In Section II. the passage is made from the popular moral philosophy
thus arising to the metaphysic of morals. He denies that the notion of
duty that has been taken above from common sage is empirical. It is
proved not to be such from the very assertions of philosophers that men
always act from more or less refined self-love; assertions that are
founded upon the difficulty of proving that acts most apparently
conformed to duty are really such. The fact is, no act _can_ be proved
by experience to be absolutely moral, _i.e._, done solely from regard
to duty, to the exclusion of all inclination; and therefore to concede
that morality and duty are ideas to be had from experience, is the
surest way to get rid of them altogether. Duty, and respect for its
law, are not to be preserved at all, unless Reason is allowed to lay
_absolute_ injunctions on the will, whatever experience says of their
non-execution. How, indeed, is experience to disclose a moral law,
that, in applying to all rational beings as well as men, and to men
only as rational, must originate _a priori_ in pure (practical) Reason?
Instead of yielding the principles of morality, empirical examples of
moral conduct have rather to be judged by these.