We have employed our semantical theory to investigate the matter of the ontological commitment of logic and semantics, and have found the following results: (i) The logical predicates, such as “or” and “entials”, are about statements not about any other objects. Moreover this is what a logical theory is supposed to characterize, namely logical objects, and only such. (ii) Tautologies refer, now to conceptual objects, now to physical ones, now to all objects. But they describe or characterize none. Hence logic is not “une Physique de l'objet quelconque” (Gonseth, 1938, p. 20). Logic is no more and no less than the theory of logical form, in particular of the form of deductive arguments. (iii) The general theory of reference we propose does not specify the nature of the referents of a construct, hence it is not ontologically committed. By contrast, any application of the theory is ontologically committed. This commitment is made the moment the predicate under scrutiny is analyzed as a function mapping n-tuples of objects into statements. Such an analysis requires the identification of such objects, which identification is based on some hypothesis or other concerning the furniture of the world. But the identification itself is a task for the special sciences not for semantics. (iv) The correspondence theory of truth is committed to the thesis that there is an external world, i.e. that there are entities that a factually true statement fits. Hence the chapter of semantics dealing with the notion of verite de fait is not ontologically neutral. And it does not overlap with model theory.