LINGUIST List 12.492

Thu Feb 22 2001

Review: Levine & Green, Phrase Structure Grammar

Editor for this issue: Terence Langendoen <terrylinguistlist.org>


What follows is another discussion note contributed to our Book Discussion Forum. We expect these discussions to be informal and interactive; and the author of the book discussed is cordially invited to join in.

If you are interested in leading a book discussion, look for books announced on LINGUIST as "available for discussion." (This means that the publisher has sent us a review copy.) Then contact Simin Karimi at siminlinguistlist.org or Terry Langendoen at terrylinguistlist.org.


Directory

  • Tania Avgustinova, Review of Levine and Green (Eds.) Studies in Contemporary Phrase Structure Grammar

    Message 1: Review of Levine and Green (Eds.) Studies in Contemporary Phrase Structure Grammar

    Date: Fri, 16 Feb 2001 15:16:13 +0100
    From: Tania Avgustinova <taniaCoLi.Uni-SB.DE>
    Subject: Review of Levine and Green (Eds.) Studies in Contemporary Phrase Structure Grammar


    Levine, Robert D., and Georgia M. Green, Eds. (1999) Studies in Contemporary Phrase Structure Grammar, Cambridge University Press, 335 pages.

    Tania Avgustinova, Saarland University

    This book contains a collection of articles on recent developments in the framework of head-driven phrase structure grammar (HPSG), all resulting from a closer investigation of a variety of non-trivial phenomena found in languages like German, Japanese, and English. HPSG - [Pollard and Sag 1987; Pollard and Sag 1994; Sag and Wasow 1999] - is a constraint-based lexicalist theory that invokes only a single level of syntactic representation. A crucial property of HPSG is the explicit assignment of linguistic objects to membership in a hierarchically organised network of types, where constraints associated with any given type are inherited by all of its subtypes. On the basis of detailed empirical analyses, the contributors suggest ways of extending and revising the current framework. Nowadays HPSG is a major theory of syntactic representation which is becoming increasingly dominant, particularly in the domain of natural language computation, cf. [Flickinger et al. 2000]. This collection of articles can be regarded as a representative reflection of the process of continuous rethinking that the theory has undergone since the appearance of [Pollard and Sag 1994].

    The book is organised as follows.

    The introductory chapter (pp.1-38) written by the editors offers, on the one hand, a basic introduction to some of the leading concepts of the classical HPSG theory and, on the other hand, outlines the enrichments, extensions, and revisions to this theory made by each of the contributions in the volume.

    Chapter 1 (pp. 39-79) by Manning, Sag and Iida considers the lexical integrity of Japanese causatives.

    Chapter 2 (pp. 80-118) by Johnston proposes a syntax and semantics for (English) purposive adjuncts.

    Chapter 3 (pp. 119-160) by Gungi is on lexicalist treatment of Japanese causatives.

    Chapter 4 (pp. 161-198) by Baker looks at "modal flip" and partial verb phrase fronting in German.

    Chapter 5 (pp. 199-222) by Fukushima presents Japanese syntactic topicalisation from a lexicalist perspective.

    Chapter 6 (pp. 223-274) by Kathol considers the problems of agreement in languages with richer morphology than English.

    Chapter 7 (pp. 275-332) by Hinrichs and Nakazawa presents an analysis of partial VP and split NP topicalisation in German.

    Finally, there is an index (pp. 333-335). The references are listed immediately after each contribution.

    The research presented in this volume focuses on four major topics.

    I. The morphology / syntax interface with respect to agreement phenomena (Chapter 6). As Kathol argues, the development of syntax-morphology interface should be driven by the analysis of more complex morphological systems than English. In particular, they could provide the foundations for a more comprehensive theory of agreement than the one developed in [Pollard and Sag 1994]. The discussion involves data from French, German, Latin, Swahili, Spanish, Italian, Russian, Serbo-Croatian, Korean, and, of course, English. Genuine agreement phenomena are analysed not in terms of selection, but rather as a co-variation or matching between relevant pieces of information available at the agreeing elements. In effect, this proposal treats agreement explicitly as a phenomenon that involves merging of information contributed by various sources in the sentence.

