Expressions like an angel of a girl show syntax-semantics mismatches; the first noun heads the phrase syntactically but not semantically. It also exhibits adjectival properties. Such properties of this type of expression are accounted for without problems pertaining to syntactic and lexical analyses, if the expression is treated as a constructional idiom, called here the adjectival noun construction, and if lexical licensing, instead of lexical insertion, is adopted. This is only possible under a correspondence view such as the one adopted in the tripartite parallel architecture model of grammar, which admits the autonomy of semantics as well as that of syntax.
This paper deals with coordinated constructions of path phrases and seeks to establish a generalization which adequately describes the relevant phenomena. What is problematic in our attempts is a case of “double specification” which does not involve a coordinator. We elucidate this intriguing phenomenon, introducing notions such as a “complementary relation” and a “part-whole relation.” A closer examination of this phenomenon reveals a second subtype which is to be subsumed under the same category. Lastly, we tackle a more fundamental issue of coordination in general-why use a coordinator to express semantic coordination? A dual function of the coordinator holds the key to the solution: a linking and a separating function.
Bowers (1987) observes that comparatives like angrier, more angry, and as angry are similar to definite noun phrases in that they form a syntactic island with respect to extraction and scope interpretation, which leads him to claim that both comparatives and definite noun phrases are members of the same syntactic category, called “Determiner Phrase” (DP). This paper argues against the DP analysis of comparatives and shows that what the DP analysis attempts to account for can be better accounted for in terms of a semantic notion like presupposition. It is emphasized that, generally, just because two or more kinds of syntactic units appear to behave similarly with respect to certain linguistic phenomena, that does not prove that they belong to the same syntactic category; there still remains the possibility of explaining the linguistic phenomena in question in terms of a semantic or pragmatic factor or factors which may range over different syntactic categories.
Chomsky's (1995b, 2000, 2001a, b) analysis of expletive there constructions involves the Caseless there assumption, whereas Lasnik's (1992, 1995a-c) analysis involves the Case-bearing there assumption and Belletti's (1988) assumption of be as a partitive Case-assigner. After showing that both Chomsky's and Lasnik's analyses are untenable, I suggest a new analysis and show that it accounts for the relevant range of data. The suggested analysis extends to other constructions, including copular be constructions and small clause constructions. This analysis is also shown to be capable of accounting for some properties of expletive it constructions.
Japanese has Multiple Object Constructions because the transitive light verb v can multiply agree with its objects. By virtue of this property of the transitive light verb v, Japanese also has ECM Constructions with finite complement clauses. When the matrix v has multiple EPP-features, the whole complement clause is first attracted to the matrix vP-Spec, followed by subsequent Attract of the embedded accusative subject to the outer-Spec of v. Rich productivity of the construction follows from the fact that it only makes use of the general devices of the language. With such devices being unavailable, English ECM Constructions are lexically-specified and void of such rich productivity.
This article explores the distribution of to-infinitival complements (to-ICs) in Belfast English (BE) on the basis of Henry's (1992) data in order to capture two generalizations about Standard English (SE) to-ICs: the complementizer for is incompatible with PRO, and complementizerless to-ICs are TPs. Importantly, these generalizations cannot be drawn from Henry's CP analysis of BE to-ICs constructed within a representational approach. Thus, I advance a new analysis within the current minimalist framework (Chomsky (2000, 2001a, b)) on the basis of the idea that BE to-ICs have for which merges in T as a part of infinitival morphology and raises to C in the same way as V raises to v. I also propose BE-specific conditions on the selection of the categorial status of to-ICs. Further, I show two ways of capturing the nature of the adjacency condition within the minimalist framework.
It is shown that the semantic properties of reflexive clauses are extended to apply to a particular type of discourse. “Reflexive discourse, ” in our terms, is marked by a type of indefinite noun phrase the properties of which are modeled after those of reflexive pronouns and related expressions. This type of noun phrase is found in both English and Japanese, and is based on two levels of reference: unique identification of a referent and a generic class of which that referent is a member. Various aspects of this type of noun phrase are discussed, including its relation to other types of noun phrases, the speech act associated with it, and the style of reflexive discourse. They are appropriately understood in terms of the Subject-Self metaphor and objectification, a pair of concepts that are originally designed for reflexive clauses.
Stowell (1981, 1982) proposes that there is a parameter that determines the availability of both preposition stranding (P-stranding) and the V-particle-NP construction. While the V-particle-NP construction requires only the positive setting of the relevant parameter, P-stranding requires several additional properties to be identified. Stowell's analysis predicts the following: English-learning children should never acquire P-stranding significantly earlier than the V-particle-NP construction. The investigation of ten longitudinal corpora for English from CHILDES has borne out this prediction. Given the recent proposal by Snyder (1995, 2001) that the verb-particle construction follows from the Compounding Parameter, which determines the availability of various complex-predicate constructions, the findings suggest that P-stranding is also dependent on this parameter, and thus support its global nature.
This paper argues for a reanalysis of one of Fiengo and May's eliminative puzzles of ellipsis. Rather than take their judgments at face value, the judgments themselves are re-evaluated and the missing reading is found to be possible in certain discourse contexts. Additionally, the missing reading is shown to be possible in examples in which certain of the pronouns cannot be given a de se interpretation. After setting out a comprehensive paradigm of examples controlling for de se interpretation, I show how this paradigm follows from analyses of pronominal anaphora developed in Tancredi (1997, 2000).
Few previous studies have systematically examined the fact that the English present perfect and the German Perfekt behave differently with respect to co-occurrence with adverbials referring to a definite time position (DTP adverbials) in the past such as yesterday/gestern and ten years ago/vor 10 Jahren. This paper aims to explain this fact systematically on the basis of the revised P (osition)-Definiteness Constraint, which is originally proposed to account for the incompatibility of the English present perfect and the DTP adverbials. It is demonstrated that the reason why the German Perfekt, unlike its English morphological counterpart, can occur with the DTP adverbials is due to the interaction between the revised P-Definiteness Constraint and certain characteristics of the German Perfekt. It is also shown that the proposed account provides an explanatory basis for the following two issues: (i) why the German Perfekt can go with DTP adverbials referring to future time, whereas the English present perfect cannot, and (ii) why the German Perfekt has a wider semantic range than the English present perfect.
Middle Mongolian has two similar constructions which express resultative ideas. One is AP-based, and the other Particle-based. The possible semantic relations the two constructions can express are very different, however. In particular, the AP construction may function only as a Weak resultative (in the sense of Washio (1997a)), whereas the Particle construction can express those ideas which are typically expressed by Strong resultatives. Mongolian is therefore similar to English in using the Particle construction to express a wide range of resultative ideas, but it is also similar to Japanese in that it has only Weak AP-resultatives. The existence of a language like this may shed light on some of the problems related to resultatives and language variation.
This paper argues that Optimality Theory gives a new insight into some aspects of the Great Vowel Shifts circa 1500. In particular, our goal is to show that the structural coherence problem raised by Stockwell and Minkova (1988) is explained in relation to the universal constraint hierarchies and that unlike previous analyses, the simultaneity of high vowel diphthongization and mid vowel raising around the 16th century is synchronically predictable. It will also be demonstrated that an approach with Local Conjunction (Smolensky (1993, 1995)) plays a vital role. As a consequence, our analysis resolves the long-disputed“chain controversy”and captures the typology of vowel shifts in a natural way.