This book presents a new theory of grammatical categories - the Universal Spine Hypothesis - and reinforces generative notions of Universal Grammar while accommodating insights from linguistic typology.
Date: Fri, 3 Oct 2003 13:31:16 -0400 (EDT) From: Edit Jakab <email@example.com> Subject: The Syntax of Hungarian
É. Kiss, Katalin (2002) The Syntax of Hungarian, Cambridge University Press.
Edit Jakab, Asymmetry Project, Department of Linguistics, Université du Québec à Montréal.
É. Kiss' work is a comprehensive and knowledgeable presentation and analysis of Hungarian sentence structure as well as its extensive literature. Although the ideas in the book are built on the theoretical and methodological assumptions of generative grammar, the approach is primarily empirical. The author's goal is to describe the theoretically relevant facts of Hungarian syntax rather than to provide the latest technical accounts for them.
In the Introduction, the author describes the general structure of the Hungarian sentence which is a hierarchical structure dividing into a topic part and a predicate part. The former contains the logical subject of the predication (the notion of logical subject is independent of the function 'grammatical subject'), while the latter contains a verb-initial propositional center and the preverbal operator positions: Negation-Distributive Quantifier-Focus. É. Kiss emphasizes that this structure was already attributed to the Hungarian sentence by Brassai in the middle of the 19th century.
Chapter 2 is devoted to the discussion of the topic position and related issues. The author argues that the Hungarian sentence has two parts: the logical subject of predication, called 'topic', and the logical predicate. It is important that the topic is not necessarily linked to the function of subject (or any particular grammatical function) despite the subject's tendency to be a topic. The topic position is generally filled by elements bearing a [+human] feature, consequently, [+human] accusative or oblique complements can often be topics. Given that the topic denotes an individual present in the universe of discourse, it must be a referring expression bearing the features [+referential] and [+specific], and can not be an argument containing an operator feature (e.g., a quantifier). A non-referential phrase such as the latter may assume these features and thus be topicalized only if it is contrasted.
The topic, occupying the specifier position of TopP, is an argument of the verb binding an argument position in the predicate part. Therefore although topic movement is A-bar movement, it is not operator movement. When the sentence contains more than one topic, the author assumes that the TopP is iterated. As É. Kiss herself points out, it is theoretically open to question what triggers the movement of more than one element to multiple specifier of TopP positions.
She distinguishes eventive (containing a stage-level predicate) and stative (containing an individual-level predicate) sentences with an invisible or visible TopP, respectively. The former have an event variable which licenses a situationally or contextually bound empty spatio-temporal element in the specifier of TopP. The latter (stative predicates), on the other hand, must contain overt material in their Spec-TopP position since they lack an event variable. However, both sentence types may be, but not necessarily are, topicless if they contain a logical operator.
In the last paragraph of this chapter, É. Kiss discusses contrastive topic. She argues that non-referential and non-specific elements such as bare nouns and quantifiers can also be topicalized, but only if they are accompanied by a contrastive intonation. She shows that these elements indeed occupy the Spec-TopP position.
The subject of Chapter 3 is the description of the minimal predicate, in particular the flat VP structure. The author raises the question of whether assuming such a structure is compatible with current theoretical assumptions. She presents the structure of the minimal predicate consisting of a VP and an AspP. Within the VP, V is obligatorily the first element after which the arguments follow in an arbitrary order. The arguments within VP, including the subject, can be only referential. On the basis of the lack of any superiority effect in wh-movement, she argues that there is no VP-external position; it is allotted to the topic. Several pieces of evidence are enumerated for a flat VP structure: the predominant lack of subject-object asymmetries in the areas of binding, in particular Binding Condition C (referring expression) and Weak Crossover. However, anaphoric binding, which seems to be thematically determined in Hungarian, suggests a hierarchical structure.
The author goes on introducing the functional projections that extend the VP: modality, tense, mood, object and subject agreement. She discusses object agreement, which is marked by a specific suffix on finite verbs taking a definite object. She argues that AgrOP is not the locus of accusative case checking since accusative assignment is not connected to an invariant Spec-AgrOP position. One piece of evidence comes from the fact that infinitives do not have an object agreement suffix, yet they are able to take an accusative direct object.
