Government-Binding Framework Research Articles

Making Sense of Corpus Data

This paper demonstrates the essential role of corpus data in the development of a theory that explains and predicts word behavior. We make this point through a case study of verbs of sound, drawing our evidence primarily from the British National Corpus. We begin by considering pretheoretic notions of the verbs of sound as presented in corpus-based dictionaries and then contrast them with the predictions made by a theory of syntax, as represented by Chomsky's Government-Binding framework. We identify and classify the transitive uses of sixteen representative verbs of sound found in the corpus data. Finally, we consider what a linguistic account with both syntactic and lexical semantic components has to offer as an explanation of observed differences in the behavior of the sample verbs.

Interaction between linguistic theory and language processing in SLA

This article examines L2 performance in three areas of French morphosyntax by English L1 learners. More particularly, it examines how coindexation as defined within the government-binding framework develops in the L2 grammar. Empirical studies relating the development of two areas of French grammar by English L1 speakers are presented. L2 performance on information questions involving qui and que in which learners have to link the wh-phrase and its trace in order to establish the syntactic function of the wh-phrase in the sentence is examined, as well as performance on the morphological phenomenon of noun-adjective agreement in French where learners have to transmit agreement features from a noun to an adjective which it governs. In both cases, learners are found to increase gradually the structural domain in which they are able to operate as their level of competence in the L2 improves, suggesting that they are faced with a parsing problem when coindexing elements in a sentence. These findings are related to a study of the acquisition of restrictive relative clauses in French L2 by English learners (Hawkins, 1989), and then discussed in the light of the current debate in SLA research about the roles played by linguistic theory, on the one hand, and language processing mechanisms on the other.

An essay on grammar-parser relations By Jan van de Koot (review)

An essay on grammar-parser relations By Jan van de Koot (review)

Focus and conditions on representation

Mainly dealing with preposition stranding cases that have been treated in Takami (1988) in terms of a functional concept, I propose a syntactic or at least formalized approach to them by taking into consideration such thematic notions as ‘+affected’ and the notion of ‘focus’. Couching my analysis in the Government-Binding framework, I show the necessity of recognizing that there are three major ways of focus formation: phonological, syntactic, and thematic ones.

Syntactic case and morphological case in the history of English (review)

146LANGUAGE, VOLUME 66, NUMBER 1 (1990) against the facts of Quechua. Even specialists in the Government-Binding approach will probably find some of the authors' allusions obscure. Finally, I found the book somewhat difficult to read. The overall structure of the argument is logical, but it can be hard to follow because of the amount of detail and because of the authors' frequent use of forward and backward references. The authors' language is abstract and vague in places—I am still not sure exactly what a 'Comp-like Case Position' is, for example. The book contains many charts, lists, tables, and diagrams; some are very helpful summaries , but others would be more useful with fuller explanation. There is also an index of names and a short subject index at the end. The book is rather sloppily printed: most of the mistakes are merely distracting (e.g., the first projection of N sometimes shows up as N' and sometimes as N'), but some interfere with comprehension (e.g., the branches of the trees do not always match up accurately with their terminals). In spite of the book's stylistic difficulties and the fact that the analysis it presents is not always completely satisfying (whose is?), the patient reader will be rewarded with a wealth of data, interesting descriptive generalizations, good examples of how to apply formal categories in an 'exotic language', and many suggestive theoretical ideas. REFERENCES Baker, Mark. 1985. Syntactic affixation and English gerunds. Proceedings of the West Coast Conference on Formal Linguistics 4, ed. by J. Goldberg, S. MacKaye and M. Wescoat, 1-11. Stanford: Stanford Linguistics Association. Chomsky, Noam. 1981. Lectures on government and binding. Dordrecht: Foris. Emonds, Joseph. 1988. The autonomy of the (syntactic) lexicon and syntax: Insertion conditions for derivational and inflectional morphemes. Seattle: University of Washington, ms. Marantz, Alec. 1979. Assessing the X' convention: Embedded sentences in English. Cambridge, MA: MIT, ms. Milsark, G. L. 1988. Singl-z'ng. Linguistic Inquiry 19.611-34. Reuland, Eric. 1983. Governing -ing. Linguistic Inquiry 14.101-36. Department of Linguistics[Received 9 August 1989.] McGiIl University 1001 Sherbrooke Street West Montreal PQ H3A 1G5 Canada Syntactic case and morphological case in the history of English. By Ans van Kemenade. Dordrecht: Foris, 1987. Pp. xi, 249. $23.90. Reviewed by Cynthia L. Allen, Australian National University This doctoral dissertation is an investigation, in a Government-Binding framework, into the workings of the abstract case-marking system in Old English (OE), and how changes in this abstract system in Middle English (ME) resulted in surface changes. The author argues that superficially unrelated facts, synchronic and diachronic, are in fact related when this abstract system is properly understood. REVIEWS147 Ch. 1 is a general introduction and review of the literature. In Ch. 2, Kemenade argues for an analysis of basic OE syntax similar to analyses which have been proposed in a GB framework for Dutch and German. Specifically, it is assumed that Old English had an underlying word order—SOV—which is obscured by the movement of the verb into second position in main clauses, and also by a fairly free process of extraposition in all clause types. The 'verbsecond ' phenomenon is explained in terms of lexicalization of infl. Infl must be lexicalized; in subordinate clauses this is accomplished by the use of a complementizer, and the verb remains in final position. But in main clauses, if infl is to be lexicalized, the verb must move into infl, which is in second position. However, K argues, it is not in S but in S'. Comp is not filled by a complementizer, but, in line with widely-accepted GB analyses of other Germanic languages, it is the position of topicalized items, including the subject in SVO sentences. This initial Topic position is not available in subordinate clauses, although K never explains this fact. In this chapter K presents an analysis in which both this topic position and infl are dominated by comp, but in Ch. 3 this structure is modified as follows (note: infl" = S'): (1)infl" COMPINFL ( = spec infl') infl Thus, comp is now solely occupied by a Topic position, and is the specifier of infl'. In Ch. 3 K considers...

