This book, which is a revised version of the author’s 1986 Ph.D. dissertation, discusses various issues in syntax and logical structure of natural language. The primary data on which theoretical claims are made is drawn from Japanese but it also contains a discussion of related phenomena in English. The book’s theoretical orientation comes from the Principle and Parameters approach, also known as Government and Binding (GB) theory. The main concern of the book is the nature of constructions which involve WH-phrases in natural language. In particular, two questions about the types of construction are addressed: First, what is the nature of the locality principle that governs the syntax and semantic interpretation of constructions involving WH-phrases? Second, what is the nature of the WH-phrase as a quantificational expression?
Within the Government and Binding framework it is a commonly shared view that the quantificational nature of WH-phrases is captured by the assumption that both WH-phrases and ordinary quantifier expressions are subject to the same type of syntactic operation. It has the effect of ‘moving’ a relevant item which occurs in an argument position (\(A\)- position) at one level of representation to an ‘operator’ position (\(A'\)- position) at the other level of representation.
The author starts from the observation that the WH-construction in Japanese contrasts with the English counterpart inasmuch as a WH-phrase in Japanese occupies an \(A\)-position in \(S\)-structure. This means that there is no syntactic movement rule in this language that moves a WH- phrase which is generated in an \(A\)-position in \(D\)-structure to an \(A'\)- position in \(S\)-structure. Although WH-phrases in Japanese do not undergo movement to an \(A'\)-position in the process mapping \(D\)- to \(S\)- structure, a discussion of grossly identical behavior of Japanese and English WH-phrases in important respects leads to an analysis of WH- constructions in Japanese, where it is argued that a WH-phrase that occurs in an \(A\)-position in \(S\)-structure is moved to an \(A'\)-position in Logical Form (LF), a grammatical level that captures ‘logical structure’ as resulting from quantifier scope and binding capability. More specifically, the WH-phrase that has moved gets governed by the \(Q\)- element that occupies Comp, i.e. the \(Q\)-elements serve as ‘scope indicators’ for WH-phrases.
Given that Japanese has a movement rule which maps \(S\)-structure to LF, the question arises as to what kind of condition or constraint restricts the movement rule in question. The discussion of this subject centers around the observation that WH-constructions in Japanese superficially exhibit some asymmetry with respect to the effects of the locality conditions which have been thought to derive from the general condition of subjacency, i.e. while they obey the WH-Island Condition effect they are free from the Complex NP Constraint (CNPC) effect. The main thrust of the discussion of chapter 2 provides arguments that the CNPC effect of subjacency is by no means irrelevant in the derivation of LF in Japanese. What is claimed is that, considering a particular class of grammatical sentences in Japanese, the WH-phrase does not move out of the complex NP but moves only within the relative clause. This triggers movement of the entire complex NP to the operator position of the main clause — a pied- piping operation at LF, the theoretical apparatus of which is substantiated in chapter 3.
In particular, it is argued that the feature associated with the WH- phrase that has been moved within the relative clause percolates up to the entire complex \(NP\), and this identifies the entire complex NP as a WH-phrase. Thus arguments are brought forward (and compared with an alternative analysis based on the Empty Category Principle by Huang) that the locality principle which governs the syntax and semantic interpretation of constructions in Japanese which involve WH-phrases must be subjacency.
Chapters 4 to 6 treat semantic properties of WH-phrases. Although previous analyses have considered WH-expressions to be (existential) quantifiers, the author demonstrates that WH-phrases are devoid of semantic content and should therefore be treated as free variables in the logical representation. The quantificational force of the WH-phrase is determined by a certain class of quantificational elements (unselective binders in Heim’s terminology, such as the \(Q\)-element “mo” in Japanese and “no matter” in English) taking into account certain structural conditions that hold with the WH-phrase that has undergone movement at LF. Furthermore, considering the nature of the pronominal coindexing in English and Japanese donkey sentences, the pronominals in question are characterized as donkey pronouns, licensed by the relations of indirect binding (in the sense of Haik) which essentially realize common principles for both languages.