Reviews

Reviews Links & Letters 1, 1994 129 cally as other objects of the sentence (direct objects, indirect objects, etc.). However, morphol~gicall~ and phonologically thev have a venr close relation to their host, like many morphemes. The last part of Spencer's book addresses one of the most important questions in morphology: what is a word? One of the phenomena that makes the answer difficult is what has been called bracketing paradoxes (discussed in Ch. 10). This term refers to the cases were the bracketing imposed on a sequence according to the phonology, for instance, does not match the bracketing imposed by the morphology, the syntax or the semantics. A very familiar example of this sort is the term atomic vhvsicist. Even thou~h the 1 -I U suff~ -ist is attached to the stem physic, the adiective atomic does not refer to ~hv- I 1 -I sicist but only to the physic part (we are talkin~ about a Derson who does atomic u physics, not a physicist who is atomic). Finallv the last cha~ter of the book (Ch. 1 1) addresses the wncluchng question: given al1 the facts discussed throurrhout the book, what place in the gram&r should morphology occupy? This is a question that has not yet received a satisfactory answer and is still the subiect of much debate. Spencer's book constitutes a longneeded summary of the most significant work that has been done in theoretical morphology in the last fifty years. In spite of very minor objeaions one can make to it (like excess of emphasis on classification issues ), it is a very valuable piece of work for anybody interested in the field of morphology. Eulalia Bonet Departament de Filologia Catalana Universitat Autbnoma de Barcelona 2. Especidy in his treatment of clitics, Spencer seems too worried about the status of certain elements as clitics or morphemes. He does not consider the possiblity that these ambiguous elements might be the result of the interaction of different types of processes, in a fashion parallel to the way passives, for instance, are viewed in the generative frarnework. ANDREW RADFORD, Syntactic Theory and the Acquisition of English Syntax: The Nature of Early Child Grarnrnars of English. Oxford: Blackwell, 1990. viii + 3 1 1 pages. Syntactic Theory and the Acquisition,ofEnzsh Syntax makes an important contriution to the study of first language t. acquisition. The main goal of the book is to analyse child speech within the Princiles and Parameters framework and within a Maturational model of language acquisition. The Principies and Parameters model (Chomsk~ 1981, 1986a, 1986b, 1988), on the one hand, is a theory of language that explains crosslinguistic similarities by assuming that there is a core of universal principies (UG) common to al1 languages and a set of parameters, with a number of open values, that account for laneua~e variation. As far as the nature of " U these parameters is concerned, two major hypotheses have been put fonvard: one (Chomsky 1986a) that associates parameters with UG principies and another according to which parametric variation is connected with functional categories (Borer 1984, Chomsky 1988, Ouhaila 1991), where functional categories include the set ofelements which in traditional typology have been referred to as closed chs, that is, Determiners, Complementisers, Inflection (Tense and Agreement), Negation.. . The Maturational model of language acquisition, on the other hand, claims that the different stages that al1 children go through are constrained by inherent maturational factors. In this respect, within the Maturation Hypothesisl two different proposals have been put fonvard: one (Felix (1984) and Borer and Wexler (1987)) that claims that the availability 1. This theory of language acquisition contrasts with the Continuity Hypothesis (Hyams (1987) and Weissenborn (1990) arnong others) according to which al1 UG ~rinci~les and parameters (however not fxed) are present throughout the language acquisition process. 130 Links & Letters 1, 1994 Reviews of UG principles is biolo 'cally determined this volume) that distinguish functional and another (Guilfoy Y e and Noonan from lexical categories, he fails to mention (1 988), Radford (1 990) and Tsimpli the grammatical features associated with (1992)) which pro oses that UG princithe former. Recent work on functional ples are available P rom the start of the categories (Ouhalla 199 1) stresses the imlanguage acquisition process and that maportance of !grammatical features -for turational processes affect functional caexample, the phi-features (person, nu '3'- tegories and their syntactic projections. ber and gender) associated with AGR - Radford's analysis is based on a wide in determining movement processes and range of data from a Corpus of more than relations. The rest of the chapter is devo- 100,000 spontaneous utterances and it deted to providing empirical evidence, monstrates that the earliest structures romainly from pronominalisation, for the duced by children (when at aroun! 