On the Typology of Syntactic Positions and the Nature of Chains: Move ? to the Specifier of Functional Projections
V. M. Deprez, 1989
This thesis proposes a redefinition of the A/A’ dichotomy, a concept central to the Principle and Parameter model of syntax developed by Chomsky (1981, 82, 85, 86, 89) and others. It is argues that in a model which incorporates recent hypotheses concerning the basic structure of clauses, such as the VP-internal subject hypothesis and the Split-INFL hypothesis, the A/A’ dichotomy is inadequate to properly account fo the distribution and properties of chains created by Move a. The central claim of this thesis (Chapter II) is that the dichotomy relevant to the identification of chains types should be expressed in terms of the distinction between [+HR] and [-HR] positions which are defined as follows:
A. a is a Head Related position ([+HR]) iff
a is a sister to X0 or to X’, i.e, a is a specifier or a complement in the X’
theoretic sense.
B. a is a non-Head-Related position ([-HR]) otherwise
Chapter I is a brief overview of the evidence for the VP-internal subject hyptohesis and the Split-INFL hypothesis, providing the theoretical background against which this thesis is set. Chapter II discusses the various properties of syntactic positions and the clusters of properties which identify two distinct types of chains. The central claim of the thesis is motivated in this chapter.
Chapter III is an in depth crosslinguistic study of the movement of objects (scrambling and Object-shift). It is shown that the chains created by object movement have properties characteristic of [+HR] chains. Object movement is analyzed as movement to the Spec of the various functional projections made available under the Split-INFL hypothesis. It is shown that object movement to [+HR] positions is constrained by the movement of verbal heads, a constraint which is argues to follow from the ECP given a dynamic notion of Minimality. A consideration of properties of the German/Dutch Mittelfeld scrambling leads us to suggest that a binary partition of positions and chains may be insufficient. We identify a third type of chain which has mixed properties and suggest that this particular type of chain is created by movement to [+HR, -Case] positions.
Chapter IV considers WH-movement ([-HR] chains) and the theory of the ECP. It is argues that contrary to recent assumptions, WH-movement is not always movement to (or through) the Spec of CP (which in our view is a [+HR] position comparable to the Spec of other functional projections). The Spec of CP is available only for subject extractions; other WH-extractions are adjunctions to CP. This proposal accounts for the fact that the que/qui alternations of the French complementizer (and similar alternations in other languages) is restricted to subject extractions and is impossible otherwise. The last part of Chapter IV is devoted to Subjacency. The difference between ECP and Subjacency violations is derived from a recursive application of the ECP after trace deletion. That is ECP applies twice: first at S-structure and then at LF, after trace deletion. The intuitive idea behind this proposal is that chains which violate the ECP both at S-structure and at LF lead to stronger judgements of ungrammaticality (i.e., standard Subjacency violations). Further distinctions (in this case, between weak and strong Subjacency violations) are accounted for in term of a calculus of barrier force. We follow Chomsky (1989) in assuming that obligatory trace deletion in non-homogenous operator-variable chains is due to the Principle of Full Interpretation. This hypothesis is shown to have empirical consequences for the distribution of floating quantifiers.
