'Did Aristotle arrive at this definition by considering and refining the ordinary use of the word? Or is it a purely technical definition which he simply introduces?' This extract from Michael Loux might help. Loux suggests, and supports the claim with argument, that Aristotle did not give ousia his own technical meaning. He seems to take the term as used (or as he understood it to be used) by predecessors such as Democritus and Plato. His novelty is to offer new criteria for what counts as ousia. (Compare in later philosophy how, say, Descartes, F.H. Bradley and Russell, don't disagree on what 'truth' means and none of them wants to give it a technical definition but they disagree over the criteria of truth - over what conditions have to be satisfied for a proposition to be true.)
WHATEVER novelty he may have claimed for his own answer to the
question "Which things are ousiai?", Aristotle viewed his attempt to
formulate a response to that question as continuous with the inquiries
of earlier thinkers; and this suggests that the sense he attached to the
term ousia is such that philosophers as different as Plato, Democritus,
and Aristotle himself can all be said to be attempting an identification
of ousiai. That Aristotle construed the term as neutral between opposing
ontologies is clear from the early chapters o? Metaphysics Z;1 and unless
we make this assumption, many of the ensuing chapters of Z and H are
likely to remain a mystery. Recent tendencies to translate ousia as it
occurs in the middle books of the Metaphysics in terms of the expression
"reality" rather than in the more traditional ways as "substance" reflect
a sensitivity to this point. But while the picture of a subject of or a thing
standing under other things plays a more central role in the Categories
than in the middle books, it is, nonetheless, misleading to read this picture
into our translation of ousia as it occurs in the earlier work; for even
there the term has a sense that is independent of the famous formula of
2a 11-12 ("neither said of a subject nor present in a subject"). That formula
does not represent the attempt to stipulate a sense for a technical term
which Aristotle wants idiosyncratically to introduce; it is rather the
attempt to provide a criterion for identifying the things he takes to be
the best candidates for an antecedently understood title. Thus, at 2a 33-2b
6 we find Aristotle effectively arguing that the things picked out by the
formula actually deserve the title ousia, a procedure pretty clearly inappropriate in the case of a stipulated definition of a technical term. But
what sense did he attach to the title? In part, he must have taken its
derivation from the verb einai as explicative of its sense. In part too, he
must have viewed its occasional use in Plato to express the idea of an
ontologically privileged entity as precedent for his own use. Finally, there
is the argument of 2a 33-2b 6, where Aristotle suggests that a minimal
condition on any attempt to identify things worthy of the title prote ousia
is that the items so identified constitute the smallest set such that we
can truly say of the members of that set, "If none of these things existed,
nothing else would either." (Michael Loux, 'Ousia: A Prolegomenon to Metaphysics Z and H', History of Philosophy Quarterly, Vol. 1, No. 3 (Jul., 1984), pp. 241-265 : 241.)
