Search

This article argues that the relegation of the binding theory to the C-I interface (LF) is theoretically undesirable and empirically unwarranted. Recent minimalist research has sought to eliminate the binding theory from UG by reducing its conditions to narrow-syntactic operations ( Hornstein 2000, 2006 ; Reuland 2001, 2006 ; Kayne 2002 ; Zwart 2002, 2006 ; Hicks 2006 ). This approach remains controversial since the canonical minimalist binding theory ( Chomsky 1993 ; Chomsky & Lasnik 1993 ) views the binding conditions as interpretive rules applying at LF, supported by evidence that condition A interacts with other interpretive phenomena assumed to be determined at LF ( Lebeaux 1998 ; Fox & Nissenbaum 2004 ). While the interaction of anaphor binding and scope relations in particular is not disputed, I show that it is attributable to factors outside the binding theory, namely the requirement that variables (including anaphors) must be c-commanded by their binders at LF. Deprived of its strongest empirical argument, the LF binding theory can then be picked apart. [ABSTRACT FROM AUTHOR]

The article presents information on the grammatical constructions referred to as raising and control. Raising and control have been central concerns of generative syntax since the 1960s. Peter Rosenbaum's book "The grammar of English predicate complement constructions," proposed the seminal generative account of raising and control constructions.

Historically, control in generative grammar has fallen within the province of syntactic theory. One primary reason for this is that Mainstream Generative Grammar (MGG) has imposed a strong uniformity criterion on analyses as a measure of their explanatory adequacy. One aspect of this uniformity criterion, which we call Interface Uniformity, holds that the syntax-semantics interface is maximally simple, in that meaning maps transparently into syntactic structure, and that it is maximally uniform, so that the same meaning always maps into the same syntactic structure. It follows from Interface Uniformity that a nonfinite VP has a syntactic subject that is assigned its external θ-role. We argue that this view is most sustainable if one does not take into account the full richness and complexity of control phenomena, but treats control strictly in terms of complementation. When a fuller range of phenomena is taken into account, it appears that it is preferable to "turn over control" to the semantics, which is better equipped to capture the facts. We outline how to formulate the syntax-semantics interface so as to get the semantic facts to line up properly with the syntactic facts. This analysis of the interface extends naturally to raising. [ABSTRACT FROM AUTHOR]

Control has played an important role in theoretical debates within the Minimalist Program. This is so because control implicates notions such as module, θ-role, the Last Resort nature of syntactic operations, movement, binding, chains, Case, complementation, and more. Hornstein (1999) has controversially claimed that control is a subspecies of movement. That is, control is just like familiar instances of raising, except that it involves movement into an additional θ-position. If correct, the movement analysis has important conceptual and empirical repercussions, some of which are examined here. [ABSTRACT FROM AUTHOR]

This article extends Howard Lasnik's commentary inSyntactic Structures Revisited(2000) on Chomsky's transformational analyses of the verbal morphology system. It reviews Lasnik's critique of the transformational machinery inSyntactic Structuresand shows how the problems he identifies are resolved under a more minimal theory of transformations that developed circa 1980. Considering the three central topics in the analysis of this system—the order and form of verbal elements, the main verb movement parameter, and the distribution of periphrasticdo—it discusses some problems that arise in Lasnik's hybrid account and proposes a lexicalist alternative that eliminates the rules of affix hopping anddo-support. [ABSTRACT FROM AUTHOR]

This paper defends three interconnected claims: (a) selection is the only licensing procedure available to UG, specifically, checking is an instance of selection; (b) selection obtains in the mutual c-command configuration; and (c) though a head does not mutually c-command its own specifier, it mutually c-commands the specifier of its complement. A head may therefore license the specifier of its complement (as well as its complement) but not its own specifier (it is not local enough). This effectively eliminates the spec-head configuration from the repertoire of syntactic configurations, in favor of a unified notion of locality strictly identifiable with mutual c-command, a symmetric configuration. The discussion shows that a theory that collapses these distinctions remains empirically discriminating. The resulting theory is therefore genuinely reductionist. [ABSTRACT FROM AUTHOR]