Williams (ELP: ch. 9; ML: ch. 11). [p. 422] thinks that, as A. W. Moore puts it, ‘the point is not to persuade anyone of anything’ in ethics (2007 [1985]: 207). The fact of the matter is, thinks Williams, moral philosophy is not for those who are uninterested in ethical considerations. In this sense, he believes that ethical philosophy is limited. However, for those persons who are within ethical life, philosophy can help self-​understanding. In this sense, we might suppose, it is unlimited. Philosophy, for persons who are open to ethical considerations, can provide a powerful critique of one’s ethical experience. Williams’s practical and positive account of ethics speaks more narrowly to persons who are within ethical life: ‘the aim of the discourse is not to deal with someone who probably will not listen to it, but to reassure, strengthen, and give insight to those who will’ (ELP: 26). Naturally, what Williams perceives to be the limits of ethics has a bearing on the scope of the internal reasons thesis. On the internal reasons thesis, the claim ‘A has a reason to x’ is true if and only if there is a sound deliberative route to x-​ing which starts from A’s existing motivations. In other words, x-​ing will be a function of what Williams calls A’s S—​her subjective motivational set. On an externalist view of reasons, alternatively, the claim ‘A has a reason to x’ can be true regardless of the contents of A’s S. External reasons are not based on an agent’s antecedent desire or other pre-​existing psychological states. Williams allows that there is a limiting condition on internal reasons. Reasons are based on an agent’s antecedent desire and other pre-​existing psychological states. McDowell takes issue with this limiting condition. Finally, an implication of the internal reasons thesis, McDowell objects, is that we must give up the normativity of ethical claims. Williams argues that ‘the only rationality of action is the rationality of internal reasons’ (1999a [1981]: 111; hereafter IER). Hence, the normative basis for criticism on the internal reasons thesis, McDowell supposes, ‘depends on there being a critical dimension to the concept of rationality’ (1995: 76). However, because that normative or critical dimension of practical rationality is practically ‘fixed by the agent’s motivations as they stand’, McDowell rejects the picture of critical rationality that internal reasons yields (1995: 77). McDowell insists that, if there is to be a normative dimension to reason claims, then it must be sufficiently independent of any agent’s psychology. Williams accepts that his account is psychologistic but denies that giving up normativity follows: ‘I insisted in my writings about this question [of external and internal reasons] that no account of “A has a reason to x” can be adequate unless it has normative force’ (WME: 191). McDowell acknowledges that this is Williams’s aim, but worries that Williams does not do enough to establish the normativity of internal reason claims. Bringing to bear a method of historical thinking will offer a fresh way to address McDowell’s worries. It can more delicately fill in a psychologistic picture of reasons, and finesse the limits of reasons internalism. [p. 423] Geraldine Ng Hard Cases’ and the Reflective Agent However vague it may initially be, we have a conception of the ethical that understandably relates to us and our actions the demands, needs, claims, desires, and, generally, the lives of other people, and it is helpful to preserve this conception in what we are prepared to call an ethical consideration. (ELP: 12) The sentence contains the seeds of much that follows. Yes, we have such a conception of the ethical. But, while we have such a conception, it is not shared by everyone. While this might adequately capture what we call an ethical consideration, in every society there will individuals who will not have a conception of the ethical. Hence, as Williams notes, ‘Ethical understanding needs a dimension of social explanation’ (ELP: 131). In Philosophy as a Humanistic Discipline, Williams makes clear that he is prepared to interpret ‘historical understanding’ under the wider rubric of ‘social understanding’ (2006d: 493; hereafter PHD). Williams, like Collingwood, has a broad conception of history. Any adequate social and historical explanation will recognize that there are some particular individuals in society for whom the ethical escapes them. An account that explains the relation of these particular individuals to the ethical will include the social and historical. Imagine someone who is rational but otherwise has no such ‘conception of the ethical that understandably relates to us and our actions the demands, needs, claims, desires, and, generally, the lives of other people’. Imagine another rational individual who has some conception of the ethical but is utterly unmoved by ethical considerations. Both imagined individuals fall under the description of what Williams calls ‘hard cases’.7 According to Williams, an agent is a hard case in this sense: ‘It is precisely people who are regarded as lacking any general disposition to respect the reaction of others that we cease to blame, and regard as hopeless or dangerous characters, rather than thinking that blame is appropriate to them’ (1998a [1995]: 43). We must cease to ethically blame them because, assuming that ethical blame is related to ethical reasons, we cannot blame someone for failing to x if we cannot claim she has a reason to x. With hard cases, ethical blame is inappropriate. A key feature of his account of ethical reasons is the space that Williams allows for the possibility of hard cases. The admission of hard cases is a practical consequence of accepting the limits of ethics. If an agent is unsusceptible to ethical considerations, then she is a hard case. Ethical discourse, including internal reason 7 See Williams’s discussion in MSH: ch. 3. For reasons of space, I do not go into the details of the argument for ‘hard cases’. However, I want to mention a limiting condition on the hard case, what he calls a ‘proleptic mechanism’ whereby an agent who may not have the reason initially might be thought to come to have it proleptically. An agent is not a hard case if a reason can be proleptically ascribed to her. [p. 424] claims, is of little meaning to those who are not open to it. To identify someone as a hard case is to recognize her refusal of an ethical life. Before I examine the hard case, it will come out more clearly if I first focus on A, under the assumption that A is not a hard case. Williams’s admission of hard cases is radical. By considering the implications of this admission in terms of the non-​ hard case, the degree to which it is radical will become evident. The admission of hard cases affords us the possibility of saying something substantive and normatively relevant about A. I propose that, in making the ethical claim ‘A has a reason to x’, we are at the same time making a basic assumption about A: that is, A is in ethical life. Of course, we could be wrong. If we are mistaken, and A is not in ethical life, then she is a hard case. An ethical reason claim about A would be mere ‘bluff and browbeating’.8 Say A is not a hard case. In making a basic assumption that A is in ethical life, we are assuming that she will have some rough conception of the ethical, some rough understanding that relates to her and her actions ‘the demands, needs, claims, desires, and, generally, the lives of other people’. In making a basic assumption that A is open to ethical considerations, we are not, of course, making the further assumption that A will act on any ethical considerations. Instead, it is a weaker claim that A is disposed to be open rather than immune to ethical considerations. Here, I want to point to a structural symmetry between reason claims about rational considerations and ethical considerations. Claims involving rational considerations in relation to reasons for action presuppose an agent’s faculty for rational deliberation. In making the rational reason claim that ‘A has a reason to x’, we are presupposing that A is rational. Claims involving ethical considerations in relation to reasons for action presuppose an agent’s disposition to ethical reflection. This is what the admission of hard cases practically amounts to. What I call the reflective disposition proposition captures this thought: the ethical claim ‘A has a reason to x’ presupposes something basic about A’s reflective disposition. To give some idea of what the reflective disposition proposition means, I appeal to the explanatory dimension of reasons. Our rational faculty moves us to search for reasons and justifications in the endorsement of our rational choices. Likewise, our reflective disposition moves us to search for reasons and justifications in the endorsement of our ethical beliefs. There are many factors that inform and contribute to an agent’s reflection on ethical considerations. In this connection, Williams refers to ‘the materials of the agent’s world’. (1996: 212). The materials of agent A’s world 8 On the one hand, Williams argues that external reason claims are nothing but mere ‘bluff ’ and browbeating (IER: 111; also 2001: 95). If the agent is a hard case, on the other hand, then one could say that for such an agent an internal ethical reason claim is also bluff and browbeating. The explanations, of course, are very different. In the first instance, it is browbeating because there are no external reasons. In the second, an ethical reason claim is browbeating because the hard case is by definition someone not open to ethical considerations. [p. 425] Geraldine Ng contribute to her practical reflection. There is considerable indeterminacy, but we can suppose that materials of A’s world that contribute to reflection will include A’s personal relationships, ground projects, social institutions, local ethical practices, and such like. These materials afford a social and historical understanding of A. The general instrumental cast of reasons is highly suggestive. Rational reasoning demands of A the use of her faculty for rational deliberation. Likewise, ethical reasoning demands of A the exercise of her disposition to reflection. On this account, we can see that an agent’s S is something not fixed but instead something that might be continually augmented with the exercise of her rational deliberative faculty and her disposition to reflection. Deliberating rationally, we exercise and sharpen our rational faculties. Practising reflection, I suggest, is the means by which we enhance our openness to ethical considerations. Rational and ethical considerations constitute the input for A’s S. The conclusion of rational deliberation and ethical reflection will satisfy some elements of A’s S, but it is not fully determined by her S.9 Now, what about the hard case? Another way of putting the hard case is to say that she is not disposed to reflection. The hard case’s refusal of ethical life stands in contrast to the agent whose openness to ethical considerations signals her access to the ethical world. The hard case is not drawn to search for ethical justifications in the endorsement of her actions and beliefs. Rather, the hard case focuses on purely personal interests, and is primarily motivated to satisfy her impulses and desires. Because her motivations are purely personal, there are no other-​regarding elements in the subjective motivational set S of the hard case to reflectively and ethically check any of her impulses or desires. By contrast, assuming A is not a hard case, we reasonably assume that A’s S, given her reflective disposition, consists of at least some non-​personal, other-​regarding elements in addition to her personal interests. The reflective self-​consciousness has a bearing on A’s S. On the assumption that she is in ethical life, we assume that A’s S is responsive to considerations that might subject personal impulses and desires to modification. The reasons in A’s S are not purely personal but, we could say, mixed, given the contribution of non-​personal elements in reflection. A’s mixed S faces her with the need and also provides her with the discretionary aptitude to arrive at an answer that can be seen to be right not just from a narrowly personal point of view but from the wider reflective standpoint of an agent whose S represents that of someone in ethical life. The reflective disposition proposition, on the one hand, makes plain the explanatory potential of the internal reasons thesis. The hard case, on the other hand, gives an indication of the limits of the internal reasons thesis. 9 ‘There is room for the imagination in deciding what to do, and correspondingly in saying what someone else has reason to do. Indeed, the stance towards the agent that is implied by the internalist account can be usefully compared to that of an imaginative and informed advisor, who takes seriously the formula “If I were you . . .” ’ (Williams 2001: 92). [p. 426] A great deal of the explanation of reasons internalism depends on Williams’s conception of moral psychology. We can interpret the ‘social’ in Williams’s remark about ‘a dimension of social explanation’ in another sense. Reflection on ethical considerations has a bearing on an agent’s self-​conscious awareness. In achieving her self-​conscious awareness, the agent who is disposed to reflection irrevocably introduces the social world. A way to describe Williams’s position is to say that he wants to do justice to, and incorporate into his account of ethical reasons, the way in which the social world might shape our reasons and augment an agent’s S. The reason we can no longer decide from the purely personal perspective within which original desires are found is that, once we become aware of the social world, our choice becomes not just what we want to do, but what we think this person in this social world should do. Reflection achieves awareness of not just the self, but the self practically in a social world. The mixed perspective, involving personal and non-​personal elements, is that view achieved by practical reflection. Call an agent who has this reflective disposition a reflective agent (RA). By definition, all agents who are not hard cases are RAs. An agent RA is confronted with a desire or other impulse that presents a certain action x as worth doing. She must decide whether to do the action or not. We can think of RA’s motive as having two aspects, the aspect under which x is presented as something RA might do and the aspect under which RA chooses to x. In supposing RA is not a hard case, we suppose that RA is aware to some degree of everyone else as well as herself. Being reflectively disposed suggests the possibility of non-​personal reasons, reasons that may have a bearing on her deliberative, reflective choices. Given her reflective disposition, RA may accord some non-​personal, social weight to at least some of her reasons for acting. These reasons may in turn inform, and sometimes constrain, what she feels justified in doing in the pursuit of her life and personal projects. In all this input of mixed reasons, we must remember that each person will come to form her own reflective point of view, her own practical self-​conception. Williams’s perspective-​dependent account of reasons is thoroughly perspectival. I grant that this implies that the normativity of reasons depends on RA’s reflective disposition. The ethical normative test is not one of external objective truth but of internal, reflective endorsement. The point is to reveal that once we see a reason as issuing from a complex web of elements in one’s S, the business of reasoning and avoiding inconsistency becomes a demanding and complex affair. If we can grasp the complexity of this business of reasoning, we can understand how it may be that ‘RA has a reason to x’ is normative, for RA. 5. The Source of Normativity ‘Morality arises from the value we find it irresistible to grant ourselves as sources of reasons once we take up the reflective attitude towards our own actions’, writes [p. 427] Geraldine Ng Nagel (2008 [1996]: 208). The interpretation of the internal reasons thesis outlined here offers an explanation of the way in which we grant to ourselves sources of reasons. Given her reflective disposition, the reflective agent RA is inclined to treat her reasons as normative. Referring to her S, the conclusion of ethical deliberation is reflective endorsement (or rejection) of an action. In reflectively endorsing an action, RA regards her reflective self as the source of her normative reasons.10 Nagel at another point remarks, ‘The reflective self is in its nature more universal than the original, unreflective self ’ (1996: 203). If ‘more universal’ is understood to mean roughly ‘less narrowly self interested’, then Nagel’s characterization of the reflective self might find support from the reasons internalist. The two explanatory accounts, of course, would be very different. On the one hand, according to Nagel, the self is more universal because it ‘achieves its self-​conscious awareness by detaching from the individual perspective’ (1996: 203). On the other hand, according to the internalist view offered here, the self achieves its self-​conscious awareness by acknowledging the social world from within which the individual perspective springs. Nagel believes that reflection achieves an impersonal view. I suggest, alternatively, that reflection achieves an impurely personal view. The reason why RA can no longer decide from her purely personal perspective within which original impulses are found is that, once RA sees herself as an agent in a social world, however tacitly or unconsciously, her choice becomes not simply what she should do but what she should do in light of her self-​conception in the social world. On the one hand, the social world plays a role in furnishing some of the elements of an agent’s S. It may contribute non-​personal elements to her set of motivations. These elements will be local, culturally determined, and historically contingent. This will be balanced against, on the other hand, the space Williams allows for an individual life.11 He emphasizes that there is sometimes nothing more to say than that this is just how a person is. Certain interests and concerns and loves are just the ones that a person has. Reflection will lay bare internal conflicts. An action serving pure self-​interest may after all be irrational, in the sense that it cannot be reflectively endorsed, given other elements in an agent’s S, as a reason to act. The agent’s S can prevent endorsement by revealing a kind of incoherence. Ethical life is pervaded through and through with reflection. It consists of a constant search for rightness and goodness in our actions. In ethical life, reflection is as ordinary as desire. Analysing the operations of reflective consciousness sheds light on the nature of the normative internal force that it generates. Our reflective self is the ground of 10 I am indebted here to Christine Korsgaard’s defence of the idea of normativity deriving from an agent’s self-​conception. See Korsgaard (2008 [1996]). 11 This thought is consistently evident in his writings. See, in particular, Williams’s discussion in ML: chs 1 and 10. [p. 428] our ethical nature and the source of normativity. But this reflective self may never be recognized. It may be ignored. It may be refuted. This is the hard case. Or, it may be only partially realized, or given only sporadic attention. This is most of us. One can take a reflective view but still be insufficiently open to ethical life. In the full light of reflection, many things will be condemned. In partial light, many actions may be endorsed which might not otherwise be. For reasons to survive reflection, dangers lie in both directions. Reasons of pure self-​interest might be revealed as incoherent in the face of certain other non-​ personal reasons which are part of one’s S. Equally, reasons drawn from social norms might wither when they conflict with reasons that support an individual’s personal desires, projects, or values. Reflective endorsement makes the normative calls in both directions. This exemplifies the workings of immanent critique in an ethical life. An agent’s S can safeguard against a bad kind of conservatism or conformism, on the one hand. It can also defeat selfishness, on the other hand. In practising reflection, an agent is continually involving the exercise of imagination as she entertains possible alternative ways of going on.12 I need to consider two objections to the picture I have just outlined. First is the rejoinder to the effect that reflective judgement and social awareness simply yield reflective endorsement of rational self-​interest. One can imagine that RA is particularly aware of the social world merely because RA desires the praise and admiration of others. Yes, this is certainly possible. It must be admitted that, nevertheless, RA is in ethical life. In addition, we cannot be certain that this is all that her reflection yields. If RA carries on with reflection, one might suppose that she may come to endorse more non-​personal social values just in virtue of her undertaking ethical life itself.13 This is merely the common-​sense notion that we begin by knowing something about the subject matter, and on that basis go on to learn more. Our knowledge deepens. Say, wanting to avoid the contempt of others and desiring praise and admiration are part of RA’s S. Without good reason, RA cannot consciously choose to act in ways that hurt others and violate social norms at the same time as rightly see herself as part of the social world. From choosing not to hurt others because she wants to avoid the contempt of others, RA might come to directly desire not to hurt others. The second worry is deeper. Much of moral philosophy is engaged in the search for a theory to justify ethical claims. Williams is looking for something entirely different. He writes: There is room for the imagination in deciding what to do, and correspondingly in saying what someone else has reason to do’ (2001: 92). 13 Of course, the opposite may be true. She may take reflection no further than rational self-​interest. But, why should we expect the world not to contain such people, given that it obviously does? I thank Paul Sagar for this point. [p. 429] Geraldine Ng Why should morality not be sustained, rather, by what is mentioned in the explanation? . . . Consider, for instance, the true explanation of (the other bits of ) her heart’s desire: what is revealed in psychological accounts of the origins of her passions may not normatively endorse them, but this does not mean that it renders them meaningless when they are considered ‘from the inside’. (1996: 211–​12) Acknowledgement of others is implicit in the reflective view. The question is this. How are considerations about others—​others, by telling us what to do, by existing, by how they live—​relevantly activated when an agent takes the reflective view? Why suppose reflection on considerations about other people to be given a hearing against her heart’s desire? Here we encounter a different aspect of the limits of internal reasons. If someone has a reason, all sorts of factors, such as insensitivity, selfishness, blind passion, and simple stupidity, may prevent her from acknowledging it. Reflection is a necessary condition for being in ethical life, not a sufficient condition for being an ethical agent. Contrast Kant’s account of how rationality or rational agency amounts to moral agency. The Categorical Imperative gives an account, from the ground up, of how an agent accepts the force of ethical considerations. The Kantian moral agent legislates for herself. Where the Kantian picture offers a determinate conception of the rational and moral agent, what I call the reflective agent is underdetermined. The adoption of an ethical social norm might be in accord with RA’s S. But its rejection might be in accord too. Precisely what the internalist account allows for is the possibility of personal discretion, and indiscretion. By acknowledging that, as Williams urges, ‘agents need to share a social world’, we are engaging historical thinking (ELP: 153). In the same way that Collingwood discovers by historical thinking ‘the thought of a friend who writes a letter’, by historical thinking we come to understand RA. By the method of historical thinking we come to understand how ‘a non-​moralized, or less moralized, psychology leaves it open, or even problematical, in what way moral reasons and ethical values fit with other motives and desires, and how far they express those other motives, and how far they are in conflict with them’ (WME: 202). Putting together Williams’s themes of the need for historically thinking about the shared social world and a less moralized, more complex psychology, what can we say? We can suppose of RA that reflection on considerations about other people, however tacitly, however unselfconsciously, will be given a hearing against her heart’s desire. What we cannot be certain of is the outcome of that hearing. 