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.
Gutierrez-Rexach, Javier, and Luis Silva-Villar, ed. (2001) Current Issues in Spanish Syntax and Semantics. Mouton de Gruyter, vi+254pp, hardback ISBN 3-11-016929-0, Studies in Generative Grammar 53 Announced at http://www.linguistlist.org/issues/12/12-2318.html#2
Miguel Rodriguez-Mondonedo, Department of Spanish and Portuguese, The University of Arizona
This book contains a collection of thirteen papers that present a wonderful overview of current linguistic theory, which is used to analyze some old problems in Spanish Grammar. Writing from a minimalist point of view, the papers deal with some features of clitics and various aspects of the structure and interpretation of Determiner Phrase (DP) and Verbal Phrase (VP). It has to be noted that, despite of the fact that they display a rigorous argumentation, all articles elucidate their contents in a very clear way. It allows considering this book not only a worth- reading update of traditional analysis, but also a useful introduction to some complex issues in Spanish Linguistics.
The book has three parts. Part 1 discusses the structure and interpretation of DP. It has three papers.
The first article, by Ignacio Bosque, is ''Adjective Position and the Interpretation of Indefinites''. Bosque introduces the idea that prenominal attributive adjectives (1b) and elatives (1c) in Spanish behave in the same way with respect to (non)specificity. For instance:
(1) a. Las cinco muchachas hab'ian conocido a un actor FAMOSO 'The five girls had met a [specific/non-specific] famous actor' b. Las cinco muchachas hab'ian conocido a un FAMOSO actor 'The five girls had met a [specific] famous actor' c. Las cinco muchachas hab'ian conocido a un actor FAMOSISIMO 'The five girls had met a [specific] extremely famous actor'
Sentence (1a) is ambiguous because each girl could have met a different famous actor (non-specific) or all of them could have met the same person (specific); but sentences (1b) and (1c) have only specific interpretation. The same contrast appears when different tests of (non) specificity are used: generic operators, quantificational adverbs, indefinite nominals in directive context, negation, subjunctive relative clauses, inter alia.
To explain these facts, Bosque assumes that prenominal attributive adjectives (i) move overtly to Specifier of Degree Phrase (DegP), and then, (ii) covertly, to Specifier of some Functional Projection (FP) above DP (''-i'' is my mark for indexes):
(2) OPERATOR.[FP (Famoso)-i [D un [DegP famoso-i [Deg] [AP* actor [AP (famoso)-i ] ] ] ]
In this way, a specific interpretation (with ''variable'' reading) is avoided, since FP blocks binding between OPERATOR and the variable provided by the indefinite.
The next paper is ''Issues in the Syntax of DP in Romance and Germanic'', by Enrique Mallen. The author observes that derived nominals in Spanish and German usually realize their arguments (as well as possessors) in a postnominal position (3a), however, genitive proper nouns (arguments and possessors) can occur prenominally in German but not in Spanish (3b). In the other hand, attributive adjectives must be prenominal in German, but they are often postnominal in Spanish (3c).
(3) a'. La descripciï¿½n del ladrï¿½n (Spanish) a''. Die Beschreibung des Diebes (German) 'the description of-the thief' b'. *De Bokan el ataque fue inexcusable (Spanish) b''. Bokans Angriff war unverzeihlich (German) 'Bokan attack was inexcusable' c'. El 'exito arrollador de la abuelita c''. Omas unubertreffliche Erfolge 'granny's superlative success'
To account for these and other similar Spanish/German contrasts, Mallen assumes that attributive adjectives (AP) are generated in [Spec, AgrP] in both languages, and then, in German, AP raises to Spec of KP (Case Phrase), because a [+strong] case feature---which will be [-strong] in Spanish. It assumes a DP structure like (4), where Agr goes upstairs cyclically and NP raises to [Spec, PredP] (Sanchez 1996).
It is possible to make the same claim regarding arguments inside DP, assuming more complex Agr projections, to explain distribution of Theme and Agent.