The chapter ends with the discussion of what Hungarian linguistics traditionally calls a prefix: a verbal modifier (VM). She argues that VM must be treated as an independent syntactic unit, a lexically selected complement, and she enumerates several reasons for this, the most important of which is the fact that VM can be separated from the verb by a particle or several particles, and it can raise to a superordinate clause. Although there are several properties that indicate the phrasal status of [the] VM, such as its ability to move across a CP boundary, to undergo topicalization and focussing or to be an elliptic sentence in itself, there is also evidence against this status in that it cannot, for example, be coordinated with another VP. Therefore, É. Kiss suggests a mixed status for VM in which it would be able to function both as a phrase and a head: it moves to Spec-AspP as a phrase after which it head-moves to V which has already raised to Asp.
In Chapter 4, the author scrutinizes the question of focus and related issues. She argues for an FP projection which immediately precedes the extended VP, thereby achieving adjacency of focus and verb. She shows that the main role of the focus is to identify the subset of a set of individuals for which the predicate holds. She also enumerates other elements, e.g., interrogative pronouns, negative adverbs of degree, frequency, and manner, negative existential quantifiers etc., whose inherent [+focus] feature makes them raise to Spec-FP. The author further demonstrates the lack of the Superiority Condition among wh- phrases in Hungarian. She also shows that multiple questions requesting the identification of two individuals and those requesting a pair-list response have different structures.
Chapter 5 examines quantification. The author discusses what positions are allotted to quantifiers and she also describes their scope properties. Accordingly, universal and positive existential quantifiers are placed in the specifier of a functional projection, DistP. Here, the quantifier assumes a distributive interpretation as well as scope over the domain it c-commands. She points out that all preverbal operator positions are scope positions. She also examines noun phrase internal quantifiers whose extraction into the verb phrase allows them to have sentential scope. Finally, she argues that certain adverbs of frequency exhibit similar syntactic properties as quantifiers.
In Chapter 6, negation is under scrutiny. The author shows that both the verb and the focus may be negated, i.e., preceded by the negative particle projecting a NegP. The negation of the universal quantifier, however, produces constituent negation.
É. Kiss describes negative concord in Hungarian by demonstrating how the negative Hungarian se-pronouns operate. Their interpretation is discussed as well: a se-pronoun receives an existential reading when it has a [-specific] feature, whereas it is universally interpreted when it has a [+specific] feature (e.g. when it precedes the focus). She also gives an account of indefinite pronouns in the scope of negation and shows their quantifier status.
The focus of Chapter 7 is the noun phrase. On the basis of their distributional differences in the sentence, three noun phrase projections are to be distinguished: the lexical noun phrase proper surrounded by operators, the indefinite numeral phrase and the definite determiner phrase.
A substantial part of the chapter is devoted to the discussion of possessive constructions. It explains how the possessive relation is realized morphologically as well as syntactically, and discusses agreement and "anti-agreement" within possessive constructions. The author also discusses the argument structure of different nominals; deverbal event nominals can keep only one argument realized as a possessor. On the other hand, the possessor expresses the object of a nominalized transitive verb, and the subject may be realized as a by- phrase. Lexically case-marked arguments and adjuncts either appear as modifiers or may be extraposed.
In Chapter 8, the postpositional phrase is discussed. Postpositions in Hungarian extend the noun phrase in a suffix-like manner, similar to case endings functioning semi-independently. It is interesting that, historically, case endings have developed from postpositions, and still exhibit postpositional characteristics in certain instances. They form a unit with the noun at the morphosyntactic level; however, morphophonologically, they are to be considered two separate entities.
When the postposition bears no suffix, it takes its complement in the dative assigned by a "possessedness" morpheme. On the other hand, the postposition is inflected when its complement is a personal pronoun.
Chapter 9 describes existing assumptions and theories about non-finite and semi-finite verb phrases. It discusses infinitival and participial phrases that have the structure of an extended VP. The author points out the salient difference between finite and non-finite verb phrases: the latter does not need to be a VP before focus or negation, and so, the verb may be preceded by a verbal modifier in AspP. She shows at length how agreeing/inflected infinitives operate in Hungarian, their distribution as well as their dative-marked subject.
She also points out how one type of adverbial participle may also take a lexical subject, which is nominative-marked. She provides proposals about the possible sources of the specific case of the subjects of these non-finite constructions.
The function of the various non-finite categories is addressed as well: the infinitive acts as a complement of the matrix predicate, the adverbial participle is an adjunct, whereas the adjectival participle is an attribute. The former two joined by an auxiliary or semi- auxiliary verb may constitute a complex predicate as well.