6. Thick Concepts Williams was a close reader of Collingwood. ‘Collingwood’, he observes, ‘respected science and based his entire philosophy on history’ (2006a: 357). Williams [p. 430] respected Collingwood and based a large part of his non-​critical, positive ethical philosophy on history, or so I am arguing. In a reply to Putnam, clarifying his view on scientific knowledge, Williams’s sympathy with Collingwood’s historicist methodology is apparent: ‘The idea was that when we reflect on our conceptualization of the world, we might be able to recognise from inside it that some of our concepts and ways of representing the world are more dependent than others on our perspective, our peculiar and local ways of apprehending things’ (PHD: 185). This recalls Williams’s notion of thick ethical concepts, or what earlier, in Morality, he refers to as ‘more substantial’ ethical concepts (1993 [1972]: 32). The more substantive interpretation of the internal reasons thesis advanced in this chapter, I suggest, finds support in Williams’s more substantive notion of thick ethical concepts.14 Thick concepts capture the social and historical diversity of what Mackie calls ‘people’s adherence to and participation in different ways of life’ (1990 [1977]: 36). Thick concepts are understood in contrast to thin ethical concepts such as ‘ “good”, “right”, “obligatory” ’ (WME: 206–​7). Thin concepts are ‘action guiding’. Thick concepts, in comparison, are both ‘action guiding’ and ‘world-​guided’. For example, if I accuse A of committing an obscenity (a thick concept), I am negatively evaluating the situation. But my judgement is subject to correction if in fact it turns out that, for example, A’s action was not in fact obscene and I misconstrued it. Supposing I was correct, and supposing A is in ethical life, and supposing A embraces the concept of obscenity, she has a reason not to commit the obscenity.15 Given their sociality and historicity, ‘Thick concepts used in different societies and at different times are not all the same’, observes Williams (WME: 206). Homosexual acts, now accepted in ethical life, were once considered obscene. Consider ‘loyalty’, another of Williams’s examples of a more complex substantive value. What in the society of the samurai was considered an act of loyalty would today be considered an act of self-​sacrifice. Hence, it is not a problem, suggests Williams, that ‘I can recognise as knowledge something which I could not, granted my outlook, myself share’ (WME: 207). Rather, the fact of the matter is that thick concepts are used differently. This is a version of a very general problem that befalls a moral philosophy that is, like his, psychologistic, non-​objectivist, and anti-​ realist. The problem concerns the relation between the local dependency of a thick concept, on the one hand, and, on the other, the claim that a thick concept represents ethical knowledge or truth. ‘One’s philosophical beliefs, or approaches, or arguments should hang together (like conspirators, perhaps)’, writes Williams (WME: 186). I want to show how two 14 See ELP: 129–​30, 140–​2; also WME: 209. 15 I am grateful to A. W. Moore for his discussion of thick concepts in Moore (2006). See especially his gloss on ‘embracing’ a concept (2006: 137). In addition, I am grateful to Edward Hall for the suggestion of ‘obscene’ as an example of a thick concept that Williams did not take advantage of himself. See [p. 431] Geraldine Ng of his philosophical approaches—​or, more carefully, the problems with them—​ hang together. I can’t give the notion of thick concepts I’m gesturing at the treatment they deserve here. What’s important for present purposes is that the problem that arises with thick concepts is of a piece with the problem that arises with a historicist interpretation of internal reasons. Both thick concepts and internal reasons are perspective dependent, the one in relation to local ways of apprehending things, the other in relation to an individual agent’s S. The challenge is that, despite their perspectivity, both have a claim to carrying ethical weight, the one in terms of ethical knowledge, the other in terms of normative force. This is the question of how an account of reflective ethical practice, the practice either of a local shared social world or of an individual’s ethical life, can be given. I should like to defend my interpretation of the internal reasons thesis, not by suggesting a complete symmetry between the two lines of thought, but by mentioning some considerations which Williams encourages us to bear in mind in connection to thick concepts, more to illuminate the argument of this chapter for historical thinking. First, Williams writes, with regard to the metaphysics of thick ethical concepts, ‘ “factual” and “evaluative” elements cannot be separated’ (WME: 206). The description of thick concepts as a mix of fact and value helps bring out, though it does not solve, the question of the relations of knowledge or truth to practice, and the normative status of practical ethical claims. Next, he continues, ‘to understand how such a concept can be applied to a new situation it is likely that one will have to grasp its evaluative point, the outlook of the people who use it’ (WME: 206). Williams bases thick concepts on the outlook of people who use them, which at once places thick concepts historically, and reminds us of genuine and pervasive cultural variation. Thick concepts give Williams’s account of ethics rich resources for refuting objectivism, since they invite us to think about the role of culture in understanding more generally. Again, ‘it is perfectly obvious . . . that the thick concepts used in different societies and at different times are not all the same’ (WME: 206).The emphasis on history, at the same time, provides a concrete sense of the variation between social worlds. Finally, Williams draws attention to the fact ‘that these concepts are not simply given, and this leaves a space’, or raises a question, about their justification when we use the concepts in discursive practice (WME: 208). Williams’s emphasis on historical dependence and the pervasive tension in normative claims meets face on the question of how ‘our agreement on truths under thick ethical concepts, or any other perspectival concept, can be coherently understood as subject to local constraints’ (WME: 209). ‘The thick concepts under which we can have some pieces of ethical knowledge are not themselves sustained by knowledge, but by confidence’ (WME: 208). The question is, confidence in what? In these connections I want to suggest that confidence is based on a presupposition that structures thick concepts and their practice. There are, among Williams’s [p. 432] philosophical beliefs, some absolute presuppositions that hang together. With reasons for action, and assuming x serves A’s S, ‘A has a reason to x’ is true on the assumption that A is rational. With reasons for action involving ethical considerations, and assuming x serves A’s S, ‘A has a reason to x’ is true on the assumption that A is in ethical life. With thick concepts, our agreement on truth under a thick ethical concept is based on the assumption of a social world whose outlook we can grasp, although not necessarily share. Now, it is useful to ask, what is involved in a thick concept’s ‘surviving reflection’ (WME: 207)? The answer Williams offers is: ‘It survives reflection just in the sense that we would not have encountered any considerations that led us to give it up’ (WME: 207). Compare this answer to what is involved in a reason surviving reflection: A has a reason to x if that reason serves some element in A’s S. In the case of a social world, a thick ethical concept must survive the test of reflection of the members of that society. In the case of an agent in ethical life, an ethical reason must survive the test of reflective endorsement of RA’s S. In both cases you might say that a kind of ethical confidence is needed: confidence either in a recognizable social world or in an agent’s active participation in ethical life. Williams plays down, where I am making much of, presuppositions which I identify as necessary in order to explain any normative ethical claims as normative at all, assuming the fact of non-​objectivity. Williams’s positive account of ethics is helped by his emphasis on history, where speaking in terms of the individual’s S, local practices, social arrangements, and ways of life as a whole permits him to mobilize these richer ethical resources and express his views in an unassuming way. Because these things are not matters of argument, the philosophical method of historical thinking succeeds where analytic philosophy alone leaves these things unexplained. 7. What Might Ethics Become? Why does philosophy need history? For Williams, the answer is clear—​without involving itself in history, philosophy is ‘likely to leave unexplained many features that provoke philosophical enquiry’ (PHD: 489). Williams put the combination of philosophy and history to varied uses. Historical understanding in its various applications runs like a thread right through all of his philosophy. This chapter has given sharper focus and expression to reasons internalism by showing why it needs history. Williams enriches the mixture, so to speak, of moral psychology and reasons internalism by stirring in some historical thinking as well. Our decisions are always made ‘from here’ (ML: 35). To capture this basic aspect of practical decision-​making, we need to tell a historical story about an agent’s reasons. Despite a huge interest in Williams, his application of history to moral philosophy has remained largely unrecognized for what it is. Contrary to Williams’s [p. 433] Geraldine Ng own testimony, his ethical philosophy is usually not read as a positive contribution to moral philosophy (ELP: 198). My primary aim has been to elucidate and amplify the ‘historicist turn’ in Williams’s moral philosophy that constitutes a significant part of its positive aspect. In light of his ‘historicist turn’, I suggest that Williams’s philosophical method is a mixture of the analytic and non-​analytic. For a definition of what is not analytic, I am helped by P. F. Strawson. Strawson begins his discussion of non-​analytic philosophy by referring to ‘that kind of more or less systematic reflection on the human situation which one finds in the work of, say, Heidegger, Sartre, and Nietzsche’ Such writings, continues Strawson, involve ‘a kind of reflection which can sometimes lead to a new perspective on human life and experience’. The analytical philosopher, by contrast, ‘promises no such new and revealing vision’ Ethics and the Limits of Philosophy is both analytic, because Williams says it is, and non-​analytic, because it is committed to raising ‘the double question of how far reflection commits us and why we should be committed to reflection’ (ELP: 21).16 This chapter has been an inquiry into those agents who are (for the most part) disposed to reflection. So-​called ‘hard cases’ supplied us with a point of departure. Philosophy, believes Williams, can help us discover ‘how truthfulness to an existing self or society is to be combined with reflection, self-​understanding, and criticism’ (ELP: 200). If Williams is an analytic philosopher, then he is not one in Strawson’s mould. Instead, like the non-​analytic philosopher of Strawson’s description, Williams is very much concerned with seeking a new perspective on human life and ethical experience. Philosophy’s ideal of reflectiveness is ‘an ideal acknowledged in the subject’s most central traditions’ (2006e: 212). Unfolding philosophy’s ideal of reflectiveness can be approached in various ways. Compare utilitarians, who also attach special weight to those of our ethical claims that are the product of reflection. Utilitarians privilege the ‘cool hour’, the abstraction that they believe reflection affords. In contrast, for Williams, reflection promises quite the opposite effect: self-​ understanding. This self-​ understanding has radical consequences.17 The heightened consciousness of ethical variation is peculiar to the modern world: ‘Modern life is so pervasively reflective, and a high degree of self-​ consciousness is so basic to its institutions’ (ELP: 3) More emphatically, ‘unselfconscious modernism is a contradiction in terms’ (Williams 2006b: 120). 16 Williams identifies himself as being ‘both deniably and undeniably, an analytic philosopher: deniably, because I am disposed to deny it, and undeniably, because I suspect that few who have anything to say on the subject will accept that denial’ (2006e: 201). Also note Williams’s admission that Ethics and the Limits of Philosophy is ‘on some broad specification, “analytical” ’ (ELP: viii). We are in an ethical condition that lies not only beyond Christianity, but beyond its Kantian and Hegelian legacies . . . we have to acknowledge the hideous costs of many human achievements that we value, including this reflective sense itself ’ (2008 [1993]: 166). [p. 434] Negatively, history tells us that ‘the demands of the modern world on ethical thought are unprecedented, and the ideas of rationality embodied in most contemporary moral philosophy cannot meet them’ (ELP: vii). Positively, Williams reminds us of the possibilities of our human agency, our disposition to reflection. Reflection can be destabilizing, but it can also be, in the light of the greater self-​ understanding it affords, a source of strength. A naturalistic psychology of ethics points to reasons reflective agents in a social world may come to hold. The question is how the relevance of the social norms should be understood. A historical understanding of internal reasons is sensitive to where an individual’s reasons come from and how her reasons are fostered. Social norms are an input into the practical decision-​making of agents assumed to be in ethical life. Assuming an agent is engaged in ethical life, we suppose no given individual will be able to dispense with other-​regarding social reasons entirely, or form the reasons in her S in complete independence of the concerns etc. of other people. Theory adherents might look on this with a cold eye, and ask: is that all philosophy can say about morality? To which Williams replies: I, for one, find it hard to resist Nietzsche’s plausible interpretation, that the desire of philosophy to find a way in which morality can be guaranteed to get beyond merely designating the vile and recalcitrant, to transfixing them or getting them inside, is only a fantasy of ressentiment, a magical project to make a wish and its words into a coercive power. (WME: 216) Williams counsels renouncing the theorists’ magical project. Instead, we should ‘recast our ethical conceptions’ (MSH: 19–​20).18 The chapter has recast reasons internalism, directing us to its more substantive, historical sense, to elucidate and amplify the insight that Williams’s method of historical thinking affords—​the idea that ethics might become so much more.19 References Collingwood, Robin George. 1994. The Idea of History. Oxford: Oxford University Press. Fricker, Miranda. 2022. ‘Bernard Williams as a Philosopher of Ethical Freedom’. In Morality and Agency: Themes from Bernard Williams. Edited by Andres Szegeti and Matthew Talbot, 265–​ Oxford: Oxford University Press. Hall, Edward. 2015. ‘Bernard Williams and the Basic Legitimation Demand: A Defence’. Political Studies 63.2: 466–​80. 18 Also see Williams (1995). 19 I am indebted to Nikhil Krishnan and Matthieu Queloz for their invitation to present an earlier version of this chapter at the Bernard Williams Virtual Workshop on 14 December 2021, to Paul Sagar for being my commentator at the workshop, and to the workshop participants for their generous discussion. [p. 435] Geraldine Ng Koopman, Colin. 2010. ‘Bernard Williams on Philosophy’s Need for History’. Review of Metaphysics 64.1: 3–​30. Korsgaard, Christine, M. 2008 [1996]. The Sources of Normativity. Cambridge: Cambridge University Press. Mackie, J. L. 1990 [1977]. Ethics: Inventing Right and Wrong. London: Penguin [Pelican Books]. McDowell, John. 1999 [1995]. ‘Might There Be External Reasons?’ In World, Mind, and Ethics. Edited by J. E. J. Altham and Ross Harrison, 68–​85. Cambridge: Cambridge University Press. Moore, A. W. 2006. ‘Maxims and Thick Ethical Concepts’. Ratio (new series) XIX.2: 129–​47. Moore, A. W. 2007 [1985]. ‘Commentary’. In Bernard Williams, Ethics and the Limits of Philosophy, 203–​21. Cambridge, MA: Harvard University Press. Nagel, Thomas. 2008 [1996]. ‘Universality and the Reflective Self ’. In The Sources of Normativity. Edited by Christine M. Korsgaard, 200–​9. Cambridge: Cambridge University Press. Strawson, P. F. 1992. Analysis and Metaphysics: An Introduction to Philosophy. Oxford: Oxford University Press. Williams, Bernard. 1993 [1972]. Morality: An Introduction to Ethics. Cambridge: Cambridge University Press. Williams, Bernard. 1995. ‘Ethics’. In Philosophy: A Guide through the Subject. Edited by A. C. Grayling, 545–​82. New York: Oxford University Press. Williams, Bernard. 1996. ‘History, Morality, and the Test of Reflection’. In The Sources of Normativity. Edited by Christine M. Korsgaard, 210–​18. Cambridge: Cambridge University Press. Williams, Bernard. 1998a [1995]. ‘Internal Reasons and the Obscurity of Blame’. In Williams, Making Sense of Humanity, 35–​45. Cambridge: Cambridge University Press. Williams, Bernard. 1998b [1995]. Making Sense of Humanity and Other Philosophical Papers Cambridge: Cambridge University Press. Williams, Bernard. 1999a [1981]. ‘Internal and External Reasons’. In Williams, Moral Luck: Philosophical Papers 1973–​1980, 101–​13. Cambridge: Cambridge University Press. Williams, Bernard. 1999b [1981]. Moral Luck: Philosophical Papers 1973–​1980. Cambridge: Cambridge University Press. Williams, Bernard. 1999c [1995]. ‘Replies’. In World, Mind and Ethics: Essays on the Ethical Philosophy of Bernard Williams. Edited by J. E. J. Altham and Ross Harrison, 185–​224. Cambridge: Cambridge University Press. Williams, Bernard. 2001. ‘Internal and External Reasons, with Postscript’. In Varieties of Practical Reason. Edited by Elijah Millgram, 77–​98. Cambridge, MA: The MIT Press. Williams, Bernard. 2002. Truth and Truthfulness: An Essay in Genealogy. Princeton: Princeton University Press. Williams, Bernard. 2006a. ‘An Essay on Collingwood’. In The Sense of the Past. Edited by Myles Burnyeat, 341–​60. Princeton: Princeton University Press. Williams, Bernard. 2006b. On Opera. New Haven and London: Yale University Press. Williams, Bernard. 2006c. Philosophy as a Humanistic Discipline. Edited by Adrian Moore. Princeton: Princeton University Press. Williams, Bernard. 2006d. The Sense of the Past. Edited by Myles Burnyeat. Princeton: Princeton University Press. Williams, Bernard. 2006e. ‘What Might Philosophy Become?’ In Williams, Philosophy as a Humanistic Discipline, 180–​99. Princeton: Princeton University Press. Williams, Bernard. 2007 [1985]. Ethics and the Limits of Philosophy. London: Fontana. Williams, Bernard. 2008 [1993]. Shame and Necessity. Berkeley and Los Angeles: University of California Press. Williams, Bernard. 2009. ‘A Mistrustful Animal’. In Conversations on Ethics. Edited by Alex Voorhoeve, 195–​214. Oxford: Oxford University Press. [p. 436] The Art of the Possible Williams on Political Judgement and the Historical Perspective Ilaria Cozzaglio and Amanda R. Greene 1. Introduction Bernard Williams’s approach to political philosophy is notable for its opposition to moralism and attention to historical context. It is tempting to see these qualities as an outgrowth of his anti-​system approach to ethics. After all, Williams presents his alternative approach to ethics in deliberate contrast to what he disparagingly calls ‘the morality system’, drawing on his critiques of Kantianism and utilitarianism. However, it would be a mistake to see Williams’s views on politics as being merely derivative of his views on ethics, and this is partly due to the different role played by history and historical thought. The framework that Williams develops for dealing with the normative questions that arise in politics is tied to a historical sensibility that illuminates the limits of political action, both now and in the past. Nevertheless, Williams insists that there are certain ethical judgements about politics that we must make, including ones about societies entirely different from ours. The key to making these judgements, however, lies in starting from political reality. The result is a historically oriented view that is as much opposed to conservatism as it is to moralism. Our aim in this chapter is to trace the contours of this view, showing why it continues to pose a challenge to contemporary political philosophy. We take up this task by way of examining Williams’s account of political judgement and its connection to the historical perspective. To have an account of political judgement, as we understand it, is to have a framework for evaluating political actors and political orders. For Williams, politics raises certain normative questions that can and should be distinguished from questions of interpersonal morality. Williams is especially interested in the pressures and responsibilities that bear down on those who wield political power, and he wants to build these factors into the theory from the start. Throughout his work, accordingly, he shifts the focus away from justice and towards legitimacy. Rather than asking about the justice of different political systems such as democracy or egalitarianism, or exploring rights and law as moral ideals, Williams instead asks about the legitimacy of past and Ilaria Cozzaglio and Amanda R. Greene, The Art of the Possible In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. [p. 437] Ilaria Cozz aglio and Amanda R . Greene present political orders, viewed in light of their historical circumstances. This deliberate incorporation of history means that when it comes to questions such as whether a certain distribution of resources is unjust, or whether a certain scheme of rights is unfair, Williams invites us to reformulate the questions in light of political circumstances. He thinks we must respect the limits of political action, where these limits emerge from certain general truths about power, as well as certain historical facts. This leads Williams to build upon the political thought of Thomas Hobbes and Max Weber, as penetrating theorists of power. Williams’s political philosophy owes as much to them as his ethics owes to Plato and Nietzsche. Moreover, Williams’s challenge to political moralism has prompted new interest in political realism, as well as in political normativity more broadly. In offering us a theory of political judgement, then, Williams is setting out the grounds and the terms according to which we are in a position to evaluate political orders in terms of their legitimacy. According to Williams, these political judgements have two aspects. One is historical and contextual, taking seriously the actual beliefs of subjects about their regime even if outsiders find the beliefs to be mistaken or objectionable. The other aspect is universalist, since it is based on a definition of a political relationship understood in contrast to domination. The normativity of a political relationship, as Williams defines it, consists of identifying the first political problem—​namely, securing peace and the conditions of cooperation—​and acknowledging that tyranny does not count as a solution to this problem, regardless of the circumstances. Ultimately, it is this very combination of contextualism and universalism that allows Williams to elaborate a political theory that is historically grounded without losing its critical edge. In so doing, he intends to refute complaints that political realism is quietist, conservative, or otherwise biased towards the status quo. Overall, Williams seeks to show that his anti-​moralism is not a matter of rejecting moral considerations in politics, but rather a refusal to take independent moral principles as a starting point without considering the historical reality that shapes them. Our discussion proceeds as follows. First, we set out the key features of Williams’s historical perspective on political theory, explaining why his approach amounts to such a deep challenge to current political philosophy (section 2). Then we examine more closely the duality that characterizes Williams’s theory. On the one hand, he leans on universal principles that distinguish political orders from systems of domination (section 3). On the other hand, he incorporates historical context in his account of legitimacy (section 4). It is the duality of universalism and contextualism, we argue, that allows Williams to condemn tyranny. A consideration of human rights, in particular, leads Williams to emphasize that there are certain judgements we must make about other societies (section 5). This anti-​quietist imperative, we argue, illustrates the subtle but unavoidable connection between political judgement and political action (section 6). It also highlights Williams’s methodological objection to a division of labour according to which judgement [p. 438] calls for philosophical analysis whereas action calls for political analysis. Instead, Williams thinks both philosophy and politics matter for both judgement and action (section 7). For Williams, the unique nature of political judgement consists in the fact that leaders are ‘politically highly answerable’ to an audience that is historically situated, a notion that sheds new light on the perennial question of political justification (section 8). 