. 30
( 38 .)


Yet, he seems to approve Ajax™s authenticity, his being true to himself, and
also his appeal “to some genuine social expectations.”58
When Ajax announces his change of heart, he appears to accept the
principle that “ought implies can”: if he can change by internalising his
husbandly and fatherly responsibilities as his “other,” he takes that change to
be required of him.59 But, in the event, his character is so ¬xed that he cannot
do so. If the ethical, as Williams would have us think, must involve some
“other-regarding” concerns, Ajax should be judged a tragically con¬‚icted
¬gure, trapped by the necessity of honouring his own internalised shame
and unable to act according to the “other-regarding” claims on him whose
emotive force he does perceive. The fact that the immensely pre-Kantian

55 Sophocles (1969a), Ajax, lines 485“524.
56 Sophocles (1969a), Ajax, lines 646“92.
57 Williams (1993), p. 101, p. 102.
58 Williams (1993), p. 98.
59 Unlike many critics, I take Ajax™s expression of his change of heart to be sincere at the time
he states it. He tries out an alternative self, predicated on his wife™s appeal, but his character
is so ¬xed that he is unable to make it actual.
Williams on Greek Literature and Philosophy

Sophocles can make us see that and see why it is so suggests that Williams™
“internalised other” may be a dangerously fragile basis for a satisfactory
criterion of “real social expectations.”
The Philoctetes, another Sophoclean play, gives Williams a much bet-
ter example of shame™s positive effects in motivating other-regarding con-
cerns. The young hero Neoptolemus initially presents himself as someone
who is ashamed to win the con¬dence of Philoctetes by lying. Odysseus
persuades him to set shame aside for one day in the interests of victory;
whereupon Neoptolemus complies.60 Later, having successfully deceived
Philoctetes, Neoptolemus experiences sharp pangs of what one can only call
conscience, reproaches himself for having departed from his own “nature,”
and describes himself as someone who has manifestly acted shamefully.61
He then reveals to Philoctetes the Greeks™ plan to take him to Troy, fully
knowing Philoctetes™ implacable resistance to that project.
Williams does not draw the connection between conscience and shame,
but he rightly observes that the episode shows how shame “can transcend
both an assertive egoism and a conventional concern for public opinion.”62
Does the shame that Neoptolemus experiences “ a shame that Williams says
“we recognise in our own world as shame” “ correspond to what we also call
“guilt”? To pursue that question, he argues that, while Greek shame may
appear to overlap with our guilt, in that shame as well as guilt may be a reac-
tion associated with indignation, reparation, and forgiveness, there is none
the less an important difference between the Greeks and ourselves. Guilt,
he suggests, buys into the distinction between “moral” and “nonmoral”
qualities, whereas shame is as liable to be felt by “a failure in prowess or
cunning as by a failure of generosity or loyalty.”63 Questioning how far we
really make use of this distinction in our lives or even fully understand it,
Williams offers the following differentiation between shame and guilt: guilt
(not necessarily having to do with the voluntary) points primarily to what
“I have done to others,” shame to “what I am,” as in falling short of what
one might have hoped of oneself.
This focus of shame on “what I am,” he argues, makes shame, as the
Greeks understood it, a more realistic and truthful conception than guilt.
If we think otherwise, that is probably because we “have a distinctive and

60 Sophocles (1969b), Philoctetes, lines 86“118.
61 Sophocles (1969b), Philoctetes, lines 902“909.
62 Williams (1993), p. 88; Cairns (1993), whose book appeared in the same year as Shame and
Necessity, makes similar points to Williams concerning internalized shame, and also argues
persuasively that we may attribute conscience to Homeric characters, albeit a conscience
based on social standards rather than personal ethical principles.
63 Williams (1993), p. 92.
174 A. A. Long

