【B.Williams】Bernard Williams: Life as Narrative

https://onlinelibrary.wiley.com/doi/full/10.1111/j.1468-0378.2007.00275.x

‘We dream in narrative, day‐dream in narrative, remember, anticipate, hope, despair, believe, doubt, plan, revise, criticise, gossip, learn, hate and love by narrative.’ The words are Barbara Hardy's, and they are quoted with approval by Alasdair MacIntyre, whose use of the idea of narrative to shed light on human transactions and human lives I take as a point of departure.1 MacIntyre himself writes: ‘Narrative is not the work of poets, dramatists and novelists reflecting upon events which had no narrative order before one was imposed by the singer or the writer: narrative form is neither disguise nor decoration’. In actual life, we are the co‐authors (but only the co‐authors) of the stories of our lives: ‘stories are lived before they are told—except in the case of fiction’.1

MacIntyre applies this idea at three distinct levels. There is the level, first, of intelligible action. An action is intelligible if, and only if, it can be situated in a narrative setting. Moreover, it is a mistake to think that some actions happen to be intelligible, while others are not: the notion of an intelligible action is the primary notion, and unintelligible actions are ones that fail to be intelligible. Thus ‘narrative history of a certain kind turns out to be the basic and essential genre for the characterisation of human actions’.1 When I perform actions, engage in conversations and other human transactions, I am a character in a narrative of which I am in part the author.

The account in terms of narrative applies, further, at the level of living. Human beings turn out to be in their actions and practice, and not just in their fictions, essentially story‐telling animals. But the narrative I construct for myself has to be part of a larger narrative enterprise, reaching beyond myself. This is one reason why I am only co‐author of the narrative. It means, too, that in deciding what to do, I must consult my narrative environment. ‘I can only answer the question “What am I to do?” if I can answer the prior question “Of what story or stories do I find myself a part?”.’ This is because we enter society ‘with one or more imputed characters—roles into which we have been drafted’, as MacIntyre rather threateningly puts it.1

The story of which I am part will offer, among other things, the story of my life. So, finally, narrative provides not merely an account of the process of living: it also provides the basis of the unity of a life. A whole life gets its sense from its success in embodying or presenting a coherent narrative; the idea of coherence appropriate to a life, in fact, is that of a narrative. When MacIntyre writes ‘the unity of a human life is the unity of a narrative quest’,1 I take it that the large ambiguity of his statement is to be resolved in favour of his meaning the quest for a narrative, rather than the narrative of a quest. For he seems to agree that a life might fail to achieve unity, and if so, his idea must surely be that it will present a failed or broken narrative, rather than the coherent—indeed, perhaps the entirely unified—narrative of a failed quest for something else.

When MacIntyre says that actions have a narrative structure, and that the structure is not merely imposed by ‘poets, dramatists and novelists’ in their telling, we have to ask why these people in particular are introduced, what specific role they represent. Are they introduced because they tell, specifically, fictions? Or because they tell in an artful way?—in which case they might have included historians as well. Or because, merely, they tell?—in which case they might include any of us story‐telling animals. When MacIntyre says that the narrative structure of actions is prior to these people's narrations, does he mean that it is prior to fictional narration, to any artful narration, or to any telling at all?

The last of these options is very hard to understand. Philosophers have disagreed, indeed still do, about the sense and the extent to which causality is prior to human explanation. With regard to many causal claims—the claim for instance that craters on the moon are due to meteorite impact—we can reasonably accept that that is how things are, whether there was anyone to tell the tale or not. We can extend that idea to some proto‐narratives, in the sense of descriptions of temporal sequences of events linked by causality, which serve as explanations: when we explain, for instance, that the mountains of the Auvergne are shaped like that because they used to be volcanoes, or why there are marsupial species in Australasia and hardly anywhere else. If the story we tell is a good explanatory tale, what makes it a good tale is independent of the telling, in the sense that if it is true, things would have been so even if there were never any narrators. But it is hard to extend that picture of things to cover claims about the essentially narrative understanding of complex human actions and the living of lives; there, what makes a given story a good story cannot be altogether prior to any telling at all—even if it may be prior to the telling of that particular story. One respect—by no means the only one—in which the narrative structure of complex human agency could not be entirely prior to narration, in the sense of mere telling, is that it could not be entirely prior to telling by agents themselves. This, surely, is what MacIntyre means. So when he says that narrative structure is prior to narrations by novelists and so forth, he does not mean that narrative structure is prior to any telling at all, but that artless telling is prior to artful telling, or factual telling is prior to fictional, or both.