2. Williams’s Anti-​moralist Account of Legitimacy In his later years, Williams came to describe himself as a political realist. He is put off by what he calls political moralism, a family of views that avow ‘the priority of the moral over the political’ (2005: 2). According to Williams, there are two prominent offenders when it comes to this dubious priority. Utilitarianism views politics as an instrumental enactment of moral principles, and political liberalism views politics as a domain that is structured by prior moral principles (2005: 1–​2). Williams has several complaints about political moralism. First, he criticizes the tendency to rely on an ideal of universal morality while neglecting the historical emergence of these standards. Williams says, ‘Moralistic liberalism cannot plausibly explain, adequate to its moral pretensions, why, when, and by whom it has been accepted and rejected’ (2005: 9). Here he has in mind liberals like Thomas Nagel (Williams 2005: 65–​7, 130–​1) and John Rawls (Williams 2005: 2).1 Second, Williams faults the liberal approach to politics for its sharp separation of principles and interests. He thinks that this separation—​on full display in the work of Ronald Dworkin—​leads them to view ‘conflictual political thought in society in terms of rival elaborations of a moral text’ (2005: 12, 115–​27). This outlook in turn encourages some moralists to seek consensus through dialogue, imagining that the problems that arise are to be handled by an ethics of disagreement. Williams thinks, however, that politics calls for something more like an ethics of competition (2005: 12). He says, ‘We should not think that what we have to do is simply to argue with those who disagree: treating them as opponents can, oddly enough, show more respect for them as political actors’ (2005: 13). Thirdly, Williams argues that moralist theories of legitimacy are not sensitive enough to the pressures that bear down on political actors, such as bare political survival (2005: 12–​13). Fourth, he thinks that the ends of political action are inherently indeterminate, subject to various developments in both history and the history of ideas. The unfolding of history is so contingent, in fact, that nothing in the shape of political moralism could provide resolution to the ethical questions that arise in politics In order to address all of these problems with political moralism, 1 On the question of whether Williams is right to view Rawls’s approach as a form of moralism, see Larmore (2020). [p. 439] Ilaria Cozz aglio and Amanda R . Greene Williams thinks, we need an approach that affords ‘greater autonomy to distinctively political thought’ (2005: 3). In his quest to move away from political moralism, Williams articulates a new way of thinking about legitimacy, encapsulated by what he calls the basic legitimation demand (BLD).2 He claims that what distinguishes a legitimate from an illegitimate state is whether the state has provided an acceptable solution to the ‘first political question’. Williams defines solving the first political question in broadly Hobbesian terms: ‘the securing of order, protection, safety, trust, and the conditions of cooperation’ (2005: 3).3 Assuming that order has been secured, Williams then claims that ‘the state has to offer a justification of its power to each subject’ in order to meet the BLD (2005: 4). And this justification, in turn, needs to ‘make sense’ to the subjects as ‘an intelligible order of authority’ (2005: 10). Moreover, the acceptance of such justification cannot be produced by the same power that is in need of justification. Williams calls this restriction the critical theory principle (2005: 6). By contrast, illegitimate states are those whose solutions only recreate the problem of securing order and protection. Insofar as states maintain their power by fear, or uphold their rule more covertly through indoctrination or manipulation, they are in a relationship of domination with their subjects. Since it is precisely this entrenched domination that politics seeks to avoid, Williams thinks, the basic legitimation demand has not been met in these circumstances. This approach is quite different from the predominant alternatives. Unlike the hypothetical choice situation that we find presented in social contract theory, it does not rely on an idealized reconstruction of individuals as having chosen their own subjection to authority. Across different versions of the social contract—​those belonging to Hobbes, Locke, Rousseau, and Kant (Williams 2005: 29–​30, 32)—​ the individuals comprising the political community are construed as deliberators whose choice to establish a common authority makes sense in terms that are uniform across individuals and across time. Williams’s view, by contrast, allows for different historical permutations of what counts as making sense to subjects as intelligible political order. This leads to political judgements that are more sensitive to the particularities of a given political community and its history. Williams’s view, therefore, can avoid the complaints made about other approaches, such as that they amount to an imposition of the philosopher’s values or that they neglect cultural differences.4 2 For the debate on whether the basic legitimation demand is actually a form of moralism, see Larmore (2013, 2020), Hall (2015), Forst (2017), and Owen (2018). 3 On the role that security and order play in Williams’s theory, see Hall (2020) and Scheuerman (2018). 4 On the role of the theorist’s perspective concerning the determination of state legitimacy, see Horton (1992, 2012). [p. 440] 3. Universalist Political Judgements While the basic legitimation demand includes some contextualist aspects that require historical sensitivity, this does not eliminate the possibility of making critical judgements that have universal applicability. For Williams, it is precisely the distinctiveness of politics that enables the judgements to have universal bearing. The basic legitimation demand has universal bearing, for Williams, in the sense that a regime’s legitimacy is to be measured against its capacity to solve the first political question in an acceptable way. This allows him to assert that ‘there are certain judgements we can and should make about the past’: that is, about societies whose circumstances and values are very different from ours (Williams Even in contexts so foreign that they test the limits of our comprehension, Williams thinks we can still ascertain whether the state has solved the ‘first political question’: namely, the establishment of peaceful order. By making the problem of order part of the definition of politics, Williams thinks he can claim that the demand for legitimation of political power does not arise because of a pre-​political morality, thereby affording it the right sort of autonomy from the rest of morality.5 Williams’s debt to Hobbes is explicit, but the points of divergence serve to highlight the duality in Williams’s account of legitimacy. On the most basic level, Williams endorses Hobbes’s insight that if there is to be anything that is absolutely essential to and definitive of politics, it must be the securing of order. He says, ‘I identify the ‘first’ political question in Hobbesian terms [and claim that] it is logically “first” because solving it is the condition of solving, indeed posing, any others’ (Williams 2005: 3). Hobbes famously contrasts the establishment of political order with the ‘ill condition which man by mere nature is actually placed in’: namely, a war of all against all (1994: 78 [XIII.13]). Williams echoes this view by saying that, as long as order has not been secured, the situation really cannot be called a political condition, but rather, something more like a state of war. And yet, Williams deliberately departs from Hobbes by denying that fear of the sovereign can be the basis of legitimate order, saying that this would be indistinguishable from tyranny. For Hobbes, it is only the fear of the sovereign that can secure peace, so the presence of this fear should be seen as a design feature of the theory rather than a design flaw. For Williams, however, meeting the basic legitimation demand requires that subjects’ obedience not be derived from fear. This departure has a universalist aspect, which we identify here, and a historical aspect, which we address in the next section. 5 For an assessment of the options available to Williams in developing a realist normativity, see Cozzaglio and Greene (2019). For critical discussions and developments of Williams’s notion of legitimacy, see Sleat (2014), Sagar (2018), and Cozzaglio (2022). On the broader debate on whether political normativity is distinctive, see Estlund (2017), Leader Maynard and Worsnip (2018), and Jubb (2019). [p. 441] Ilaria Cozz aglio and Amanda R . Greene Williams departs from Hobbes while retaining a certain universalism when he says, ‘One of the few necessary truths about political right is that it is not merely might’ (2005: 135). Therefore, it cannot be considered politics—​as opposed to domination—​when the power of some over others rests on sheer might: If the power of one lot of people over another is to represent a solution to the first political question, and not itself be part of the problem, something has to be said to explain . . . what the difference is between the solution and the problem, and that cannot simply be an account of successful domination. (2005: 5) This political axiom holds, Williams thinks, regardless of the value system of a particular society. Accordingly, he explains the category of what ‘makes sense’ in a way that ties it to the securing of order without introducing tyranny: A given historical structure can be (to an appropriate degree) an example of the human capacity to live under an intelligible order of authority . . . It makes sense (MS) to us as such a structure . . . [While] situations of terror and tyranny MS, . . . [the] question is whether a structure MS as an example of authoritative order. This requires . . . that there is a legitimation offered which goes beyond the assertion of power. (2005: 10–​11)6 Williams is well aware that a given system of order can make perfectly good sense as tyranny. However, this would not count as making sense as intelligible authority. While the specific terms in which a system of order comes to make sense are relative to a given society, the category of making sense as intelligible order is not. According to Williams, then, authoritative order can be distinguished from domination and tyranny—​always, everywhere, and for everyone (2005: 69). This is why there are certain political judgements that we can make about other societies despite the difference in context. 4. Contextualist Political Judgements The contextualist aspect of Williams’s approach, however, places considerable limits on making political judgements about other societies. Basically, this is because what makes sense in one political community as a justification of political order may not make sense in another one. In other words, the historical factors that determine the particular value systems present among a group of subjects also serve to constrain what can count as ‘making sense’ to those subjects. This 6 For further discussion of ‘making sense’ and its normativity, see Williams (2002: ch. 10, esp. 234–​5). [p. 442] historical indexing of what counts as making sense is the second departure from Hobbes. What counts as an acceptable solution, for Hobbes, is something that can be put in the same terms for anyone, in any historical context. For Williams, however, the standard of acceptability is itself a function of certain historical developments. As such, it requires the theorist to engage in historical interpretation. This departure from Hobbes reflects the influence of Max Weber. Weber foregrounds the link between legitimacy and people’s beliefs in his very definition of legitimate rule: that is, when ‘the manifested will (command) of the ruler or rulers is meant to influence the conduct of one or more others (the ruled) and actually does influence it in such a way that their conduct to a socially relevant degree occurs as if the ruled had made the content of the command the maxim of their conduct for its very own sake’ (1978: 946). Thus, for Weber, a system of rule becomes legitimate whenever subjects take the authority to be valid by their own lights (Greene 2017). Williams adopts this sensitivity to the subjects’ perspective in claiming that legitimacy depends on what makes sense to a certain group of people in a particular historical and cultural environment. Williams’s claim that political judgements are partly dependent on historical and cultural circumstances is made concrete when he proposes that ‘LEG +​ Modernity =​Liberalism’ (2005: 9). This formulation means that when the conditions of modernity are in place, only the principles of liberalism can justify a legitimate political order. Given those historical and cultural conditions, particularly the understanding of freedom and agency, other legitimation stories would not make sense (Queloz 2021: 238–​41). Hence, the specific requirements for a state to be legitimate can vary in time and space. In this way, the Williamsian category of making sense incorporates a consideration of the historical context, including the value systems, of a particular political order. It is here that Williams introduces a distinction between evaluative and normative assessments. He claims that the category of making sense is not normative in the sense used by contemporary political philosophers. In other words, it is not a matter of justice, rights, and morality where these are defined independently of the context: One can say, as I have said, that ‘MS’ [making sense] is itself an evaluative concept; certainly, it is not simply ‘factual’ or ‘descriptive’. . . . What it certainly is not, is normative: we do not think, typically, that these considerations should guide our behaviour, and there is no point in saying that they ought to have guided the other people’s behaviour . . . But when we get to our own case, the notion ‘MS’ does become normative, because what (most) MS to us is a structure of authority which we think we should accept. (Williams 2005: 11) As we understand Williams, he thinks that when it comes to evaluating our political order, liberal principles make sense to us as suitable and appropriate, and [p. 443] Ilaria Cozz aglio and Amanda R . Greene are normative in our case (assuming arguendo that we are liberals). But in coming to see that other systems of authority make sense to others, such as those based on honour or tradition, the conclusions we reach are evaluative rather than normative. The reason, it seems, to call them evaluative is to mark the fact that such assessments always go beyond mere description. What is being evaluated, in fact, is whether the values that evidently guide the political order correspond in a suitable way to the values that are shared in that political community; otherwise it would not make sense. This sort of congruence, then, is the object of evaluation. This shows the sense in which, for Williams, we are able to make judgements about other societies that count as evaluative. But no judgement that depends on our value system, or that of others, can be normative in the sense of being universally applicable. Williams thinks that there is ‘no problem’ in drawing such a distinction between evaluative and normative judgements (2005: 11), and later we come back to the grounds for this confidence. For now, it is enough to see that this distinction means that it is not appropriate for us to judge the political legitimacy of other states, especially those that existed in the past, solely on the basis of our current value system, of what makes sense to us, here and now. As Williams admits, we ‘may indeed think, in light of our entzaubert state, that some of what makes sense to them does not MS to us because we take it to be false, in a sense that represents a cognitive advance’, but this should not lead us to conclude that those states were illegitimate 5. Relativism and Human Rights To some, it might sound like relativism to claim that what counts as a justification of power depends on the contingencies of history and culture. To be sure, Williams’s account is sensitive to historical context by means of the ‘making sense’ standard. And yet, it also contains elements that are universal, as we have seen, since some verdicts transcend the frame of meaning for a particular society.7 Therefore we think it is better to think of the apparently relativist elements in Williams’s account as a form of anti-​moralism, where this does not amount to an overall rejection of moral considerations in politics. In fact, what makes sense to us includes a set of moral principles we believe to be true. Instead, the anti-​moralism that we find in Williams corresponds to a rejection of viewing ‘one’s morality as universally applicable to anyone’ (2005: 67). 7 Williams foreshadows this nuanced view in his essay on the relativism of distance: ‘One can defend a relativistic view of justice. There is some pressure, if one thinks historically at all, to see modern conceptions of social justice . . . as simply not applying to hierarchical societies of the past . . . Yet there are strong pressures for the justice or injustice of past societies not merely to evaporate in the relativism of distance’ (1985: 165–​6). [p. 444] Williams’s unique brand of anti-​moralism is illustrated in his discussion of human rights. He draws again on the idea of the first political problem, and what would count as a solution: Our most basic conceptions of human rights are connected with our ideas of what it is for the supposed solution, political power, to become part of the problem. Since—​once again, at the most basic level—​it is clear what it is for this to happen, it is clear what the most basic violations of human rights are. In the traditional words of the Catholic Church, the most basic truth on this matter is quod semper, quod ubique, quod ab omnibus creditum est. (2005: 63) For Williams, basic human rights and the basic legitimation demand are correlates. As with judgements of legitimacy, when it comes to assessing whether human rights have been violated we should not refer to our own value system as a touchstone. We should rely only on what is acknowledged always, everywhere, and by everyone. So the verdicts in both domains—​legitimacy and human rights—​ present us with the same duality of contextualism and universalism.8 Williams accordingly limits his list of human rights to protection from clear abuses of power—​ namely, ‘torture, surveillance, arbitrary arrest, and murder: the world of Argentina under the junta, the story . . . of those who disappeared’ (2005: 69). Absent from Williams’s list are demands that are not typically recognized outside a modern liberal state, such as gender equality, abortion, and assisted suicide (2005: 65), as well as religious tolerance (2002: 138) and freedom of speech (2002: 212–​19). He also excludes the social and economic demands now considered by many to be human rights, since governments should not be viewed as tyrannical if they fail to meet these demands: Any modern state will seek to forward [freedom from unemployment and deprivation] by political means. But its unsuccess as such does not necessarily show that it is a despotic government, whereas its misuse of its own powers or its failure to curb people’s use of their powers to subordinate others is directly that, for that is what the state first is. (2005: 60–​1) In confining human rights to a more limited range of government responsibilities, therefore, Williams relies again on contrasting politics with domination. This leaves room, however, for a political context in which social and economic rights are recognized and upheld in accordance with the local legitimation. This way of thinking about a standard of legitimacy, including the place of rights within it, stands in stark contrast to alternative views that embrace equality and 8 Williams thinks that any attempt to base human rights on our own value system would be not only inappropriate but also fruitless in terms of our moral and sociological understanding (2005: 66). [p. 445] Ilaria Cozz aglio and Amanda R . Greene justice. Liberals like Nagel and Dworkin consider failures of distributive justice to be failures of legitimacy because they fail to manifest equal concern. But for Williams, the important political questions about the scope of individual rights are a matter of historical context (we return to this below). When it comes to deeming a regime illegitimate, there is an important difference between failing to guarantee security and failing to guarantee justice. Nevertheless, it should be said that Williams later associates his critical theory principle with an ideal of emancipation that he thinks reflects ‘the most basic sense of freedom’—​freedom from being ‘in the unrecognized power of another’ In this respect his view might seem closer to those theorists who think legitimacy is about freedom, such as Philip Pettit (2012). Still, Williams’s position remains distinct insofar as it is derived from axioms about political power rather than ideals of justice or freedom. Focusing on the question of rights, therefore, allows us to see just how much Williams’s theory of political judgement differs from the alternatives. With this perspective, one might now wonder: why does the universalist aspect of political judgement, when applied to the case of human rights, not amount to reintroducing the problem of ideological moralism? Recall that Williams wants his political realism to retain critical traction without relying on particular value orientations, such as liberal democracy. But how can this be done? Here is how Williams envisions it: It seems to me sensible, both philosophically and politically, to make our views about human rights, or at least the most basic human rights, depend as little as possible on disputable theses of liberalism or any other particular ideology. We should rely, so far as we can, on the recognition of that central core of evils (quod semper, quod ubique, quod ab omnibus . . .) (2005: 74) Here Williams claims that when it comes to human rights, our verdicts should combine what is universal with what is contextual. In order to do this, he identifies three ingredients: a recognition of core political evils, a critical understanding of legitimation, and a consideration of modern conditions (2005: 74). Together these give rise to political judgements that can avoid the charge of ideological moralism, according to Williams. Lest someone dismiss this as easier said than done, Williams points to a success story: Judith Shklar’s liberalism. For Shklar, our safety is always at risk of being threatened by those exercising political power. According to Williams, Just as [Shklar’s liberalism] takes the condition of life without terror as its first requirement and considers what other goods can be furthered in more favourable circumstances, it treats each proposal for the extension of the notions of fear and freedom in the light of what locally has been secured. It does not try to determine [p. 446] in general what anyone has a right to under any circumstances and then apply it. It regards the discovery of what rights people have as a political and historical one, not a philosophical one. (2005: 61) It is important to see that the approach Williams praises is not construed as merely considering feasibility at the implementation stage. The idea is not, first, to identify the eternal guiding principles of political morality, and then, secondly, to revise or suspend them in light of the circumstances. Rather, the idea is to start from ‘what locally has been secured’ in elaborating those principles in the first place. This is related to what Williams calls a ‘bottom-​up, not top-​down’ approach At the bottom, presumably, is the question of what makes sense to actual subjects, where this depends on interpreting and analysing their beliefs and values. If we were to start at the top with abstract principles of justice, then we would lose sight of the way in which order is maintained over time through a local legitimation. After all, it may turn out that the ‘sense-​making’ resources possessed by a given set of subjects is limited. Their overall world view may be impossible to shift in certain directions or by certain means, perhaps because the influence would be perceived as external and thus suspect. For Williams it is important to have sincere respect, borne of genuine understanding and appreciation, for the idiosyncratic terms on which a non-​tyrannical order has been established before seeking to improve it. This is why Williams praises Shklar’s historical approach, since it starts from what makes sense locally and builds upon it: [T]‌he particular arguments that carry forward liberal policies in particular situations must be not just practically but conceptually a matter of those circumstances. This is the truth in anti-​universalism, insisted on by some of liberalism’s opponents. The liberalism of fear can combine this with its own universalism. It does so in the form of its constant reminder of the reality of politics, that there is a political reality out there. (2005: 61) Again, Williams recommends combining universalism with the element of truth in anti-​universalism. But what does Williams have in mind when he speaks of the ‘reality out there’, calling it a form of universalism to bear this reality in mind? Certainly the reality includes the ever-​present danger of the abuse of power. According to Shklar, ‘One must put cruelty first and understand the fear of fear’ One of the most important lessons of history—​according to Shklar and Williams—​is the fact that there is always the possibility of sliding back into tyranny, even for liberal democracies. But the reality also includes the fact that political actors operate under certain constraints. Their actions are curtailed—​often severely—​by the limits of political possibility. We will address the importance of understanding the limits of political possibility in the next section. For now, it is [p. 447] Ilaria Cozz aglio and Amanda R . Greene enough to observe that invoking the universal human experience of political fear allows Williams to defend his view against the charge of ideological moralism. So far, we have shown that on Williams’s theory, judgements about human rights, as with judgements about legitimacy, have both universalist and contextualist aspects. This explains why, according to Williams, we can make judgements about societies in the past without falling into ideological moralism or neglecting the importance of history. However, Williams gestures towards a further requirement: not only is it possible to make judgements about other societies, but there are cases in which we must do so (2005: 67). We call this the anti-​quietist imperative. Where does this injunction come from? Why, for example, can we not just remain silent in the face of human rights violations? Would this quietism not be more consistent with a respect for history? 