linguistic: the metaphysician isn't saying anything meaningful. Of course, he thinks he is,
or he wouldn't speak and write at such length. But in fact, there's no meaning to what he
says. And since his talk is without meaning, the questions of whether there are good
grounds for believing what he says, whether there could be good grounds, and so forth,
cannot even arise.
The question that now cries out for answer is obvious: What's the test for whether a piece
of discourse is or is not meaningful? As their answer, the positivists proposed their now
famous criterion for meaning: a sentence is meaningful if and only if it's either
analytically true or false, or empirically verifiable.
The positivists, criterion in hand, energetically set about trying to clarify the distinction
they were employing between analytic and synthetic sentences, trying to develop an
account of the nature of analytic truth and falsehood, and trying to refine and articulate
the concept of verifiability so that all sentences of reputable natural science, and no
sentences of “metaphysics,” satisfied the concept. Even by their own lights, they failed in
all three endeavors; it was especially their inability to devise a criterion of verifiability
satisfactory for the purpose of demarcation that proved a bitter pill to swallow. One does
still notice positivist yearnings in
end p.254
certain philosophers, but as a doctrine affirmed up front, logical positivism has
disappeared among philosophers. Its disappearance is one of the most dramatic examples
of the disappearance of an ism in the entire history of philosophy. It was in its heyday,
however, when Wittgenstein was doing his work.
Wittgensteinianism
I think the best way to understand Wittgenstein's philosophy of religion generally, and his
epistemology of religious belief in particular, is to see him as exploiting a qualification
that the logical positivists early on attached to their criterion of meaning. Shortly after
they had propounded their starkly elegant criterion—a sentence is meaningful if and only
if it's either analytically true or false, or empirically verifiable—the positivists found
themselves forced to qualify it. The qualification was there already in A. J. Ayer's
youthfully exuberant Language, Truth, and Logic. When it's said of a person that what he
said or wrote lacks meaning, the clear implication is that he should retract his words and
try again—unless, perchance, he is entertaining children with jabberwocky. But whereas
it was noted soon after the positivists first issued their criterion that moral judgments do
not satisfy it, the positivists were not so revolutionary as to recommend that we stop
making moral judgments. Instead, they said that the criterion should be understood as a
criterion for just one kind of meaning.
It proved unexpectedly difficult to say what kind that was. Rather than cataloguing the
suggestions, let me just mention the one that eventually pretty much won the day: the
criterion of meaning is to be understood as a criterion for whether one has said something
true or false. Since to say something true or false is make an assertion, we can say that the
criterion was to be understood as a criterion for assertoric meaning. Sentences not
satisfying the criterion, such as sentences expressing moral judgments, may have a
positive role in human life; of many of them it may be appropriate to say that they are
meaningful. But the acceptance of the positivist criterion of assertoric meaning imposes