Yet isn't there a problem even with the relatively unambitious claim that the ideas of complex human actions and of the unity of a life have to be explained through the possibility of narrative? A narrative is a story about a temporal sequence of happenings. There can be many different ways in which a story, although about different happenings, may yet have a unity or be (in the most extended sense) about one thing. For instance it might be about a sequence of happenings at one place, or it might be the story of a penny. In the case of the marsupials, it will be a story about a piece of the earth's surface. Among these stories are some, each of which could claim to be the story of one person's life. But we could not identify those narratives unless we had a conception of a person's life, and recognized from that conception what the story of such a thing might be like: any more than we could recognize a story about a penny unless we knew, more or less, what a penny was. Hence our conception of a person's life cannot be derived from such a narrative.

‘Story of one person's life’ cannot pick out a narrative by purely formal or syntactic criteria. It could not be a type of narrative, for instance, defined by the recurrence of a personal proper name: ‘personal proper name’ is not a syntactic category, and the question of the appropriate reference of such a name raises the same problems again. In any case, if the idea of such a narrative is supposed to imply some appropriate notions of coherence, we need more than the recurrence of what we have somehow identified as a personal proper name: perhaps ‘HCE’ in Finnegans Wake is a personal proper name, but its use there does not invite those notions of coherence.

It seems that we could not use the idea of narrative to model a person's life unless we could independently pick out a person; and since what is in question is a person's life, a sequence (at the very least) of what happens to him or her, we need to have that notion as well. Indeed, it is reasonable to think that this is not a separate requirement, and that we would not even have a notion of a person unless we had some idea of what it is for a person to lead a life.1

This is the first part of an argument against MacIntyre's claim. But it is tempting to go a stage further, and extend the argument. It is not merely that we need to have already some idea of a person's life; we need to have already an idea of the coherence of a person's life. (This is the second part of the argument.) How, it may be asked, could we gather from certain kinds of narrative certain forms of coherence supposedly appropriate to people's lives—as opposed to other ways in which narratives can hang together or make sense—unless we were guided by some antecedent notion of coherence, given to us by our understanding of our own and other people's lives? If we get the same result with that question as we did with earlier questions of the same kind—that is to say, if it turns out that we can understand the coherence given in the narrative of a life only because we already understand the coherence of a life—the analogy of narrative is not going to provide anything very important. We shall have had to assume before we get to it most of what it is supposed to provide.

MacIntyre sees that some questions of this kind are raised, but he thinks that they are answered by speaking of ‘mutual presupposition’ among these conceptions: we understand the idea of a narrative about a person because we understand what a person is, but we also understand what a person is because we have the idea of a narrative about them. But this seems to me to fail in explanation: we need more structure than this gives us, if we are to see what these dependencies might be. Moreover, and more substantively, unless we have a better grasp of how these ideas hang together, we shall not know how to face—in particular, we shall not be able to decide how far we should accept—some very sceptical consequences that are often associated with these uses of the idea of narrative. It is a banality to say, with regard to any complex subject matter, that there are no definitive narratives; that each is perspectival and perhaps incommensurable with others; that narratives are constructed, not discovered. To the extent that these consequences roll in, the supposed priority of fact over fiction will turn out not to have much substance to it. There are some particular worries (I am going to suggest later that they are largely justified) about the status of the coherence that is given to a life by a narrative structure; we need to ask whether the ‘inevitability’ that may be conveyed by a narrative, and the capacity of narratives to represent some developmental sequences as coherent while others seem arbitrary or inexplicable, may not express some other, external, kinds of constraint, tacitly appealing to a power which is not simply that—if there could be such a thing—of the truth. MacIntyre himself does not discuss these concerns, and shows a remarkably robust faith in the determinacy of narrative interpretations of at least some people's lives. But we will not know how far we can agree with him unless we do some more to spell out these dependencies.

I think that MacIntyre can up to a point answer the argument against him, but only at the cost of some substantial concessions. The first part of the argument, as I called it earlier, can be met by conceding that the narrative paradigm does not reach all the way down to the most elementary levels of understanding what a person is and what it is for a person to act. The narrative paradigm applies more strongly to human and personal affairs to the extent that they are more socially and temporally complex. Persons are human beings, and human beings are animals, and while every one of their intentional actions is touched by culture, some are simpler and invite (or may permit) less interpretative depth than others. So in the order of explanation, and to some extent in the order of learning as well, we start with a simple idea of a person and a simple idea of an action, which we can identify without an appeal to narrative structure. These help to form ideas of complex actions which indeed demand and rest on the idea of a narrative; but we already have a schematic conception of what such a narrative would look like, and of how it would present a complex action by one person over time, such as building a house.1 (The whole thing is an example of the familiar process of boot‐strapping.)

