. 4
( 38 .)


this volume. Long contextualizes this monograph, contrasting it with work
that Williams produced both before and after its publication on ancient
moral psychology and ethics. He brings out very clearly that Shame and
Necessity, for all its meticulous philological and literary scholarship, was
clearly intended to further a philosophical agenda and that this explains
some of the readings of ancient philosophy, literature, and history that Long
and others have found to be lacking in Williams™ usual nuance and ¬nesse.

49 “Kalokagathia” is a term resistant to translation, meaning something akin to a combination
of moral nobility and beauty.
50 Williams (1993), p. 4.

As Long points out, the most general programmatic statement of that
which Williams hoped to “recover” from ancient ethical thought occurs in
his earliest comprehensive treatment of it published in 1981:

It has, and needs, no God. . . . It takes as central and primary questions of
character, and of how moral considerations are grounded in human nature: it
asks what life it is rational for the individual to live. It makes no use of a blank
categorical imperative. In fact “ though we have used the word ˜moral™ quite
often for the sake of convenience “ this system of ideas basically lacks the
concept of morality altogether, in the sense of a class of reasons or demands
which are vitally different from other kinds of reason or demand . . . The
questions of how one™s relations to others are to be regulated, both in the
context of society and more privately, are not detached from questions about
the kind of life it is worth living, and of what is worth caring for.51

In these most general terms ancient ethical thought receives, as it were,
a blanket endorsement. However, by the time of Shame and Necessity the
in¬‚uence of Nietzsche on Williams™ views had made itself felt. In a paper,
to which I have already referred, that usefully supplements Long™s contri-
bution to this volume, Raymond Geuss analyzes at length the precise form
of this in¬‚uence.52 He focuses in particular on Nietzsche™s claim that the
“philosophical” elevation of Plato™s account of human psychology over that
of Thucydides was tacitly moralistic. Nietzsche argued that Thucydides
is a more reliable guide to the understanding of other people because of
an uncompromising realism about other people™s motivations that is not
tainted by a prior commitment to understanding psychology in moralized
terms. This is taken up, and endorsed, by Williams, who writes in Shame
and Necessity that:

Thucydides™ conception of an intelligible and typically human 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.53

This endorsement of Thucydides as opposed to Plato formed part of
Williams™ commitment to a general philosophical naturalism applied to
the speci¬c case of explanations of the ethical.54 Philosophical naturalism
is the claim that one need only ever cite natural properties and facts in one™s
explanations, a plausible but very weak thesis that does little other than rule

51 Williams (1981c), p. 251
52 Geuss (2005).
53 Williams (1993), pp. 161“162.
54 Williams (1995b).
16 Alan Thomas

out any appeal to the supernatural. To put the idea to work one needs more
than this; however, it is unclear what more is available. An inventory of
natural properties can only look to the sciences, but if this account is to be
plausible the social sciences will have to be included alongside the natural
sciences and they will, as Williams himself emphasizes, have to re¬‚ect the
fact that living in a convention-governed culture is itself natural for the kind
of animal that we are.55 The problem does not stop there: do we take the
inventory of the natural properties from natural and social science as they
are now? Even a modest fallibilism makes that look hubristic. Adverting to
an ideally completed scienti¬c world picture, however, seems to lose the
constraint that naturalism was supposed to place on actual explanations in
the here and now.
Williams suggested that at least in the case of the ethical one could come
up with a plausible account of ethical naturalism that avoided this general
dilemma. In a “minimalist moral psychology” inspired by Nietzsche, one
seeks to explain distinctively moral ideas, those that seem to have placed
themselves beyond any explanation (not simply beyond reductive explana-
tion) by appealing to other ethical motivations that one is committed to
anyway, that are continuous with (but not restricted to) concepts used in
the explanation of the rest of human psychology.56 However, to avoid circu-
larity, one cannot have begun one™s investigation by tacitly moralizing basic
psychological categories: hence the preference for Thucydides over Plato.
Thus Nietzsche™s in¬‚uence on the arguments of Shame and Necessity
leads, as Long points out, to a “strongly critical posture . . . in relation to the
moral psychology of the Greek philosophers, especially Plato.” However,
it is equally true that in his search for a Thucydidean ethical naturalism that
does not moralize basic psychological categories, Williams supplemented
this account with a “sympathetic engagement with the implicit ethics and
psychology of Homer and the Greek tragedians.”57 In the context of a
discussion of Thucydides™ views alongside those of Sophocles, Williams
identi¬ed that which he found attractive in their ˜joint™ outlook:

Each of them represents human beings as dealing sensibly, foolishly, some-
times catastrophically, sometimes nobly, with a world that is only partially
intelligible to human agency, and in itself is not necessarily well adjusted to
ethical aspirations.58