6. Judgement and Political Action Williams could have ended his account here: that is, after explaining why there are certain judgements that we can make. But he goes on to say that there are judgements that we must make: [W]‌e need to emphasize a particular kind of judgement which we can, indeed must, make about the past: those that we make in virtue of what I called the first question of politics, the question of order, and the danger related to that question, that the solution may become part of the problem. (2005: 69) Evidently the judgements that we must make pertain to those situations in which the solution has become part of the problem. But what does Williams mean when he says we must say whether a particular political situation counts as a case of domination? If we were to pose the question about a certain situation and fail to draw the right conclusion, or fail to draw any conclusion at all, what kind of failure would that amount to, according to Williams? Granted, Williams’s account is less clear on the imperative to judge than it is on the ability to judge. Still, we think that his remarks give us some indication. Here it is crucial to look at the connection between political judgement and political action—​in particular, the responsibilities that follow from recognizing the limits and hazards of political action. For Williams, when it comes to politics, there is an intimate connection between judgement and action. However, it is important to avoid two ways of misunderstanding this connection. The first misconstrual would be to think that there are judgements that we must make only in so far as, and because, there are actions that we must take. Stated so broadly, this idea would be misleading. For Williams, there are some judgements that we must make that are about the past, where our actions [p. 448] can change nothing. But there is a grain of truth in the misconstrual, which is that our political judgements are bound to reflect considerations that are related to action. For example, sometimes our judgements about whether a particular political action counts as an abuse of power depend on the gravity of the threat that is to be averted. Williams thinks of this as a question that cannot be settled simply on the basis of principle: [Governments] may use . . . ruthless methods against subversives or threatening revolutionaries. Are such measures in themselves violations of human rights? If they are, are they violations justified by emergency? . . . Any state may use such methods in extremis, and it is inescapably true that it is a matter of political judgement, by political actors and by commentators, whether given acts are part of the solution or of the problem. (2005: 70) Williams’s point is that there may come a time to set aside restraint in order to maintain order. He even mentions Carl Schmitt’s (2007) argument that as sovereignty becomes eroded under liberal parliamentary systems, it must be reasserted through emergency executive power. Whether such emergencies count as an abuse of power is a matter about which commentators can make judgements, Williams thinks, even if political action is not open to them. The point is that while judging need not always be for the sake of acting, judging always involves assessing and comparing the likely effects of political action (or inaction, as the case may be).9 Here Williams is echoing Weber’s idea that a responsible political leader is bound to use coercive means only within the limits of a sense of proportion. For Weber, an irresponsible politician is one who overreaches in their exercise of power, taking decisions and actions that are unsuitable in light of the political goals that are to be achieved (2004: 76). Another potential misconstrual is to think that whenever we arrive at a suitably contextualized judgement, we straightaway have a basis for action. Stated so broadly, again, it would miss the subtlety of Williams’s view. As Williams emphasizes, judgements about human rights violations do not and should not necessarily lead to action. For example, when it comes to foreign intervention, we may arrive at a judgement that the solution has become part of the problem and, yet, still decide not to intervene, thereby treating the question of what to do as distinct. It is important to see that, for Williams, the decision not to intervene would not undermine the judgement that the solution has become part of the problem. 9 Williams is also interested in the psychology and dispositions of politicians who must do morally disagreeable things if they are to pursue the moral aims of politics (1978: 64–​5, 68–​9). Since we are focused on the nature of justification and evaluation, we do not have space to address the psychological dimension of Williams’s views on political judgement, though we think it is interesting and important. [p. 449] Ilaria Cozz aglio and Amanda R . Greene The correct way to see the connection between judgement and action, instead, is to notice that the special nature of political action bears on judgements about whether a regime is legitimate and on judgements about how to respond, though it does so in different ways. Consider first the judgements about whether a regime is legitimate. Here the special nature of political action is relevant in the sense that we must consider the limits of politics. As evident already in Williams’s comments about liberal restraint, judgements about what to think must take into consideration the limits of political possibility, which includes what people think of their regime. This is because whether a government counts as making sense to subjects is constrained by what subjects actually believe, what they actually value, and how they actually see the world. How subjects view their regime and its actions affects what we conclude about its legitimacy. These sorts of considerations also matter when it comes to deciding how to respond. However, Williams here cautions against applying the principles pertaining to the morality of private relations. In an essay on humanitarianism, Williams says: [My] question . . . is how far decisions to engage in international rescue can be helpfully modelled on private decisions under the moral principle of rescue . . . My answer is: not very far . . . We should conclude that we obtain the right slant on these questions by seeing such decisions [to intervene] as political decisions. In the analysis that follows, Williams points out that, for any country powerful enough to intervene effectively, the decision to do so cannot but be political. For Williams, this is basically because intervention is fundamentally the use of might to counter might. As he observes elsewhere, ‘government is in the first instance the assertion of power against other power’ (2005: 63). But as Williams goes on to explore the disanalogies between humanitarian intervention and private moral rescue, it is notable that they do not all stem from whether the decision is to be made by a government. What makes the decision political arises from a constellation of factors: someone has to decide to do it, the action has to be collective and coordinated, the action necessarily involves questions of power and authorization, and most importantly, those who intervene undertake a certain political responsibility (Williams 2005: 148–​50). All of this means that if we were to think of intervention chiefly in terms of the moral principle of rescue, we would be ignoring the limits of political possibility. Therefore, a judgement about political intervention is bound up with our sense of political possibility. Here the special nature of political action is relevant in the sense that it generates an ethic of responsibility. For Williams, we can view responsibility in the domain of politics as implying its own ethos, its own standard of correct action. He likens it to Max Weber’s ethic of responsibility: [p. 450] I want a broader view of the content of politics, not confined to interests, together with a more realistic view of the powers, opportunities, and limitations of political actors, where all the considerations that bear on political action—​both ideals and, for example, political survival—​can come to one focus of decision . . . The ethic that relates to this is what Weber called Verantwortungsethik, the ethic of responsibility. (Williams 2005: 12)10 Here Williams is developing a special sense of responsibility that applies to political decisions simply in virtue of their being political decisions.11 In other words, the evaluative standard for political action is coeval with political decision-​making, where consequences loom large. This is especially clear in matters of intervention, where Williams warns against cases in which ‘intervention will make it worse’ Taking responsibility, however, also means responding appropriately when the situation is particularly outrageous: that is, when the violation of human rights is so severe that the political order approximates a state of war (2005: 73). A responsible consideration of consequences, therefore, may require or prohibit intervening, depending on the circumstances. But the scope of consequences that responsible political actors must attend to is much wider than the effectiveness of the intervention. It includes the risks related to ‘involvement in foreign wars, loss of prestige in case of failure, antagonizing other states who think that their interests are involved, and so forth’. Any responsible government must weigh these issues, says Williams, since ‘that is part of what it is for this to be a political decision’ (2005: 150). In addition, the relevant consequences of political action go beyond questions of harm and geopolitical risk to encompass questions of leadership. When the intervener is a state, a government must justify to the public the decision to expose some of its citizens to death (Williams 2005: 150). The sending of troops, for example, must be something the government can ‘get away with’, in the sense that it will not lead to its undoing (2005: 150–​1). A government must consider how its actions will be perceived by the subjects over whom it exercises power and claims authority. While noting that former US President ‘Nixon was often criticized for the cynicism of this’—​that is, calculating what he could get away with and still hold onto power—​Williams himself approves of Nixon’s attention to perception as a matter of political responsibility: ‘The question of how it will play in Peoria [a town composed of ordinary voters] . . . can involve a consideration of political right, as well 10 The reference is to his 1919 lecture, Politics as a Vocation (Weber 2004). 11 Geoffrey Hawthorne sees Williams as developing an ethic of political responsibility that goes beyond Weber. He says, ‘[Weber,] apart from warning of the demonic force of violence, did not say what the political responsibility he talked of was, and to whom. For Williams, these were the central questions’ (Hawthorne 2005: xii). While we agree that these are the questions that interest Williams, we do not agree that Weber lacked answers. We have argued elsewhere that Weber’s account of political normativity is superior to that of Williams, in fact, on precisely those dimensions that drive Williams to be a political realist. See Cozzaglio and Greene (2019). [p. 451] Ilaria Cozz aglio and Amanda R . Greene as of expediency’ (2005: 151). The perception of the victims also matters, since the intervention would be occurring ostensibly on their behalf. It would be bad, for example, if the victims did not see themselves as victims, because then foreign intervention would be hard to distinguish from ‘ideological imperialism’ (2005: 73). 7. On the Supposed Division of Labour between Politics and Philosophy When Williams urges caution concerning humanitarian intervention, we should not lose sight of the underlying methodological mistake he is highlighting. The deeper point is that it would be a misleading distortion to accept a division of labour between philosophy and politics. In this way he again affirms the Weberian maxim that politics is not free from ethical considerations, but rather infused with a certain sort of normative reasoning (Wolin 1981). According to Weber, the point is precisely to understand ‘what is the true relationship between ethics and politics’, and the response to such a question provides the ground for establishing the standards of a good politician (2004: 80). Here is how Williams puts it: Whether it is a matter of philosophical good sense to treat a certain practice as a violation of human rights, and whether it is politically good sense, cannot ultimately constitute two separate questions. The first question that we have to ask, I said, is: what is actually going on? Which includes: how is it to be interpreted? The second question is: what, if anything, can we do about it? It should be obvious that this must be on every occasion a political question. (2005: 72) When Williams says these are not two separate questions—​that is, whether it is philosophically sensible and whether it is politically sensible—​should we understand these as corresponding to the two questions we must ask ourselves: what is happening and what should be done about it? No, it would be a mistake to understand Williams that way. The ‘not two separate questions’ claim that Williams makes should not be taken to imply that the question of what is happening and the question of what to do are not distinct. The questions of what to think and what to do are two different questions, and each of them involves a philosophical and a political sensibility. For example, with regard to a foreign intervention case, we can arrive at a judgement—​say, that what is going on is that the solution has become part of the problem—​and yet still decide not to intervene, thereby treating the question of what to do as distinct. So the two things that are not separate, according to Williams, are philosophical good sense and political good sense. What Williams is objecting to is a division of labour between the domains of philosophical and political analysis, where each one excludes the other. In other words, he refuses to accept the view that the domain of judgement is philosophical [p. 452] but not political, whereas the domain of action is political but not philosophical. This simple contrast between philosophical-​not-​political judgement, on the one hand, and political-​not-​philosophical action, on the other hand, is what Williams urges us to reject. He thinks that it is misguided—​dangerous, even—​to accept this division of labour. Instead, Williams urges us to recognize that the appraisal of reality, including what is politically achievable in a particular time and place, is not something that comes after the normative judgement. It is part of the normative judgement. Having registered this methodological point, it is now possible to return to the ethic of responsibility and ask if it pertains only to action. Though tempting, we think that restricting questions of responsibility to the domain of action would be too simple, missing the nuance of the view. For Williams, there is a sense in which responsibility also pertains to judgements about what to think—​in terms of answering the question: what is going on?—​apart from deciding what to do. This comes out clearly when he says, ‘Franklin D. Roosevelt famously said of Somoza, the ghastly dictator of Nicaragua, “He is a Son of a Bitch, but he is our Son of a Bitch.” This can, on some occasions, be the correct attitude to take’ (2005: 73). Here Williams speaks of a standard of correctness for judgements, apart from a standard of correctness for action. Suppose that Williams imagined Roosevelt calculating the limits of American power abroad and concluding that all of the American government’s alternatives to working with this dictator were worse than working with him. Then the responsible way to proceed in these circumstances would be to think of the dictator as an ally for certain purposes. By asking us to imagine an actor mindful of constraints while still taking responsibility, Williams illustrates the way in which facing up to political reality is bound to shape our judgements as well as our actions. The methodological intervention Williams makes, then, lies in showing that there can be no division of labour between philosophical and political analysis when it comes to either political judgement or political action. Ultimately, this is why Williams thinks there are some judgements that we must make. When it is clear that a regime is tyrannical, or that intervention is called for, then we simply cannot remain quiet or passive. Therefore, on the question of criticizing the abuse of power, Williams’s view is anti-​quietist rather than quietist. 8. Politics as Answerability Earlier we suggested that political judgements are related to action in certain ways. The special nature of political action bears on judgements about what to do and what to think, and it is impossible to separate the political and philosophical aspects of judgement. Bearing in mind these unique aspects of Williams’s approach, now we want to suggest that at the most basic level, Williams’s account [p. 453] Ilaria Cozz aglio and Amanda R . Greene arises from a conception of politics as answerability. In this final section, we elaborate on answerability by looking at how power’s aptness for justification and for critique depends in important ways on the audience. In other words, when it comes to justifying power or criticizing power, Williams thinks that we must take seriously the political circumstances of the audience. Not all criticism of power can have a point, and not all justification of power can count as such. In both of these cases, it will depend partly on the target audience. In developing the notion of answerability to an audience, our aim is to clarify the distinction that Williams employs between evaluative and normative assessments. This will, in turn, show why his political anti-​moralism is not just derivative of his anti-​system approach to ethics. For Williams, it is important to consider to whom political philosophy addresses itself, and how this conception of an audience differs from moral philosophy: [There is a question] of the audience of political philosophy . . . Who does the author suppose needs to know this philosophy, and for what purpose? This point arises also with moral philosophy . . . who needs such a theory? What for? What relation might it have to someone’s life? But that is not the sort of question to be pressed with political philosophy . . . there may indeed be some clearer answers about what some kinds of theory might do in some kinds of circumstances. But we have to be clear about these circumstances . . . [since it] is very relevant to an understanding of what political philosophy might and might not achieve. Williams thinks that political philosophy involves a distinction between the listener and the audience, which can be viewed as a mirroring of the distinction between author and narrator. Williams brings up this distinction in order to observe that much of contemporary political theory exhibits some misplaced assumptions about both the listener and the audience. He says, ‘The audience is almost always supposed to be the public at large; the public, indeed . . . of more than one state However,] the text seems to address itself to the attention of someone who has power [i.e. the listener], who could enact what the writer urges on him’ (2005: 57). Williams wryly observes that we could make sense of this if we took the writers to be assuming that their listener is empowered. But in that case, we could not make sense ‘of the fact that they address a . . . listener with . . . very few political restrictions on what he can do’ (2005: 58). Williams regards this as a confusion because it is senseless to address a listener who is both empowered and unconstrained. Anyone who is truly in a position of power is thereby embedded in an ongoing contest over power. To pretend otherwise is to take the politics out of political philosophy: It is typical of such political philosophy that the others are not there. This is related to the absence from much political philosophy of any sense of political contest . . . [p. 454] The displacement of politics is not merely the effect of the listener’s being taken to be empowered. For consider [Machiavelli’s] The Prince . . . it addresses itself precisely to the politics of his situation, . . . of how he should remain empowered. The [alienation of ] politics from political philosophy rather involves this: that [it] deals in ideals . . . and also addresses a listener who is supposedly empowered to enact just what such considerations enjoin. And no actual audience, no audience in the world, is in that situation. (Williams 2005: 58) This is the confusion that we must avoid, Williams thinks, if we are to be clear about what political philosophy can be expected to achieve. In order to speak meaningfully to an audience of people empowered as political actors, an author of political philosophy cannot abstract away from the ongoing struggle for power. To address the listener in such a way would manifest just the methodological mistake Williams has identified: to suppose there can be a division of labour between philosophical and political analysis, as we said above. In this sense, Williams endorses Weber’s view that power needs ‘to justify itself ’ and, therefore, that legitimacy depends on the subjects’ perception of the validity of the commands issued by the political authority (Weber 1978: 953–​4).12 But what, then, can political philosophy be expected to achieve, according to Williams? What can be said that will make a difference? If there must be critiques of power, yet they can only go so far, can they still be meaningful? The answer is yes, provided that we take up critique in the right mode. Here Williams proposes the critical theory test, a way of challenging a manufactured acceptance of power. He writes: ‘We can introduce the following test of a belief held by a group: If they were to understand properly how they came to hold this belief, would they give it up?’ (2002: 227). Williams imagines this test operating in a way that sidesteps the problem of an external moral standpoint. Instead, it takes up an internal perspective by starting with what subjects actually believe. According to Williams, we are to imagine them noticing certain patterns of disadvantage, then observing that these patterns are upheld by a ‘story’ told by those in power, and then asking whether they have any reason to accept the story other than its being propagated by those in power. The critical theory test seeks to verify, via a hypothetical process of self-​reflection, that subjects’ beliefs are not merely the product of manipulation. As Williams envisions it, the critical theory test has the capacity to influence subjects by pushing them to think about the grounds on which they accept political decisions and orders, perhaps coming to reject what initially made sense to them. When this happens, a political order can start to lose its legitimacy. For this 12 Williams hopes to go beyond Weber on the point of self-​justification. Williams claims that while some forms of justification may be accepted, they are not acceptable. Thus, he introduces the ‘critical theory test’, designed to detect the problematic case in which power has generated its own acceptance. On the question of whether Williams can depart from Weber in this way while retaining a realist approach, see Cozzaglio and Greene (2019). [p. 455] Ilaria Cozz aglio and Amanda R . Greene reason, the inducement of critical reflection can be likened to political action: it should be undertaken with a sense of responsibility. In the end, the demand of subjects for justification of power is the essential defining characteristic of politics. The notion of what can count as a justification of power to those who are subject to that power, therefore, is absolutely crucial for Williams. To return to the context of humanitarian intervention, he closes his analysis by pointing out that all political decisions involve a certain kind of answerability. Something must be said, both to those sponsoring the intervention and those affected by it, about why the exercise of power is justified. Otherwise it will not make sense to those involved as an instance of upholding an intelligible order, but instead will seem more like tyranny. Williams thinks that even the most idealized form of a moral rescue agency that we could imagine would raise this question: The obvious objection is that [the agency] is not answerable to anyone. But that, in a sense, is the point. It is, above all, the fact that the governments which intervene, as things are, are politically highly answerable that makes the decision to intervene inescapably and essentially political . . . [a decision that has] to be justified to people like other political decisions. (2005: 152–​3) For Williams, all exercises of power trigger a demand for justification. In so far as those exercising power claim to represent an intelligible system of authority, they are ‘politically highly answerable’. The fact that politics is about answerability is not so much the conclusion of an argument for Williams as it is the corollary of a political axiom: namely, that might does not make right. Politics is defined by such an axiom, for Williams, whereas there is no equivalent in moral philosophy. While it is true that consequences can matter in moral philosophy, it is not irreducibly about answerability to a concrete audience. Politics is fundamentally about answerability because it is inescapably about the exercise of power. This notion of answerability as definitive of politics is, ultimately, what gives political judgement the duality that Williams describes: namely, of being both contextualist and universalist. Since power must always justify itself, the category of making sense applies everywhere, but the terms in which an order makes sense differ according to the circumstances. The notion of answerability is also what allows for there to be a distinction between the normative and the evaluative when it comes to politics. Williams thinks that the question of whether a political order makes sense to some other group is evaluative, whereas the question of whether a political order makes sense to us is normative. That is because their political leaders are answerable to them, whereas our political leaders are answerable to us. Now we are in a position to see that what we earlier presented as the complex connection between judgement and action in politics is just a consequence, for Williams, of taking it to be an axiom that politics is essentially about answerability. [p. 456] As long as political decisions involve a high degree of answerability, this introduces unavoidable constraints. It imposes on political actors, and to some degree commentators, a responsibility to consider the limits of political possibility and the likely consequences of asserting power against power. Whether there is a point in criticizing or justifying this assertion of power always depends on the target audience of the critique or the justification. And that is because, for Williams, answerability belongs to the essence of politics. 9. Conclusion In this chapter, our aim was to outline Williams’s account of political judgement in order to clarify some of its subtleties and show why it poses a deep challenge to contemporary political philosophy. First, we showed how Williams’s basic legitimation demand combines elements that are universal with elements that are contextual. For Williams, what is universal is the distinction between order and tyranny, whereas the contextual asks when political order ‘makes sense’, an inquiry which is necessarily tied to the beliefs and values that subjects hold in a particular historical context. To illustrate how Williams intends his theory of political judgements to work, we looked at his framework for identifying human rights violations. Like the legitimation demand, the threshold for human rights violations is based on a distinction between political order and tyranny. We highlighted how this approach gives Williams a different perspective on humanitarian intervention, one that emphasizes the limits of political action and an ethic of responsibility. In the discussion of humanitarian intervention, we saw that Williams thinks political judgement and political action are intimately related. This holds for judging what we are to think as well as what we are to do. This view amounts to a methodological challenge to other theorists. Pushing back against the commonly accepted division of labour between philosophy and politics, one that is based on separating principles from feasibility considerations, Williams diagnoses that this comes from a misguided methodology according to which political judgements belong to the domain of philosophical analysis, whereas political actions belong to the domain of political analysis. According to Williams, political judgements and political actions each require both a philosophical and a political sensibility. Ultimately this is because adequate judgements about what to think and what to do must be sensitive to the limits of political possibility and the responsibility borne by anyone engaged in politics. Such inescapable relation to power is what ultimately distinguishes political and moral philosophy, in Williams’s view. Our aim was also to present the best case for why Williams’s approach to political judgement allows for legitimacy assessments to be realistic and have a critical edge. Since the possibilities for critique depend on an audience and what makes sense to them, we should be circumspect about what political philosophy [p. 457] Ilaria Cozz aglio and Amanda R . Greene can achieve. While this means there are limits to criticizing political actions and political orders, there are none the less imperatives to criticize them when power is being abused. Thus, the imperative to criticize itself incorporates a sense of its limits: namely, it must recognize just when the criticism can have a point and make a difference. When we see why he thinks there are some judgements that we must make, but that these are nevertheless limited, we can appreciate that Williams is neither a quietist, nor a conservative, nor otherwise biased towards the status quo. Rather, by asking and answering the crucial question concerning the circumstances in which political philosophy can achieve something, Williams brings into sharp relief the fact that very few others have so much as posed this question for themselves, much less offered an answer that takes history seriously. References Cozzaglio, Ilaria. 2022. ‘Legitimacy between Acceptance and Acceptability: A Subjects-​First View’. Social Theory and Practice 48.1: 61–​88. Cozzaglio, Ilaria, and Amanda Greene. 2019. ‘Can Power Be Self-​Legitimating? Political Realism in Hobbes, Weber, and Williams’. European Journal of Philosophy 27.4: 1016–​36. Estlund, David. 2017. ‘Methodological Moralism in Political Philosophy’. Critical Review of International Social and Political Philosophy 20.3: 385–​402. Forst, Rainer. 2017. Normativity and Power: Analyzing Social Orders of Justification. Oxford: Oxford University Press. Greene, Amanda R. 2017. ‘Legitimacy without Liberalism: A Defense of Max Weber’s Standard of Political Legitimacy’. Analyse & Kritik 39.2: 295–​324. Hall, Ed. 2015. ‘Bernard Williams and the Basic Legitimation Demand: A Defence’. Political Studies 63.2: 466–​80. Hall, Ed. 2020. Value, Conflict, and Order: Berlin, Hampshire, Williams, and the Realist Revival in Political Theory. Chicago: University of Chicago Press. Hawthorne, Geoffrey. 2005. ‘Introduction’. In Bernard Williams, In the Beginning Was the Deed, xi–​x x. Princeton: Princeton University Press. Hobbes, Thomas. 1994 [1668]. Leviathan. With selected variants from the Latin edition of 1668. Edited by Edwin Curley. Indianapolis: Hackett. Horton, John. 1992. Political Obligation. London: Macmillan. Horton, John. 2012. ‘Political Legitimacy, Justice and Consent’. Critical Review of International Social and Political Philosophy 15.2: 129–​48. Jubb, Rob. 2019. ‘On What a Distinctively Political Normativity Is’. Political Studies Review Larmore, Charles. 2013. ‘What Is Political Philosophy?’ Journal of Moral Philosophy 10: 276–​306. Larmore, Charles. 2020. What Is Political Philosophy? Princeton: Princeton University Press. Leader Maynard, Jonathan, and Alex Worsnip. 2018. ‘Is There a Distinctively Political Normativity?’ Ethics 128.4: 756–​87. Owen, David. 2018. ‘Realism in Ethics and Politics: Bernard Williams, Political Theory, and the Critique of Morality’. In Politics Recovered: Essays in Realist Political Theory. Edited by Matt Sleat, 73–​92. New York: Columbia University Press. Pettit, Philip. 2012. On the People’s Terms: A Republican Theory and Model of Democracy. Cambridge: Cambridge University Press. Queloz, Matthieu. 2021. The Practical Origins of Ideas: Genealogy as Conceptual Reverse-​ Engineering. Oxford: Oxford University Press. [p. 458] Sagar, Paul. 2018. ‘Legitimacy and Domination’. In Politics Recovered: Essays in Realist Political Theory. Edited by Matt Sleat, 114–​39. New York: Columbia University Press. Scheuerman, William E. 2018. ‘Getting Past Schmitt? Realism and the Autonomy of Politics’. In Politics Recovered: Essays in Realist Political Theory. Edited by Matt Sleat, 270–​95. New York: Columbia University Press. Schmitt, Carl. 2007 [1932]. The Concept of the Political: Expanded Edition. Chicago: University of Chicago Press. Sleat, Matt. 2014. ‘Legitimacy in Realist Thought: Between Moralism and Realpolitik’. Political Theory 42.3: 314–​37. Shklar, Judith. 1998. ‘The Liberalism of Fear’. In Judith N. Shklar, Political Thought and Political Thinkers. Edited by Stanley Hoffmann, 3–​20. Chicago: Chicago University Press. Weber, Max. 1978 [1922]. Economy and Society: An Outline of Interpretive Sociology. Berkeley: University of California Press. Weber, Max. 2004 [1919]. The Vocation Lectures. Indianapolis, IN: Hackett. Williams, Bernard. 1978. ‘Politics and Moral Character’. In Public and Private Morality. Edited by Stuart Hampshire, 55–​73. Cambridge: Cambridge University Press. Williams, Bernard. 1985. Ethics and the Limits of Philosophy. Cambridge, MA: Harvard University Press. Williams, Bernard. 2002. Truth and Truthfulness. Princeton: Princeton University Press. Williams, Bernard. 2005. In the Beginning Was the Deed. Princeton: Princeton University Press. Wolin, Sheldon S. 1981. ‘Max Weber: Legitimation, Method, and the Politics of Theory’. Political Theory 9.3: 401–​24. [p. 459] A Project of ‘Impure’ Enquiry Williams’s Historical Self-​Consciousness Miranda Fricker 1. Styles in Philosophy Bernard Williams wrote perhaps more than any other philosopher in the broadly analytic tradition about the importance of history, and particularly the importance of the history of philosophy, to the activity of philosophical reflection in the now. He often cast this in terms of the importance of philosophy mixing itself with history, and so embracing ‘impurity’.1 It is hard to credit that two decades have elapsed since his death, so present and vivacious remains his philosophical voice in our philosophical conversations. It is poignant that we must now reluctantly begin to think of his work as part of the fabric of the history of our discipline. Doing so is, I find, far from straightforward, not least because there are a number of special challenges in doing exegetical work on Williams that stem from his deceptively conversational style of writing, and his avoidance of those -​isms that might have emboldened one to fix him in this or that philosophical mould.2 Of course, that was precisely why he avoided them, as he preferred to proceed in a more piecemeal and suggestive fashion. I suspect this is one of the features of his style that means his work will endure, outliving this or that vogueish taxonomy of positions. Aside from these features of his style, another of the difficulties in starting to think of Williams’s work as part of the history of our subject is simply a matter of personal philosophical history. I knew him as a teacher, as I had the good fortune of having Williams as one of my supervisors for my doctoral thesis.3 Although one can engage perfectly well in exegesis of the work of a living philosopher, as we often 1 In a discussion of ethical objectivity, for instance, Williams says: ‘The fate . . . of the theoretical issue of objectivity reminds us in one way of the impurity of philosophy; if it is to have anything to say about that question, it will have to address a lot more than philosophy’ (1995: 148). 2 I explore these features of his style in Fricker (2020). 3 I had the double good fortune of being supervised in my doctoral work on feminist philosophy both by Sabina Lovibond—​whose work at the avant-​garde of feminist philosophy was my reason for applying in 1991 to go back to Oxford to see if I could, in my decidedly half-​baked state, write a doctoral thesis on what I was calling ‘perspectival realism’—​and by Bernard, who was now dividing his time between Oxford and Berkeley. With every year that goes by, I appreciate each of their distinctive, brilliant, and benevolent influences on me with increasing clarity and gratitude. Miranda Fricker, A Project of ‘Impure’ Enquiry In: Bernard Williams on Philosophy and History. Edited by: Marcel van Ackeren and Matthieu Queloz, Oxford University Press. © Oxford University Press 2025. [p. 460] do, still in cases where the philosopher whose work one hopes to interpret was relatively recently such a central and lively contributor to the conversation, it can feel like it is forever too soon to be taking up the focused exegetical stance. One may for this reason resist for a while, uneasy with the prospect of actively pressing them into the past, as if in fear of prematurely stilling an animated body of thought. The impulse to avoid active participation in any such process of intellectual stilling is particularly acute in relation to a philosopher like Williams, whose thinking and conversation and writing were somehow essentially animated: agile, funny, erudite, candid, rather free with delicious anecdote or ironical comment. (Perhaps this explains why so few of the photographs on the dust jackets ever seem to actually look like him—​they are too still.) Williams’s sceptical and sometimes satirical mood when it came to certain schools of thought in moral philosophy is not a superficial feature of his philosophical style, for it speaks of a Voltairean instinct to puncture that which struck him as misguided because overblown, or moralistic, or phony, or ‘academic’ in the pejorative sense of detached from the lively reality of human culture. This spirited sense of the human personality of different schools of thought in moral philosophy, and an occasionally mischievous stroke of the caricaturist’s pen when it came to representing them on the page, was a quintessential feature of his philosophical style, whether he was portraying the moral purism of the Kantian ‘morality system’, the petty moral accountancy of act utilitarianism, or the ‘government house’ forms of rule utilitarianism whose moral colonialism he found ‘deluded’.4 If this spirited scepticism was a characteristic of his philosophical style, we might ask how did it square with the attitudinally neutral style of modern analytic philosophy and its penchant for certain kinds of purity? Williams’s views about the analytical style are instructive. He found that it had a number of important virtues: first, a carefulness regarding matters of accuracy (an epistemic virtue which he later cast, in Truth and Truthfulness, as one of the two cardinal ‘virtues of truth’);5 secondly, a sound aspiration to conceptual clarity; thirdly, a penchant for somewhat plain prose, though in this connection he also noted the tendency for excessive technicality, along with an occasional self-​congratulatory narrowness about what philosophical clarity can involve. In his essay ‘What Might Philosophy Become?’, Williams takes up this last critical point and expands upon it throughout, describing the main purpose of the paper by declaring: ‘what I want to call in question is the idea that there is a style which defines fairly clearly and uniformly across the range of philosophy what counts as clarity and precision and getting it right, and that this style has been defined by the typical procedures of 4 ‘Government House utilitarianism had at least the merit of one kind of realism, inasmuch as it tried to find the theory an actual social location. It placed it in a particular body of people, the utilitarian elite (though it had a deluded idea of what such people might be like in reality)’ (Williams 1985: 110). 5 ‘I shall call Accuracy and Sincerity the two basic virtues of truth’ (Williams 2002: 44). [p. 461] Miranda Fricker analytic philosophy’ (2006a: 203). He found this particular one-​size-​fits-​all assumption moreover to manifest a scientism that he abhorred. He called this the ‘scientism of style’—​a species of purism he characterized particularly in relation to the idea that the form of a philosophical text makes no contribution to its philosophical content. Finding this to be quite false, he entreated us to consider the writings of key figures in the broadly analytic tradition and ask ourselves whether they can reasonably be read as putting forward content wholly independently from features of style. Williams put the point as a rhetorical question: When we turn, in particular, to moral and political philosophy, and we look at the canon of past philosophy that even analytic philosophy agrees on, does it look like this? Plato, Hobbes, Hume, Rousseau, indeed John Stuart Mill, not to go into more disputed territory: do we really suppose that their contributions to the subject are independent of the imaginative and expressive powers of their work? a: 205) In contrast, Williams argued, what is needed in philosophy is not merely a focus on whether or not the content is accurate and true, but on whether it ‘rings true’ a: 206), and this question calls upon style in the sense described, of employing techniques of persuasive power and imagination to constitute the philosophical vision as a compelling one.6 In order to ring true and convince the reader of its argumentative force, a philosophical text may, quite properly and ordinarily, appeal to metaphors, thought experiments, and other fictional scenarios, as a proper part of its arguments for the picture it aims to imprint on its reader’s mind. This is actually, if we’re honest about it, the normal mode of analytic philosophy, whose canonical texts might invoke, in epistemology, for instance, the captivating Cartesian figure of the Evil Demon, or the brain-​in-​a-​vat scenarios dreamed up by Harman or Putnam; or in moral philosophy, for example, the many and fabulous trolley problem scenarios conjured up by Foot or Kamm. We might reflect too on the power simply of metaphor: for instance, in Rorty’s attempt to smash the idea of philosophy as a ‘mirror of nature’. The thoroughly fictional scenario of the State of Nature is another literary-​philosophical conceit that involves the reader’s imagination in explicit and inexplicit ways. And this particular example reminds us—​if I may digress for a moment—​that part of analytic philosophy’s denial about its own uses of socially loaded metaphors and other imaginative mechanisms long had the effect of actively obscuring the sometimes crudely gendered nature of those imaginary tropes, to name just one dimension of imaginative exclusion. I vividly remember, when I was a graduate student in the early 1990s, hearing Michèle Le Doeuff talk 6 On this issue, see also Williams’s remarks in ‘Descartes and the Historiography of Philosophy’ c: 258) and in Shame and Necessity (1993: 13). [p. 462] about Hobbes’s State of Nature, and how the philosophical imaginary he employed seemed to picture people—​all apparently ready-​made adult men—​sprouting spontaneously like mushrooms from the ground overnight, so that the women who in any halfway realistic imagining would have given birth to them and nurtured them to sustain their lives were not imagined as co-​populating the State of Nature at all, but at best implicitly imagined as part of its natural resources (Le Doeuff 1990). Williams, I believe, understood these subterfuges—​or he was ready to. One might characterize them as tricks of the purist fantasy that philosophical works are not texts, their argumentative structures being imagined as free-​floating from all that being a text implies, such as the socially and historically situated author, an imagined reader with certain specific social characteristics, and then some real readers too, who may or may not resemble the imagined one. In ‘What Might Philosophy Become?’ Williams gave a brief but stout defence of feminist philosophy against Mary Warnock’s comments in the Preface to her 1996 edited collection, Women Philosophers (1996), where she echoed a standard analytic philosophical trope of the time, asserting the alleged impossibility of feminist philosophy. In a nutshell, she wrote that philosophy as such addressed human universals, and feminism as such did not, ergo there can be no feminist philosophy. Williams wrote of this entrenched and, to my mind, self-​deceiving piece of purism: ‘There is a rather rapid reply to this argument in more or less its own terms—​that among things which are humanly universal are the sorts of practices and attitudes that are the concern of feminist philosophy’ (2006a: 211). Quite.7 Williams practised what he preached when it came to the literary form of our philosophical writing sometimes contributing to its intellectual message. A moment’s reflection quickly provides examples: his use of the State of Nature in Truth and Truthfulness, his invocation of the fictional Gauguin’s moral luck, his dilemmas of moral integrity where the protagonist might be a botanist on an expedition or a research scientist facing an impossible choice. I have been suggesting that this mingling of form and content, this interweaving of argumentative structure and imaginative attitudinal colour in his writing, constitutes a first important dimension of impurity in his philosophy—​which he self-​consciously embraced, as opposed to doing it while pretending to some fantasized pure persuasion of reasons. But let’s now ask what, if anything, we might find in Williams’s style that is especially pertinent to his lasting fascination specifically with the borderlands of philosophy and history.8 In order to explore this question, let’s take as our starting point his conception of the history of philosophy—​his conception, that is, of the 7 Similarly robust points regarding the coherence of the idea of feminist philosophy were made in some reviews of the Warnock collection (e.g. Fricker 1996). I vividly recall a review by Jennifer Hornsby which rounded off with a wry observation that it was fitting for a collection with that editorial premise that it should be appearing in its publisher’s ‘Everyman’s Library’ series—​though alas neither she nor I have been able to trace it. 8 For a discussion of the importance of historical understanding in Williams’s philosophy, with an emphasis on the meaning-​laden nature of the subject matter of both disciplines, see Moran (2016). [p. 463] Miranda Fricker very fabric into which his own writing must now gradually be woven. Doing so will not only lead us to the borderlands we wish to explore; it will also enable us to ground our exploration firmly in Williams’s own conception of the history of philosophy. In the Preface to his 1978 book, Descartes: The Project of Pure Enquiry, Williams contrasted the history of ideas with the history of philosophy. The first sort of endeavour focuses on the work considered in its own historical context, and aims to reconstruct, albeit inevitably through a veil of textual indeterminacies, what it meant at the time. For this sort of enterprise, the central interpretive question is ‘what did it mean then?’ He describes such a project as proceeding ‘horizontally’ through time, being first and foremost one of historical enquiry, so that the genre of the resulting study is, as he puts is, ‘unequivocally history’ (Williams 1978: 9). Naturally, the history of philosophy also focuses on the work itself, and again considers its significance to be under-​determined by its own time, but recognizes in this an opportunity to creatively uncover what the work can mean to us in the present. For this sort of enterprise, the central interpretive question is ‘what does it mean now?’9 Williams’s study of Descartes’ philosophy firmly self-​identified as taking this latter approach, and he described his project as preferring ‘the direction of rational reconstruction of Descartes’ thought’ (1978: 10). In this, he was a modernist, not only as a matter of scholarly focus but also on account of the fact that his approach to Descartes’ text exemplified the energetic modernist spirit of historically self-​conscious creativity in the now.10 I hope that his embrace of this robust modernist attitude may transfer to us a certain aesthetic permission to rationally reconstruct Williams’s work in similar spirit, and to use the historical self-​ consciousness that it bequeaths us as a way of augmenting our own sense of the past, as one that is coming to contain Williams’s philosophy. We have seen that Williams’s conception of the history of philosophy was as a rational reconstruction that contributes to a conversation in the present, free-​spirited but disciplined by the obligation to remain carefully accurate regarding the facts of the historical work. In this conception we may recognize a two-​tier structure: one of basic and epistemically plain facts in the first tier, overlain by intrinsically contestable interpretive activity in the second. This two-​tier model of interpretation was a constant in Williams’s philosophical imagination. Having put it forward in the Descartes book as his conception of what he was doing with Descartes’ texts, it received at last its fullest expression in his final monograph, Truth and Truthfulness, no longer specifically in relation to the history of philosophy, but rather in relation to the discipline of history quite generally. There the two-​tier structure is how he 9 Williams reaffirms this conception in his 1994 paper ‘Descartes and the Historiography of Philosophy’, posthumously reprinted in Williams (2006c: 257). There Williams also talks of the history of philosophy in relation to the aim of ‘making the familiar seem strange, and conversely’ (2006c: 259). For a discussion of this aspect of Williams’s approach, helpfully labelled as the ‘alienation effect’, and for a defence of methodological pluralism, see van Ackeren (2019). 