. 8
( 38 .)


indicate what makes them both true.49 This is implicit in the following
extract from the original argument:
[If A and B each has knowledge, and their representations differ,] then it
seems to follow that there must be some coherent way of understanding why
these representations differ, and how they are related to one another. . . . [A]
story can be told which explains how A™s and B™s can each be perspectives
on the same reality.50

Williams does elsewhere insist on the “non-additive” nature of knowledge.
He writes, “Not all pieces of knowledge can be combined into a larger body
of knowledge.”51 But this relates to what I said in the previous section about
directness. Again I see no con¬‚ict. What his argument and its conclusion
require is the possibility of telling some story that gives an indirect indica-
tion of what A and B between them know. What the “non-additive” nature
of knowledge prevents, or may prevent, is directly conjoining what A knows
with what B knows, without dilation or manipulation of any kind.52 The

49 This is basically what I call the Fundamental Principle in Moore (1997a): see pp. 21“22.
50 Williams (1978), p. 64.
51 Williams (1985), pp. 148“149.
52 See again n. 42. The kind of dilation or manipulation that I have in mind is illustrated in
the case in which A knows that it is humid and B, in the very same place six months later,
knows that it is snowing: in order to indicate what A and B between them know (which is
Realism and the Absolute Conception

basic idea, then, is that all knowledge ultimately answers to reality “ the
same reality in every case “ and this is an idea that means nothing unless it
means that, for any given items of knowledge, there is some single way of
indicating, however indirectly, how reality is thereby known to be.
What are the alternatives to this? Here is one.
Where we have . . . , [two true representations] that con¬‚ict . . . their realms
are . . . less aptly regarded as within one world than as two different worlds,
and even “ since the two refuse to unite peaceably “ as worlds in con¬‚ict.

This is Goodman, summarizing the chief motif of his iconoclastic work
Ways of Worldmaking.53 As we can see, he is using conceptual apparatus that
differs from that of Williams in precisely the radical way envisaged. Not that
Goodman would mind admitting that the resultant apparent disagreement
between him and Williams is essentially terminological. On the contrary, to
admit this would itself be to signal a striking example of the very pluralism
he is advocating. Earlier in the same book he writes:
While I stress the multiplicity of world-versions, I by no means insist that
there are many worlds “ or indeed any; for . . . the question whether two
versions are of the same world has as many good answers as there are good
interpretations of the words “versions of the same world.” The monist can
always contend that two versions need only be right to be accounted versions
of the same world. The pluralist can always reply by asking what the world
is like apart from all versions.54

But if we do adopt the pluralism that Goodman advocates, then we shall
not say, with Williams, that for any two items of knowledge, there is some
single way of indicating what is thereby known. We shall say instead, with
Goodman, that some items of knowledge answer to different worlds, worlds
that cannot be combined into one, and that, in such a case, there is no single
way of indicating what is thereby known.55
To understand better what is at issue here, let us look at two applica-
tions of these ideas. The ¬rst is to the question whether tense is real. On
one way of construing that question, it is, precisely, the question which of
these alternatives is the right one for tensed knowledge (knowledge from a

not, of course, that it is both humid and snowing) we make explicit reference to the dates
and times concerned.
53 Goodman (1978), p. 116.
54 Ibid., p. 96. (Actually, the pluralist had better do better than that. The monist can just as
easily acknowledge the force of that rhetorical question.)
55 Cf. ibid., p. 115.
40 A. W. Moore

temporal point of view). Those who claim that tense is real, on this con-
strual of the question, adopt a Goodmanian pluralism. Their position is
this. Given some tensed item of knowledge, there is no indicating what
makes it true except from the same temporal point of view. This is because
what makes it true contains something directly corresponding to its tense
that does not obtain from any other temporal point of view. Thus, suppose
I know that it is humid today. Then what makes my knowledge true is the
fact that it is humid today. But this is a fact that I can indicate only today. If
I say tomorrow, “It was humid yesterday,” that will not indicate the same
fact. At best it will indicate some intrinsically related fact, a fact about (as
it were) hesternal humidity which can itself be indicated only tomorrow.56
Reality fractures into different temporal worlds, then. Each temporal point
of view carries its own world with it. The facts that peculiarly constitute
one of these worlds can be indicated only from the corresponding temporal
point of view. It immediately follows that there are some items of knowl-
edge, namely, items of knowledge from different temporal points of view,
for which there is no single way of indicating how reality is thereby known
to be “ a consequence that is, of course, in direct violation of Williams™
The second application of these ideas “ an application more relevant
to Williams™ concerns “ is to the analogous question about social reality.
The analogue of the view that tense is real is the view that reality fractures
into different social worlds, and that the facts that peculiarly constitute one
of these social worlds can be indicated only from the corresponding social
point of view. Williams himself uses the expression “social worlds,” as I did
in echo of him when outlining his view in §1.58 And sometimes he makes
claims that suggest that he adopts precisely this Goodmanian view about
them. He writes:
[An observer] cannot stand quite outside the evaluative interests of [a] com-
munity he is observing, and pick up [one of their thick ethical concepts]
simply as a device for dividing up in a rather strange way certain neutral
features of the world. . . . [Their ethical knowledge is] part of their way of
living, a cultural artifact that they have come to inhabit.59