false picture of the moral life, according to which the truly moral self is
characterless,” with “reason” supposedly supplying moral norms and lim-
ited attention paid to other people™s opinions of us: otherwise “morality is
thought to have skidded into the heteronymy that . . . we recognised as a
familiar charge against the mechanisms of shame.”64
I am not sure what to say about this proposal, to the extent that I under-
stand it. Williams™ point that Greek shame and our shame can motivate
other-regarding concerns is very well taken. However, I think that his
interest in stigmatizing the modern concept of moral guilt and its Kantian
and Christian associations muddies the waters as far as understanding early
Greek ideas is concerned. He says that the Greeks “had no direct equivalent
for our word guilt.”65 True, they had no noun equivalent to “moral guilt,”
but they had the adjectives aitios and anaitios, which they used to express
guilty or responsible, and not guilty or not responsible. Does Neoptole-
mos in Sophocles™ play feel shame as distinct from guilt? The answer, in my
opinion, is that he feels both of these, and precisely in Williams™ terms: guilt
at what he has done to Philoctetes and shame at not living up to his own
self-image as someone who should set a premium on honesty. And while
he is plainly distressed at being “shown” and “caught” as a cheat, there is
no suggestion that his distress is motivated by what others will say about
him or even by an internalised other: it appears to spring entirely from his
sense of having besmirched his own ethical norms.
This example “ and more could be given “ shows that Williams does
not need to devote so much space to defending the Greeks from the charge
that their ethics was “unacceptably heteronomous.”66 His pursuit of this
anti-Kantian agenda in Greek contexts can be more distracting than illu-
minating. A similar objection may be made to his alignment of Plato with
Kant on the grounds that both philosophers have a conception of “the moral
self as characterless,” meaning a self that “is simply the perspective of reason
or morality.”67 In making this claim about Plato, Williams is contrasting
him with the earlier Greek view he has approvingly developed concern-
ing the ethical value of shame and its attention to the internalised other.
Plato, we are asked to suppose, makes the “implausible” proposal that rea-
son, abstracting from everything that one contingently is, suf¬ces by itself
to distinguish good from bad.68 Such an idea of autonomy, if it lacks “an

64 Williams (1993), p. 94.
65 Williams (1993), p. 88.
66 Williams (1993), p. 97.
67 Williams (1993), p. 159.
68 Williams (1993), p. 100.
Williams on Greek Literature and Philosophy

internalised other that carries some genuine social weight . . . may become
hard to distinguish from an insensate degree of moral egoism.”
As I have already remarked, the social weight of an internalised other is
hardly as promising an idea as Williams suggests. Moreover, so far as Plato
is concerned, which for Williams means Plato™s Republic, it is hardly true that
he regards the “moral self” as either characterless or lacking construction
through a social process.69 For one thing, the Platonic candidates for “moral
self,” if they are the guardians of the ideal state, have been acculturated
from childhood by a system of education that requires them to internalise
good role models. Second, their selves are not restricted to their rational
faculty but include nonrational faculties, each with its own desires, whose
normative condition, when guided by reason, contributes to the good of the
whole person. Third, the entire procedure of Socratic dialectic is premised
on the idea that ethical deliberation, so far from being a solipsistic or purely
introspective exercise, requires social interaction and the testing of beliefs
in an interlocutory context with people of quite various characters. Platonic
ethics, though it does not exclude the motivation of shame, undoubtedly
places a radically new premium on the guidance of reason. I sympathize
with many of Williams™ worries about the ef¬cacy of reason on its own to
deliver appropriate ethical motivations, but his juxtaposition of Plato and
Kant is too strained to show that the earlier Greeks were better off (and by
implication we ourselves would now be better off) without Plato™s ethical


In the ¬nal chapters of Shame and Necessity Williams continues to develop
his anti-progressivist argument, now focusing on ideas of external necessity,
bad luck, and coercion. Given the cultural differences between the Greeks
and ourselves, including especially their beliefs in “supernatural necessity,”
it is easy to suppose that they inhabited a quite different mental universe.
What, for instance, do we have in common with people who largely accepted
the institution of slavery without demur and who took themselves to be sub-
ject to the arbitrary power of mysterious divine beings? Notwithstanding
these historical parameters, about which Williams shows keen understand-
ing, he makes a powerful case for the proposition that modern liberalism

69 For further objections to Williams™ claims about Plato™s characterless self, see Irwin (1994),
pp. 60“62, and Woodruff (1996), p. 180.
176 A. A. Long

has a long way to go before it can congratulate itself on having supplanted
Greek necessities with fully effective ideas of justice and their social imple-
Greeks in general, as he says, neither questioned the legitimacy of slavery
nor sought to justify it.70 They knew that it rested on coercion and was often
terrible for those who experienced it, but it fell outside most debates about
justice because it was seen as a social and economic necessity. Because “we
have economic arrangements and a conception of society with which slavery
is straightforwardly incompatible,” it may appear that the Greeks had a
radically defective idea of justice.71 To ward off that comforting conclusion,
Williams powerfully observes:

We have social practices in relation to which we are in a situation much like
that of the Greeks with slavery. We recognise arbitrary and brutal ways in
which people are handled by society, ways that are conditioned, often, by
no more than exposure to luck. We have the intellectual resources to regard
the situations of these people, and the systems that allow these things, as
unjust, but are uncertain whether to do so, partly because we have seen the
corruption and collapse of supposedly alternative systems, partly because
we have no settled opinion on the question . . . how far the existence of a
worthwhile life for some people involves the imposition of suffering on

