ective mental habit, it follows that all
knowledge of nature which goes beyond mere observed fact is not knowledge
(neither demonstrative knowledge nor knowledge of fact), but belief.[1] The
probability of our belief in the regularity of natural phenomena increases,
indeed, with every new verification of the assumptions based thereon; but,
as has been shown, it never rises to absolute certainty. Nevertheless
inferences from experience are trustworthy and entirely sufficient for
practical life, and the aim of the above skeptical deliverances was not
to shake belief--only a fool or a lunatic can doubt in earnest the
immutability of nature--but only to make it clear that it is mere belief,
and not, as hitherto held, demonstrative or factual knowledge. Our doubt
is intended to define the boundary between knowledge and belief, and to
destroy that absolute confidence which is a hindrance rather than a help to
investigation. We should recognize it as a wise provision of nature that
the regulation of our thoughts and the belief in the objective validity
of our anticipation of future events have not been confided to the weak,
inconstant, inert, and fallacious reason, but to a powerful instinct. In
life and action we are governed by this natural impulse, in spite of all
the scruples of the skeptical reason.

[Footnote 1: Hume distinguishes belief as a form of knowledge from
religious faith, both in fact and in name. In the _Treatise_--the passage
is wanting in the _Enquiry_--our conviction of the external existence of
the objects of perception is also ascribed to the former, which later
formed Jacobi's point of departure. Religious faith is referred to
revelation.]

In Hume's earlier work his destructive critique of the idea of cause
is accompanied by a deliverance in a similar strain on the concept of
substance, which is not included in the shorter revision. Substances are
not perceived through impressions, but only qualities and powers. The
unknown something which is supposed to have qualities, or in which these
are supposed to inhere, is an unnecessary fiction of the imagination. A
permanent similarity of attributes by no means requires a self-identical
support for these. A thing is nothing more than a collection of qualities,
to which we give a special name because they are always found together. The
idea of substance, like the idea of cause, is founded in a subjective habit
which we erroneously objectify. The impression from