Abstract
The ninth and final lesson began by further clarifying the contributions to our project of this brief foray into the philosophy of chemistry, based on the Lavoisier case, on the epistemological level first of all: the harmonious combination, in the approach, of conceptual or analytical and empirical elements constitutes a promising advance of what the discovery of a posteriori necessities might consist in. Lavoisier also helps us to see that, in order to achieve this kind of epistemic access, we don't have to conceive of the scientific method as based solely(contra LaPorte [1]) on the principle of generalization and predictive success; we also have to make room for hypotheses, imagination and perhaps even mistakes: after all, we also learn a lot from our prejudices by correcting them, by falsifying them. Which already tells us something about how we might benefit from looking at certain models to achieve a genuine "metaphysical knowledge" of ENs.
Turning to the question of referential stability, Lavoisier's case certainly begs the question of what exactly he's referring to when he uses the term "element", and whether the question of stability is posed differently for different species ("oxygen" is a new term, which is not the case for copper, iron or sulfur, as known to the ancients). Are we dealing with discoveries or simple decisions linked to interests, and therefore conventional, in which case the reference of terms and nomenclatures would be the pure product of the historical and sociological context (LaPorte). We have presented some arguments recently put forward (Hendry [2]) to qualify such a position, aiming instead to support both the Putnamian hypothesis of referential stability (and its usefulness) and the validity of a realistic rather than instrumentalist reading of the "discoveries" (rather than "decisions") made in the natural species to which the history of chemistry bears witness.