''The Syntax and Semantics of Preverbal Topical Phrases in Spanish'', by Eugenia Casielles, addresses the features of topical phrases in Spanish. Contreras 1991, Olarrea 1996, among others have proposed that preverbal subjects in Spanish are always adjuncts and they do not use any specifier position. Casielles challenges this idea observing that Spanish does not allow bare nouns (BN) as preverbal subjects (5a) but it does allow them as preverbal topics (5b):
(5) a. *Niï¿½os jugaban en el parque 'Children were playing in the park' b. Libros hay en la biblioteca 'Books there-are in the library'
In addition, since BNs cannot rise (left-dislocated BNs do not involve movement), they have to stay inside VP and cannot get a generic interpretation (only an existential one). Therefore, there must be two different positions for subjects (6a) and dislocated phrases (6b).
(6) a. [IP SPECIFIER [I' I [VP ] ] b. [IP XP [IP XP [IP XP [IP ] ] ] ]
However, subjects can be dislocated, as in:
(7) [A 'el] [su madre] [el coche] no se lo dejar'ia nunca [to him] [his mother] [the car] not would-lend never 'His mother would NEVER let him borrow the car'
This suggest that (5a) could be grammatical if [Niï¿½os] is topical. Casielles tells that this is not true, and proposes to use the distinction between two points of view: Topic- Comment and Focus-Background (basically, the idea that topic is not equal to non-focus). According to her, BNs cannot be Topics, but Backgrounds. In addition, (7) has no topic, because the ''appearance of a dislocated elements means that the sentence has a Background-focus structure'' (p.75). Then, we can conclude that (5a) will be grammatical if the sentence has a Focus-Background Structure.
Part 2 discusses the distribution of clitics. It has five papers.
''On the Doubling of Overt Operators'', by Jon Franco, addresses the fact that, in Basque Spanish and other Spanish dialects, clitic-doubling of wh-words and quantifiers is possible, as in:
(8) ?A quiï¿½n-i le-i viste? to whom ACC-CL3sg saw-2sg Who did you see?
However, (8) is possible only if the wh-word is linked with an element inferred from the discourse. Franco argues that a clitic can double only presupposed entities. To encode presuppositionality without using an ad hoc feature, the author assumes that these structures have a referential pro, which, in some dialects, can be licensed by discourse:
(9) ?[ A qui'enes pro-i ] les-i deportarï¿½n? to which-pl ACC-CL-3-PL deport-fut-3-pl Which ones of them will be deported?
To obtain a discourse referent from presuppositions, Franco uses Rizzi's 1997 structure of left periphery, and supposes that Topic Phrase (TopP)'s head licenses pro (under c- comand and coindexation), and then, the clitic-doubled element raises to [Spec, TopP], covertly, forcing a discourse referent. Unlike Spanish Basque, the dialects where (8) is not grammatical do not accept Top as a licenser of pro.
''Interface Conditions and the Semantics of Argument Clitics'', by Javier Gutierrez-Rexach, proposes that clitics are Determiners, with D feature. They may select an overt DP (the doubled element) and can saturate a verbal argument. Since clitics are prosodically weak, they need to move ''to a host that will be the prosodically strong element'' (p. 111), namely, V. Since some times there is not a doubled element, we can conclude that ''the selectional feature D of the clitic is optional'' (p. 113). Then, (11) is a derivation for (10), leaving aside the feature structures:
(10) Mar'ia le dio el libro a Pedro Maria him gave the book to Pedro 'Maria gave the book to Pedro'
(11) a. MERGE: <le> <a Pedro> => <le a Pedro> b. MERGE: <dio el libro> <le a Pedro> => <dio el libro le a Pedro> c. MOVE: <le dio el libro a Pedro> d. MERGE: <Mar'ia> <le dio el libro a Pedro> => <Mar'ia le dio el libro a Pedro>
This analysis allows to consider a clitic as a ''function mapping sets--the context set argument---to arity reducers'' (p. 118). The denotation of clitic will depend on semantics properties of its case (an interpretable feature, according to Gutierrez-Rexach). Accusative Clitics require that the doubled element will be semantically definite (a principal filter), excluding, therefore, the non-specific interpretation of doubled existential quantifiers. Dative clitics have not such requirement.