    II. The syntax / semantics interface with respect to purposive adjuncts in English (Chapter 2). The analysis is a nice illustration of how the parallel representation of linguistic information in HPSG linguistic expressions supports and facilitates modelling the syntax - semantics interaction. A crucial theoretical role is played by the type hierarchy which assigns them to appropriate ontological statuses bearing on their grammatical behaviour. Johnston identifies three separate interpretations of English "for" PP adjuncts, namely, "recipient", "benefactive", and "acquire". All of them introduce the purpose of the parameterised state-of-affairs described by the clause they modify. A number of striking parallels between these "for" PPs and infinitival purposives are systematically considered, enriching thus the overall picture.

    III. Alternative accounts of basic constructions in Japanese, with special attention to causatives (Chapters 1 and 3) and topicalisation (Chapter 5). Agreeing on a substantial number of points, these contributions explore different theoretical technologies to capture the range of facts adduced.

    There exists a consensus that the treatment of Japanese causative constructions entails some kind of dual structure. The two approaches presented in the volume start from different premises with respect to where the crucial mismatch among the linguistic representational components is located.

    Manning et al. sketch a strictly lexical theory of Japanese causatives that makes the phonology, morphology, and syntax parallel, while a mismatch occurs with the semantics. Japanese causatives are treated as single verbal forms with complex morphological structure. Crucially, they are associated with hierarchical ("nested") lexical argument structures, which are acquired derivationally (i.e. by means of derivational types), and the resulting constituent structure is clearly mono-clausal. In contrast, Gunji proposes a different kind of lexical approach in which a kind of bi-clausal (VP-embedding) constituent structure is mapped onto a string in which elements belonging to different syntactic constituents form units with respect to morphological / phonological rules. In other words, the duality-relevant mismatch is located between morpho/phonology and constituent (syntactic and semantic) structure, i.e. between a "phenogrammatical" and a "tectogrammatical" levels of representation.

    Fukushima's approach to topicalisation employs topic- substitution and topic-addition lexical rules to the effect that the topic is interpreted as an element of the argument structure. The author thoroughly examines empirical and theoretical consequences of such a lexical account of what is traditionally regarded as a syntactic phenomenon.

    IV. Argument structure and constituency in German (Chapters 4 and 7). The two contributions on this topic share a number of common perspectives on the analysis of German clausal structure as well as the nature of the lexical entries for auxiliaries. They focus on a complex of interrelated phenomena involving the topicalisation of incomplete verb phrases and the multiple ordering possibilities of sequences of verbs with respect to certain governing auxiliaries. Both analyses take advantage of the expressive power of lexical rules.

    Aiming at integrating several analyses of German verbal phenomena - verb second (V2), modal flip (MF), and partial verb phrase (PVP) fronting - Baker assumes a common phrase structure for PVPs and MF contexts. The PVPs are interpreted as unsaturated constituents whose valence is related - in a complex and non-trivial way - to the valence of the head and its complements. In contrast, Hinrics and Nakazawa, who build on their previous work in argument composition, treat the fronted PVP as a completely saturated constituent, with the missing material being extracted (hence, recorded in the SLASH specification of the V head). They also show how the same type of lexical approach can account for the so-called split NP topicalisation. What PVP and split NP topicalisation have in common, on their approach, is that the material missing from the topic appears as a complement in the non- topicalised position.

    Summing up: most of the issues explored in this book are central to the HPSG theory and tackle very specific problems in considerable depth. The style of the contributions is primarily intended for specialists, but the collection reflects a high quality linguistic research which should (for the most part) be accessible to advanced students trained in HPSG. The concise and well-organised overview provided by the editors is very helpful in navigating through the volume.

    REFERENCES

    Flickinger, D., S. Oepen, J. Tsujii and H. Uszkoreit, Eds. (2000). Journal of Natural Language Engineering. Special Issue on Efficient Processing with HPSG. Cambridge, Cambridge University Press.

    Pollard, C. and I. Sag (1987). Information-Based Syntax and Semantics. Stanford, California, Center for the Study of Language and Information, Stanford University.

    Pollard, C. and I. Sag (1994). Head-Driven Phrase Structure Grammar. Chicago, University of Chicago Press.

    Sag, I. and T. Wasow (1999). Syntactic theory: a formal introduction. Stanford, CSLI.

    ******* Dr. Tania Avgustinova Computational Linguistics, University of Saarland Postfach 151150, 66041 Saarbruecken, Germany taniacoli.uni-sb.de http://www.coli.uni-sb.de/~tania/