Why did Greek science search for its own achievement in the axiomatic form?
We try to answer this question by confronting Aristotelian epistemological
program of axiomatisation of the sciences in the Posterior Analytics with
its realisation in Euclid's Elements. It appears that the different Euclide
an first principles fit quite exactly Aristotelian requirements. But the co
nverse is not true. Moreover, it is not possible to draw a coherent concept
of first principles from the text of the Posterior Analytics. We propound
to resolve both types of difficulties by a single assumption. The historica
l constitution of a deductive system stands from formalised procedures of t
ransmission of knowledge in the discourse of an ideal teacher (who know eve
rything) talking to an ideal pupil (who knows nothing). The system of first
principles in the Posterior Analytics is thus unstable and intermediate be
tween two coherent systems. The first one comprises terms to be understood
(definitions), and propositions held to be true, either by necessity (axiom
s) or contingently; these, in turn, can further be divided into hypotheses
and postulates, depending on assent or dissent of the hearer. The second sy
stem comprises axioms stating unprovable properties, definitions of terms a
nd hypotheses of existence. Thus, the Posterior Analytics seek to reconcile
the two Greek concepts of science. According to the first one, science is
a system of interlocutive transmission; according to the later one, it is a
system of objective knowledge. Euclid's Elements are at the end of this pr
ehistorical development of ancient axiomatic.