10 For a landmark articulation of this modernist spirit, see Berman (1982). [p. 464] pictured the whole enterprise of truthful historical interpretation and explanation. He argued that history must rely on a ‘chronicle’ of events: that is, a series of events ordered in time but without causal relations or explanatory structure of any other kind (Williams 2002: 238). One might speculate that many of the items in such a chronicle would be what he had earlier introduced as ‘plain truths’ (2002: 45–​ facts which are epistemically plain, not in the sense of necessarily being simple or immediately perceptible facts, but in the sense of being evidentially plain in the context. At any rate, on Williams’s conception, history proceeds with a series of agreed event-​facts in chronological order, but then the interpretive business of history gets started, positing causal and explanatory priorities and orderings, thereby already catapulting us into the realm not of truth—​for there can be no single historical truth—​but rather of truthfulness. As he expressed the point: [T]‌here is no such thing as ‘the truth’ about the historical past, though as with the universe or Caesar there are many truths about it . . . [This] means that while we must demand that interpretations of the past should tell us the truth, in the sense that they should not lie or mislead, what we need them for is not to tell us something called ‘the truth about the past’. We need them to be truthful, and to make sense of the past—​to us. (2002: 257–​8) In this way, Williams’s conception of history of philosophy now shows up as all of a piece with his conception of history quite generally. They have in common a two-​tier structure: a first tier of plain truths constituting the historical ‘chronicle’ or, alternatively, the philosophical text to be interpreted; and a second tier at which some plurality of truthful and essentially contestable interpretations is generated. This two-​tier structure, considered as a general model for philosophical interpretation, strikes me as yet another example of a kind of impurity in philosophy that Williams explicitly championed. When it comes to philosophical interpretation, there is no one true story. The project of Truth and Truthfulness marks the culmination of Williams’s conception of the intertwined relation of history and philosophy, and in particular what this can mean for philosophical method. The method of the book combines them, and so stands out—​as does his brilliant exploration of ancient Greek ethical culture in Shame and Necessity—​as an exemplar of what impure philosophy can be like. I should make clear that, unlike Nietzsche’s genealogy, Williams’s State of Nature scenario with which Truth and Truthfulness begins is not itself one iota historical, but a thought experiment in fictional abstraction, which hands over to vignettes of intellectual history only after the fact, once the universally necessary proto virtues of Accuracy and Sincerity have been established in the State of Nature. Williams’s distinctively hybrid State of Nature genealogy combines in one philosophical method the a priori speculative pursuit of universal necessity that is so habitual in analytic philosophy, with the historical exploration of contingent development that is so distinctively characteristic of the European tradition. [p. 465] Miranda Fricker The definitive agenda of the book, and in an obvious way the historical dimension of its method, is Nietzsche’s. Williams was entirely explicit about that. The book’s animating concern is our fundamental interest in achieving truthfulness with ourselves concerning the nature of our value commitments. Truth and Truthfulness is an experiment in how far our deepest ethical commitments (and in particular our commitment to truthfulness itself ) can withstand transparency about their instrumental origins.11 For this reason, and many others besides, Nietzsche’s presence will be felt throughout this chapter, though only as a spectre at the feast, for space demands that we reserve the riot of impurities served up in Nietzsche’s genealogical text for another occasion (see Fricker forthcoming). I propose instead to read Williams—​to rationally reconstruct him, perhaps—​in relation to two very different philosophers: Descartes and Wittgenstein. I find this a surprisingly fruitful combination for accessing some of Williams’s deepest motivating commitments, and ultimately his conception of what philosophy needs to be. My hope is that by retracing the ways in which these two historical figures shaped Williams’s thought, rather less explicitly than the way Nietzsche did, some previously hidden contours of his own historical self-​consciousness as a philosopher will show themselves to us. 2. Descartes and Williams’s ‘Absolute Conception’ In a late essay on Janáček,12 Williams said: ‘Unselfconscious modernism is a contradiction in terms’ (2006b: 120). I have so far suggested that Williams—​a classicist by first training, and an ancient scholar whose philosophical understanding of the Homeric imagination was second to none—​presented as a modernist in at least three ways. One was his literarily self-​conscious attitude to producing a philosophical text, and the insistence that if philosophy was to ‘ring true’ to its readers, it had better make more honest use of the fact that form can contribute to content. Another was his self-​consciously bold rationally reconstructive stance regarding the practice of history of philosophy. And the third—​now using ‘modernist’ in the more simply historical sense—​was that he had written a scholarly book about a philosopher of the early modern period. This last was mentioned only in passing, so let me now take up this aspect of his modernism and ask: what did Williams take from Descartes’ ‘project of pure enquiry’ that shaped his own thinking? The first thing I would like to mention about the Descartes book, but will quickly set aside, is that in it Williams explores the (hardly Cartesian) idea that the concept of knowledge naturally arises in a very basic human social situation of needing to 11 On the social construction of truthfulness as an intrinsic, rather than instrumental, value, see Williams (2002: 90–​2). 12 A short piece originally published in 2003; reprinted in 2006 in On Opera. [p. 466] gain information from others: who can tell me whether p? (Williams 1978: 37–​47 and passim).13 Williams himself did not further elaborate this idea until the final stage of his career, some years after Edward Craig had produced his immaculate monograph Knowledge and the State of Nature (1990), in which he offered a ‘practical explication’ of the concept of knowledge using (as Craig makes clear) the same starting conception. A decade or so after that, when Williams was writing Truth and Truthfulness, it was Craig’s epistemic State of Nature that he used to ground the project. The point I wish to bring out here is simply that the seed-​thought for Truth and Truthfulness was already sown in the Descartes book. And this presents us with a salient example of what a surprisingly synchronous thinker Williams was, in that the arc of his whole philosophy over time can seem to be already contained, prism-​like, in any suitably chunky moment of it.14 However, it is Williams’s underlying conception of mind and world that I wish to focus on for present purposes. This conception is drawn far more directly from Descartes’ own thinking, and it is anchored in the idea of ‘the absolute conception of the world’. Williams diagnoses Descartes’ sceptical interrogation of knowledge as indicating his commitment to a basic idea: namely, that if we have knowledge, then it is at some level knowledge of what is there anyway. That is, it is knowledge of something mind-​independent—​albeit chez Descartes, ultimately guaranteed by a benevolent God. This is represented as implicitly driving Descartes’ sense that the category ‘knowledge’ has an intrinsically problematical character conducive to sceptical anxiety. Williams describes this anxiety as stemming from: a very basic thought, that if knowledge is what it claims to be, then it is knowledge of a reality which exists independently of that knowledge, and indeed (except for the special case where the reality known happens itself to be some psychological item) independently of any thought or experience. Knowledge is of what is there anyway. One might suppose this thought to be incontestable, but its consequences can seem to be both demanding and puzzling. (1978: 64) Demanding and puzzling indeed.15 In a natural progression of propositions that Williams finds to be recurrent in Western thought, he rehearses the idea that if A and B each have some representation of a patch of the world, where these are 13 Williams’s phrase ‘our original primitive truth-​gatherer’ (1978: 47) in particular anticipates Craig’s more fully articulated State of Nature scenario of inquirers and ‘good informants’—​on further aspects of this anticipation, see Queloz (2021: 135–​7). When Williams first handed me his copy of Craig’s book, kindly lending it to me as his student, he said with an ironical smile (just because this was a funny thing to be sincerely saying of any philosophy book), ‘Here, read this. Everything in it is true.’ 14 I have pursued this idea as an interpretive conceit in Fricker (2020). 15 For an independent and fully articulated theoretical elaboration, see Moore (1977). For purposes of our understanding of Williams, it is also worth noting that in ‘Philosophy as a Humanistic Discipline’, he describes Moore’s book (in a singularly unusual-​for-​him superlative) as ‘an outstanding discussion’ a: 185n8). For Moore’s later discussion of Williams’s idea, see Moore (2007a). [p. 467] Miranda Fricker both items of knowledge, then there must be some more inclusive representation in which A and B and their different representations both feature, and which explains how these differing items of knowledge are related. Individuals (or groups) with different colour vision would be a case in point, where the more inclusive representation reveals what it is about the different parties’ genes (as per the typical case) that explains their cognitive differences. This process can, as he says, be iterated, and the maximally inclusive conception of the world as containing what is there anyway plus all knowledge-​representations of it, related coherently to one another, is the absolute conception of the world. As Williams expressed it: ‘The absolute conception should explain, or at least make it possible to explain, how the more local representations of the world can come about—​it is this that would enable us to relate them to each other, and to the world as it is independently of them’ To be sure, at this point Williams is primarily presenting something he finds implicit in Descartes’ view rather than expounding his own. On the other hand, it is clear that Williams finds compelling some, only slightly slimmed down, version of the absolute conception discerned in Descartes’ project of radical doubt. Indeed, he says that almost as much is ‘involved in the distinction between primary and secondary qualities’ (1978: 246), understood so that it is only primary qualities that ‘characterize the material world as it really is’. And he adds, ‘I believe that these ideas are not incoherent, and have some faith that they are correct. But certainly they involve extensive intellectual commitments, not easy to fulfil’ (1978: 246). Later, in Ethics and the Limits of Philosophy, Williams employs his own idea of the absolute conception—​one liberated both from Descartes’ condition of absolute certainty (1978: 247) and also from what Williams dismissed as the ‘positivist fantasy’ of a ‘cumulative, convergent, self-​ vindicating unified science of man and nature’ (1978: 302). The version of the absolute conception he advanced there, however, is clearly still an evolution of the idea first discerned in the Cartesian engagement with the preconditions of knowledge. In the Ethics and the Limits discussion, Williams’s purpose was no longer to problematize the concept of knowledge, but rather to emphasize and develop a more general positive metaphysical and epistemological picture, which we might cast as a kind of perspectivalism (with apologies for the superimposed -​ism).16 At any rate, on this issue his main purpose by this time was to advance an ethical cognitivism according to which we can know many perspectival things—​for instance, that the grass is green, or the action cruel—​by employing concepts and required 16 Almost. ‘Relativism’ was a notable exception, though it was always partnered with a signature qualifier: Williams called his view the ‘relativism of distance’. [p. 468] aspects of sensibility which are more or less local, more or less perspectival. His strategy is to disaggregate: we should stop talking about knowledge en masse as if it related mind to world in a uniform way, and instead allow that some kinds of knowledge are more perspectival than others. Metaphysically and epistemologically, Williams effectively posits a continuum rather than a dualism, and he arrives at this continuum of perspectives as the way out of an apparent dilemma, which he formulates as follows: On the one hand, ‘the world’ may be characterized in terms of our current beliefs about what it contains; it is a world of stars, people, grass, or tables. When ‘the world’ is taken in this way . . . our conception of the world as the object of our beliefs can do no better than repeat the beliefs we take to represent it. If, on the other hand, we try to form some idea of a world that is prior to any description of it, the world that all systems of belief and representation are trying to represent, then we have an empty notion of something completely unspecified and unspecifiable. So either way we fail to have a notion of ‘the world’ that will do what is required of it. But the dilemma is only apparent, for Williams next offers a third way, a way out of the impasse of choosing either a conception of the world that is a mere function of our beliefs or, alternatively, a conception of a world whose radical independence renders it wholly unreachable in cognition. Instead, he observes: Each side of this dilemma takes all our representations of the world together But there is a third and more helpful possibility, that we should form a conception of the world that is ‘already there’ in terms of some but not all of our beliefs and theories. . . . We can select among our beliefs and features of our world picture some that we can reasonably claim to represent the world in a way to the maximum degree independent of our perspective and its peculiarities. The resultant picture of things, if we can carry through this task, can be called the ‘absolute conception’ of the world. (1985: 138–​9; emphasis added) In this switch to (what I find to be) a new use for the phrase ‘absolute conception’, he moves from the previously maximally inclusive use—​a conception that contains all the features of the world, and all our representations of them, coherently related to one another—​to an apparently minimalist use—​a conception that contains only those features of the world that find representation in our concepts and beliefs that are maximally non-​perspectival, non-​local, or non-​‘peculiar’, to use another term he often employed in this connection. As I tend to think of it, this more minimal absolute conception will represent only primary qualities (though this is probably best treated as a rough approximation only, given the complexities [p. 469] Miranda Fricker of the primary–​secondary distinction).17 At any rate, he seems to be switching to a use of ‘absolute conception’ as a conception of a substratum of the world that is maximally mind-​independent, and to which we become implicitly committed as soon as we entertain the idea of knowledge, no matter how peculiar or local the perspective of that knowledge: for instance, the knowledge that the grass is green, or (moving up the continuum) the knowledge that the action was cruel. The absolute conception, then, is now a conception of a substratum of material features of the world, which will feature at the base of any explanation how two different knowledge-​representations are both perspectives on the same world. The absolute conception has become a conception of that underlying same world—​and therefore a conception, as Williams put it, ‘of the world that might be arrived at by any investigators, even if they were very different from us’ (1985: 139). The main pressure on us all to form the absolute conception is explanatory. We need to explain the differences between different knowledge-​perspectives as being perspectives on the same world (otherwise they could not both be knowledge). However, all this being said, Williams’s chief point and purpose in this discussion is not in fact to affirm our necessary commitment to a shared, mind-​independent stratum of reality, but rather to indicate and emphasize an exhilarating implication. What’s emphatically indicated is that the vast extent of the world that we know is perspectival to a greater or lesser degree. Most of our knowledge is enabled by more or less perspectival concepts and accompanying sensibility, and the features that figure in such knowledge—​such as colours, or values, or social norms and institutions, or aspects of personal relationships—​definitively do not figure in the absolute conception, though Williams’s view is that they must be intelligibly related to it. What Williams was aiming to achieve in this discussion in Ethics and the Limits is really two things: he aimed to reject any dualism of fact and value, while conserving and vindicating what he held to be important differences between the knowledge delivered by the physical sciences and other kinds of knowledge, such as historical or ethical knowledge. The great asymmetry between scientific knowledge and ethical knowledge is of particular importance to Williams, because the same scientism he found repellent as a matter of philosophical style—​where it involved a pretence that philosophy in general is, or ought to be, like the empirical sciences—​is at work in the idea that the only real knowledge, the only knowledge proper, is knowledge gained of worldly features figuring in the absolute conception. Not so, says Williams. And against this scientism, he affirmed science as distinctive, difficult, and important; but argued that there are other kinds of knowledge that are also distinctive, difficult, and important, since there are great expanses of the world—​our world—​that cannot be known through scientific enquiry. 17 Putnam, however, seems to take it as exactly what Williams intended (Putnam 1992: 98 and passim). For a helpful complicating discussion of the history of the primary–​secondary distinction in this connection, and a partial defence of Williams against Putnam’s overall critical view, see Blackburn (2009). [p. 470] Collapsing the dualism of fact and value in this way, by effectively positing a continuum of increasingly perspectival types of representation of the world, is another way in which I interpret Williams as insisting on a kind of impurity, a mingling of elements often kept separate. Whereas positing a binary fact–​value dualism insists on black and white, positing a continuum insists on a graduation of subtly changing shades in between. A perspectival representation of the world—​whether a colour perception or an ethical cognition—​involves a cognitive grasp of some feature of the world discerned using a concept (and its associated sensibility) that is local. It might be species-​local to most humans, as in the case of colour perception; or it might be highly culturally and historically local to some us, where ‘us’ might indicate a small ethical or political sub-​community within a moral culture. In the particular case of ethical knowledge, Williams offered us an account of how the very concepts in which such items of knowledge were couched were themselves inseparable admixtures of fact and value. ‘Thick’ ethical concepts, as he called them, are both ‘action-​guiding’ and ‘world-​guided’, both descriptive and evaluative.18 Williams expressed agreement with John McDowell with regard to the inseparability of the two aspects.19 The thought is, one cannot pretend that a thick ethical concept is an item of merely descriptive content to which we have added a positive or negative moral valence—​as if it were ‘guided round the world by its descriptive content, but [with] a prescriptive flag attached to it’, as Williams critically cartooned Hare’s prescriptivism (1985: 141). To the contrary, Williams believed that the ability to discriminate situations in which one might properly apply an evaluative concept, such as ‘selfish’ or ‘vain’ or ‘generous’ or ‘churlish’, requires already also grasping the evaluative point of the concept in question. This leads us directly to an important stretch of philosophy’s border with history. In a case where critical reflection, or indeed any other influence, has alienated us from concepts that were formerly our ‘own’, in the sense that we lived by them and had internalized their action-​guiding authority, they cease to be the sort of concept that couches ‘ethical knowledge’ for us. Instead, the knowledge of how to get around our social world—​perhaps ‘you have blasphemed’, or ‘she is not chaste’, or whatever it may be—​is no longer action-​guiding for us, and, as Williams theorizes it, therefore no longer ethical knowledge for us. If philosophy seeks to understand our ethical concepts—​including their role in motivating action and framing our ethical evaluations—​then it had better take a historicized conception of them. 18 Philippa Foot had advanced such a view of ethical claims and beliefs in her ‘Moral Arguments’ 19 Williams notes, in this connection: ‘McDowell is above all concerned with the state of mind and motivations of a virtuous person, but I understand his view to have the more general implications discussed in my text. The idea that it might be impossible to pick up an evaluative concept unless one shared its evaluative interest is basically a Wittgensteinian idea. I first heard it expressed by Philippa Foot and Iris Murdoch in a seminar in the 1950s’ (1985: 240n7). For some critical discussions, see, for instance, Moore (2006), Heuer (2012), the papers in Kirchin (2013), and for a full-​length study, Kirchin (2018). [p. 471] Miranda Fricker As Williams puts it in ‘Philosophy as a Humanistic Discipline’: ‘philosophers cannot altogether ignore history if they are going to understand our ethical concepts at all. One reason for this is that in many cases the content of our concepts is a contingent historical phenomenon’ (2006a: 191). In a case where it is indeed reflection that has alienated someone from a concept—​for instance, when Oscar came to the view that ‘blasphemy’ was not one of his concepts, or when Nora perhaps came to the view that ‘wifely obedience’ was no longer one of hers—​what we are confronted with is the destruction of ethical knowledge by reflection. This is one of Williams’s more notorious ideas, but I do not think it should be as problematical as all that, given what he meant by ‘ethical knowledge’. The relevant idea of ‘ethical knowledge’ is a product of Williams’s particular form of moral cognitivism, which is non-​objectivist. He argues that we can have ethical knowledge by correctly applying thick ethical concepts whose values we have internalized and live by in practice. If we hadn’t internalized them as practical reasons for us, they would (trivially) have no intrinsic action-​guiding force for us. For Williams, ethical knowledge is knowledge within a moral outlook—​with any luck, an outlook that offers us some tools to think critically about what aspects of it we want to retain, which concepts we want to continue to live by. Coming to a sufficiently collective critical attitude towards a thick ethical concept presses that concept out of general ethical service. Elsewhere I have described the resulting situation as analogous to what happens after a currency such as shillings and crowns is put out of circulation (Fricker 2000). One can still count costs using them if one wants to—​we don’t suddenly lose the relevant arithmetical competency—​but doing so no longer has the same social function or meaning. Functionally speaking, it’s just not our money anymore; we can’t buy anything with it. The knowledge we once had in using such concepts is no longer knowledge of what things cost around here, but only of what they used to cost. Thus, ethical knowledge whose concepts are pressed out of service becomes (I surmise) some other kind of knowledge—​presumably, historical knowledge (Williams does not say). Thus, the destruction of ethical knowledge involves no damage to our grasp of the evaluative point of the now-​redundant thick ethical concept, but only to its action-​guiding force. A nineteenth-​century woman who grew up instilled with a concept of ‘wifely obedience’ may rebel against it at some point, but in so rebelling she loses no conceptual competence and no grasp of evaluative point—​indeed, she may feel she is only now seeing what its more complex evaluative point really is. What she loses in rejecting the concept, in dissociating herself so far as possible from its hold on her, is its action-​guiding and wider normative force. On this conception, moments of moral progress will likely involve some destruction of ethical knowledge and its automatic conversion into historical knowledge. Thus (I further surmise), there is no net loss of knowledge when we make improvements to our patterns of moral motivation and evaluation. [p. 472] Williams’s general conception of thick ethical concepts and what it takes to have competency in using them is one of the areas of his philosophy that has a secret passageway from Wittgenstein. In a footnote in Ethics and the Limits, Williams observes: ‘The idea that it might be impossible to pick up an evaluative concept unless one shared its evaluative interest is basically a Wittgensteinian idea’ n7). (In this respect we see it is Wittgenstein who provided Williams and McDowell with a rare point of meta-​ethical agreement.)20 The eschewal of any dualism of fact and value is something Williams admired in Wittgenstein, whom he elsewhere described as producing ‘some of the deepest insights of modern philosophy’ (2006a: 181). It is time to consider Williams’s critical engagement with Wittgenstein directly, so that we may reflect anew not only on issues of style, scientism, and the conception of the world, this time through the prism of Wittgenstein’s extraordinary philosophical writing; but ultimately also on the broadest question of what, from Williams’s point of view, philosophy might become. 