Category change in English with and without structural change

This paper offers a partial taxonomy of changes of category (word class), exemplified with recent English data. The paper takes as its starting point a structuralist syntax which employs constituent structure and conventional category labels but which lacks empty categories or elaborate functional structure. No fixed, universal inventory of categories is assumed. Three types of category change are distinguished: those where only the affected node and its phrasal projection change labels; those where the topology of the syntactic tree is altered as well; and those where a wholly new category enters the grammar. Most but not all of the examples of category change involve grammaticalization. There is evidence of gradience, and semantics may lead syntax. A distinction is drawn between ambiguous and equivocal syntax, where the latter is under-­determined. I suggest that WYSIWYTCH (‘What you see is what your theory can handle’) militates against the recognition of syntactically equivocal strings, and I conclude that for handling grammatical change of the kind surveyed, a rigidly structuralist syntax may turn out to be unrevealing.