Thesis Supervisor: David M. Pesetsky
Title: Associate Professor of Linguistics
<br/>
Table of Contents
Chapter I Theoretical Background
1.1 The VP-internal subject hypothesis
1.1.1 Theoretical argument for the VPS
1.1.2 Empirical argument for the VPS
1.1.2.1 INFL as a raising category
1.1.2.2 Languages with S-structure VP subjects
1.1.2.3 Floating quantifiers
1.2 The Split-INFL hypothesis
1.3 Conclusion of the chapter and overview of the thesis
Chapter II Properties of positions and chains
2.1 Introduction
2.2 Properties of positions
2.2.1 The syntactic relevance of positions
2.2.1.1 q-marked positions
2.2.1.2 Case-marked positions
2.2.1.3 Base generated and derived positions
2.2.1.4 Adjunct structure
2.2.1.5 Adjunct at S-structure
2.2.1.6 Base generated adjuncts
2.2.1.7 Summary
2.3 Properties of chains
2.3.1 Standard properties
2.3.2 Floating quantifiers
2.4 Towards a definition of the relevant dichotomy
2.4.1 Properties of the VP external subject position
2.4.2 Properties of the intermediate specifiers
2.5 Summary of chapter II
Chapter III Movement to Intermediate Specifiers
3.0 Introduction
3.1 Hindi scarmbling
3.1.1 Word order in Hindi
3.1.2 Sentence internal scrambling
3.1.3 Long distance scrambling
3.1.3.1 Summary
3.2 Japanese scambling
3.2.1 WCO in Japanese
3.2.1.1 Word order in Japanese
3.2.1.3 The disappearance of WCO effects
3.2.1.3 Against a PG analysis of WCO repair
3.2.2 Scrambling as an instance of AN-chain
3.2.2.1 Vbl-chains vs AN-chains and reconstruction
3.2.2.2 Vbl-chains vs AN-chains: reconstruction and proper
binding
3.2.2.3 Anaphoric binding
3.2.2.4 Numeral quantifiers
3.2.2.5 Functional projections in Japanese
3.3 Object shift
3.3.1 Object shift in Mainland Scandinavian
3.3.2 Object shift in Icelandic
3.3.3 Evidence against the Head Determination Hypothesis
3.3.4 Object shift as movement to a [+HR] position
3.3.4.1 Verb movement in Scandinavian
3.3.4.2 Universal properties of verb movement
3.3.4.3 Clausal structure in Scandinavian
3.3.5 Restrictions on Object Shift
3.3.5.1 Dynamic minimality
3.3.5.2 Pronouns vs DPs
3.3.6 Summary
3.4 German and Dutch Mittelfeld scarmbling
3.4.0 Scrambling vs. base generation
3.4.1 Vbl-chain properties of German scrambling
3.4.2 AN-chain properties of German scrambling
3.4.3 Towards a ternary partition of positions
3.4.4 On the nature of the positions which head chains with mixed
properties
3.4.5 Functional projections in German
3.4.6 Bare quantifiers in French
3.4.7 Conclusions of chapter III
Chapter IV The Spec of Comp
4.1 The problem
4.1.0 Introduction
4.1.1 Short distance WH and NP movement
4.1.2 Long distance WH and NP movement
4.1.2.1 Long distance WH movement and intermediate traces
4.1.2.2 Weak cross over
4.1.2.3 Super raising
4.2 ECP: a first approach
4.2.1 What counts as a barrier?
4.2.1.1 The barrierhood of IP
4.2.1.2 The barrierhood of VP
4.2.2 What can be adjoined to? Towards an elimination of the ban on
CP adjunction
4.2.2.1 WH-islands
4.2.2.2 Adjunct islands
4.2.2.3 Subjects islands
4.2.2.4 CNPC islands
4.2.2.5 Conclusion
4.3 Towards a definition of the ECP
4.3.1 C-command vs M-command
4.3.2 Relativized Minimality vs Dynamic Minimality
4.4 Case study
4.4.1 Extraction out of sentential complement
4.4.2 WH-islands
4.4.3 Adjunct islands
4.4.4 Subject islands
4.4.5 CNPC islands
4.5 Consequences
4.5.1 Solutions to 4.1.2
4.5.2 Super raising
4.5.3 Agreement in CP across languages
4.6 Case marking into CP
4.6.1 Infinitival relatives
4.6.2 Variables and Case-marking
4.7 Subjacency
4.7.1 Trace deletion
4.7.1.1 Referential indices
4.7.1.2 Quantifier float and operator variable chains
4.7.2 Subjacency
4.7.2.1 ECP vs subjacency
4.7.2.2 Weak vs strong subjacency
4.8 Conclusions