3. Wittgenstein—​and Williams’s Ambivalence Earlier I suggested a number of respects in which we might see Williams as a modernist, including with respect to style in philosophy. In a wonderful essay on the composer Leoš Janáček, Williams offered some reflections on the modernist style in music, though swiftly moving to a comparative discussion with philosophy. Here Williams explores the significance of the ellipses in Janáček’s minimalist style of musical composition, making a revealing and somewhat disobliging comparison with Wittgenstein’s own disjoined cadences and cumulating motifs. The comparison goes like this: In some modernist philosophers there is an analogy to Janáček’s ellipses, the elimination of transitions: with them, juxtaposition and repetition provide the structure. But it is not clear how well this can work, indeed how well it does work even in the greatest of such philosophers, Wittgenstein. The problem perhaps lies in this, that even if the loss of large-​scale development in Janácek is counted a loss, the result is unquestionably music, and very powerfully so; but philosophy without argumentative development, philosophy in which all the therefore’s have been replaced by and’s or by spaces, is dubiously philosophy at all. Or rather, if it is philosophy, as Wittgenstein’s quite certainly is, it is so because it is about earlier philosophy, even if it does not say so: in his case, the earlier philosophy was often his own. That other aspect of modernism, historical self-​consciousness, has to sustain the identity of the enterprise altogether. (2006b: 119) 20 For a discussion of Williams’s relation to McDowell and Wittgenstein, see Lang (forthcoming). [p. 473] Miranda Fricker Here we glean something of Williams’s deep interest in Wittgenstein’s philosophy, and equally some insight into his still deeper ambivalence towards it. This ambivalence, I believe, was partly about its gnomic style, but more importantly about two independent concerns: the conception of the world, as delivered in Wittgenstein’s philosophy of language; and, more generally, the conception of philosophy as quietist—​as an enterprise which, properly understood, leaves everything as it is. Let me, first, address the ambivalence regarding style. Williams, as I read him, respects Wittgenstein’s stylistic modernism, and seems even to consider its self-​ conscious concern with written form as distinctively Wittgensteinian. In ‘What Might Philosophy Become?’, having just expounded the need for philosophy to seek to ‘ring true’ to its readers by embracing literary techniques that allow form to contribute to philosophical content, Williams remarks that ‘The demand on philosophy that it should listen to what it says leads to the concerns of Wittgenstein’ a: 207). Williams doesn’t explain exactly what he has in mind here in the idea that philosophy ‘should listen to what it says’, but I interpret him as alluding to Wittgenstein’s anti-​systematic, fragmented cadences of philosophical prose as an example of a philosopher who does listen to what he says, in the sense of deliberately producing a text whose form manifests the content of his philosophical view. In Wittgenstein’s writing, one might say the in-​between gaps that separate the short, aphoristic paragraphs are as integral to the work as the rests in a musical composition. On the other hand, it is with a certain caution that I air this thought, conscious that the explanations of Wittgenstein’s fragmentary texts are surely multiple. The Tractatus, comprising a series of short numbered paragraphs ordered to ‘indicate the logical importance of the propositions’ is one thing;21 but the form and presentation of the posthumously published Philosophical Investigations are quite another, having a more complicated editorial backstory.22 We should pay attention, for instance, to Wittgenstein’s seeming disclaimer in his 1945 Preface to the Investigations, which reads: Originally it was my intention to bring all this together in a book whose form I thought of differently at different times. But it seemed to me essential that in the book the thoughts should proceed from one subject to another in a natural, smooth sequence. After several unsuccessful attempts to weld my results together into such a whole, I realized that I should never succeed. The best that I could write would never be more than philosophical remarks . . . (Wittgenstein 2009: 3) 21 Wittgenstein’s note reads: ‘The decimal figures as numbers of the separate propositions indicate the logical importance of the propositions, the emphasis laid upon them in my exposition . . .’ n*). 22 See, in particular, the Editorial Preface to the fourth edition by Hacker and Schulte (Wittgenstein 2009). [p. 474] On the other hand, however, this sentence (which itself may be as much ironical pretext as sincere admission) continues with the thought that the explanation why the task proved impossible was the very nature of its subject matter: ‘my thoughts soon grew feeble if I tried to force them along a single track against their natural inclination.—​And this was, of course, connected with the very nature of the investigation. For it compels us to travel criss-​cross in every direction over a wide field of thought’ (2009: 3). Perhaps, then, my suggestion that Williams appreciated in Wittgenstein a remarkable self-​conscious harmonization of literary form and philosophical content is sufficiently attuned to the editorial facts, even allowing for the possibility that the gappy modulations of the text may have been partly in spite of, rather than because of, Wittgenstein’s avowed best efforts. In addition, I think we have reason to speculate that Williams found in Wittgenstein’s fragmentary form an apt implicit entreaty to do what, at paragraph 66 of the Investigations, he famously instructs philosophers to do more generally: namely, to ‘look and see’ what our various different ‘language games’ are actually like, rather than proceeding on the basis of habitual essentialist assumptions.23 This chimes with Williams’s own anti-​theoretical, anti-​systematic stance—​though I would also emphasize that, in Williams’s case, he adopted this stance specifically in ethical and to some extent political philosophy for the reason that these particular areas of philosophy take highly historically contingent spheres of human life as their subject matter. He did not dogmatically assume that the same anti-​ theory stance was called for all across the different areas of philosophy. In this particular, one might find Williams to be the more consistent adherent of the injunction to ‘look and see’, for he did not assume that all areas of philosophy call for the same approach. These two moments of harmony being noted, however, it is clear from the slightly unfavourable comparison with Janáček’s musical ellipses that Williams found Wittgenstein’s style rather too elliptical. This was not owing to any undiscriminating distaste for the rhythms of aphorism. Williams after all relished similar in Nietzsche, but found that Nietzsche successfully ‘mined or booby-​trapped his works against being assimilated’ (Williams 2006a: 208). And when it comes to other Wittgensteinian tropes such as compression and allusion, Williams himself embraced these conceits in his own philosophical writing.24 Rather, what Williams found problematic in Wittgenstein’s style was that in his particular use of these conceits he left his text so wide open, so minimal in discernible argumentative development, that it was left vulnerable to misappropriation. Not unrelatedly, he also detected a certain self-​absorption in Wittgenstein’s historical self-​consciousness, Don’t say: “They must have something in common, or they would not be called ‘games’ ”—​but look and see whether there is anything common to all’ (Wittgenstein 2009: para. 66). 24 On the question of compression as a stylistic feature common to Williams and Nietzsche, see Babiotti (2020). [p. 475] Miranda Fricker given that the critical responsiveness seems largely to be in relation to his earlier self. Indeed, Williams described the later work as Wittgenstein ‘talking to himself ’. More a literal description than an objection: I would suggest we hear Williams as offering a mischievously expressed yet genuinely problematizing observation about the distinctive voice of Wittgenstein’s work, and the unusually self-​ referential diachronic self-​consciousness it embodied. In this, Williams was not really saying anything that Wittgenstein might not have agreed with, and indeed almost said himself. There is, for instance, that charming moment in the Preface to the Tractatus where he writes: How far my efforts agree with those of other philosophers I will not decide. Indeed what I have here written makes no claim to novelty in points of detail; and therefore I give no sources, because it is indifferent to me whether what I have thought has already been thought before me by another. (1922: 23)25 We also know that Wittgenstein thought the vision of the Investigations could not come upon one except by seeing it as a rejection of the picture theory he had presented in the Tractatus. This was why he wanted them published together. Seen from this angle, it is manifestly true of Wittgenstein that he was, in a straightforward if limited sense, ‘talking to himself ’; but on the other hand, of course, one can hardly miss Williams’s gently lampooning turn of phrase, which was, as ever, a sonorous proper part of his philosophical point. What of the metaphysical and epistemological differences between the two thinkers—​their conceptions of mind and world? Williams was also ambivalent—​I think deeply so—​about Wittgenstein’s philosophy in this regard, and getting clear about why will ultimately lead us to perhaps their greatest disagreement, which concerns the very nature of philosophy and what it can do. But let’s start with their conceptions of our cognitive relation to the world. Here Williams’s ambivalence stems significantly from what he considers to be Wittgenstein’s stance towards science. On the face of it, there is some harmony between the two thinkers in this regard, since Wittgenstein was vehemently anti-​scientism, detesting any effort to assimilate the philosophical enterprise to the scientific one. In this particular, Williams was ready to recognize something of a kindred spirit, and it is principally for this reason that he described Wittgenstein as an ‘anomalous’ figure in analytic philosophy (2006a: 207). On the other hand—​and here comes the source of the ambivalence—​in ‘Wittgenstein and Idealism’ (Williams 2006c).26 Williams describes Wittgenstein as having a ‘hatred of the cockiness of natural science’ but adds the heavy qualifier that this is ‘something which seems to me not easy in his 25 Wittgenstein continues: ‘I will only mention that to the great works of Frege and the writings of my friend Bertrand Russell I owe in large measure the stimulation of my thoughts’ (1922: 23). 26 Reprinted from Moral Luck (Williams 1981) [p. 476] case to distinguish from a hatred of natural science’ (2006c: 375).27 Williams continues: ‘to adapt a remark of Kreisel’s, when the child asks why the people on the other side of the world don’t fall off, many would give an explanation in terms of gravity acting towards the centre of the earth, but Wittgenstein would draw a circle with a pin man on it, turn it round and say, “now we fall into space” ’ (2006c: 375).28 Williams was suspicious, then, that Wittgenstein’s anti-​scientism was a veneer that concealed, perhaps even from himself, a metaphysically cheap anti-​scientific attitude. This, as we have already seen, Williams most certainly did not share, not least because it encouraged an aprioristic mindset in analytic philosophy which served to banish history from the other direction. In this regard, apriorism was and remains the equal and opposite prejudice to scientism, and both are the enemies of the sorts of humanistic impurity that Williams prized and advocated in philosophy. Williams had no doubt about the importance of science, and its genuinely overlapping relation to certain areas of philosophy. His respect for the real challenges of attaining scientific knowledge saw him make a firm but fair distinction between philosophy that honestly engages with relevant scientific research, and philosophy that merely displays a vain pretension to the manners of science. Williams quipped, for example, that some ‘People can perhaps persuade themselves that if they fuss around enough with qualifications and counter-​examples, they are conducting the philosophical equivalent of a biochemical protocol’ (2006a: 184). His respect for science was also surely a driving force behind his commitment to the absolute conception as populated exclusively by the objects of physical scientific knowledge. For the absolute conception was, I would emphasize, a conception of something physical, not metaphysical. (This is one of the respects in which Putnam mischaracterizes the absolute conception in Renewing Philosophy, where he describes Williams as viewing ‘physics as giving us the ultimate metaphysical truth’ (Putnam 1992: 108), something Williams explicitly denied.)29 27 But see Mulhall (2009) for some defensive counterpoint in this regard, and indeed with regard to Williams’s broader interpretation of the later Wittgenstein as containing elements of a transcendental idealist position, which Mulhall suspects of being ‘in the service of constructing an objection to those who might wish to take Wittgenstein’s later conception of philosophy with any seriousness’ For an alternative perspective—​one more sympathetic with Williams’s interpretation—​ see Moore (1985, 2007b); and for a robust rebuttal of Mulhall’s interpretation of Williams’s attitude to Wittgenstein, see Queloz and Krishnan (this volume, Chapter 14). 28 Georg Kreisel, the celebrated Austrian logician and mathematician, was born in Graz, 1923. There is a widely circulated anecdote according to which, in his years teaching at the University of Reading, he would habitually take a certain train to London to go to the theatre. That train stopped running, but Kreisel discovered there was an equally well-​timed train from Bristol which paused at Reading station strictly to take on water, not passengers; and yet (the story goes) he repeatedly snuck on board. When, on one occasion, he was accosted by the guard telling him, ‘This train doesn’t stop at Reading, Sir’, he is reputed to have replied, ‘If it doesn’t stop here, then I didn’t get on here.’ I do not know if it was this particular storied remark that Williams had in mind in his own adaption. 29 Williams retorts: ‘Now I have never held any such view, and I agree entirely with Putnam in rejecting it. However, I have entertained the idea that science might describe the world “as it is in itself ”, that is to say, give a representation of it which is to the largest possible extent independent of the local perspectives or idiosyncrasies of enquirers, a representation of the world, as I put it, “as it is anyway”. [p. 477] Miranda Fricker Williams’s science-​led realist instinct is what drives his critical concern with Wittgenstein’s conception of our relation to the world. In ‘Wittgenstein and Idealism’, he argues that Wittgenstein’s language-​oriented conception is haunted by the spectre of idealism. Wittgenstein famously declared in Remarks on the Foundations of Mathematics: ‘Not empiricism and yet realism in philosophy, that is the hardest thing’ (Wittgenstein 1978: 325 VI-​23). Agreed, and Williams’s concern was that the desired realism was not obviously achievable given other of Wittgenstein’s commitments. Williams agreed with Peter Hacker30 that the anti-​ solipsistic guarantees of the Private Language Argument successfully fend off phenomenalism and any merely aggregative species of idealism whose conception of the world simply adds my language-​world to yours, and yours . . . and so on. Such a view, says Williams, would amount to a non-​transcendental or ‘empirical’ idealism, committed to the incoherent idea that the world is dependent on minds that are themselves part of the empirical world. However, he goes on to argue that there remains a residual worry: even on the compelling non-​private conception of language, we still have a view that seems very close to some other kind of idealism. It is tempting to re-​express Williams’s concern here in terms of the worry that Wittgenstein’s philosophy is unable to sustain something that Williams after all believed to be entailed by the very idea that we know anything: namely, the absolute conception. But in doing so, I admit such a rational reconstruction would overreach the textual evidence for the form of his worry, and certainly would not make it any more compelling to those of Wittgensteinian sensibility. Recording the temptation, however, is useful if only to mark the fact that in Williams’s own thought the absolute conception is the imaginative anchor for his realist commitments as regards our knowledge of the material world, and particularly scientific knowledge, and that it underpins the essential contrast between scientific and ethical knowledge. Be that as it may, the explicit form of Williams’s worry concerning the conception of reality available in Wittgenstein’s philosophy was that the world came out looking too much like a function of linguistic practice. Specifically, he argued, albeit in a markedly qualified manner, that the later-​Wittgensteinian idea that the limits of our world are the limits of our language still contained distinct elements of a transcendental idealist position. If there is nothing that could count for us as the world except the world as it is intelligible to us in language, then that at least seems very close to a transcendental idealist position, or may be indistinguishable from one.31 Williams summed up his concern as follows: Such a representation I called in my jargon “the absolute conception of the world”. Whether it is attainable or not, whether the aspiration to it is even coherent, are of course highly disputable questions’ (Williams 2006a: 184–​5). 30 Williams references Hacker (1972). 31 But see Stephen Mulhall’s critical discussion (Mulhall 2009), particularly his suggestion that Wittgenstein’s purpose may have been rather to show that we need to move beyond the fundamentally misleading contrast between realist and idealist positions—​‘that the very idea of a distinction between [p. 478] The fact that in this way everything can be expressed only via human interests and concerns, things which are expressions of mind, and which themselves cannot ultimately be explained in any further terms: that provides grounds, I suggest, for calling such a view a kind of idealism (and not of the stupid ‘aggregative’ Another way of bringing out Williams’s science-​inspired realist inclinations in relation to Wittgenstein’s philosophy is to consider that Wittgenstein effects a metaphysical levelling of all domains of knowledge—​notably, scientific and ethical. If all regions of knowledge are equally to be understood in relation to the contours of our ‘language games’, then there are no significant asymmetries when it comes to their metaphysical status. This levelling, impressively built upon by Wittgensteinian ethical realists who effectively consider it a levelling up for ethics,32 is precisely what Williams believed we needed to resist in order to make sense of the mind-​independence of the basic physical features of the world, and indeed to locate the contrast with the historical contingency of ethical thought. This concern finds a more straightforwardly normative expression in Williams’s paper ‘Pluralism, Community, and Left Wittgensteinianism’: [On the later Wittgensteinian picture] whether in mathematics, in our ordinary language, or in ethics, everything equally is a matter of practice, of what we find natural. It is mistaken, on this picture, to try to ground our practices, whether ethical or cognitive; we must rather recognize that our way of going on is simply our way of going on, and that we must live within it, rather than try to justify it. But Williams took this to be unsustainable, because in the modern era the domain of the ethical and political is characterized by an array of alternative actual and possible ways of going on, so that the injunction to keep on going on is hopelessly under-​determinative. This normative point finds expression once again in relation to the Wittgensteinian insistence on the universal use of ‘we’ and ‘our’: When Wittgensteinians speak of ‘our’ form of life, they characteristically use that expression in what linguistics calls an inclusive rather than a contrastive way: the idealism and realism . . . turns out, when strictly followed through, to be simply and straightforwardly empty’ (2009: 388). 