55 Williams (1995b), p. 67.
56 Williams (1995b), pp. 68“74.
57 Long, this volume.
58 Williams (1993), p. 164.

As Long points out, Williams does not take ancient tragedy as in some
way expounding philosophical theses, but as treating universal human expe-
riences in a way that is free from philosophical distortion. Homer and the
Greek tragedians avoid the “moralisation of basic psychological categories,”
particularly those categories and concepts drawn on in the explanation of
action. More speci¬cally, the target is any dichotomy between reason as
aligned with the good and mere “desire” and hence the elevation of a virtue
of rational self-control to central importance in moral psychology, with a
corresponding downgrading of individual moral character and its contin-
gencies. In his particular treatment of ancient tragedy, Williams also draws
attention to the representation of the experience of practical necessities
understood as distinct from the idea of moral obligation, and a comple-
mentary emphasis on shame, where the latter is freed from the progres-
sivist assumption that it represents a less developed moral emotion than
guilt. Throughout Shame and Necessity, Williams fought a running battle
with those progressivist interpretations of ancient philosophy and litera-
ture, primarily exempli¬ed by the work of Arthur Adkins and Bruno Snell,
that fails to ¬nd the characteristic concepts and concerns of the modern
morality system in these writings and for that very reason regards them as
faulty, lacking in a developed and mature form of ethical consciousness.59
Long challenges the argument strategy of Shame and Necessity in two
main respects: although not defending the “˜progressivist” theses of Snell
and Adkins in their entirety, he does show that understanding the moti-
vations for their respective projects casts them in a more charitable light
than Williams™ highly critical account of their work. Second, he questions
the details of Williams™ Nietzschean project of isolating an ethical psychol-
ogy free of moralizing prejudice in Thucydides, Homer, and the tragedi-
ans, from the moralized psychology of Plato and Aristotle. He does so by
pointing out that Plato™s psychology, in particular, is not that far removed
from that of Homer. Where Williams sought a radical difference of kind,
Long detects an explicable difference of degree. Overall, Williams™ “anti-
Kantian agenda” can, it seems to Long, “be more distracting than illumi-
nating.” This is brought out most clearly in the treatment of Plato. In his
major publication before Shame and Necessity, Williams™ treatment of Plato™s
ethics and moral psychology is largely approving; however, by the time of
Williams™ later treatment of the subject, the critique of the morality system
had been presented in the intervening Ethics and the Limits of Philosophy. That

59 Williams objects to the accounts of ancient ethics and psychology presented in Snell (1953)
and Adkins (1960).
18 Alan Thomas

signi¬cantly colours Williams™ treatment of Plato in Shame and Necessity.
This seems, to Long, a retrograde step: Plato™s moral psychology is assim-
ilated to Kant™s in a misleading way, structured around the claim that both
take one™s rational nature to be that which aligns one with the good, inde-
pendent of the particularities of one™s character, making the moral self “char-
acterless.”60 Long questions whether this is a fair representation of Plato™s
account of moral agency; however, he also notes that one of Williams™ very
last publications on ancient philosophy reverses this negative evaluation of
Plato™s overall achievement and acknowledges that in Plato, just as much
as in his own work, there is an emphasis on the “social and psychological
insecurity of the ethical.”61
Is it possible to draw any general moral from the disparate treatment of
Williams™ critique of the morality system in Skorupski™s, Louden™s, Long™s,
and Stocker™s contributions to this volume? Perhaps that a sceptical view
gains in plausibility in so far as it becomes more local and focused: clearly
identifying Williams™ target makes his critique more plausible. The ¬rst
point to be noted is that Williams did not deny that there are clearly many
aspects of our modern moral ideas that do survive re¬‚ective scrutiny; one of
the aims of his last monograph, Truth and Truthfulness, was to demonstrate
that genealogy can supply vindicatory explanations at least of some key
epistemic values and correlative virtues. The critique of the morality system,
then, has to identify a restricted target and to be interpreted as using some of
our ethical (and philosophical) ideas to criticize others. Our commitment to
truthfulness undermines a set of ideas grounded partly on an ethical ideal of
purity, partly on mistaken philosophical views about agency, voluntariness,
and emotion and partly on our continuing use of some practically useful
¬ctions that have, as is characteristic of key concepts of the morality system,
done a very good job of concealing their tracks (as Craig puts it later in this
Second, Williams seems to have believed that the morality system was
largely represented in the re¬‚ective and theoretical accounts of the ethical
produced by moral philosophers and other intellectuals that they in¬‚uence,
such as Snell and Adkins, and that this re¬‚ective account genuinely repre-
sented part of the common sense moral outlook of nonphilosophers, too.
However, if Kant has theorized part of the outlook of all of us, it is life
itself and the full range of ethical experience available within it that falsi¬es
that outlook, combined with drawing attention to a less dominant tradition