''Adverbial Weak Pronouns: Derivation and Interpretation'', by Javier Gutierrez-Rexach, analyses weak adverbial pronouns (temporal and locative ones) as clitics, disregarding the idea that they are expletives. Some Spanish forms for this category are ''ahi'' ['aj] ('there'), ''ahora'' ('now'), ''ya'' ('already'), aun ('still'). All of them have a corresponding strong form, which some times has a different stress, for instance, ah'i [a'i] ('there'). Only the later ones can be focused:
(12) a. Anda por ['aj] / *[FOCUS 'aj] walk-2p.sg over there 'Go away!' b. Anda por [a'i] / [FOCUS a'i] walk-2p.sg over there 'He is walking over there'
As nominal clitics, they can be doubled and the derivation of this structure will follow similar steps than nominal pronoun (see (11) above), that is, clitic and doubled element merge, and, in some moment, the clitic moves to a verbal host because of a prosodic feature (13a). However, also it can be morphologically attached to the doubled element (13b) or other non-verbal hosts.
(13) a. ['Ahi] te lo pongo [sobre la mesa] there to you it put-I over the table b. Te lo pongo ['ahi sobre la mesa] to you it put-I there over the table 'I put it on the table for you'
A main difference between adverbial and nominal clitics is that the former ones has semantics restricted to properties of events. However---since thematic roles can be properties of events (Parsons 1990)---adverbial clitics can satisfy thematic properties of a verb. They have also interesting interactions with aspect, ergativity, negation and telicity.
''Universal Constraints on 'Superfluous' Elements: The Case of Galician 'Arb Che' '', by Vï¿½ctor Longa and Guillermo Lorenzo. This paper analyses the behavior of ''che'', a special ethical dative in Galicean. This form (always in second person singular) express ''an arbitrary (or supra- individual) interest in the subject matter of the sentence'' (p. 177); for that reason, the authors call it 'Arb CHE'.
(14) As rapazas de hoxe sonCHE moi descaradas the girls of today are-ARB very impudent 'Today's girls are very impudent'
The authors assume that 'Arb CHE' is the head of a tau Projection (tP), between VP and Dative Phrase (DatP), which is also the position for reflexive anaphoric clitics like ''se'', since they have complementary distribution.
'Arb CHE' is base-generated, which explains why it cannot appear in non-clitic form (it is non-argumental); also, there is a PRO in [Spec, tP], to account for its arbitrary reading. Since [Spec, tP] is an inherent case position (PRO could get a Null Case), a minimality effect occurs when a dative with Experiencer Role raises to [Spec, DatP] (another inherent case position) to check case in constructions with psych- verbs: it crosses two case positions of the same kind. It explains why 'Arb CHE' cannot appear with psych-verbs.
''Clitic Doubling and the Acquisition of Agreement in Spanish'', by Marta Lujï¿½n and Claudia Parodi, also addresses clitic-doubling constructions, but in the context of languages in contact, namely, English (E) and Spanish in Los Angeles (LAS), and Spanish and Quechua (Q) in South America (AS, Andean Spanish). Surprisingly, the same kind of ''free doubling'' (with no agreement) occurs in both places, regardless of the difference between the other language in contact with Spanish, as in (16):
(16) LAS a. LO veo LA NIï¿½A 'I see the girl' b. LA dejï¿½ EL COCHE en la esquina 'S/he left the car at the corner'
(17) AS a. Me LA han roto MI COMETA 'They've ruined my kite' b. No LO vi A SUS HERMANITOS 'I didn't see his brothers'
To explain this fact, Lujan and Parodi, following Chomsky 1995, assume that Objects are checked in [Spec2, vP], after the adjunction of V to v. We can see the derivation in (18).
Since doubled constructions are interpreted as focal in Standard Spanish (SS), but not in LAS or AS, we can assume that the doubled object does not erase AGR features in v; then, the clitic can check AGR when V raises to v (as in 18). In LAS and AS, AGR and Case can be checked separately, therefore, the clitic can double Objects freely (with no agreement). A similar approach explains other doubling constructions, like double genitive. Sets of features with parametric strong/weak values take care of these differences, making unnecessary the notion of 'transfer' to account for this phenomenon.
