Date: Fri, 18 Oct 2002 15:44:31 +0200 From: "Vlachou, Evangelia" Subject: Syntax: Frank, Robert (2002), P hrase Structure Composition and Syntactic Dependencies
Frank, Robert (2002) Phrase Structure Composition and Syntactic Dependencies, MIT Press, xiv+308pp, hardback ISBN 0-262-06229-1, $40.00.
Reviewer: Evangelia Vlachou, University of Utrecht (UiL-OTS) and University of Paris IV.
GENERAL SYNOPSIS "Phrase Structure Composition and Syntactic Dependencies" explores the role of the Tree Adjoining Grammar (TAG) which provides operations for composing pieces of tree structures to form larger syntactic structures. Many different approaches could be taken when attempting to formalize in syntax. TAG formalism was first defined in 1975 by Joshi, Levy and Takahashi. Kroch and Joshi (1985) highlighted the importance of TAG for linguistic theory. Chomsky's (1993) paper on the Minimalist Program and the introduction of generalized transformations into the syntactic theory made TAG compatible with a more widely adopted approach. Frank opts for marrying these two approaches of transformational grammar, by presenting the TAG syntax, based on his own work growing out of Kroch and Joshi's earlier proposals. Robert Frank is a prominent scholar in the field of TAG. He is the author of "Syntactic locality and Tree Adjoining Grammar" (1992) and co-author (with Kulick and Vijay-Shanker) of "Monotonic c-command: a new perspective on tree-adjoining grammar" (2000).
Because of this marriage of all approaches presented above, Frank's book constitutes a great point of reference for every syntactician who works in any field of theoretical linguistics. It is therefore intended for readers with advanced linguistic knowledge. However, the way in which the author starts his book, explaining what "mental grammar" and "linguistic expression" are, makes it a strong candidate for being used as a supplement for graduate courses intended to provide syntax-oriented students with an overview on TAG.
CHAPTER CONTENT The book consists of six chapters: introduction ("setting the stage"), the nature of elementary trees, a case study: raising, local constraints and local economy, a case study: wh-dependencies, conclusion ("looking onward").
Chapter 1 introduces the author's main claim after a detailed analysis of Chomsky's approach on the dependencies across boundaries: "there is in fact a way to maintain the idea that the locality of movement-derived syntactic dependencies stems from the derivational independence of structural units that are bound in size". He proposes further that the most optimal way to succeed in this is to make use of the derivational machinery of TAG. He concludes that this new system is tightly related to Chomsky's system and other more recent derivational systems. Based on these assumptions, Frank provides the reader with a panel of the basic notions of TAG like "structure of TAG derivations" and the relation between "formal grammar and human grammar".
Chapter 2 is dedicated to defining which structural domain can be represented as an elementary tree. The pilot subject for doing so is the idea that basic predications and clauses form the foundation of semantically local domains. The chapter concludes with a number of well-formedness conditions on elementary trees and their effect on the TAG derivations. Chapter 3 constitutes an application of the theoretical frame presented in the previous chapter on a construction: subject-to-subject raising. Through the demonstration of possible approaches for this construction, like the conditions on the locality of movement or the well-formedness of empty categories, the author ends up arguing that the TAG view is the most optimal for this case. He gives examples from nonlocal licensing and thematic assignment, scope reconstruction, contraction, 'there'-insertion, non raising in nominal and gerunds, raising passives, raising from small clauses, raising and inversion. In his argumentation, he refers to other possible solutions given by Kayne (1984), Moro (1997) without giving an extended account of these proposals. His argumentation on these points is rather fast. Furthermore, in order to prove the "puzzling" character of Moro's account, he claims that languages like Russian and Hebrew allow predications to be expressed without the copula. He doesn't give however any example from these languages. Finally, he concludes that a distinction has to be made between two types of raising: lexical raising predicates and copula. To this aim, he refers to Hegarty's (1993) approach claiming that even in that model we should involve Adjoining. However, no detailed account is given on this work, though it seems to be very interesting for the linguist's argumentation.
Chapter 4 turns to the "there"-insertion and more precisely discusses the impossibility of "partial raising" in English like in the following example:
*There seemed a man from the CIA to be at the meeting.
Firstly, the author relates these cases with the raising analysis presented in chapter 3: the embedded subject does not receive nominative case within its infinitival nominative tree. Secondly he gives a reformulated definition of the ?-criterion (part 2) as applied to chains and proves in a convincing way that in French, English and Dutch (see examples below) this criterion would lead to two incorrect predictions: a. simple sentences involving such verbs lack subjects in the specifier of TP position and b. elementary trees headed by argumentless predicates are impossible, as the EPP and this criterion impose conflicting demands.
*(It) is raining.
