n not moving must needs be at rest, is not more certainly
demonstrable than that inferences cannot possibly be drawn except in one
or other of these two ways. Is not every inference a species of belief,
and must not every belief be either innate within us, or have been
acquired artificially; and, as in the latter case, it is a mental
acquisition, must it not in that case have been acquired by an operation
of the mind or understanding? Is it not clear, then, that inferences
must always be either intuitive or ratiocinative; and is it not strange
that Hume should deny that they ever are so? Yet stranger still is it,
that even while denying them to be either one or the other, he, almost
in the same breath, pronounces them to be both. For, after having on one
page denied that they are founded on reason, or any process of the
understanding, he describes them on the next page as being not simply
founded on, but as being themselves 'processes of the mind,' 'processes
of thought,' and immediately afterwards 'arguments,' nay, '_reasonings_
from experience;' and, yet again, after as short a pause, these very
same 'reasonings,' and 'arguments,' and 'processes of the mind and
thought,' he concludes by styling 'natural instincts which no reason or
process of the thought or understanding is able either to produce or to
prevent'--'operations of the soul as unavoidable' when the mind is
placed in certain circumstances, as it is 'to feel the passion of love
when we receive benefits, or hatred when we meet with injuries.'

What are we to say to a description of mental operations which are and
are not 'arguments' and 'reasonings,' which are and are not 'processes
of thought,' which are not 'intuitive,' and which yet are 'instincts?'
How are we to account for such amazing inconsistencies in an exposition
of one of the greatest of philosophers? With all humility, I submit the
following as a possible solution of the enigma.

The one solitary ground on which Hume denies the argumentative and
ratiocinative character of what he nevertheless terms arguments and
reasonings, is the impossibility of producing the _chains_ of argument
or reasoning of which they are composed. But this impossibility can at
most only prove that the reasonings are elementary, and have,
consequently, no component parts into which they can be resolved. But
reasoning is not the less reasoning for being elementary, or for being
only a single link in a chain, instead of being