Movement-Resistant Aspects of Control

Abstract

This paper is a contribution to the ongoing debate about the nature of obligatory control (OC) in recent syntactic theorizing.* Although the debate has seen many participants and approaches, I focus here on two opposing views in particular: The view represented in Landau 1999, 2000, 2003, 2004a, 2006, and the view represented in Hornstein 1999, 2001, 2003 and Boeckx and Hornstein 2004. According to the former, OC is formed by an abstract Agree relation, mediated by functional heads. On the latter view (the movement theory of control (MTC)), OC is formed by A-movement, an instance of raising. Landau 2003 provides numerous independent arguments showing that the MTC as developed in Hornstein 1999 fails to account for the most significant generalizations about OC and non-obligatory control (NOC). In response, Hornstein (2003) addresses a variety of empirical problems and offers novel, sometimes ingenious, analyses for them. It also criticizes key aspects of the proposal in Landau 1999. Boeckx and Hornstein 2004, in turn, claims that most of the arguments in Landau 2003 against the MTC do not survive upon closer scrutiny. These are welcome developments. As the debate proceeds, theoretical positions are continually sharpened, bringing to light more and more empirical consequences. The purpose of the present paper is threefold. It first examines Hornstein’s (2003) treatment of various empirical challenges to the MTC and evaluates how well his solutions meet these challenges. It next clarifies and defends certain aspects of Landau 1999 that are criticized (and misdescribed) in Hornstein 2003. Lastly, it shows why the reply in Boeckx and Hornstein 2004 fails to address the essence of the critique expressed in Landau 2003. The general organization of this paper more or less follows the discussion in Hornstein 2003 and then turns to some issues exclusively treated in Boeckx and Hornstein 2004. One cautionary note to the reader: Because this is an ongoing debate with a considerable history, it becomes increasingly cumbersome to elaborate this history on each new installment. Therefore, much material – both data and arguments – that is already well-represented in the previous stages of the debate has been omitted. The discussion to follow thus presupposes some familiarity with the relevant literature. It is my hope that the arguments below will be appreciated and judged against the background of this knowledge.