Spans and words

1. The problem: Words are a pervasive unit of syntax and yet the dominant theory of them, the X0 theory, is problematic, predicting more parallels between phrasal and head movement than are observed. Phrasal movement approaches to word formation fare even worse on that score. Mirror Theory (MT) also has shortcomings, for example in relying on an unmotivated notion of specifier. 2. The solution: A theory of how syntactic structures are mapped onto functional and lexical words, positing syntactic features w for lexical access points and @ for linearization points. The theory draws on the late insertion of DM, the cycles of phase theory, the direct linearization principles of MT, and the non-terminal spell-out of Nanosyntax, separating word formation from linearization and appealing to spans (head-complement sequences) as the units of cyclic lexical access and storage.