Edge-Based Clausal Syntax: A Study of (Mostly) English Object Structure

1. Introduction

1 Introduction 1.1 Concept Barrels A framework for theorizing about the grammar of a natural language (NL) can be thought of as an abstract barrel containing linguistic concepts and principles applicable to formulating accounts of particular constructions. For example, the long-dominant barrel representing the various generative transformational grammar frameworks emanating from the work of Noam Chomsky has included the concepts subsumed in (1.1), not of course all at the same time.1 (1.1) Elements of the Chomskyan generative syntactic barrel {abstract case, atomic node labels, atomic traces, binding principles based on c-command (Principles A, B, C), c-command, complex node labels composed of sets of feature specifications, configurational definitions of grammatical relations (see (2.4) of chapter 2), constituent structure trees, copy traces, derivations, economy principles, empty nodes, feature checking, Greed, Last Resort, lexical entries, lexical rules, phrase structure rules, Procrastinate, reanalysis, Relativized Minimality, Subjacency, y-roles, the A-over-A Principle, the Case Filter, the Chain Condition, the Cyclic Principle, the Doubly Filled Comp Filter, the Empty Category Principle, the Extension Condition, the Head Movement Constraint, the Least E¤ort Principle, the Minimal Distance Principle, the Minimal Link Condition, the Minimality Condition, the Opacity Condition, the principle of full interpretation, the principle of recovery of deletion, the Projection Principle, the Specified Subject Condition, the structure-preserving hypothesis, the Superiority Condition, the y-Criterion, the Tensed-S Condition, the Visibility Condition, the Wh-Island Constraint, transformations, X-bar theory} Hereafter, I refer to the ideas of (1.1) as Barrel A. It seems correct to identify Barrel A with what Culicover and Jackendo¤ (2005) call mainstream syntax. But nothing hinges on that. One can have reasonable confidence that adopting a barrel actually containing the ideas requisite for a correct account of a given NL construction will ultimately yield its right description. But if a barrel fails to contain the appropriate elements, no matter how much time and e¤ort is devoted to choosing diverse combinations of its elements, and to refining and explicating them, a right description will never emerge. As time passes and work assuming that the right answers lie somewhere in a particular barrel fails to yield proper descriptive accounts, it becomes plausible to suspect that the barrel simply does not contain the correct and necessary concepts; the plausibility of such a negative conclusion increases along with the time involved. Since work on the part of hundreds, perhaps thousands of linguists assuming subsets of the ideas in Barrel A has now proceeded for half a century, absence of insight and descriptive failures on NLs that have been extensively studied in its terms (especially English, surely the principal target of such work) must cast into doubt the correctness of Barrel A itself. Much of the rest of this work can be taken as an argument, though mostly an implicit one, for the failure of Barrel A ideas to provide reasonable levels of insight into English object structure. As far as I can see, almost none of the facts dealt with in this work could be argued to have received an adequate account in Barrel A terms. Whether one accepts this conclusion about Barrel A or not, almost none of its devices beyond nodes and node labels will be exploited here for descriptive purposes or taken as objects for theoretical discussion or criticism. For better or worse, the present study of certain aspects of English grammar is based on a sharply di¤erent collection of concepts, assumptions, and principles than those represented by the array of ideas in (1.1). The account here is developed in terms of a distinct conceptual barrel overlapping only minimally with Barrel A. Some of its distinctive syntactic aspects or at least related ones are spelled out in works such as these: (1.2) Works characterizing relational frameworks related to that adopted here Aissen 1987; Johnson and Postal 1980; Perlmutter 1983b; Perlmutter and Rosen 1984; Postal 1985, 1986, 1989, 1990a, 1990b, 1992a, 1992b, 1996, 1998, 2004, chap. 1; Postal and Joseph 1990 2 Chapter 1 I sketch some general features of a personally current version of these ideas in sections 1.2–1.6. 1.2 Languages and Grammars One deep di¤erence between the work discussed here and most of the work that appeals to Barrel A ideas involves foundational assumptions about NL. Barrel A research has mostly developed in a...