not directly known? To assert the 'correspondence' must become a
groundless postulate about something which is defined to transcend all
knowledge. The correspondence theory, then, does not _test_ the
truth-claim of the assertion; it only gives a fresh definition of it. A
'true' thought, it says, is one which _claims to correspond_ with a
'reality.' _But so does a false,_ and hence the theory leaves us as we
were, puzzled to distinguish them.[D]



Yet the theory is not wholly wrong. Many of our thoughts do claim to
correspond with reality in ways that can be verified. If the judgment
'There is a green carpet in my hall' is taken to mean 'If I enter my
hall, I shall _see_ a green carpet,' perception tests whether the
judgment 'corresponds' with the reality perceived, and so goes to
validate or disprove the claim. But the limits within which this
correspondence works are very strait. It applies only to such judgments
as are anticipations of perception,[E] and will test a truth-claim only
where there is willingness to act on it. It implies an experiment, and
is not a wholly intellectual process.



The superiority of the 'correspondence' theory over the belief in
'intuitions' lies in its insistence that thought is not to audit its own
accounts. Its success or failure depends upon factors external to it,
which establish the truth or falsehood of its claims. No such guarantee
is offered by the next theory, which is known as the 'consistence' or
'coherence' theory. In order to avoid the difficulty which wrecked the
'correspondence' theory, that of making the truth of an assertion reside