This book offers a comprehensive study of relative clause syntax and explores a model of syntax in which Merge is feature-driven and projection is determined by selection. The analysis is based on new data from Moksha Mordvin, a Finno-Ugric, Uralic language. With three relativization strategies – including the cross-linguistically rare inverse case attraction – Moksha provides an ideal testing ground for theories of relative clause syntax. Drawing on the correlation between standard connectivity diagnostics (idioms, binding, Condition C) and case, the book argues that the raising derivation must be part of natural language syntax and must coexist with the head-external derivation. It then shows that several long-standing issues in the syntax of raising are resolved by projecting movement of the head noun, which follows from once feature-driven Merge and the projection-by-selection approach to labeling and combined with the possibility of upward search. Finally, the book demonstrates that further properties of relative clauses can be captured if syntactic selection can target not only a category but other features of the selected object as well.