So also with the idea of a person's life. Starting with the ideas of a person's simple and complex actions, and a purely structural idea of their life, given in the first instance by the extent of their biological existence, we can conceive a space which will indeed be filled only by narrative structures.

What about the constraints of coherence? What I called earlier the second part of the argument insisted that we needed some idea of the coherence of a life before we could know how to deploy narrative in order to articulate it. This second part of the argument, however, seems to me not to follow. The most that we can properly demand is that, before we apply narrative's conceptions of coherence to a life, we must have some grasp of the material to which those conceptions are to be applied. We must have a conception of what invites interpretation in these terms. This material will itself no doubt involve complex actions and in various ways use the resources of less ambitious, shorter term, narratives. But the fact that we have this material may do very little in itself to determine what we expect the coherence of a life to be; it may leave it very largely open what the resources of interpretation are, and where they come from.

This picture is enough, I think, to answer the claim pressed by the second part of the argument against MacIntyre: the claim, that is to say, that it cannot possibly be true that the idea of the coherence of a life is to be found essentially in narrative, since we could not identify the right kind of narrative without already having the idea of a coherent life. The answer is that we do not need to recognize the coherence of a life in advance of its interpretation in narrative, because we are not simply given a life and its narrative coherence. We are given (or, by the earlier processes, construct) the material about which questions of coherence can be asked, and those questions may indeed be answered in terms of a narrative and its coherence.

But just because we are not simply given a life and its narrative coherence, we are also not automatically provided with an answer to the sceptical doubts that are readily raised by accounts in terms of narrative, particularly at this level. It may emerge that in giving MacIntyre an intelligible theory, we have at the same time left it open to the sceptical doubts, and also, perhaps, taken away from him any substantive claim that fact is prior to fiction. Here, at the level of narrative interpretation of a whole life, the most interesting questions are about the sources of these interpretations; their standing; and their relations to fiction. The rest of this discussion will be concerned with these questions.

At three different points in his book The Thread of Life Richard Wollheim refers to an admirable remark of Kierkegaard's, from his Journal for 1843:

It is perfectly true, as philosophers say, that life must be understood backwards. But they forget the other proposition, that it must be lived forwards. And if one thinks over that proposition, it becomes more and more evident that life can never really be understood in time simply because at no particular moment can I find the necessary resting point from which to understand it—backwards.

Kierkegaard says that I am never in a position to understand my life in the only possible way, backwards, and hence life cannot be understood in time; but that cannot follow, unless we make another assumption (congenial indeed to Kierkegaard) that if anyone could understand my life in time, it would be myself. Perhaps I, for Kierkegaard's very powerful reason, can never understand my life properly in time, because I can never be in the position to give the retrospective narration of it; but others might. But then a fundamental question arises, of what the relation is supposed to be between the coherence of my life and my way of living it. How does the final story of my life stand to the questions I ask myself and the reflections I make in the course of living it?

MacIntyre, to turn back once more to his account, suggests an account of this relation which is radically, and revealingly, mistaken. In a passage that I quoted earlier, he suggested that I can answer the question ‘What am I to do?’ only if I know of what story or stories I am part. But how am I supposed to know what stories I belong to, except by deciding what I am going to do? If it is my story, that is the way in which I decide how it is to continue. MacIntyre's implication was that there can be, to some extent, an independent answer to the question, because society had cast me: deciding what I should do is discovering who I am supposed to be, what character in fact I am.

An example that MacIntyre gives of the narrative significance of a life (and incidentally of his confidence that there are right answers about such things) does not make this answer seem appealing, or even fully intelligible. He points out1 that there have been several accounts of Thomas à Becket's career, which assign it, in effect, to different genres. In some medieval versions, it belongs to hagiography; in an Icelandic poem Becket appears as a saga hero; while Dom David Knowles treats his life as a tragedy, that of his relations to Henry II. Knowles, according to MacIntyre, is clearly right. There are several odd features of this account: for instance, that although fact is supposed to be prior to fiction, these possibilities for interpreting the life of an actual man are drawn from fiction. Moreover, it is unlikely that those possibilities would all have been understood by Becket. But the most basic point is that even if Becket's life were uniquely well described in Knowles' terms, that could tell us nothing at all about how he lived it. He certainly did not live it by asking, when considering what to do, how to carry on the tale of one locked in a tragic relation to his king. Perhaps someone, not Becket, could do that—but his story would not be a tragedy, but rather (in the way described in another work referred to by MacIntyre, The 18th Brumaire of Louis Napoleon) would have already moved on to farce.

