id, a timorous method of proceeding. We must
not attempt to fly, when we can scarcely pretend to creep. In
considering any complex matter, we ought to examine every distinct
ingredient in the composition, one by one; and reduce everything to the
utmost simplicity; since the condition of our nature binds us to a
strict law and very narrow limits. We ought afterwards to re-examine the
principles by the effect of the composition, as well as the composition
by that of the principles. We ought to compare our subject with things
of a similar nature, and even with things of a contrary nature; for
discoveries may be, and often are made by the contrast, which would
escape us on the single view. The greater number of the comparisons we
make, the more general and the more certain our knowledge is likely to
prove, as built upon a more extensive and perfect induction.

If an inquiry thus carefully conducted should fail at last of
discovering the truth, it may answer an end perhaps as useful, in
discovering to us the weakness of our own understanding. If it does not
make us knowing, it may make us modest. If it does not preserve us from
error, it may at least from the spirit of error; and may make us
cautious of pronouncing with positiveness or with haste, when so much
labor may end in so much uncertainty.

I could wish that, in examining this theory, the same method were
pursued which I endeavored to observe in forming it. The objections, in
my opinion, ought to be proposed, either to the several principles as
they are distinctly considered, or to the justness of the conclusion
which is drawn from them. But it is common to pass over both the
premises and conclusion in silence, and to produce, as an objection,
some poetical passage which does not seem easily accounted for upon the
principles I endeavor to establish. This manner of proceeding I should
think very improper. The task would be infinite, if we could establish
no principle until we had previously unravelled the complex texture of
every image or description to be found in poets and orators. And though
we should never be able to reconcile the effect of such images to our
principles, this can never overturn the theory itself, whilst it is
founded on certain and indisputable facts. A theory founded on
experiment, and not assumed, is always good for so much as it explains.
Our inability to push it indefinitely is no argument at all against it.
This inability may be owing to our