As the culmination of a distinguished philosophical career, Bernard Williams's latest production would conventionally be expected to receive shining reviews. This convention is happily redundant. Truth and Truthfulness is subtle, agile, diligent in its treatment of detail, yet always with an eye on the "big questions". Williams's manner is assured, worldly and, detractors might add, somewhat haughty. His long experience in the business allows him to dismiss certain questions with a wave of the hand, while pursuing others in a spirit of pure showmanship.
Although densely argued, the book is never dull or nerdy. It is suffused by a sly Oxonian humour and a keen feeling for pleasures of philosophical argument. The spirits of Hume and Diderot preside over it. Yet this playfulness does not detract from its underlying seriousness of purpose: this is a defence of the value of truth against those modern sceptics who deny its existence.
One of the chief accomplishments of Truth and Truthfulness is its seamless synthesis of two very different styles of philosophical argument. Williams acknowledges these dual influences indirectly in his choice of prefatory quotations. The first, taken from Proust, runs: "I have always had a high regard for those who defend grammar and logic. One realises 50 years later that they have warded off great dangers." This is a nod to Williams's mentors, the Oxford philosophers of the Fifties and Sixties, whose austere schooling is still very evident in the more analytic sections of this book, as well as in the general care with which Williams handles language.
But, in other respects, Truth and Truthfulness marks a departure from the analytic tradition. Williams acknowledges this in the second of the two quotations, taken from Nietzsche: "Lack of an historical sense is the hereditary defect of philosophers. . . . So what is needed from now on is historical philosophising, and with it the virtues of modesty." Analytic philosophy is implicitly accused of being residually Platonic, conceiving its task to be the elucidation of an unchanging set of linguistic essences. But, argues Williams, such essences are fewer than is often supposed. Some concepts, such as that of truth itself, are timeless because their possession is a necessary condition of any use of language. These can be investigated using the methods of analytic philosophy. But other concepts associated with truth, such as those Williams identifies as Accuracy and Sincerity, are subject to historical change. To make sense of them, philosophy must incorporate history. Like Alasdair Macintyre, a philosopher with whom he has little else in common, Williams has striven mightily to overcome analytic philosophy's traditional historical blindness.
The immediate, polemical purpose of Truth and Truthfulness is to try to refute the arguments of a group of people whom Williams calls "the deniers". The deniers - they are not identified by name, although Foucault would be an obvious example - are those who deny the existence of truth, ironically or otherwise, while at the same time manifesting a passion for truthfulness, a "pervasive suspiciousness, a readiness against being fooled, an eagerness to see through appearances to the real structures and motives that lie behind them". As Williams points out, this is an unstable combination. If there is no such thing as the truth, then "what is the passion for truthfulness a passion for"? If the unmasking of power is not guided by a commitment to the truth, then it is itself nothing more than an assertion of power, and can in turn be unmasked as such. It is particularly unwise, adds Williams, for disadvantaged groups to try to reduce intellectual conflicts to conflicts of power, for power is precisely the commodity they lack. "It is always a mistake for a minority . . . to reduce things to the bottom line, for on the bottom line they are simply a minority."
Yet if Williams is critical of the deniers, neither is he particularly impressed by what he calls "the party of common sense". He means those (mainly analytic) philosophers who attempt to refute the arguments of the deniers by pointing to the indispensable role of truth in defining concepts such as belief and assertion. Such arguments are all very well as far as they go, but they do not even address, let alone answer, the deniers' many legitimate anxieties. The point at issue is not the existence of a property called "truth", even if some of the deniers confusingly present it in those terms; it is rather the value of truthfulness. The real question is not logical but ethical. The task Williams sets himself, then, is that of demonstrating the intrinsic value of truthfulness - thus refuting the deniers' claim that it can be understood solely in instrumental terms - without drawing on the assumption that this value can be derived from the logical role of truth.
Williams's approach to this problem is self-consciously naturalistic. The intrinsic value of truthfulness is not self-evident; it is not "sent down by the gods". One is obliged to show how it might have arisen from a hypothetical earlier state of society in which truthfulness was valued merely as an instrument for the attainment of more basic human goods. Williams is here following the Enlightenment tradition of justifying the institutions of civilisation by tracing their emergence from a "state of nature". Yet such narratives (or "genealogies", as Williams, following Nietzsche, calls them) are double-edged tools. They can be used not only to justify, but also - as Nietzsche himself demonstrated in his Genealogy of Morals - to debunk.
