内容

目次

Acknowledgments ix Abbreviations xi Introductory Notes and References 1 1 Doing Science with Language: Introductory Concepts 5 This chapter introduces hypothesis formation and testing in the realm of human language and discusses the paradox of language acquisition. It offers an initial sketch of the Principles&Parameters approach and the innateness hypothesis. 2 The Structure and Classification of Words 12 Words are analyzed into roots and affixes. A system of generative word formation is introduced involving morphemes and word formation rules. Also discussed are criteria for identifying the lexical class of roots, stems, and words. Finally, a discussion of the “meaning” of particular affixes leads to the conclusion that affixes do not have “simple” meanings, but instead participate with a constellation of other factors to determine meaning, something referred to as “compositional” semantics. 3 Determining the Structure of Sentences 29 Tests of phrasehood are introduced, indicating the presence of hierarchic structure within sentences. Also presented is some of the core terminology of syntactic relations among phrases. 4 Rules of Sentence Structure: A First Approximation 38 Phrase structure rules are introduced as a means of explaining the presence of hierarchic structure within sentences. Beyond basic phrasal structure, key concepts such as structural ambiguity and recursion are presented as further evidence of the efficacy of the phrase structure approach to the analysis of sentences. Recursion is noted as the key to explaining “linguistic creativity.” 5 Assigning Meaning in Sentences 53 Presented here is the system of determining grammatical function (subject, object, or adjunct) based on structural position. Building on this, theta roles and argument structure are introduced, offering an explanation both of how arguments (subjects, objects, etc.) get their explicit meanings, and how verbs “choose” the correct complementation pattern. 6 Some Category–Neutral Processes 63 Here, the notion of “category–neutral” processes is first introduced, paving the way for the generally category–neutral system of X–bar syntax presented later. The processes discussed here are coordination and proform insertion. 7 How Structure Affects Pronoun Reference 71 This chapter introduces c–command and some of the phenomena that ccommand has been crucial for explaining, including the distribution of negative polarity items, and the Binding Principles, the distribution and semantics of anaphors and pronominals, and referring expressions. The presence of such mechanisms as the Binding Principles in the theory of syntax points offers further support for the innateness hypothesis. 8 Complex Verb Forms 82 The case is made here that auxiliary verbs each head a VP, so that sentences with multiple verbs involve a recursive VP architecture. Also, the first transformation, Affix Hopping, is introduced, opening the discussion of transformational grammar, and the levels deep structure and surface structure. 9 Real vs. Apparent Sentence Structure 90 Tense affixes are argued here to originate in the same position as modal verbs do, leading to the claim that deep structure is “abstract,” that is, consistently different in its alignment of elements from that seen in surface forms. Also discussed is the position of negation and the head movement rule V–to–T, which raises an auxiliary verb to the position of tense. All of this expands the transformational view of syntax. Arguments are presented for the presence of a “null” tense affix in sentences like “They like beans,” making the system of affixation fully general. 10 Generalizing Syntactic Rules 104 Arguments are advanced that phrases headed by themajor lexical categories NP, VP, AjP, and PP share the same internal architecture, pointing toward the conclusion that the rules of the syntactic system are category–neutral rather than category–specific – instead of having separate rules for NP or VP, a single, general rule set explains the internal architecture of all major phrase types. 11 Functional Categories 116 The category–neutral analysis is extended here to functional categories such as T and C, leading to the conclusion that the system of syntax is completely category–neutral. The rules of syntax are few and simple. The specific details of derivations are largely driven by the features and argument structure of the words/morphemes employed in the derivation. The concept of parameter setting is developed further. 12 Questions, Relative Clauses, and WH Movement 127 A number of apparent anomalies raised in the detailed consideration of WH questions and relative clauses are resolved by addition of the transformation “WHmovement.”WHmovement exemplifies phrase movement to a non–argument position. Apparent “long”WHmovement is shown to be composed of series of “short” moves. The WH Island Effect is introduced in connection with this discussion. The syntactic system is argued to be “constructionless”, since its rules apply broadly, across different construction types. 13 NP Movement 144 Arguments are advanced for the VP–internal subject hypothesis, the idea that the subject of a sentence originates low, in SpecVP, rather than in its higher surface position, SpecTP. This indicates the existence of a rule,NP movement, which searches for an NP low in the structure to fill the SpecTP position. This leads easily into the analysis of passive sentences, where no subject appears in SpecVP (due to theta role suppression), so that Move NP must find another (non–subject) argument to fill the SpecTP position. NP movement is also central to explaining subject–to–subject–raising constructions, where a higher clause may “steal” the subject of a lower clause. Like WH movement, NP movement participates in deriving a range of constructions, supporting further the view that the syntactic system is both category–neutral and construction–neutral. 14 Things to Come: Various Aspects of “Current Theory” 160 Here, three further significant aspects of syntactic analysis are sketched out, anticipating further studies in syntax. These include the unaccusative hypothesis (the idea that the subject of certain apparently intransitive verbs actually starts as an object), theVP shell hypothesis (the idea thatmultiple complements are not “flat” but involve asymmetrical c–command), and the DP hypothesis (the theory that “traditional” NPs are in fact DPs, phrases headed by the functional category D). Appendix 1: Minor Grammatical Categories 167 Appendix 2: Argument Structures 171 Index 174