Part 3 discusses some properties of Verbal Phrase. It has five papers.
''The Causee and the Theory of Bare Phrase Structure'', by Luis Lï¿½pez, analyzes Spanish causatives like (19), where the causee (in capital letters) is interpreted as receiving two thematic roles: <affectee> from HACER, and <agent> from REPARAR (repair):
(19) Yo le hice reparar mi coche A MI MECANICO FAVORITO I CLITIC made repair my car DAT my favorite mechanic 'I made my favorite mechanic repair my car'
Lopez assumes that the causee is merged with REPARAR, receiving its first role, and then moves to [Spec, HACER] to receive its second role. This is a problem for traditional analysis because a role cannot be assigned via Move. Lopez states that Move would have to work with no semantic limitations. As evidence, he argues that the causee cannot check case in the lower predicate, and it needs to raise upstairs. Since his proposal violates the Theta Criterion, he needs to find an advantage to his hypothesis: it eliminates ''asymmetry between Merge and Move with respect to Theta Theory'' (p. 238)
''Temporal Modification, the 24-Hour Rule and the Location of Reference Time'', by Gerhard Brugger, is a study of the Present Perfect (PrP), which is defined in Reichenbach's (1947) system as E_R,S----with Event Time (E) preceding (''_'') Reference Time (R), which is simultaneous ('','') with Speech Time (S). However, Brugger finds variation in the meaning of PrP cross-linguistically (also, it could be E,R_S). For Spanish, a 24-hour rule can be formulated (''PrP can be modified by a definite adverb, but only if the adverb denotes an interval that is part of today''--p. 247), but it does not work in other languages as Italian, French, German or English. The author proposes that different forms have to account for differences in meaning.
He assumes that E, R and S are represented syntactically in Tense Projection (TP), to explain their temporal order, as in (20):
(20) [TP S T R [TP2 E ... VP...] ]
This allows explain cross-linguistic differences, if we postulate an Agreement-Tense Correlation (ATC):
(21) ATC ''The PrP-auxiliary distinguishes person (in all tenses) or has semantic content or both'' (p. 267)
According to ATC, if PrP-auxiliary has no personal distinctions, must express a temporal relation semantically, with a [-PAST] feature which sets R,S; otherwise, if PrP has personal distinctions, since it may not express such relation, the feature [+PAST] sets R_S.
''Ergative Patterning in Spanish'', by Clancy Clements, examines constructions where Spanish, traditionally a Nominative-Accusative (Nom-Acc) language, shows ergative patterns. The first case is the position of bare plurals and mass nouns, which can be Direct Object (DO) with transitive verbs but must occur postverbally in passives and intransitive verbs; in addition, a bare plural subject will be always focus and, therefore, cannot be in sentence- initial position. The second case is the higher frequency of clitic-doubling when the order is not SVO. To explain these facts, Clements proposes that Spanish has a Nom-Acc marking for the subject, but an Ergative-Absolutive (Erg- Abs) one for the topic-focus distinction.
The Erg-Abs marking is particularly evident in the pronominal system, since that, in Castilian Spanish, the same clitic LE is used to double DO and Indirect Object (IO)---this phenomenon is called ''leï¿½smo''. This marking is expression of a Primary Object-Secondary Object pattern in Dryer's 1986 sense, which the author identifies with an ergative pattern.
''Morphological Underspecification and Overt Subjects in Child Catalan and Spanish'', by John Grinstead, raises the question about what determines the presence of overt subjects in children's Spanish and Catalan. The author offers evidence showing that number and tense morphology is unavailable in children's early stages of language acquisition; therefore, no overt subject is used, since Nominative Case cannot be checked. However, the covert subject used by children cannot be pro, given that pro is licensed, essentially, in the same conditions of overt subjects, that is, under a context of rich agreement. Therefore, PRO is the ideal candidate to explain the null subject being used.
