HUME: NECESSARY CONNECTION AND CREDENCE
 Hume’s Problem of Induction: There is no rational support for inductive inferences. (Hume
argues that there is no way to support the claim that the future will be like the past without
begging the question.) We can clarify Hume’s Problem by contrasting it with some other well-
known problems with causal reasoning: (a) ﬁnite sample problems and generalization worries;
(b) interpreting probabilistic or statistical claims, e.g. “There is a 90% likelihood that A will be
followed by B.” The usual charge is that we fail to have knowledge in such cases. However, the
skeptic and the defender of causal reasoning alike usually grant that past instances are relevant
to future prediction. Hume denies this common premiss.
 In Enquiry 5, Hume readily admits that we do have beliefs about the future, based on past
experience. (He even thinks we’re better off for having such beliefs, and furthermore that such
beliefs are largely correct; those aren’t the point at issue. Instead, Hume is worried about our
epistemic warrant for these beliefs, which is another matter altogether.) Since we do not acquire
these beliefs though conscious or unconscious reasoning processes, we must get them merely by
habituation—through the inﬂuence of ‘custom’
A 1ollowed by B 1 ⇒ an impression-of-A ,1an impression-of-B 1
A followed by B ⇒ an impression-of-A , an impression-of-B
2 2 2 2
. . .
. . .
A n ⇒ an impression-of-A n
[the imagination moves to a belief in Bn]
A n+1 ⇒ an impression-of-A n+1
[the imagination moves to a belief in B n+1 ]
. . .
. . .
The question that occupies Hume in Enquiry V–VII is: How does habituation work in this case?
Given past experience of a constant conjunction between A and B, and given the present
impression of an A, how do we come to have a belief in the (prospective) existence of
Hume’s strategy in these chapters is to explore the nature of belief and, once equipped with an
analysis of belief, to work backwards: he identiﬁes the key component in the causal judgment
“A causes B” as the idea that A and B are necessarily connected. Then, by the ‘empiricist theory
of meaning’ (to any legitimate idea there must correspond an impression in some fashion), he
goes looking for the impression-of-necessary-connection.
 In Part One of Enquiry 7, Hume argues that we have no direct experience of the ‘power’