distinguish between reason and experience,
and to suppose that these species of argumentation (_sic_) are
entirely different from each other. The former are taken for the
mere result of our intellectual faculties, which, by considering _a
priori_ the nature of things, and examining the effects that must
follow from their operation, establish particular principles of
science and philosophy. The latter are supposed to be derived
entirely from sense and observation, by which we know what has
actually happened from the operation of particular objects, and are
thence able to infer what will for the future result from them....
But notwithstanding that this distinction be thus universally
received, both in the active and speculative scenes of life, I
shall not scruple to pronounce that it is at bottom erroneous, or
at least superficial.'

Hume, it will be observed, is not here bent on vindicating the rational
character of direct inference from observation: he had set out in the
text by disparaging customary thinking as non-rational; and he is now
claiming for the 'reasoning' man that experience goes a long way to
generate his reasoning processes. 'The truth is,' he says in his final
paragraph, 'an inexperienced reasoner could be no reasoner at all, were
he absolutely inexperienced.' It is a fragmentary note to a hasty
passage; but at least it concedes that reasoning _is_ largely a matter
of inference from experience, and thus decisively gainsays the assertion
in the text that no inference from experience is an 'effect of
reasoning,' inasmuch as it says such inference is reasoning; that
reasoning is a working of the mind on the facts of life; and that the
common distinction between reasoning and [beliefs derived direct from]
experience 'is at bottom erroneous, or at least superficial.'[12] If, he
says in the fourth paragraph of the Note, 'If we examine those
_arguments_ which, in any of the sciences above mentioned, are supposed
to be the _mere_ effects of reasoning and reflection, they will be found
to _terminate_ at last in some general principle or conclusion for which
we can assign no _reason_ but observation and experience.' If an
argument be not a process of reasoning, neither word is intelligible. If
an argument terminates (=has one end) in a conclusion founded on
observation, and if that observation be a 'reason' for a proposition,
then arguing is reaso