ABSTRACT

In much recent grammatical thinking, the assumption is made that morphosyntax exhausts the subject of morphology. By “morphosyntax” I mean the rules for realizing phrasal syntax as morphemes. I will present an argument here that this is incorrect. In particular I will argue that some clear generalizations about prefixes across languages cannot be coherently treated in a theory in which those prefixes occupy slots in the Functional Structure of the clause. Rather, these prefixes are added in a word system that is separate from phrasal syntax, but which supplies phrasal syntax with (complex) items for insertion into phrasal structure. The character of the arguments is such that it only holds for what are traditionally called derivational prefixes, and so rests on the traditional distinction between derivational and inflectional morphology. This distinction has been widely rejected by both lexicalists (myself included) and non-lexicalists (e.g., “Distributed Morphology”), but the considerations presented here suggest it should be looked at again. The problem has always been in defining the distinction, and that problem remains, though the generalizations exposed here might turn out to be helpful.