60 Williams (1993), p. 159.
61 Long, this volume.

of thinking about ethical psychology that is exempli¬ed by the ¬gures of
Williams™ counternarrative, and that is also, implicitly, part of the outlook
of all of us. Third, it is undoubtedly true that Williams™ target is simply
too large and diffuse for his critique consistently to hit the mark, but if one
identi¬es it more narrowly as that which Taylor calls an “ethics of obligatory
action,” particularly in the local context of the Oxford moral philosophy of
Ross, Prichard, and Hare to which Williams critically responded, then we
have both a clear target and a powerfully focused critique of the “peculiar
institution” of the morality system. This may make Williams™ critique seem
too much a matter of theory; too much a matter of a philosopher correcting
the mistakes of other philosophers. However, we would not be tempted to
identify in the views of Ross, Prichard, and Hare an accurate representation
of our ordinary ethical outlook if there was not something in that outlook
itself that made such a theoretical misrepresentation attractive to us.
Throughout his later work Williams cited the in¬‚uence of Nietzsche;
as I have noted, this is both a matter of methodological self-consciousness
about that which philosophy ought to become and in Williams™ last book
also an attempt to apply this method to a central question for Nietzsche, that
of the value of truth. Truth and Truthfulness addressed itself to the question of
the value of truth, or, more accurately, the value of those virtues that attach
to truthful representations, their production and dissemination. Williams
also explicitly adopted Nietzsche™s method of genealogy, partly in¬‚uenced
by what he took to be the highly successful example of a synthetic account
of the concept of knowledge in Edward Craig™s Knowledge and the State of
In his later work, Williams became highly self-conscious both about
the practice of philosophy and its methods.63 Concerned to defend the
integrity of philosophy as a humanistic discipline and to resist the scientism
of a range of contemporary approaches to philosophy, Williams also came to
see philosophy as intimately connected with another humanistic discipline,
history. If philosophy is concerned to retain its purity, or its special standing
as a wholly a priori form of enquiry, it will yield results too indeterminate
to help with the central task of ˜making sense of what we are trying to do in
our intellectual activities™.64 Certainly in the case of modern ethical ideas,
extending that term broadly to cover the norms surrounding truth telling,
there is an essential role for history in explaining why our actual practices

62 Craig (1990), discussed in Williams (2006b).
63 For example, in Williams (2006a) and (2006b).
64 Williams (2006a), p. 186.
20 Alan Thomas

take the shape that they do. In the ¬nal paper of this volume, Craig examines
Williams™ use of his later “synthetic” method and of the various things
that can be understood as “genealogical” projects. Craig concedes that his
use of the term “genealogy” is very broad, ranging from the ˜vindicatory
explanation™ of a particular belief to developmental narratives that, either
essentially or accidentally, subvert or vindicate a concept, belief, or practice.
Craig considers, in particular, how a highly abstract and indeterminate
set of apparently historical claims can be put to philosophical purposes,
assessing the question by examining a broad range of genealogical projects
in philosophy from Hume, through Nietzsche, to Foucault, Nozick, and
Williams. In the course of a perceptive discussion, Craig notes that genealo-
gies typically work by uncovering a function where it did not seem to exist.
He also argues that appeals to “states of nature,” a species of the wider
genus of genealogical arguments, can be defended when understood not
as speculative prehistory but essentially as a rhetorical device functioning
as part of an argument about one™s current beliefs, concepts, and practices.
What grounds them and gives them their justi¬catory power are that which
Craig calls “factual claims about human nature.” The degree to which these
claims can withstand scrutiny depends, Craig argues, on how far we under-
stand this form of genealogy, one that utilizes the state of nature vocabulary,
as explanatory. His genealogy of the concept of knowledge, like Williams™
genealogy of the virtues of truthfulness, accuracy, and sincerity, aims to be
both explanatory and vindicatory. There is no appeal, in any of this, to imag-
inary histories: the distinction is rather that state of nature theorists seem
to begin from very general and hence less determinate claims about human
beings and genealogists draw on a more speci¬c set of claims about people
situated in more particularized and determinate historical circumstances.
In the ¬nal section of his chapter, Craig turns to the question of how
genealogies do not simply identify function where there seemed to be none,
but also to account for that fact itself “ the capacity of some functionally
justi¬ed concept, belief, or practice to conceal its own origins. Conceding


. 4
( 38 .)