People can of course live their lives by reference to fiction, and there are many more, and less, interesting ways of coming to grief in that project than Emma Bovary's or Don Quixote's. But that could not provide the way of living a life. Nor is it merely that we cannot impose narrative coherence upon our lives by consciously referring to existing fictions: the point goes much farther than that. Although at first it seemed surprising that MacIntyre should appeal, in Becket's case, to narrative paradigms drawn from fiction, when he has insisted that fact is prior to fiction, this appeal in fact reveals what his idea is. He indeed believes that the unity of an actual life is like the unity of a fictional life—that is what makes narrative the grounding conception for both of them. The priority of fact over fiction consists only in this: that the unity is found first in life, and is carried over from life to the construction of fiction. But this conception must be wrong: there is a deep disanalogy between the situation of a person living his life and any fictional character. It lies in the ingenuous but significant point that fictional characters are not living at all. MacIntyre engagingly says that fictional characters share with us the limitation that they do not know the future. It is of course true that they do not (standardly) know the future, for the boring reason that that is how they are represented. But there is a more basic reason why they do not know their future, and this they do not share with us at all: that they have no future, that all of them is already there. When the reader starts, and in that sense when they start, they are already finished.1 Moreover, the fact that we are conscious of this, and indeed conscious of how the characters finish, presents no obstacle to our reading. Our experience of them, and in works of any real interest, our fullest experience of them, is of characters who will finish in that way.

The life of a fictional character is necessarily something that our lives are not, a given whole. However coherent or incoherent in everyday terms their lives may be represented as being, they have a special unity that no real life can have, that the end of them is present at their beginning. This peculiar unity of their lives cannot help us in trying to find coherence in our own. It is essential to fictional lives that their wholeness is always already there, and essential to ours that it is not.

It is tempting to put this by saying that for a fictional character there are no unrealized possibilities.1 There is a boring falsehood in that, which mirrors the boring truth that in the fiction a character is represented as having a future. In the sense that Pip, when he first runs into Magwitch, is a boy to whom many things are going to happen, he is a boy to whom many different things may happen, and equally at the end he is a man to whom other things might have happened. But in the fundamental sense in which the Pip of Great Expectations is a given whole, there is nothing else (it is tempting to say) that he might have been except what he is. Yet this, again, seems not quite right. Even though it may be an unhelpful kind of speculation, there is sometimes room for the idea of how the fiction might have gone otherwise; and in the case of serials, and a succession of works presenting the same character, there is practical room for it. To that extent, Pip, or (less interestingly) Sherlock Holmes, might have been rather different.

One may say, of course, that in that case there would have been no such character as Pip, and Dickens would have created a different character. (It is what Leibniz said about God's creation of an actual person: it is significant that, for Leibniz, actual people are a species of possible people, and the end of a possible person's life (as it is called) is, in the real order of things, timelessly co‐existent with its beginning.) It is not clear to me how helpful it is to insist on that, but whether we say it or not, it is clear that this dimension of literary possibility gives us no help at all in developing any notion of coherence that can be applied to our own lives. We shall say that Pip might have gone a different way only if the different way would have been in Pip's style; and our notion of Pip's style can be derived, again, only from the Pip we have got, the completed whole, and that is exactly what, in our own case, we do not have.

In some cases more than one style may be associated with one and the same character, appearing in different works. The Odysseus of Sophocles' Ajax displays characteristics different from those of Odysseus in Sophocles' Philoctetes. They are, in a sense, the same character, because they are both, in ways that the audience understood, Homer's Odysseus, and that is an important fact about each of them. But they are not one character with a variable style, because it makes no sense to add these works together, even to the extent (and that is problematical) that it makes sense to see the Oedipus at Colonus as presenting us with the old age of the Oedipus who was Tyrannus. This shows again the primacy in literature of the given whole: the difference in style of those two figures of Odysseus is given to us only by the given whole that is each of them.

There is nothing in any of this for understanding our life as we live it. There may seem to be various possibilities, but none of them is helpful. To consult the style of an existing literary character is an eccentric possibility, but it is not a paradigm of living one's life. To consult, rather, a style which will be, finally, one's own personal style, in just the sense that a fictional character may have a style, is impossible, precisely because, unlike the lives of those characters, my life is not a given whole. All that is left is that one should consult what (as it seems to one) has so far emerged of one's final style. This may be a recognizable description of some reflective thought, but it would have to be a delusion to suppose that it embodied the basic reflection involved in living. For if any such style has begun to emerge, then it has done so in one's actions and in one's first‐order considerations of what to do; it has not (and necessarily not) emerged up to now in reflections on the style in which one does those things. If that same style is to continue to emerge, then it might be expected to do so in the same way, in further actions and first‐order considerations, rather than be altered by this reflection into the clearly different style that consists principally of thinking about one's style.