But of course, Williams does not adopt Goodmanian pluralism about social
worlds. This quotation is another illustration of the distinction that arises

56 Is this not self-stultifying? Is not my very reference to this fact that I say can be indicated only
tomorrow an indication of it today? No. Cf. earlier, n. 38: when I talk about “indicating” a
fact, I mean something like “expressing” it, not making reference to it.
57 See Moore (2001) for a fuller account of the material in this paragraph.
58 See, e.g., Williams, (1985), p. 150.
59 Ibid., pp. 142 and 147.
Realism and the Absolute Conception

within his realism between, on the one hand, directly expressing what mem-
bers of the community know “ that is, what they know through exercise of
their thick ethical concepts “ and, on the other hand, telling a story that
indicates indirectly what they know. The observer cannot do the former
because, as someone who lives outside their social world and who does
not share the values that sustain exercise of their thick ethical concepts,
he cannot himself make use of those concepts. Nor, crucially, does he
have the wherewithal to construct neutral equivalents of those concepts:
that is the point that Williams is making in the quotation.60 By contrast,
he may be able to do the latter. He may be able to understand enough
about the community, about their social world, and about its history to be
able to see how their use of these concepts enables them to live in that
world, and he may be able to say, in the light of that, how their circum-
stances warrant the exercise of the knowledgeable judgments they make
using these concepts. Admittedly, if he does succeed in doing this, he might
still not have carved out the same chunk of logical space as they do in
making any of the relevant judgments. But carving out the same chunk
of logical space must not be confused with carving out a chunk of the
same logical space. He will have done the second of these, which is all that
Williams requires. In particular, he will have said enough to entail what they
know, in the relaxed sense of entailment that I alluded to in the previous
If the choice between Williams™ realism and Goodman™s pluralism is
essentially terminological, then can it matter which we adopt? It can. The
word “terminological” has pejorative overtones. But an issue can be both
terminological and of considerable moment. It can be, in effect, an issue
about which of two conceptual structures is better equipped to meet certain
theoretical and/or practical needs. (Think again about the choice between
Euclidean geometry and its non-Euclidean variants.)
Very well, then; what is there to be said in favour of either of these
alternatives? There are different things to be said in different contexts, but
one thing that I think must be said quite generally in favour of Williams™
realism is that it is our natural starting point. Goodman™s pluralism is an
unintuitive departure from the way we normally think and speak.61
Still, the way we normally think and speak is not sacrosanct. This advan-
tage can be outweighed.62 Let us turn back to the question whether tense
is real. Here the Goodmanian alternative might be thought to have an

60 See further ibid., pp. 141“142, and the references therein. See also McDowell (1981),
pp. 144“145.
61 Cf. the parenthetical remark in n. 48.
62 Cf. in this connection Wittgenstein (1974), Pt I, §402.
42 A. W. Moore

advantage in relation to the idea that the future is open. It is far beyond the
scope of this essay to say fully what I have in mind. But here is a sketch.
Suppose, what many people intuitively think, that the future is open. That
is, roughly, suppose that nothing is the case at any given time about what
is contingently the case at later times. And suppose I said yesterday, “It
will be humid tomorrow.” Then not only did this fail to indicate the fact
that veri¬es my knowledge that it is humid today. It was not even true. For
it was not (then) the case that it would be humid today. By far the most
natural way to capture this idea, and arguably the only way to do so, is
by appeal to different temporal worlds; more speci¬cally, by appeal to a
sequence of temporal worlds such that those later in the sequence contain
details corresponding to gaps in those earlier in the sequence.63
But even if Goodman™s pluralism has the whip hand in this context
(which actually I doubt), there are other contexts in which Williams™ real-
ism retains all sorts of advantages over it, and indeed over any of its rivals.
In particular, it retains signi¬cant advantages in the context of moral phi-
losophy. Paramount among these is something that will serve to bring my
essay full circle: namely, the way in which Williams™ realism, by sustaining
the argument for the possibility of the absolute conception, creates an ideal
framework for expressing his opposition to realism about ethics. It is impor-
tant here to be clear just what the relation between his realism tout court
and his opposition to realism about ethics is. The latter does not of course
follow from the former (in the way that the belief that the absolute concep-
tion is possible does). Rather, the former is part of a conceptual structure
that facilitates expression of the latter. If we repudiated Williams™ realism in
favour of Goodman™s pluralism, or even in favour of some species of ideal-
ism, this would not undermine any of his arguments against ethical realism;
but it would make those arguments harder, if not perhaps impossible, to
Williams™ realism gives us, I think, a richer understanding than its vari-
ants do of the nature of our ethical experience. By enabling us to see that our
ethical knowledge is from a point of view that admits of equally legitimate
and incompatible alternatives, in a way in which our scienti¬c knowledge
is not “ and by enabling us to see how history, psychology, and/or anthro-
pology are needed to explain why we have the ethical point of view we

