A new theory of labeling that sheds light on such syntactic phenomena as relativization, successive cyclicity, island phenomena, and Minimality effects.

When two categories merge and a new syntactic object is formed, what determines which of the two merged categories transmits its properties one level up-or, in current terminology, which of the two initial categories labels the new object? In (Re)labeling, Carlo Cecchetto and Caterina Donati take this question as the starting point of an investigation that sheds light on longstanding puzzles in the theory of syntax in the generative tradition. They put forward a simple idea: that words are special because they can provide a label for free when they merge with some other category. Crucially, this happens even when a word merges with another category as a result of syntactic movement. This means that a word has a "relabeling" power in that the structure resulting from its movement can have a different label from the one that the structure previously had. Cecchetto and Donati argue that relabeling cases triggered by the movement of a word are pervasive in the syntax of natural languages and that their identification sheds light on such phenomena as relativization, explaining for free why relatives clauses have a nominal distribution, successive cyclicity, island effects, root phenomena, and Minimality effects.