Abstract

Every fifteen years or so Stephen Schiffer writes a state of the art book on the philosophy of language, with special emphasis on belief ascriptions, meaning, and propositions. The latest is his terrific new book The Things we Mean. It is again full of ideas, insights, arguments, expositions, and theories. For us, however, who believe that that-clauses are first and foremost clauses, not referring expressions, and that they thus do not refer to propositions or anything else, The Things we Mean brings home the news that our champion, the author of Remnants of Meaning, has, alas, crossed over to the dark side. Although Schiffer’s earlier book defended one of the best versions of the no-reference theory, and brought up many of the issues that need to be addressed to defend such a theory, he now has recanted and switched sides. His new theory holds that propositions do exist after all, and that-clauses do refer to them. However, some of the motivation for the no-reference theory is incorporated into his new theory. In Remnants of Meaning one of the main reasons for rejecting the reference of that-clauses was the apparent impossibility to compositionally assign that-clauses their referents, and thus to give a compositional semantics for natural language. In The Things we Mean Schiffer still finds fault with any way to compositionally determine what things propositions are. But now the conclusion is not that they are not things, but that they are things that are not reducible to certain other things: they are sui generis entities. But they are not just any kind of sui generis entities, they are pleonastic entities. The use of the term ”pleonastic” might be slightly confusing, though, since propositions according to the new theory are neither pleonastic in the sense of redundant, nor pleonastic in the sense of the pleonastic it, which suggests a no-reference theory. Rather they are pleonastic in a certain technical sense. Simply put, pleonastic entities are the ones that i) can be introduced by 1 something-from-nothing transformations, and ii) the statement that there are such entities doesn’t imply anything about other entities that wasn’t implied before..