This paper presents several arguments for Right Node Raising Analysis of Gapping-like phenomena in Japanese, Korean, and German. Close examination of English data shows that movement is not involved in the derivation of English Gapping constructions, contrary to the recent analyses. And also, the movement analysis is untenable for Gapping-like constructions. On the other hand, under the Right Node Raising analysis, a number of properties of Gapping-like sentences follow straightforwardly. The analysis of this paper offers a key to the mechanism of ellipses in English Gaping and Right Node Raising.
This paper studies the syntactic relation of c-command in derivations involving adjunction within the framework of Chomsky's minimalist program. First, we re-examine how Reinhart's notion of c-command defined on representations applies to structures containing adjuncts. We then extend Epstein's idea of derivational c-command to incorporate derivations involving adjuncts. By assuming the syntactic operation Adjoin and by revising the notion of cycle in a derivation, we show that the concept of derivational c-command provides for an explanatory account of various types of data containing adjuncts.
This paper presents an analysis of so-called parasitic gap constructions within the framework of the Minimalist Program advanced in Chomsky (1995). Basically adopting the null operator analysis of Chomsky (1986b), I will demonstrate that the Anti-C-Command Condition and the Overt Licensing Condition are reducible from the general mechanism of feature checking, coupled with the assumption that the operator involved in this construction has a defective referential feature which must be checked by an antecedent. It will be shown that no construction-specific device or S-Structure condition is necessary in accounting for the data, which suggests that the Minimalist Program is not only conceptually desirable but descriptively adequate.
Given Kayne's (1994) Linear Correspondence Axiom (LCA), SOV languages are classified into overt object shift languages. On the basis of this classification, this paper first proposes the following generalization: object-oriented floating quantifiers, multiple object construction, multiple subject construction, scrambling, and extraction from within subject are permitted only in overt object shift languages. Then, claiming that only Agr with a strong D-feature can license multiple specifiers, we derive this generalization from the existing conditions on movement and feature checking in a minimalist theory. Since the distinction between Agr and other functional categories is crucial and has a wide range of consequences, Chomsky's (1995) attempt to eliminate Agr should not be tenable.
This paper concerns some syntactic constraints imposed on Extraposition from NP. It has been pointed out in the literature that Extraposition is possible from D-structure object NPs. The present study examines applicabilities of Extraposition to object NPs in double complement constructions, and shows that while the second complement NPs tolerate Extraposition, the first ones reject Extraposition. I claim that this constraint imposed on the first complement NPs follows immediately from the Larsonian shell analysis of phrase structures. This in turn means that the present study motivates the shell analysis of phrase structures in terms of extraction phenomena.
It has sometimes been pointed out in the generative literature that a semantic restriction is imposed on depictive predicates: stage-level predicates are acceptable as depictive predicates, but individual-level predicates are not. In this paper, we argue that as an answer to the question of why such a restriction is imposed on depictive predicates, Maruta's (1995) explanation is basically correct, but that the status of depictive predicates is not as simple as he supposes: they are, so to speak, half-fledged adverbials with adjectival nature rather than full-fledged adverbials. We therefore present an analysis which can provide an adequate account for such peculiar status.
This paper discusses the antecedent-contained deletion (ACD) construction and examines under what conditions the acceptability of the construction is decided. First, the paper critically examines previous syntactic analyses such as May (1985), Hornstein (1994, 1995) and Diesing (1992). Then, it alternatively argues that the acceptability of the ACD construction is primarily controlled by the functional notion of focus. It subsequently shows that the present functional analysis can deal with the fact that there is pure asymmetry concerning the distribution of determiners between the ACD construction on the one hand and wh-extraction from NP and extraposition from NP on the other.
This paper aims to explicate, from a unified point of view, the differences in meaning and cooccurrence restriction between be going to and will, and a diversity of their uses in contemporary English. It presents the temporal structures of be going to- and will-sentences which provide an explanatory basis for these phenomena. It is argued that the notion of temporal focus is necessary in order for the temporal structures to have a more explanatory power. It is demonstrated that these temporal schemata not only systematically explain both synchronic and diachronic aspects of be going to and will, but also account for why be going to in the past tense tends to express unfulfilment of the infinitive part while will in the past tense does not.
This paper proposes a unified analysis of the cognate object construction and the intransitive unergative sentence. Based on the assumption that the cognate object construction is an alternation of the unergative construction, the lexical and syntactic structures of the two constructions are presented. It is argued that the three readings of the cognate object construction correspond to the three-way distinction between DP and NP proposed in Borer (1994). This correspondence enables us to explore not only the cognate object construction but also the cognate object itself, focusing on referentiality and event structure.
This paper deals with the English cognate object constructions in the framework of cognitive grammar established by Langacker (1987, 1991). Based on the seminal analysis by Langacker, the analysis here not only accounts for the characteristics of the phenomena but also reveals that the cognate object's usual construal as an instance of the activity designated by the verb varies by virtue of modification. Moreover, it is proposed that there are at least two possibilities for the construal of cognate objects, construal as an instance of an activity (i.e. as an event) or else as a type of activity. Furthermore, it is shown that the present analysis will enable us to account for some problems related with construal and transitivity.
This paper provides a unified account of the occurrence of possessive forms with two types of nominals, namely derived nominals with-tion, -sion, -al suffixes and action nominals in -ing forms. Based on the claim by Langacker (1987a, 1991) that the two differ in terms of construal of boundedness, it is argued that their semantic difference triggers various behavioral discrepancies. Having argued that the basic characteristics of the possessive construction in general is the reference-point function, two types of possessives are investigated, i. e. objective possessives and temporal possessives, from the point of view of the semantics of both the construction and the nominals in it. Regarding the former, a constraint which presupposes bounded construal of the event is formulated in terms of Croft's (1991) causal chain representation. Concerning the latter, it is argued that a temporal possessive in general requires boundedness of its identifying target.
In NEG first (NEG1) constructions, which are frequently observed in Old English, the negating particle ne occupies a “sentence initial” position and is immediately followed by a finite verb. Since the negating particle is analyzed as cliticized to the finite verb, NEG1 constructions apparently seem to be instances of verb first sentences, which are occasionally observable in Old English. This article will show that in spite of their appearance NEG1 constructions should be analyzed as verb second, but not verb first constructions. Specifically, it will be argued that in the constructions the initial position is occupied by an empty operator and that the combination of the negating particle and finite verb is in the second position.