A minimalist theory of feature structure

This chapter has two intertwined aims. One is to outline, within a broadly minimalist framework, a fairly explicit proposal for a theory of feature structures, that is, the structure of features and the syntax of how features combine to form lexical items; the other is to explore the consequences of the idea that structure embedding in human language is only ever syntactic (that is, that there is a single engine for the generation of structure and that engine is the syntax—see Marantz 1997, Borer 2005; also compare the ideas in Hauser, Chomsky, and Fitch 2002, Fitch, Hauser, and Chomsky 2005). The two aims are connected, since if structure embedding is only syntactic, then the feature structures that are the basic atoms of syntax (i.e. lexical items) cannot involve embedding of one feature inside another. This approach contrasts rather starkly with work in other approaches which take lexical items to have rich featural structure (e.g the chapters using HPSG, LFG and FUG in this volume). Before launching in on features, a few words about what a grammar in Minimalism looks like. Much like categorial grammars, Minimalist grammars can be seen as lexically Many thanks to Ash Asudeh, Pavel Caha, Annabel Cormack, Gabi Danon, Justin Fitzpatrick, Daniel Harbour, Terje Lohndal, Michal Starke, Peter Svenonius, the editors of this volume (Anna Kibort and Grev Corbett), the participants of the Features Workshop and the Tromso Thursday Seminar for extremely helpful discussion and comments. The research reported here was funded by a Leverhulme Trust Major Research Fellowship