238 J.J. Haldane
can know nothing, or to statements about them, the truth conditions of
which may be inconceivable? Recalling the via negativa we might hope to
provide some content by way of saying what such entities and states of affairs
arenot; but this is unlikely to escape the charge of vacuity pressed by Berkeley
when he writes, somewhat ironically, ‘I do not find that there is any kind of
effect or impression made on my mind, different from what is excited by the
termnothing’.^18 The idea of a ‘something’ of which nothing can be thought or
said fails to distinguish itself from ideas of indefinitely many purportedly
distinct such bare ‘somethings’, and thus reveals itself to be no significant idea
at all.
One response to this argument is to distinguish realism as a thesis about
what exists independently of ourconception of it, from realism as a claim
about mind-independence per se. Thus it might be conceded that the
claim that there may be entities or states of affairs that are in principle
unknowable and even inconceivable, is an empty one. Nonetheless this con-
cession allows for the thought that entities may exist independently of our
capacity to know or to conceive of them. But once the general point about
the vacuity of conception-transcendence is granted, what can then sustain the
weaker position? Why not, in short, be ‘humanistic idealists’, saying with
Protagoras that man is the measure of all things? Putting the point in
Dummettean terms the problem with realism is that it requires that we have
a conception of truth such that a statement may be true though the condi-
tions of its truth are in principle inaccessible and even inconceivable. But this
makes no more (or less) sense than the requirement that we have a concep-
tion of entities such that they may be in principle inaccessible and inconceiv-
able. The idea of an entity has to permit there being a possible conception of
it, and the idea of truth has to be such that its obtaining is in principle
determinable. Once one allows these points it is hard to halt the retreat from
realism at the stage where conceivability passes from what is generally to what
is humanly the case. What could justify the claim that while truth cannot
transcend the possibility of its recognition as such, yet it may entirely tran-
scend the possibility of our recognizing it?
The answer ‘nothing’ is embraced by anti-realists of post-modern and neo-
pragmatist orientation who welcome the complete collapse of realism in the
name of constructive humanism. What, though, if one finds Protagoreanism
not only uncongenial but fantastical, both in being incredible but also as
licensing fantasy in place of disciplined pursuit of the real? Is there any way of
combining acceptance of the idea that reality has to be knowable, with the
thought that what we do or can know is not the measure of the real? For all
that Michael Dummett has come to be associated with anti-realism, he has
long insisted that he doubts whether this position is actually credible or even
intelligible, and he has occasionally speculated about a theistic alternative.