The given whole of a fictional character does present us, I have suggested, with a peculiar unity, which consists in its end being there with its beginning; just for this reason, that unity is not available to us. However, in more everyday terms that may indeed invite comparison with reality, the fictional life need not be to any notable degree coherent. A life narrated in fiction may be even more picaresque, episodic, disjointed and arbitrary than an actual life, and, unlike an actual life, there is no chance of looking for further material to make more sense of it. Even if we lack any overall picture that is notably coherent, we may of course think that we understand to some extent how the character is supposed to have got from one stage to the next, and that we have some insight into what it would be to live such a life. But those insights do not primarily rely on ideas drawn from completed narrations. They demand only a conception of what it is to be at a certain point and to think, if only on a small scale, about how to move on from there. That is a conception of living a life—indeed, it is theconception of living a life; but it does not specially depend on conceptions of narration, whether fictional or not.

What this has shown, if it is right, is that the idea of a completed, unified, or coherent narration is of no help in leading a life. The idea of living as a quest for narrative is baseless. Yet it is no doubt true that in understanding people's lives—above all, other people's lives—narratives that give them a certain direction or meaning are very important: the story of someone's life might for instance be the narrative of a quest. But this invites, finally, the sceptical problem. If a particular and significant narrative structure can plausibly be applied to a life retrospectively and from outside, and yet the person whose life it was could not, typically, have lived it with the aim of its embodying that structure, where does the plausibility, the fit, come from? It seems like magic.

There are only two possible lines of answer to this problem. One is that, although the agent could not consciously direct his life to live out that narrative, nevertheless the considerations that appealed to him at various stages of his life and which moved him to act were appropriate to such a narrative: they were drawn from a repertoire of stories defining recognizable, perhaps acceptable, lives. The other line is that the narrative is a fiction, the power and plausibility of which lies in such things as its resemblance to familiar stories, its capacity to help us make sense of some larger set of events, or its reassurances that there can be some immanent meaning, if no more, in the totality of our variously improvised moves from one set of circumstances to the next. The two lines cannot consistently both be pressed all the way together, but they can be mixed, even in the case of one life.

The types of scepticism invited by these two lines are rather different. The first is primarily a scepticism about the considerations that help to shape one's life, and the extent to which they are part of a conventional repertoire of which one is not fully conscious. To the extent that they are, a narration of one's life as unified in these terms may well be appropriate, but its appropriateness might come as an unpleasant surprise to one. The second form of scepticism leaves the considerations that shape one's life in the disorderly state that is natural to them, and attacks the supposed coherence and unity that narration can give to people's lives.

There is surely a good deal of room for both forms of scepticism, but less room, as I have already said, for both at once. My own hope is that the second form is more important than the first. We have a much greater interest, as it seems to me, in living a life that is our own and in having an adequate grasp, at the time, of the considerations that at various stages direct it, than we do in the ambition that it should genuinely present a well shaped tale to potential narrators of it.1

NOTES

• 1 MacIntyre 1981: ch. 15: quotation at p. 197 from Hardy 1968.
• 2 MacIntyre 1981: 197.
• 3 MacIntyre 1981: 194.
• 4 MacIntyre 1981: 201.
• 5 MacIntyre 1981: 203.
• 6 Wollheim 1984 is built round this claim, in the particular form of emphasizing the living of a life.
• 7MacIntyre is wrong to reject the discussion of basic actions, by Davidson and many others, as misguided given the importance of narrative. My point is that, on the contrary, the latter can be got going only given (something like) the former. It is true, however, that some work in basic action theory has neglected the many ways in which complex actions go beyond conjunctions of basic actions.
• 8 MacIntyre 1981: 198.
• 9Genuine serials (as opposed to serializations), such as TV soap operas, are a partial exception, and it is a commonplace that they can generate a startling ambivalence in some viewers, who care passionately about what is to happen to a character and show it sometimes by trying to persuade the writers to take the story in one direction rather than another.
• 10Alexander Nehamas presents a very strong version of this claim in his application of related ideas to the thought of Nietzsche: see Nehamas 1985, esp. ch. 6.
• 11[This is the unrevised text of a talk which Bernard Williams gave in the late 1980s, in Berkeley, to the editorial board of the journal Representations and which he never prepared for publication. We are grateful to David Heyd, who obtained a copy of the talk, for drawing it to our attention.—Adrian Moore and Patricia Williams (Bernard Williams' literary executors).]