believe in causation in a sense which, in general, he condemns. Let us take an illustration. I see an
apple, and expect that, if I eat it, I shall experience a certain kind of taste. According to Hume,
there is no reason why I should experience this kind of taste: the law of habit explains the
existence of my expectation, but does not justify it. But the law of habit is itself a causal law.
Therefore if we take Hume seriously we must say: Although in the past the sight of an apple has
been conjoined with expectation of a certain kind of taste, there is no reason why it should
continue to be so conjoined: perhaps the next time I see an apple I shall expect it to taste like roast
beef. You may, at the moment, think this unlikely; but that is no reason for expecting that you will
think it unlikely five minutes hence. If Hume's objective doctrine is right, we have no better
reason for expectations in psychology than in the physical world. Hume's theory might be
caricatured as follows: "The proposition 'A causes B' means 'the impression of A causes the idea
of B.'" As a definition, this is not a happy effort.
We must therefore examine Hume's objective doctrine more closely. This doctrine has two parts:
(1) When we say "A causes B," all that we have a right to say is that, in past experience, A and B
have frequently appeared together or in rapid succession, and no instance has been observed of A
not followed or accompanied by B. (2) However many instances we may have observed of the
conjunction of A and B, that gives no reason for expecting them to be conjoined on a future
occasion, though it is a cause of this expectation, i.e., it has been frequently observed to be
conjoined with such an expectation. These two parts of the doctrine may be stated as follows: (1)
in causation there is no indefinable relation except conjunction or succession; (2) induction by
simple enumeration is not a valid form of argument. Empiricists in general have accepted the first
of these theses and rejected the second. When I say they have rejected the second, I mean that they
have believed that, given a sufficiently vast accumulation of instances of a conjunction, the
likelihood of the conjunction being found in the next instance will exceed a half; or, if they have
not held exactly this, they have maintained some doctrine having similar consequences.
I do not wish, at the moment, to discuss induction, which is a large and difficult subject; for the
moment, I am content to observe that, if the first half of Hume's doctrine is admitted, the rejection
of induc-