He grants that liberalism, unlike the Greek outlook, hopes that considera-
tions of justice will mitigate the effects of necessity and chance on individual
lives, but he is surely right to insist that something like the Greek concepts
of economic or cultural necessity and individual bad luck are still very much
alive in the modern world.
To investigate what we can make of the Greeks™ “supernatural neces-
sity,” Williams returns to tragedy and Homer. In this chapter, more than
in any other parts of the book, he offers sustained and insightful readings
of passages that have been greatly discussed in modern scholarship. When
Agamemnon, in Aeschylus™ play of that name, faces the awful dilemma “
should he sacri¬ce his daughter in order to secure his ¬‚eet™s sailing to Troy,
or should he abandon that military expedition? “ the play™s Chorus tells
us that “he put on the harness of necessity” and proceeded to the sacri¬ce
in a frenzied state of mind. Williams, quite correctly in my opinion, ¬nds

70 Williams (1993), pp. 110“115, includes a trenchant criticism of Aristotle™s defence of
“natural” slavery.
71 Williams (1993), p. 125.
72 Williams (1993), ibid.
Williams on Greek Literature and Philosophy

no invitation in this remarkable text to ask questions about Agamemnon™s
freedom of will or to apportion moral blame. The “harness of necessity”
invokes the idea that Agamemnon™s decision is not simply what he thinks
he has to do but also what he thinks he is externally required to do in order
to placate the goddess Artemis.
With reference to this and some similar passages, including those that
draw upon oracular pronouncements, Williams asks us to analyse the notion
of being subject to supernatural necessity. What it invokes, he suggests, is
“the image of being in someone™s power” where the someone, especially in
the contexts of Greek tragedy, “has no characteristics except purpose and
power.”73 Yet, if we think that our own rejection of supernatural necessity
has liberated us from much that oppressed the early Greeks, we had better
think again:

Human beings are metaphysically free in the negative sense that there is
nothing in the structure of the universe that denies their power to intend, to
decide, to act, indeed to take and receive responsibility in the fundamental
and intelligible sense that were found . . . already in Homer. But metaphys-
ical freedom is nothing “ at any rate, very little.74

It is little or nothing because what really threatens anyone™s freedom is not
supernatural necessity but the constraint that the Greeks called anangke, as
manifested in the psychological, social, and political situations that place
coercion on choice, making us subject to the will of another.
The historian Thucydides, though contemporaneous with Sophocles
and Euripides, gave no thought to supernatural necessities. Acknowledg-
ing Thucydides™ remarkable capacity to explain events in terms of typical,
nonmoralistic motivations, Williams ¬nds common ground with Sophocles
in spite of the latter™s theological assumptions. What he ¬nds them sharing,
and moreover sharing with himself, is no belief that

in one way or another the universe or history or the structure of human
reason can, when properly understood, yield a pattern that makes sense
of human life and human aspirations. . . . Each of them represents human
beings as dealing sensibly, foolishly, sometimes 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.75

73 Williams (1993), p. 151.
74 Williams (1993), p. 152.
75 Williams (1993), p. 164.
178 A. A. Long

It does not greatly matter whether this assessment of Sophocles and Thucy-
dides needs quali¬cation; for in the end, Shame and Necessity is most illumi-
nating and provocative for what it shows us about Williams the philosopher
and his remarkable ability to cut through the hackneyed distinction between
thoughts that are strictly “philosophical” and ideas that are only “literary.”
Ever since he published the seminal essay “Moral luck,” with its subtle dis-
cussion of Anna Karenina and an imagined Gauguin, we have known his
gift for drawing cogent insights from “literature,” and it is superbly evident
in his account of Rousseau and Diderot in Truth and Truthfulness.76 What
many readers of Shame and Necessity, including myself, ¬nd most impres-
sive about this book is Williams™ philosophical engagement with great texts
conventionally called literary.77 In the copious notes he attaches to each
chapter he discusses relevant linguistic details with all the skill and sensiv-
ity of an expert philologist. Under his guidance, which stands as a model
for the practice, the leading characters in Homer and Greek tragedy offer
material for ethical and psychological re¬‚ection without losing their con-
textual identity; and that, I take it, is precisely what Williams intended to
achieve in his mission to make moral philosophy an enterprise that is true
to the complexity of human life as it is actually lived or brilliantly imag-
ined. The passage quoted above captures the essence of the mind at work in
this book and its author™s characteristic qualities “ ruthlessly honest, utterly
secular, humanely imaginative, impatient of anything that smacks of con-
ceptual fuzziness or pietistic edi¬cation or the claims of reason to deliver
more than a cool experience of human life can warrant.


. 30
( 38 .)