There is a paradox in the mantle of the enfant terrible. One wearing it is expected to cut up rough, to kick over the traces, to create a stir, épater les bourgeois. In being contrary, he thus obediently fills his expected role – but only if he is acceptably contrary. If we stand back a bit from the fray, we might sometimes find that the dust raised has made far less of a storm than expected. For a long time, in my view, this was so with Williams’s philosophy. Williams was in the vanguard of opposition to many things that needed opposing, especially in moral philosophy. He criticised moral philosophy’s predilection for producing moral theories, its preoccupation with what, in one essay here, he calls ‘the situation of the rational agent intending to change the world’, ‘Christian’ readings of Aristotle’s ethics, philosophy’s neglect of the ways contingency shapes our lives, and a good deal else. Much of Williams’s earlier work in this vein was genuinely telling, and invigorating. But even so, the way Williams invited us to go on from what he opposed often struck me, at least, as less creative than the initial opposition. The significant local disturbance he created tended to leave in place the restrictive tyranny of the way a certain issue was conceived. (I spell out one example below.)

But these essays show that Williams came more and more into his own in later years. The discovery of Nietzsche helped him here. Nietzsche provided him with a model for a genuinely distinctive philosophical voice, one with an historical accent almost entirely absent from British philosophy. The historical essays are in my view what is best in these collections. (In calling the contents of The Sense of the Past ‘essays’, the editor, Myles Burnyeat, is making the point that these are not philosophical ‘papers’ or ‘articles’ in the usual mould, but, as Williams himself puts it, ‘something that a grown-up, concerned, intelligent person might say to another about these subjects’.) But the historical accent also belongs to many of the essays that are not focused on historical figures. In Williams’s view, the specific form in which philosophical problems present themselves – to us or to past philosophers – is not a local dress that needs removing to reveal the real, timeless problem underneath, but essential to what the problem is, and to how it needs exploring. Williams writes: ‘Nothing that has a history can be defined, as Nietzsche rightly said, and our virtues and our values certainly have a history’. The philosophical task of understanding ourselves, then, by understanding ‘our virtues and our values’ and also much else, has to be pursued along an historical dimension. One theme of various essays, explicitly or implicitly, is the limitations of an ahistorical Anglophone philosophy too preoccupied with the analysis of concepts.

Such limitations also show when this philosophy turns its attention on philosophy’s own history. Plato might then be read – or rather misread – as if what he wrote had just appeared in the most recent issue of Mind. Williams thought we needed constant reminding of how different, in many ways, were the problems and preoccupations of our philosophical predecessors. Not only would that help us be true to them, but our thinking about what concerns us here and now would also be most stimulated, enriched and deepened by contact with the past if we let its difference make the familiar strange to us, and the strange familiar.

In addition to Nietzsche, R.G. Collingwood, the neglected English idealist philosopher of the first half of the twentieth century, helped Williams to his appreciation of the historical dimensions of philosophy. (Williams rightly observes that, with the partial exception of Isaiah Berlin, Collingwood was the only twentieth-century English philosopher for whom philosophy was essentially historical. He is one of only two of Williams’s Oxbridge predecessors to earn an essay of his own here. The other is the late nineteenth-century Utilitarian Henry Sidgwick.) This was a gradual development. Williams’s earlier books – Morality: An Introduction to Ethics (1972) and Ethics and the Limits of Philosophy (1985), for instance – certainly show a social and psychological aware ness lacking from most other English-language moral philosophy of the period. But even so, their attacks on ‘moral theory’ have a tendency to lapse back into a ‘timeless’ perspective of their own, a roughly Humean sort of empiricism that is as ahistorical as what it opposes. Shame and Necessity (1993) is perhaps the first book in which Williams’s philosophising is self-consciously historicised.