Surprisingly, in English, French and German (overt subject languages), children use null subjects as well as overt subjects from the very beginning, although no morphological specification of tense and number is found. The author makes an interesting generalization regarding the arrival of tense and agreement morphology in child language. In null subject languages, this arrival indicates the beginning of overt subject use (and of pro), whereas in overt subject languages, this same arrival indicates the end of null subject use.
''Verbless Derivations in Historical Syntax: A Case Study of Northwestern Iberian Languages'', by Luis Silva-Villar, studies verbless sentences in some Northwestern Iberian (Languages Spanish, Galician, Portuguese and Leonese), like in (22)
(22) a. Mi vieja leal DOLA? (14th C Spanish) My old-woman loyal where-her 'I wonder where my loyal old friend is'
b. Mira la lleï¿½a secu.- ULO? (Leonese) look the firewood dry where-it-neuter 'Look at the dry firewood. Where is it?'
In traditional terms, these ULO-sentences are structures where the verb ''to be'' is no overtly present, whose overt subject is an enclitic form, and whose general meaning is ''Where is X?'' To explain them, Silva-Vilar assumes a distinction between ''regular'' and ''irregular'' syntax (Chomsky 1998) to account for the fact that these sentences can only be derived partially, since some kind of precompiled syntax needs to be added to the Numeration.
ULO is conceived as a multicategorial form [C+D], a verbal- like item with T-feature but with an inert [+V], whose internal syntax is not accessible by the Computational System. This implies that we need to see ''irregular'' syntax from a representational point of view. According to Silva- Villar, this insight helps to understand the syntax of Spanish HE (from Latin ECCE) and Brazilian Portuguese distinction CADE^/QUEDE^.
My brief comments on this book will not be focused in details from the articles, but in some evident relations among the papers.
It is worth noticing the extreme importance that the discussions about Topic and Focus have in the syntactic research in Spanish (and other languages, as well). Almost no paper escapes from this issue in this book. Some address the subject directly, but even these that make no mention of it need to be revised from this perspective because the grammatical judgments they make (that is, the asterisks they put) are very sensitive to Topic-Focus distinction.
Another feature of the papers is that they present (as is common practice in the field) very issue-oriented solutions to the problems they address. No attempt is made to make compatible their solutions, not even in the Introduction. It is possible that such propose would be unattainable, given the very dissimilar explanations for similar phenomena (on DPs and clitics, especially). Of course, it is not necessarily a failure, but it is evidence that we lack general working frames in formal Spanish Linguistics. That means that a lot of work needs to be done yet. For the moment, this book represents an extraordinary contribution that will help to create the big picture.
REFERENCES Chomsky, Noam. 1995. The Minimalist Program. Cambridge, Massachusetts: MIT Press.
Chomsky, Noam. 1998 ''Minimalist Inquires: The Framework''. Ms. MIT
Contreras, Heles. 1991. On the position of subjects. Syntax and Semantics 25:63-79.
Dryer, Matthew. 1986 Primary object, Secondary objects, and antidative. Language 62, 808-845
Olarrea, Antxon. 1996. Pre and Postverbal Subject Positions in Spanish: A Minimalist Account. University of Washington: PhD Dissertation.
Parsons, Terence. 1990. Events in the Semantics of English. Cambridge, MA: MIT Press.
Reichenbach, Hans. 1947. Elements of Symbolic Logic. New York: MacMillan.
Rizzi, Luiggi. 1997. The fine structure of the left periphery. In Elements of Grammar: Handbook in Generative Syntax, ed. Liliane Haegeman, 281-338. Dordrecht: Kluwer.
Sanchez, Liliana. 1996. Syntactic Structure in Nominals: A Comparative Study of Spanish and Southern Quechua. USC: PhD Dissertation
ABOUT THE REVIEWER:
I'm a graduate student in the Department of Spanish and Portuguese, in The University of Arizona, Tucson. I want to become a formal linguist and I have strong interest in syntax, phonetics, and syntax-pragmatics interface. Right now, I'm working in a thesis about modality in Spanish to get my MA. Also, I'm looking for a place to get my Ph.D in Linguistics.