32 As Sabina Lovibond showed in her Realism and Imagination in Ethics (1983), part of the appeal of constructing a realist view of both science and ethics by way of this levelling is that is highlights our inescapable responsibility for how we go on in all regions of intelligibility. In the relevant discussion in Ethics and the Limits of Philosophy, Williams notes Lovibond’s book. Other key figures in this connection are John McDowell (e.g. McDowell 1979), Cora Diamond (see e.g. Diamond 1995), and, more recently, Alice Crary (see e.g. Crary 2007). [p. 479] Miranda Fricker ‘we’ represents not us as against others, but an ‘us’ that embraces anybody with whom we could intelligibly hold a conversation. It is obvious why this should be appropriate to matters of meaning and of understanding, and why also it is plausible to say in those connections that we have no way of standing outside ‘us’. But we cannot simply take this idea over into ethical and social thought . . . (Williams Here Williams is talking about ‘Wittgensteinians’ rather than Wittgenstein himself; but in this particular I think it makes little difference. The metaphysical levelling effected by Wittgenstein’s philosophy of language is found problematic by Williams on at least two scores: it sells short the objective reality of the physical world, thus by implication also underselling the difficulties of science; and it trades in an insufficiently ‘historically and socially realistic’33 conception of the ethical, gliding over the potentially intractable difficulties and lacunae in determining how to go on in the midst of conflicting value commitments.34 We saw in Williams’s discussion of the absolute conception that he took us to be confronted with an apparent dilemma concerning the idea of a mind-​independent reality—​an all-​or-​ nothing dilemma that we could, however, escape by embracing his perspectivalist ‘third way’. He said there that (as I put it) we need to disaggregate our representations of the world and see that some, but only some, of them are the minimally perspectival representations of features in the absolute conception. For Williams, there is plenty of our world round and about us, knowledge of which is in some degree perspectival; but given we know anything at all, this must be intelligibly related to a substratum which is there for any knower—​‘there anyway’. This metaphysical picture is obviously quite anathema to the levelled Wittgensteinian landscape where the two regions of thought are on a par. The issue reveals a fundamental metaphysical difference between the two thinkers, and one that is, moreover, expressive of two utterly different philosophical temperaments: Williams’s asymmetrical commitment, on the one hand, to a science-​oriented realism about the physical world of which we are a part and, on the other, to a historicist social 33 Williams does go on to identify the possibility of a ‘Left Wittgensteinianism’, which would not be attended by these problems: ‘What we are left with, if we reject foundationalism, is not an inactive or functionalist conservatism that has to take existing ethical ideas as they stand. On the contrary, once the resultant picture of ethical thought without foundationalism is made historically and socially realistic, in particular by registering in it the categories of modernity, it provides a possibility of deploying some parts of it against others, and of reinterpreting what is ethically significant, so as to give a critique of existing institutions, conceptions prejudices, and powers’ (2005: 37; emphasis added). But, as I understand Williams, this is considered to require renouncing the blanket use of the universal or inclusive ‘we’ as part and parcel of an acknowledgement that the scientific and the ethical are not on a metaphysical par. The envisaged Left Wittgensteinianism would thus embody many of Williams’s philosophical commitments; but it is not clear it would still embody all that many of Wittgenstein’s. 34 For a discussion of William’s ideas of Right and Left Wittgensteinianism, and a defence of an expanded version of the latter using pragmatic resources rooted in Wittgenstein, see Queloz and Cueni (2021). [p. 480] constructionism concerning the ethical is the heart of the second respect in which I am suggesting we should see Williams as deeply ambivalent about Wittgenstein’s philosophy. The third and final feature of Wittgenstein’s thought that I wish to address from Williams’s point of view concerns the idea that sometimes our processes of reasoning and justification run out, so that we come to a point where we can but affirm that our spade is turned, ‘This is simply what I do.’35 With this thought, Williams regards Wittgenstein as having offered us one of the greatest insights of modern analytic philosophy. However, once again the universal application of the idea meant that Williams was not in any straightforward agreement. He thoroughly agreed with its application in value contexts, where he felt we urgently still needed some way of making sense of the fact that some values are for us unhintergehbar—​ he used the German word as a label to crystallize the ethical-​psychological impossibility of renouncing certain of our most deeply held values. Take the fundamental ethical idea that all human life has equal moral worth. For such a conviction, philosophy needs to be able to articulate (for some ‘us’ at a time) the idea that ‘for us, it is simply there’ (2006a: 195), and the Wittgensteinian insistence that the reasons and justifications for such commitments inevitably soon run out so that we arrive at bedrock is a game-​changing philosophical advance. Williams not only agrees with it, and considers it a great insight, he also explains how this very fact that we arrive at justificatory bedrock in such matters is part of what makes a historical perspective in philosophy so valuable, since we need it to help us understand which values have this special status for us, and why this might have come about. He says, in ‘Philosophy as a Humanistic Discipline’: Above all, historical understanding . . . can help with the business . . . of distinguishing between different ways in which various of our ideas and procedures can seem to be such that we cannot get beyond them, that there is no conceivable alternative. This brings us back to Wittgenstein. Wittgenstein influentially and correctly insisted that there was an end to justifications, that at various points we run into the fact that ‘this is the way we go on’. (2006a: 195–​6) Once again, however, Williams would not have agreed with the blanket application of this idea, and in particular, he explicitly disagreed with the implication that since all our ways of going on are as they are, there is no significant normative or explanatory job left for philosophy to do. On the contrary, he asserted that ‘Even with regard to those elements of our outlook for which there are no further justifications, there can still be explanations which help to locate them in relation to their rivals’ (2006a: 195). Philosophy need not and should Once I have exhausted the justifications, I have reached bedrock, and my spade is turned. Then I am inclined to say: “This is simply what I do” ’ (Wittgenstein 2009: 91, para. 217). [p. 481] Miranda Fricker not fall quiet, leaving everything as it is, even in those cases where we have run out of justifications for the proposition, for instance, that all human lives possess equal moral worth. For there is other work to be done besides tracing direct lines of justification. There is, for instance, critical work to be done in relation, as Williams puts it, to comparing our ethical ideas with their rivals, or refining exactly what it is that we find unhintergehbar. Moreover, his historicism allows—​insists—​that what is unhintergehbar, for a given ‘us’ in a moral culture, can change over time. The associated Wittgensteinian idea that there are no genuine deep philosophical problems, but only problems conjured into existence by philosophy itself, is also an element that Williams found problematic because overgeneralized. He was not averse to drawing this verdict in particular cases. He himself argued, for instance, that the problem of free will was exactly like this. In Shame and Necessity, he effectively advocated quietism on the free will issue, though sourcing the relevant attitude in Nietzsche rather than Wittgenstein, and praising the Greeks for having the right idea: The traditional metaphysical problem of the freedom of the will . . . exists only for those who have metaphysical expectations. Just as there is a ‘problem of evil’ only for those who expect the world to be good, there is a problem of free will only for those who think that the notion of the voluntary can be metaphysically deepened. In truth, though it may be extended or contracted in various ways, it can hardly be deepened at all. What threatens it is the attempt to make it profound, and the effect of trying to deepen it is to put it beyond all recognition. The Greeks were not involved in those attempts; this is one of the places at which we encounter their gift for being superficial out of profundity. (1993: 68)36 But the point is that free will was, for Williams, a special case; indeed, an exception proving the rule.37 As ever, he resisted the universal, proceeding in a more piecemeal fashion—​insisting, in his own terms, on ‘looking and seeing’. Wittgenstein, of course, had his own distinctive reasons for coming to his quietist conception of philosophy, and they were undoubtedly general in nature. But Williams believed that Wittgenstein’s dramatically chastened conception of what philosophy can do for us stemmed significantly from the idea—​the false idea—​that it is an essentially 36 Here Williams is alluding to Nietzsche’s comment ‘Those Greeks were superficial—​out of profundity’ (Nietzsche 1974: 38). 37 For a slightly different angle on this point, see Queloz and Krishnan (forthcoming), who suggest that Williams’s critique of the ‘morality system’ in Ethics and the Limits of Philosophy can itself be seen as a piece of philosophy-​as-​therapy, albeit with a powerfully critical bent (ms: 29). On this, wider, conception of philosophy-​as-​therapy, one would naturally see a large portion of Williams’s ethics as therapeutic—​perhaps we should say ‘critical-​therapeutic’ to make explicit the anti-​quietist, critical-​ theoretic purpose. [p. 482] and radically peculiar intellectual activity. Worse, this essentializing conception of philosophy as radically peculiar is possessed of an unfortunately self-​f ulfilling power. Williams says: [S]‌ome of the deepest insights of modern philosophy, notably in the work of Wittgenstein, remain undeveloped—​indeed, at the limit, they are rendered unintelligible—​precisely because of an assumption that philosophy is something quite peculiar, which should not be confused with any other kind of study, and which needs no other kind of study in order to understand itself. Wittgenstein in his later work influentially rejected essentialism, and spoke of family resemblances and so on, but at the same time he was obsessed—​I do not think that is too strong a word—​by the identity of philosophy as an enterprise which was utterly peculiar compared with other enterprises. (2006a: 182) By contrast, when it came to the nature of philosophy and its relation to other forms of enquiry or intellectual activity, Williams was basically on Descartes’ side. Unlike Wittgenstein, Descartes saw philosophy as continuous with other forms of enquiry, both science and, of course, history. In the Descartes book, just after recounting Descartes’ comment to Princess Elizabeth that philosophical enquiry should be undertaken at most only once in a lifetime, and by most not at all, Williams explains: [T]‌his is not at all because Descartes holds, as some modern views hold, that philosophy has a quite special subject-​matter, or is in some other way quite discontinuous from the sciences; on the contrary he thinks rightly, that philosophy and the sciences are continuous with one another, and he expresses one version of that idea in his image of the tree of knowledge, of which the roots are metaphysics, the trunk physics, and the branches, the other sciences . . . (1978: 34) 4. Conclusion Williams believed in philosophy’s obligations to science, for all the reasons we have been discussing; but above all, he advocated that philosophy needed history. This is the chief impurity that our enquiries need in order to make sense of our world as it has come down to us, including the modes of understanding it that we have inherited into the bargain. History can assist philosophy by preventing us from reinventing the wheel, or indeed the square wheel,38 but Williams was In general, one must take extremely seriously Santayana’s warning, that those who are ignorant of the history of philosophy are doomed to recapitulate it (not just reinventing the wheel, but reinventing the square wheel)’ (Williams 2006a: 204). [p. 483] Miranda Fricker equally aware, of course, that history can undermine our philosophical efforts by discrediting the conceptions they employ.39 In this connection, he most often mentioned Rorty, because Rorty so vehemently and explicitly argued against the so-​called ‘epistemological’ conception of philosophy as a ‘mirror of nature’, which he rejected wholesale, presuming it essential to the enterprise (which is strange coming from a champion of contingency). Rorty’s own alternative conception of philosophy, or at least of ethical and political thought, was that height of peculiarity: ‘liberal ironism’. This conception required the philosopher to be possessed of dual personae, operating as public liberal ‘out there’ in life, and private sceptic when in the study. But Rorty aside, the point I want to make is that Williams might just as naturally in this connection have referred to Wittgenstein’s equally disillusioned conception of philosophy as turning out to lack (and after such a promising debut, too) any genuine puzzles to solve, a match only for those of its own fantasized making, so that the best place we can find for philosophy is, if not shut in the study exchanging self-​undermining reflections with the ironist, perhaps sitting in the corner simply keeping quiet. Williams found that both Rorty’s and Wittgenstein’s conceptions of what philosophy should be were catastrophically misguided. I would suggest that they both have the form of a counsel of despair after a disillusionment with philosophy that would never have happened were it not for misguided ambitions for it in the first place. As ever, the way to avoid disillusionment is to start out with a more realistically conceived idea of how philosophy is related to other forms of enquiry, most particularly (for Williams) science and history. Both scientism and apriorism squeeze out historical self-​consciousness, and so squeeze out the prospect of philosophy being able to really help us make sense of our practices in the present in the light of the past. That is not something we can do in science, and it is not something we can do exclusively a priori either. So let me end with Williams’s own way of stating the present challenge for philosophy in its self-​ conception as a humanistic discipline, for it seems to me that over twenty years after he wrote these words of warning, they are, if anything, more urgently applicable in our time: We run the risk . . . that the whole humanistic enterprise of trying to understand ourselves is coming to seem peculiar. For various reasons, education is being driven towards an increasing concentration on the technical and the commercial, to a point at which any more reflective enquiry may come to seem unnecessary and archaic, something that at best is preserved as part of the heritage industry. If we thought that our outlook had a history which was vindicatory, we might to that extent ignore it, precisely as scientists ignore the history of science. . . . But if we do not believe that the history of our outlook is vindicatory, then understanding the history of our outlook may seem to interfere with our commitment to it, and in particular with a philosophical attempt to work within it and develop its arguments (Williams 2006a: 192). [p. 484] If that is how it is preserved, it will not be the passionate and intelligent activity that it needs to be. (2006a: 199)40 References Babiotti, Paulo. 2020. ‘Nietzsche, Williams, and the Problem of Style’. European Journal of Philosophy 29.4: 937–​47. Berman, Marshall. 1982. All That Is Solid Melts into Air: The Experience of Modernity. New York: Verso. Blackburn, Simon. 2009. ‘The Absolute Conception: Putnam vs. Williams’. In Reading Bernard Williams. Edited by D. Callcut, 9–​23. Abingdon and New York: Routledge. Craig, Edward. 1990. Knowledge and the State of Nature: An Essay in Conceptual Synthesis. Oxford: Clarendon Press. Crary, Alice. 2007. Beyond Moral Judgement. Cambridge, MA: Harvard University Press. Diamond, Cora. 1995. The Realistic Spirit: Wittgenstein, Philosophy, and the Mind. Cambridge, MA: The MIT Press. Foot, Philippa. 1958. ‘Moral Arguments’. MIND 67.268: 502–​13. Fricker, Miranda. forthcoming. ‘ “Williams” “Philosophical Anthropology”: A Humean and Nietzschean Synergy’. European Journal of Philosophy Special Issue ‘Philosophy and History’ eds. Wayne Martin, Andrew Norris, David Stern. Fricker, Miranda. 1996. Review of Mary Warnock’s Women Philosophers. Times Higher Education Supplement. Fricker, Miranda. 2000. ‘Confidence and Irony’. In Morality, Reflection, and Ideology. Edited by E. Harcourt, 87–​112. Oxford: Oxford University Press. Fricker, Miranda. 2020. ‘Bernard Williams as a Philosopher of Ethical Freedom’. Canadian Journal of Philosophy 50.8: 919–​33. Reprinted in Morality and Agency: Themes from Bernard Williams. Edited by András Szigeti and Matthew Talbert, 265–​89. Oxford: Oxford University Press, 2022. Hacker, P. M. S. 1972. Insight and Illusion: Wittgenstein on Philosophy and the Metaphysics of Experience. Oxford: Oxford University Press. Heuer, Ulrike. 2012. ‘Thick Concepts and Internal Reasons’. In Luck, Value, and Commitment: Themes from the Ethics of Bernard Williams. Edited by U. Heuer and G. Lang, 219–​46. Oxford: Oxford University Press. Kirchin, Simon, ed. 2013. Thick Concepts. Oxford: Oxford University Press. Kirchin, Simon, ed. 2018. Thick Evaluation. Oxford: Oxford University Press. Lang, Gerald. forthcoming. ‘Williams’s Wittgenstein: Who are “We”?’ In Wittgenstein and Other Philosophers: His Influence on Historical and Contemporary Analytic Philosophers. Edited by A. Hosseinkhani and G. Kemp. London: Routledge. Le Doeuff, Michèle. 1990. The Philosophical Imaginary. Stanford: Stanford University Press. Lovibond, Sabina. 1983. Realism and Imagination in Ethics. Minneapolis: University of Minnesota Press. McDowell, John. 1979. ‘Virtue and Reason’. The Monist 62.3: 331–​350. Reprinted in McDowell, Mind, Value, and Reality, 50–​73. London: Harvard University Press, 1998. 40 A version of this chapter was first published as the 2023 Mark Sacks Lecture in the European Journal of Philosophy 32.2 (2024: 301–​20), and I am most grateful to Joseph Schear, as editor of the journal, for kind permission to reprint here. Iterations of the talk were given at the Edinburgh Workshop on Bernard Williams, where my commentator was Ulrike Heuer, and at the College of Wooster, Ohio, as the Lindner Lecture in Ethics, where my commentators were Marcel van Ackeren and Claire Kirwin. I am immensely grateful for all the comments and discussion. [p. 485] Miranda Fricker Moore, A. W. 1985. ‘Transcendental Idealism in Wittgenstein’. Philosophical Quarterly 35.139: Moore, A. W. 1997. Points of View. Oxford: Clarendon Press. Moore, A. W. 2006. ‘Maxims and Thick Concepts’. Ratio (new series) 19.2: 129–​47. Moore, A. W. 2007a. ‘Realism and the Absolute Conception’. In Bernard Williams. Edited by A. Thomas, 24–​46. Cambridge: Cambridge University Press. Moore, A. W. 2007b. ‘Wittgenstein and Transcendental Idealism’. In Wittgenstein and His Interpreters: Essays in Memory of Gordon Baker. Edited by G. Kahane, E. Kanterian, and O. Kuusela, 194–​218. Oxford: Blackwell. Moran, Richard. 2016. ‘2015 Mark Sacks Lecture: Williams, History, and the “Impurity of Philosophy” ’. European Journal of Philosophy 24.2: 315–​30. Mulhall, Stephen. 2009. ‘ “Hopelessly Strange”: Bernard Williams’ Portrait of Wittgenstein as a Transcendental Idealist’. European Journal of Philosophy 17.3: 386–​404. Nietzsche, Friedrich. 1974. The Gay Science. Translated by Walter Kaufmann. New York: Vintage. Putnam, Hilary. 1992. ‘Irrealism and Deconstruction’. In Putnam, Renewing Philosophy, 108–​33. Cambridge, MA: Harvard University Press. Queloz, Matthieu. 2021. The Practical Origins of Ideas: Genealogy as Conceptual Reverse-​ Engineering. Oxford: Oxford University Press. Queloz, Matthieu, and Cueni, Damian. 2021. ‘Left Wittgensteinianism’. European Journal of Philosophy 29.4: 758–​77. van Ackeren, Marcel. 2019. ‘Williams (on) Doing History of Philosophy: A Case Study on Ethics and the Limits of Philosophy’. In Ethics Beyond the Limits: New Essays on Bernard Williams’ Ethics and the Limits of Philosophy. Edited by Sophie Grace Chappell and Marcel van Ackeren. Ethics Beyond the Limits: New Essays on Bernard Williams’ Ethics and the Limits of Philosophy. London and New York: Routledge. Warnock, Mary. 1996. Women Philosophers. London: Orion. Williams, Bernard. 1978. Descartes: The Project of Pure Enquiry. London: Penguin. Williams, Bernard. 1981. Moral Luck: Philosophical Papers 1973–​1980. Cambridge: Cambridge University Press. Williams, Bernard. 1985. Ethics and the Limits of Philosophy. London: Fontana. Williams, Bernard. 1993. Shame and Necessity. Berkeley: University of California Press. Williams, Bernard. 1995. ‘Saint-​Just’s Illusion’. In Williams, Making Sense of Humanity and Other Philosophical Papers 1982–​1993, 135–​50. Cambridge: Cambridge University Press. Williams, Bernard. 2002. Truth and Truthfulness: An Essay In Genealogy. Princeton: Princeton University Press. Williams, Bernard. 2005. ‘Pluralism, Community and Left Wittgensteinianism’. In Williams, In the Beginning Was the Deed: Realism and Moralism in Political Argument. Edited by Geoffrey Hawthorn, 29–​39. Princeton: Princeton University Press. Williams, Bernard. 2006a. Philosophy as a Humanistic Discipline. Princeton: Princeton University Press. Williams, Bernard. 2006b. On Opera. New Haven: Yale University Press. Williams, Bernard. 2006c. The Sense of the Past: Essays in the History of Philosophy. Princeton: Princeton University Press. Wittgenstein, Ludwig. 1922. Tractatus Logico-​Philosophicus. London: Kegan Paul, Trench, Trubner & Co. Wittgenstein, Ludwig. 1978. Remarks on the Foundations of Mathematics. Cambridge, MA: MIT University Press. Wittgenstein, Ludwig. 2009. Philosophische Untersuchungen. Philosophical Investigations. Edited by P. M. S. Hacker and Joachim Schulte. Translated by G. E. M. Anscombe, P. M. S. Hacker, and Joachim Schulte. Oxford: Blackwell. [p. 486] Index For the benefit of digital users, indexed terms that span two pages (e.g., 52–​53) may, on occasion, appear on only one of those pages. action-​guiding 207, 457–​58 Aeschylus 57, 59–​60, 63, 65, 68 agency human 254, 365–​66 moral 176, 416, 442 agent-​regret 70, 252–​53 analytic genealogy 383–​84, 385–​86 8, 451 Berlin, Isaiah 361, 369–​72, 373, 375 Blackburn, Simon 156–​57, 164, 171 Brandom, Robert 396–​99 categorical imperative 206, 209–​11, 413–​14 Cavell, Stanley 285–​86, 301, 309 Collingwood, Robin 260–​82, 338–​39, conception of reality 123, 269–​70, 464 scientific 275 consequentialism 130, 270–​71, 307, 394 critical theory principle 426 culture 167, 190–​91, 266 76, 451 moral 244–​45 ethical agent 59, 189–​90 considerations 38–​39, 408–​9, 411–​12 culture see culture dispositions see dispositions knowledge 417–​18, 456 reality 129, 131 74, 407 thought/​thinking 102, 129, 166–​67, evolutionary anthropology 384–​85, 396 fact-​defective 222–​23, 224, 387–​88 Foucault, Michel 223, 256–​57, 385 reflective 201–​4 Frege, Gottlob 350–​51, 390–​91 functionalist 293–​98, 301, 390, 398 Fundamental Assumption 327–​28, 331 Descartes, René 117–​33, 134–​55, 272–​73, 321–​ determinism 52, 178, 181 dispositions 163 genealogical 80–​81, 82–​89, 177, 254, 346–​47, genealogy fallacy 383–​84, 390 imaginary 223, 383–​84, 387–​88 vindicatory 73, 193, 240, 366, 387 Greek philosophy 33, 109, 249, 329 emotions 45, 179–​80, 401–​2 epistemic imperialism 260–​61, 271–​72 ethic of responsibility 436, 443 Hacker, Peter 401–​2, 464 Hegel/​Hegelian(s), 60–​61, 201, 298, 299 historically contingent 366, 414, 461 [p. 487] Inde x history of ideas 117–​18, 264–​65, 319–​20, 324, historical reality 275, 329, 424 truth 343, 346, 451 understanding see understanding Hobbes/​Hobbesian 178, 225–​26, 426, 427–​29 human purpose(s), 294–​95, 309, 367 humanitarian intervention 438, 442, 443 post-​Humean 278 imaginary genealogy see genealogy individualism 164, 166 formal 163 internal reasons thesis 409, 412, 413–​14, 417 interpersonal dialogue 97 virtue of 39–​40, 62, 224 anti-​, 163 ethics 198–​99, 203–​4 law-​defective 222–​23, 387–​88 laws of nature 74, 275, 279–​80 of concepts 391 Leiter, Brian 244–​46 local phenomenon 365–​66, 376 McDowell 408, 409, 457 metaphilosophical 107, 275–​76, 300–​6, 388–​89 method of doubt (Descartes), 127–​28, 137, 140 mind-​independent 453, 465, 466–​67 modernist 450, 452 Moore, Adrian 112, 274–​75, 278–​79 moral address 184–​90 beliefs 233–​34 culture see culture obligation(s), 176, 205–​6, 213, 262–​63 responsibility 176–​78, 179–​80, 182–​83, 185–​ and human nature 160–​61, 165–​66, 168–​69 (inter)personal 80–​81, 423–​24 and practical reason 43, 158 and rationality 242, 249–​50 Nagel, Thomas 119, 121, 127, 183–​84, Nietzsche, Friederich 172, 190–​91, 226–​34, normative power 212 principles 212–​13 Nozick, Robert 222–​23, 387–​88 Nussbaum, Martha 59–​61, 66, 68–​69, 157, 288–​89 objective reality 121, 127, 147–​48, 466–​67 distinct 121, 125, 134–​55 philosophical understanding see understanding pre-​platonic 37, 38 political theory, 110–​11, 440 power abuse of 431, 433–​35 justification of 439–​43 power 427, 432 primary qualities 153, 272–​73, 455–​56 psychological facts 131–​32, 179–​80 Putnam, Hilary 273–​74, 416–​17 rational agency see agency deliberation 207, 412 reconstruction 322–​23, 335, 341–​42, 343, 450 rationality/​reason irrationality 56–​57, 108, 162–​63, 184–​85, 414 [p. 488] Inde x Rawls/​R awlsian 201, 372–​73 Reginster, Bernard 227, 230, 231, 233–​34 religious metaphysics 117 Rorty, Richard 86, 288, 323, 448–​49 scientific inquiry 124, 129, 268–​70, 275 self-​conscious(ness), 201, 228–​29, 412, 446–​72 single human nature 171–​72 see also Hume State of Nature 73, 224–​26, 230–​31, 255–​56, Strawson, Peter 179–​80, 386, 420 philosophical 171–​72, 349–​54, 395 unhintergehbar 303–​4, 467–​68 utilitarianism 68–​69, 85–​86, 179, 239, 270–​71, thick (ethical) concepts 273, 293–​98, 304, 416–​ Thucydides 48–​49, 253 truthfulness 82, 85, 254–​56, 296–​97, 347, Truth Rule 135, 142 value instrumental 85–​86, 297, 327–​28 of truth 255–​56, 293 view from nowhere 119, 154, 274–​75 see also conception of reality, absolute vindicatory genealogy see genealogy virtue(s) altruistic 88–​89 and genealogy 72–​91, see also genealogy and Hume 165, 180–​84 intellectual 331 of justice 224 and knowledge 108 and life 41, 89, 101 theoretical 292 of truth(fulness), 72, 256, 301–​2, 366, 388, underdetermination 104, 137–​38, 191 understanding Weber, Max 429, 435, 438 Western 104, 119, 362 philosophy 33, 319, 329, 342, 348 Wittgenstein, Ludwig 283–​315, 386, 459–​69