The sixth lesson began by highlighting the precautions that need to be taken if the project of a metaphysical knowledge of nature is to be successfully carried out, and to make a clear distinction between "scientific" and "scientistic" metaphysics (S. Haack), before recalling the prevalence of the antinomies that all scientific realism must face (hereafter: SR). A first group of definitional properties of SR reveals tensions. On the one hand, we have :
SR theses: 1) Observable facts provide data indirectly confirming the existence of unobservable entities, and our theories describe this unobservable reality. 2) The theoretical entities postulated by the sciences are indispensable to our explanations and ineliminable. 3) The success of our scientific theories (especially in their predictions) can only be explained because they are true.
On the other hand, we have the theses of anti-realism (instrumentalism, conventionalism, pragmatism):
The antithesis of SR : 1a) Observable facts do not allow us to infer the existence of unobservable entities, and our theories are merely instruments for our predictions. 2a) Theoretical entities can be reduced and eliminated in favor of constructs referring to observations. 3a) Our scientific theories can succeed in their predictions without being true of an independent world.
Arguments have been put forward in favor of SR: theunification of its theories; the explanatory character of the theories; the existence of new predictions. The "non-miraculous" argument: only SR can explain the non-miraculous "successes" of science. We then presented three new objections to SR, in favor of anti-realism: the argument that theories are "under-determined by empirical data" (the Duhem-Quine thesis); the falsity of theories in the course of history; the impossibility of saying which is the "best" theory. To the question of whether we are therefore doomed to relativism or neutralism, we countered Bas Van Fraassen's constructive empiricism with new positive arguments in favor of SR, based on its metaphysical, semantic and epistemic aspects (S. Psillos). Betting on a certain epistemic optimism, we presented arguments based in particular on: the non-miraculous argument; the epistemic value of abduction or "inference to the right explanation". Finally, we presented the reasons why a project of metaphysical knowledge of nature presupposes not only a commitment to scientific realism (S. Psillos), but also a metaphysical commitment on the part of the scientific realist himself.