*(Il)a été tiré sur le bateau. It has been fired upon the boat "The boat was fired upon"
After having pointed out the problem of such an approach, Frank restricts the application of the ?-criterion to semantically contentful DPs. In a quite quick way which makes his argumentation rather obscure, the linguist proposes that this solution is not the best one and turns quickly the discussion to the most optimal one, based on Chomsky's analysis (1995) on expletive sentences, comparing the two following phrase structures (in the form of trees):
[T' [T VP [V T']]]
[T' [T VP [DP V' [V T']]]]
His claim is the following: there is an "extended projection principle" following which "a TP projection in an elementary tree ? must have a specifier if and only if there is some otherwise licensed element within ? that can be moved to the specifier of TP" (p.117). He investigates then how these schemata can be applied to languages like Icelandic which possess a raising construction analogous to the English one. This language differs however from English in that the presence of a dative argument to the Icelandic "seem" makes raising of the lowest subject impossible. In this point Frank's argumentation becomes loose and floating using examples from English in order to reinforce the value of the extended projection principle and going back again to Icelandic in order to verify the results of the English paradigm. Basically, his aim is to show that elementary trees do not uniformly abide by the standard conception of the EPP, according to which every TP must have a specifier. Frank claims that the complexity in the statement of the EPP is only apparent and shows in a nice way that there is no need for the complex formulation of the EPP once general processes of elementary-tree-local feature checking governed by the principle of local derivational economy.
Chapter 5 turns back to another type of dependency as the one presented in the third chapter related to transformational movement: wh-movement. Frank claims that from the TAG point of view some dislocations of wh-elements are treated as the result of the Adjoining. In the beginning, he gives the basic properties of a TAG-based analysis of wh-dependencies based on Kroch (1987, 1989) and Frank (1992). He then shows how the structural hypotheses related to the elementary trees derive from properties of the system of feature checking as presented in the previous chapter. The TAG analysis of wh-movement restricts on movement in Subjacency and the Adjunct Condition. With a brief demonstration of languages that violate the principle of wh-islands, he explains it in terms of crosslinguistic differences. At the last step of his argumentation Frank sheds light to cases of grammatical extraction from weak islands and pursues a possible way to deal with them in the framework of TAG.
Chapter 6 is the final chapter, which constitutes an overview of the study and closes the book with two open, one empirical and one theoretical, questions. The empirical one concerns a class of phenomena that exhibit dependency structures that the TAG approach cannot account for and the theoretical one touches the interface of the derivations used throughout the previous chapters.
CRITICAL EVALUATION The first impression that the book gives is appealing with the tree-schemata that Frank uses which makes it easier for the reader to follow the argumentation. The titles that the author gives to every chapter reveal the deductive character of his argumentation; starting from an introduction to the elementary trees, he reinforces his argumentation by shading some light on a very interesting case where the TAG can be applied, the wh-dependencies. Finally, instead of giving to the last section of his book the stereotypical character of a "conclusion", he invites the reader to go back again, to his initial question and argue again using possible counter-arguments to the analysis he has presented so far. The author avoids extreme positions and gives a remarkably neutral account although it is made clear, from the preface, that he presents his own idea to incorporate TAG into syntactic theory. By the same token, he doesn't avoid presenting other possible theories and tries to demonstrate how and to what extend his own perspective would be the most optimal. What is very positive about the book is that it provides the readers with new areas of investigation. However, they are not extensively elaborated. For instance, he gives no account concerning the process of interpretation that applies during the derivation of an elementary tree and requires therefore reference to structure outside the limits of elementary trees. He explicitly claims that his work leaves open the problem of "derivationally anaphoric dependencies" and he doesn't attempt to give an answer. It would be appealing however to see how the linguist would deal with the semantic consequences of his approach. It would help him to support his initial claim that "the system of connections between form and meaning afforded by human language is taken to derive by mental grammar".
Chomsky, Noam. 1993. A minimalist program for linguistic theory. In Kenneth Hale and Samuel Jay Keyser, eds., The view from Building 20, 1-52. Cambridge, Mass: MIT Press.
Chomsky, Noam. 1995. Categories and transformations. In The Minimalist Program, 219-394. Cambridge, Mass.: MIT Press.
Frank, Robert. 1992. Syntactic locality and Tree Adjoining Grammar: Grammatical, acquisition, and processing perspectives. Ph. D. thesis, University of Pennsylvania.
Frank Robert, Seth Kulick, and K. Vijay-Shanker. 2000. Monotonic C-Command: A New Perspective on Tree Adjoining Grammar, Grammars, 3, 151-173.
Hegarty, Michael. 1993. Deriving clausal structure in Tree Adjoining Grammar. Manuscript, University of Pennsylvania.
Joshi, Aravind K., Leon Levy, and Masako Takahashi. 1975. Tree adjunct grammars. Journal of the Computer and System Sciences 10, 136-163.
Kayne, Richard. 1984. Connectedness and binary branching. Dordrecht: Foris.
Kroch, Antony and Aravind K. Joshi. 1985. The linguistic relevance of Tree Adjoining Grammar. Technical report MS-CS-85-16, Department of Computer and Information Sciences, University of Pennsylvania.
Kroch, Anthony. 1987. Unbounded dependencies and subjacency in a tree adjoining grammar. In Alexis Manaster-Ramer, ed., The mathematics of language, 143-172. Amsterdam: John Benjamins.
Kroch, Anthony. 1989. Asymmetries in long distance extraction in a tree adjoining grammar. In Mark Baltin and Anthony Kroch, eds., Alternative conceptions of phrase structure, 66-98. Chicago: University of Chicago Press.
Moro, Andrea. 1997. The raising predicates; Predicative noun phrases and the theory of clause structure. Cambridge: Cambridge University Press. ç
ABOUT THE REVIEWER:
ABOUT THE REVIEWER Evangelia Vlachou is a Ph. D. student in the University of Utrecht (UiL-OTS) and University of Paris IV. Her research interests are formal semantics, syntax, polarity and free-choice indefinites in Greek and French. á