Chapter 10 concludes the book with the formal treatment of movement across boundaries and, thus, subordination as well. It talks about how subordinate clauses are structurally identical to matrix clauses except for the presence of a complementizer/subordinating operator in the former. When embedded clauses constitute arguments, they are associated with a pronoun which bears the case assigned by the matrix predicate to the argument, in this case the embedded clause. This pronoun also, in some sense, takes the place of the subordinate clause in the matrix operator positions that are unable to take clausal complements.
Embedded interrogative clauses are introduced by the complementizer /hogy/ 'that' occupying C, whereas the wh-word itself is in the specifier of the focus phrase, thus it does not create an island for extraction. When a subordinate interrogative operator is associated with a matrix interrogative pronoun, it may take matrix scope.
She also shows that complex sentences containing a bridge verb as a matrix may represent a single domain for operator movement. Long operator movement, complementizer deletion and the licensing of parasitic gaps are discussed as well.
É. Kiss' book is a thorough and informative study as well as an overview of Hungarian syntax, a work that reveals the author's vast knowledge of the subject from both the traditional and the generative perspective. In her analyses, she relies on books, articles, talks as well as papers on Hungarian from all sources including her students, teachers and colleagues from all over the world. Consequently, her bibliography is extensive and well-informed.
She provides various proposals for all aspects of Hungarian syntax, and then she amends them, so to speak. She uses comparison with English throughout the book. For example, she draws parallel between Hungarian and English subject-object asymmetries in Chapter 3, or between Hungarian thematic hierarchy and English dative shift, which, if thematically motivated, supports the former.
Lack of space allows me to concentrate on only a few issues for discussion. In chapter 4, on the basis of the free postverbal order of adverbs, the author argues against Brody's (1995) V-to-F movement in sentences expressing identificational focus (p. 86). It seems to me that the elimination of an adverb's possible occurrence between the focus and the verb by just simply stating the Focus-V adjacency requirement does not appear to be sufficient. This requirement does not explain how Focus and V end up next to each other.
Another issue to select is the interesting phenomenon of inflected infinitives. On the basis of Toth's (2000) thesis, the author claims that in the case of an agreeing infinitive with a dative/ablative case- marked subject it is actually the lack of agreement between the controller of its subject and the matrix verb that allows for the agreement. Landau (2003) demonstrates that the absence of matrix agreement is only a necessary but not a sufficient condition for licensing inflected infinitives. He argues that the other relevant factor is the Tense of the infinitive, namely, its lack of Tense. Since agreeing infinitives are untensed, they are not able to define a tense domain distinct from the matrix. On the other hand, tense mismatch is marginally possible when the infinitive is not inflected, in other words, when it is tensed (cf. Jakab 2003 which connects the agreement and tense features of Hungarian infinitives to their restructuring properties).
Finally, I should mention that I found a minor but rather confusing mix-up in the numbers of examples on page 18, 3-4 lines below example (19): "In Kratzer's approach the crucial difference between the stage- level (19a,b) and the individual-level (12)-(13)..." -- should be stage-level (12)-(13) and the individual-level (19a,b).
All in all, É. Kiss' book is an excellent source for both non-Hungarian linguists who will find various analyses of all syntactic issues in Hungarian, and for Hungarians whom it enables to brush up on or extend their knowledge of Hungarian sentence structure.
Brody, M. 1995. Focus and checking theory, in I. Kenesei, ed., Approaches to Hungarian 5: 29-44. Szeged: JATE.
Jakab, É. 2003. Agreement, Tense, Restructuring. The Licensing of Hungarian Inflected Infinitives. Paper presented at CLA, Halifax.
Landau, I. 2003 (To appear). The Scale of Finiteness and the Calculus of Control. NLLT.
Toth, I. 2000. Inflected Infinitives in Hungarian. Doctoral dissertation. Tilburg University.
ABOUT THE REVIEWER:
ABOUT THE REVIEWER Edit Jakab did her undergraduate studies at Lajos Kossuth University in Debrecen, Hungary, her graduate work at Indiana as well as Princeton, and received a Ph.D. in theoretical and Slavic linguistics from Princeton University. Her interests include comparative syntax, syntax/semantics interface, the clausal structure of nonfinite constructions, in particular, (Non-)Agreement, Tense, Mood and Modality in Finno-Ugric and Slavic languages as well as in English. She is currently a Postdoctoral Research Fellow in linguistics in the Asymmetry Project at UQAM.