63 For a thorough discussion of the issues and complications that arise here, a discussion which
attempts in a quite different way to capture the belief that the future is open, see Tooley
(1997), passim. (Note: whether we are to regard the future as open or not may itself be a
terminological issue of the very kind I have tried to identify. See further Cockburn [1997],
esp. Ch. 9.)
Realism and the Absolute Conception

have, where this is not itself a matter of our knowing anything “ Williams™
realism gives us a much ¬rmer grip on what people are getting at when
they make clumsy appeal to the fact/value distinction. Up to a point, their
clumsy appeal is apposite. As Williams observed long ago,

in the factual case, there is a possible thought. . . . “I am convinced that p,
but it is possible for all that that not-p,” . . . [which] registers the impersonal
consideration that how things are is independent of my belief; however they
are, they are, whatever I believe. . . . But . . . [there is] no parallel thought
possible on the moral side: . . . there just is no content to “I am convinced
that racial discrimination is intrinsically wrong, but it is possible for all
that, that it is not,” except things like “How convinced am I?” or “I suppose
somebody might make me change my mind.”64

However, as Williams™ critique of thick ethical concepts clearly shows, there
is more to it than that. The concept of being intrinsically wrong is not
itself a thick ethical concept. Its applicability is not “world-guided” in the
way that, say, the concept of being a racist is. My conviction that racial
discrimination is intrinsically wrong is not an item of knowledge. But “ and
this is the point “ it does enable me to know such things as that Wagner
was a racist. The clumsy appeal to the fact/value distinction obscures this.
Williams™ more layered view makes it very clear. It also makes clear what
kind of thing I need, or more generally what kind of thing we need, if we
are to maintain our point of view and continue to have such knowledge,
from that point of view. We need con¬dence.65 Not that Williams™ realism
itself gives us con¬dence. On the contrary, it contributes signi¬cantly to
undermining our con¬dence “ not least, by making us aware of our need
of it. But that is the predicament that we must learn to face if we are to
live in light of the truth, something we have every reason to do. It would
be a serious mistake to think that we would be better off if we had never
re¬‚ected in this way on our ethical experience; if we had never thereby
known what it is to be, in a phrase borrowed from Wallace Stevens and
quoted by Williams at the beginning of Ethics and the Limits of Philosophy,66
“shaken realists™.67

64 Williams (1972), p. 49.
65 Williams (1985), pp. 170“171; and Williams (1995b), pp. 205“210. For further discussion,
see Fricker (2001).
66 This phrase is taken from Stevens™ poem “Esth´ tique du Mal,” in Stevens (1954); Williams
quotes it as part of the frontispiece of Williams (1985), p. x. See further Williams (1985),
pp. 167 ff.
67 I am very grateful to Anita Avramides for many helpful discussions on these matters.
44 A. W. Moore


Blackburn, Simon (1994). “Enchanting Views,” in Peter Clark and Bob Hale (eds.),
Reading Putnam (Oxford: Basil Blackwell).
Child, William (1994). “On the Dualism of Scheme and Content,” in Proceedings of
the Aristotelian Society 94.
Cockburn, David (1997). Other Times: Philosophical Perspectives on Past, Present and
Future (Cambridge: Cambridge University Press).
Dancy, Jonathan (1993). Moral Reasons (Oxford: Basil Blackwell).
Davidson, Donald (1984). “On the Very Idea of a Conceptual Scheme,” reprinted
in Inquiries into Truth and Interpretation (Oxford: Oxford University Press).
Davidson, Donald (2001). “Indeterminism and Antirealism,” reprinted in Subjective,
Intersubjective, Objective (Oxford: Oxford University Press).


. 8
( 38 .)