Clitic movement and the ECP: Evidence from berber and romance languages

This paper deals with the phenomenon of clitic-placement within the Government-Binding framework as formulated in Chomsky (1981, 1986b) and references cited therein. The basic arguments are, first, that clitics are head categories in the sense of X-bar theory, and, second, that clitic-placement is a syntactic process of head-movement in the sense of Baker (1985b) and Chomsky (1986b). It is argued, furthermore, that clitic-movement inside the same clause or from an embedded clause to a matrix clause operates directly, thereby violating the restriction on head-movement processes expressed by the HMC. In view of the demonstrated fact that despite the long nature of the movement no violation of the ECP arises, it is concluded that the HMC is basically wrong and therefore should be dispensed with.

On the German infinitival marker zu and the analysis of raising constructions

This paper argues, within the Government-Binding framework, that the German infinitive marker zu is a complementizer morpheme, not an INFL morpheme, as is standardly assumed; and that subject-Raising is possible in the presence of a complementizer, again contrary to the standard view. Treating zu as a complementizer allows for a simple and unified account, not previously available, for the absence of interrogative infinitives, of accusative subjects, and of verb-second effects in infinitives containing zu. The existence of subject-Raising in constructions with zu-infinitives creates a problem insofar as subject-Raising entails S'-deletion — i.e. the absence of a complementizer — according to the standard version of the theory. Since, however, by contrast with other languages, there is no independent evidence for S'-deletion in the relevant constructions of German, it is argued here that the relevant part of the theory — the Empty Category Principle — should be revised so as to permit Raising without S'-deletion in just the German case. The second part of the paper makes a specific proposal to this effect within the terms of Chomsky's (1986) ‘Barriers’ framework.

Passives in modern greek

Descriptive and pedagogical grammars of Modern Demotic Greek distinguish at least five different uses of the passive construction. In this article these constructions are analyzed assuming the Government-Binding framework of Chomsky (1981, 1986). Some of these passive construction seem to be in violation of the Uniformity Principle and seem an exception to the analysis for Passives proposed in Chomsky (1981) and in Jaeggli (1986). The analysis proposed in this article distinguishes two kinds of passives and shows the exceptions to be apparent and all of the different constructions noted in the descriptive and pedagogical grammars are explained in terms of these two passive constructions.

Prepositional phrases as subjects and objects

The positions of subject and object in simple active sentences, and object of a preposition are normally filled by NP's but they can also be filled by PP's. There are reasons for analyzing such PP's as just PP's and not as PP's embedded in NP's. Yet such an analysis appears to be problematic for the analysis of raising and passive sentences assumed within the government-binding (GB) framework, which is formulated with exclusive reference to NP's. Assuming that the basic mechanisms at work in sentences with PP's in typical NP positions are identical to those in sentences with NP's, PP's must sometimes require Case to avoid a violation of the Case Filter. It is possible to incorporate this idea into the GB framework given a particular view of Case and Case assignment, and a revised Case Filter. The analysis receives some further support from a range of related data.

Why Irish raising is not anomalous

M (1984) documents persuasively the existence of Irish raising constructions which, viewed from an ordinary constituent structure or transformational perspective, seem unique and bizarre and which, viewed from certain particular theoretical perspective, e.g., that of the Government-Binding framework, seem anomalous. The present remarks have aimed to show that viewed in relational terms, Irish raising constructions can unproblematically be taken to manifest a type of raising identical to that of better known languages like English and French. The relevant differences between Irish and English or between Irish and French have to do in these terms not with raising per se but with the interaction of the raising structure with other logically independent grammatical phenomena. In particular, it has been suggested that Irish raising structures manifest the same sort of dummy structure found in French examples like (18b), as represented in (19), with the chief difference being that the dummy arc determines that the earlier 2 demotes to 3 in Irish, rather than to 8, as in French. Finally, the relevant dummy 1 arcs in Irish self-erase, so that the Irish construction contains no phonologically visible dummies, while those in French in general do not and manifest as il. Looked at in these terms, the apparently unique Irish raising phenomenon is a combination of independently recognized elements. From the APG perspective, Irish raising constructions satisfy all relevant relational grammatical principles, e.g., the Relational Succession Law, the Final 1 Arc Law, the Stratal Uniqueness Law, the Demotion Ban, the Ghost Arc Law, and other conditions on arcs whose heads are dummy nominals; for the latter, see Johnson and Postal (1980, Chapter 9). Hence, while relatively superficial assumptions suggest that Irish raising requires broadening the class of phenomena allowed by grammatical theory (to permit raising to object of a preposition), internal to the relational framework sketched here, Irish raising fits into the narrowest constraints on natural language grammatical structure independently permitted by what (little) is so far known about languages other than Irish. This provides an argument for adopting the sort of perspective capable of reducing the apparent anomaly to the interaction of independently known grammatical features. A good deal of the force of the argument depends, however, on providing a similar analysis of the Modern Greek case mentioned in footnote 1. Unless some way can be found to reduce the apparent raising of 1s to object of a preposition in Greek to more widely documented phenomena, there is little reason to reject the idea that Irish might manifest such raising. Unfortunately, an analysis of the Greek situation reducing it to better-known structural features seems far less apparent than was the case for Irish and no one seems to have even tried to develop such an account to this point.