Secondary extensions, meanings and non-null terms (Q2545138)

Nelson Goodman's claim that the ``difference of meaning between any two terms can be fully accounted for without introducing anything beyond terms and their extensions'' is built upon the well-known distinction between primary and secondary extension. This article challenges ``the extended extensionalist thesis'' that secondary extension can determine meaning. Goodman's entire argument for secondary extensions determining meaning is built upon null classes, and there are several difficulties in extending this argument to nonempty classes. Criticism is raised against the claim that compounds formed from the original term are completely ``arbitrary labels''; it is argued that there must be some relationship between the original term as it is used to refer and the term as it appears in the compound. Finally, it is claimed that a circular argument is involved in saying that no two terms ``P'' and ``Q'' will ever have the same meaning because the inscription ``a P that is not a Q'' is a P-description and not a Q-description. The extended extensionalist thesis fails since secondary extensions fail to determine meaning.