If Williams’s philosophical thinking continued to ripen, it is as it should be that he still leaves us plenty to quarrel with. And despite the trajectory of Williams’s development, these things for me remain much what they were. Here are two such issues, the first of which I have already mentioned. This is Williams’s view about philosophy’s failure – almost its temperamental incapacity – to appreciate the depth to which luck and contingency penetrate our human lives. He says, for instance: ‘The very plain fact that everything that an agent most cares about typically comes from, and can be ruined by, uncontrollable necessity and chance is no part of [most modern moral philosophy’s] concern.’ In ‘The Legacy of Greek Philosophy’, Williams contrasts Plato’s and Aristotle’s outlook with that of Greek tragedy: ‘Socrates’ saying, that the good man cannot be harmed … expresses an ideal of rational self-sufficiency, of freedom from the damage of contingency’. This ideal expresses the aim ‘that what is of highest value, what matters most, should be entirely under the self’s control’. Greek tragedy, by contrast, expresses ‘a sense of exposure to fortune’. Its ‘sense … that what is great is fragile and that what is necessary may be destructive … has disappeared from the ethics of the philosophers, and per-haps altogether from their minds’. This may be true enough of Aristotle, but it certainly is not true of Plato. Williams may also be right that much modern moral philosophy misses the significance of luck and contingency, but it is a different matter whether he is himself right about the implications of that significance. He seems to think that any ‘ideals’ of a roughly Platonic kind must be self-delusory rather than, as Plato himself recognised, inevitably insecure. The closest Williams comes to acknowledging this is when he says that Plato’s ‘dialogues preserve a sense of urgency and of the social and psychological insecurity of the ethical’. But that acknowledgment is undercut by this remark: ‘Plato is constantly aware of the forces – of desire, of aesthetic seduction, of political exploitation – against which his ideals are a reaction.’ The problem lies in that last word ‘reaction’. It makes those ‘ideals’ sound as if they are simply psychological defences against our fear of the brutally destructive forces of contingency – a kind of whistling in the dark. That diagnosis is familiar enough – we hear it, for example, in Freud – but many will find it inadequate to Plato, as to some of his philosophical heirs.

The second issue is closely linked to this first one; perhaps it is just a way of pressing the first issue further. In ‘Pagan Justice and Christian Love’, Williams highlights Socrates’ conviction that the good man cannot be harmed, and says that ‘this aspect of Socrates’ conclusion can make sense only if it is completed, as it was by Plato and by Christianity, with a belief that this is not the only or the most important life we lead’ (i.e. because there is a more important ‘afterlife’). While the Platonic Socrates did express the conviction that the good man cannot be harmed, his emphasis was more often on the conviction that nothing worse can befall someone than to do evil, that if forced to choose he (Socrates) would always choose to suffer evil rather than to do it, and that, come what may, he would never knowingly commit evil. These are the Socratic convictions at the centre of the Gorgias, for example, and they astonish not only the figures of Polus and Callicles in that dialogue, but Plato himself. They would not do so if Plato thought them simply derived from belief in an afterlife. What astonished Plato – but also so moved him that its influence is evident throughout his own philosophical life – was that Socrates’ whole life was a living testament to those convictions – to the absolute importance of living one’s life justly and decently, simply because of what it means to do that. Attention is deflected from that more important issue by focus on the claim that the good man cannot be harmed. (That claim may be a hyperbolic extension of those other convictions; perhaps it registers the faith that one who lives justly, come what may, will at least always be able to avoid despair. I should say that I do not myself believe that.)

Williams also thinks it straightforward that Plato had a ‘belief’ in an afterlife. But any discussion of this, as of the existence ‘in another realm’ of those famous Platonic ‘Forms’, had better take account of how often Plato signals that in talking about these things he is going to speak the language of mythos (roughly ‘story’) rather than logos (roughly ‘a rational account’). In different ways, Plato provides such signs in the Republic, the Gorgias and the Phaedrus; and they are surely meant to make us beware of taking what is said there as expressing any straightforward ‘beliefs’. (Exactly how we are meant to take it remains a crucial question about how to read Plato.) In my view, a reductionist tendency in Williams’s thinking sells Plato short here, as it also sometimes limits Williams’s discussions elsewhere.

Many will not agree with this judgment, but even those who do can recognise and be grateful for the genuine and substantial merits of these essays, for their range, fertility and vigour of thought, and for the real illumination they offer. There is ample reward here for philosophers and interested non-philosophers alike. In her preface to Philosophy as a Humanistic Discipline, Williams’s widow, Patricia Williams, says that ‘Bernard had no faith in his, or any philosopher’s, ability to predict whose work would be of any lasting interest to their successors. That was for the future to decide.’ Whether Williams’s oeuvre constitutes a mosaic that will prove greater than the sum of its variously striking parts is a question which, taking my cue from Williams himself, I will not try to answer.