Band 33, Heft 1 - Phase Extension Contours of a theory of the role of head movement in phrasal extraction

1. The Program Starting out from the central ingredients of the analysis of the locality restrictions on Predicate Inversion presented in Den Dikken (2006a), in this paper I will lay the foundations for a theory of syntactic locality and the relationship between phrasal extraction and head movement that is predicated on the premises below: (1) Phase Impenetrability syntactic relationships (Agree) and processes (Move) are constrained by the Phase Impenetrability Condition (PIC) of Chomsky (2000 et passim ): in phase α with head H, the domain is not accessible to operations outside α, only H and its edge are accessible to such operations (2) Inherent Phase an inherent phase is a predication (subject–predicate structure) (3) Phase Extension syntactic movement of the head H of a phase α up to the head X of the node β dominating α extends the phase up from α to β; α loses its phasehood in the process, and any constituent on the edge of α ends up in the domain of the derived phase β as a result of Phase Extension

1. In his target paper, Den Dikken [hereafter DD] aims at “lay[ing] the foundations for a theory of syntactic locality” that makes use of the notion of phase .

1. The ingredients of locality Since the early 1960s, one of the major goals of work in generative syntax has been to explain the locality properties of syntactic dependencies. Current accounts in the Minimalist framework derive locality from two basic notions. First, the syntactic relations that give rise to agreement and displacement are subjected to the Minimal Link Condition , which stipulates that such relations can obtain only in the absence of relevant intevening elements.

1. Introduction Den Dikken's paper (henceforth DD) explores an interesting alternative to Chomsky's conception of Phase Theory to account for different phenomena involving Predicate Inversion (PI). In particular, DD concentrates on cases of so-called copular, locative, and dative inversion (the respective instances in (1)): (1) a. The #1 best-seller in the country is this book. b. On the president's desk lay this book. c. I gave my students this book.

Phrases cease to be phases once the head raises out of the phrase. This is a clear claim. So, in the simplest case, one would expect a direct correlation between overt head-movement and phrasal transparency properties for extraction constructions in typical cases of head-movement, as for instance V movement.

In this short note, I point out some potential problems of the way den Dikken (this volume) understands the mechanism and timing of spell-out. I then test the consequences of his understanding, but do not conclude anything on the basis of the presented English data. I then test the prediction on a case of NP internal predication in Slovenian and explain the problems they pose for the theory presented in den Dikken (this volume).

1. Introduction The main proposal of den Dikken's paper consists of two elegant assumptions, which propose to derive phase theory in a dynamic way and to explain a number of extraction constraints. The first assumption is that a projection immediately containing a predication (i.e., a subject and its predicate) is a phase inherently. The second assumption is that a phase is extended to the maximal projection that its head moves to.

1. Truth in unpackaging In his paper, den Dikken argues that several syntactic and semantic phenomena provide evidence for a package of proposals that includes the following four claims: (1) Phase Extension “Syntactic movement of the head H of a phase α up to the head X of the node β dominating α extends the phase up from α to β; α loses its phasehood in the process, and any constituent on the edge of α ends up in the domain of the derived phase β as a result of Phase Extension”. (den Dikken's (3)) (2) Phase Impenetrability Condition as in Chomsky (2000, passim) [PIC] “[I]n phase α with head H, the domain [of H] is not accessible to operations outside α, only H and its edge are accessible to such operations”. (den Dikken's (1)) (3) Adjunction Prohibition “[A]djunction to meaningless categories is disallowed.” (den Dikken's (18)) (4) Inherent Phase “[A]n inherent phase is a predication (subject–predicate structure).” (den Dikken's (2))

1. Introduction Marcel den Dikken's exploratory study on locality based on Chomsky's phase-based model of syntax diverges from Chomsky's (2000, 2001, 2004) phase theory in two crucial regards. First, the particular notion of phasehood he puts forward is defined by the semantic property of predicationality (rather than Chomsky's loosely interpreted propositionality). Second, the paper (re)introduces dynamicity to the theory of locality domains by proposing that inherent phases are extended through head movement (HM), i.e., phasehood can be inherited (compare Chomsky 1986, Bobaljik and Wurmbrand 2003).

Phase Extension is a theory of ‘domain widening” resulting from movement of the head of a phase (either an inherent phase or a derived phase that is itself the result of Phase Extension) to the next head up the tree. Phase Extension is the automatic and unavoidable result of movement of such head movement. It saves some operations that would be ungrammatical without its application, but it blocks others that would have been perfectly legitimate if no Phase Extension had taken place. In my target paper, I presented a variety of empirical domains in which Phase Extension provides an explanatory account for otherwise often quite recalcitrant data. In the following pages, I address many of the comments and questions raised by my commentators regarding some of the empirical proposals, the technical ingredients of the theory, and their conceptual underpinnings. In the closing section, I will consider Pesetsky's alternative ‘property delay’ model, and point to some areas where it makes predictions that are dierent from, and as far as I can see less accurate than, those made by the Phase Extension theory.