While both schools sought to trace out the identity of the practical
principles of virtue and happiness, they were not agreed as to the way
in which they tried to force this identity, but were separated
infinitely from one another, the one placing its principle on the side
of sense, the other on that of reason; the one in the consciousness of
sensible wants, the other in the independence of practical reason on
all sensible grounds of determination. According to the Epicurean, the
notion of virtue was already involved in the maxim: "To promote
one's own happiness"; according to the Stoics, on the other hand,
the feeling of happiness was already contained in the consciousness of
virtue. Now whatever is contained in another notion is identical
with part of the containing notion, but not with the whole, and
moreover two wholes may be specifically distinct, although they
consist of the same parts; namely if the parts are united into a whole
in totally different ways. The Stoic maintained that the virtue was
the whole summum bonum, and happiness only the consciousness of
possessing it, as making part of the state of the subject. The
Epicurean maintained that happiness was the whole summum bonum, and
virtue only the form of the maxim for its pursuit; viz., the
rational use of the means for attaining it.
Now it is clear from the Analytic that the maxims of virtue and
those of private happiness are quite heterogeneous as to their supreme
practical principle, and, although they belong to one summum bonum
which together they make possible, yet they are so far from coinciding
that they restrict and check one another very much in the same
subject. Thus the question: "How is the summum bonum practically
possible?" still remains an unsolved problem, notwithstanding all
the attempts at coalition that have hitherto been made. The Analytic
has, however, shown what it is that makes the problem difficult to
solve; namely, that happiness and morality are two specifically
distinct elements of the summum bonum and, therefore, their
combination cannot be analytically cognised (as if the man that
seeks his own happiness should find by mere analysis of his conception
that in so acting he is virtuous, or as if the man that follows virtue
should in the consciousness of such conduct find that he is already
happy ipso facto), but must be a synthesis of concepts. Now since this
combination is recognised as a priori, and therefore as practically
necessary, and consequently not as derived from experience, so that
the possibility of the summum bonum does not rest on any empirical
principle, it follows that the deduction [legitimation] of this
concept must be transcendental. It is a priori (morally) necessary
to produce the summum bonum by freedom of will: therefore the
condition of its possibility must rest solely on a priori principles
of cognition.
I. The Antinomy of Practical Reason.
{BOOK_2|CHAPTER_2 ^paragraph 10}