If the higher has arisen in gradual degrees from the lower, then perhaps it is nothing more than the lower in disguise. Whether a genealogy is "vindicatory" or not depends, argues Williams, on whether we can accept it as true without thereby being forced to abandon the beliefs and practices it seeks to explain. Hume's explanation of the origins of justice is not intended to undermine our respect for justice; whereas no one who accepts Nietzsche's explanation of the origins of Christianity can go on being a Christian. Williams describes his own genealogy of truthfulness as vindicatory. The reader may not be so sure.
Williams breaks truthfulness down into two separate virtues, which he names Sincerity and Accuracy. Sincerity is the virtue of not misleading people with respect to what one believes; Accuracy is the virtue of making sure, to the best of one's abilities, that what one believes is true. Williams then employs his genealogy to demonstrate that these two virtues have an intrinsic, and not merely an instrumental, value. His starting point is a "state of nature" scenario in which Sincerity and Accuracy are valued simply as a means of pooling information necessary for the community's survival. Yet such a crudely instrumental justification is not self-sustaining; knowing that the community has a general interest in Sincerity does not give me - in a particular case where I would be better off lying - a reason to be sincere. This presents us with the following paradox: Sincerity and Accuracy, in order to fulfil their purely instrumental role, must be believed by members of the community to have more than a purely instrumental role. They must be believed to have an intrinsic value. "We need people to have dispositions of Sincerity, and this implies that people treat Sincerity as having an intrinsic value."
But here an obvious doubt arises. Has Williams really demonstrated anything more than the utility of a general belief in the intrinsic value of truthfulness? Surely this is simply an instance of what Nietzsche called "the useful lie". Such a genealogy is not, as Williams puts it, "stable under reflection". Someone who came to understand that the belief in the intrinsic value of truthfulness had arisen in the foregoing manner would no longer be able to maintain it. The genealogy, at this point, is not vindicatory but debunking.
Williams recognises the problem. To save his genealogy, he must demonstrate not merely that it is necessary for basic human purposes that people treat truthfulness as an intrinsic good; he must also show how it is possible for them coherently to treat it as an intrinsic good. For this, truthfulness must be shown to have "an inner structure in terms of which it can be related to other goods". It is at this point that Williams's argument descends from its abstract heights to engage with history. The goods in question turn out to include trust, honour and justice, all of them historically and geographically relative. Certain people "deserve the truth", and one may feel disgraced "in one's own eyes, and in the eyes of people whom one respects and who one hopes will respect oneself", if one does not give it to them. Williams is here gesturing to the ethics of honour, which he elaborated and defended in his earlier work, Shame and Necessity. His idea seems to be that truthfulness can be said to have an intrinsic value so long as it is central to the historically prevailing structures of mutual respect. Crudely put: if lying disgraces me in the eyes - in the imagined eyes - of those whom I respect and whose respect I desire, then truthfulness can be said to have an intrinsic, and not merely an instrumental, value for me.
The trouble with this kind of explanation is that it yields only a very limited notion of intrinsic worth. What if the prevailing structures of mutual respect no longer give any encouragement to truthfulness? In this case, it appears, the individual has at best instrumental reasons to tell the truth. Williams seems to accept this consequence: "One might say that Sincerity, at least in good surroundings, is a good thing in its own right." But it is, surely, precisely in "bad surroundings" that we most admire an unconditional attachment to truthfulness. The point can be illustrated by means of an example. What reasons, according to Williams, might the defendants in the Moscow show trials have had for telling the truth?
Those whom they respected and whose respect they hoped to win were on the other side of the bench, encouraging them to lie. And they had nothing to gain by truthfulness. One cannot, in view of the methods used, condemn them for having succumbed, but one would surely admire them more for having held out. How can Williams explain this, given his premises?
There is much more than this in Truth and Truthfulness; there are virtuoso discussions of ancient Greek historiography, of Rousseau, Diderot and of Nietzsche. The book makes few concessions to the reader; it is densely argued and poorly signposted. But for those willing to make the effort, it offers the rare pleasure of a first-rate philosophical mind at work.