18 ciaim that once children have developed months they start combining words) are lexical word categories they have develolexical-thematic, that is, they are characped the corresponding phrasal categories teriied by the absence of funaionai categories. as well. The bookis divided into ten chapters. Chapters 4, 5, 6, and 7 are largely The first chapter, Aims and approache~, concerned with substantiating the claim sets out with an introduction to the ultithat functiond categories are missing in mate goal of a language acquisition the lexical stage. Radford gives a detailed theor~: namel~ that of explining how presentation of data from imitative and when, and why children develop their spontaneous speech to show the absence initial grammars. With this goal in mind, 0f D(eterminers) (chapter 4: Absence of fi the author goes on to provide an overview Determiner System in Early ChildEnglish), of the current methodolog~ and theoreti- C(omplementizers) (chapter 5: Absence of cal framework. He points out the limita- a complementizer system in ~~~ly child tions of SPontaneous language sam~les English ), 1 (nflection) (chapter 6: Absence and examines the problem of establishing ofan inyection srstem in ~~~ly chiu the productivity of structures. glish ) and Case (chapter 7: Absence of n In cha~ter 2, Categori=tion earb' Case System in Early Child Eng/&). In chik Enghb, Radford addresses the tenchapter 7, he elaborates on interrelatral claim 0f the boOk. He Presents mortion between the absence of Determinen, phological and syntactic evidence to show ~~fl~~~i~~ and case: the clim hat there is that at about the age of 18-24 months no case system operative in child children have developed a SYStem of le''- grammars follows from the assumption cal categories: nouns, verbs, adjectives, that case is an intrinsic property of the and some prepositions. This stage in the Determiner phrase and that nominative develo~ment of children's s~eech is reassigned by (both being functional ferred to as bxical categonal (as opposed categories and hence missing in early to the pre~ate~orial stage (0-18 months) grammars). where no categorization has taken place 1, chapter 8, írhe grammar ofmissing yet)> Or P~&~~~~~~~~ (as 'pposed arguments in Enghh, Radford later functional stage (24-30 discusses alternative analyses of missing characterized by the emergence of funcarguments, he smdy ofwhiCh has received a tional categories). great deal of attention in the literature. In The crucial distinction between functhis sense, almost all modern work in syntional and lexical categories is dealt with tactic theory recognizes that sentence in cha ter 3, Lexical categoy systems in structures involve ccinvisiblen parts, that early c 1 ild Englih. Although the author discusses the morphological and nonthematic or c(ategorid)-selectional proper- 2. Abbreviation for the agreement features in the ties (See the review of Ouhall& book in INFLECTION category in a finite clause. Reviews Links & Letters 1, 1994 131 is, syntactic positions which are not Continuity Hypothesis and argues for a fleshed out by words and which aid in the Maturational approach whereby functioexplanation of the relations between nal-nonthematic structures come "on non-adjacent elements. The author arline" at a specific point of maturation. gues first against Hyams' (1986, 1987) One issue that K~dford does not fully account of missing arguments as instances bcw, however, is whether maturation only of pro (a null pronominal element found affetcs functional categoria or whether it 3 in governed positions ) it alternates with affects princi~la of UG aswell. overt pronouns which occur in governed Syntactic theoy and the acquisition o positions since they must be assigned Englisb Yntax provides the reader wit f case), and then against D~ H~~~ and helpful and comprehensive information Tuijnmans (1988) analysis of rnissing arconcerning not onl~ earl~ child 'peech guments as variables (an empdy category but also recent research on theoretical bound by a null topic phrase 1. Radford linguistics. This book is likely to stimulate finally explores the claim that missing investigations which will seek to test arguments could be analyzed as null NPs hpotheses about language acquisition in other languages. (np). However, a very plausible analysis that he does not take into consideration is the structural realisation of missing arerences guments as PRO (a null element which Ref can occur only in ungoverned positions), as suggested by Tsimpli (1992). BAKER, M.C. (1 988). Incorporation: A Chapter 8, The overallstructure ofearly Theory of Grammatical Function chikdgrammars ofEnglish, contains a fairly changing. chicago universis. press. detailed summary of al1 the previous BORER, H. (1984). parametrjc syntax: chapters and examines the range of mo- 5 CaseStudiesin SemiticandRomance Landules that are operative during the earguages. Dordrecht: Foris Publicaliest stages of language acquisition. tions. The tenth and final chapterj Explana- BORER, H. and K. WEXLER . (1 987). tions and implications, seeks to explain "The Maturation of SyntaxV. Roeper, why lexical-thematic systems should T and E. Williams (eds.) Purameter come into operation before functiond Settjng, 123-172. Dordrecht: Reidel. non-thematic systems. He challenges the CHOMSKY, N. (1981). Lectures on Government und Binding. Dordrecht: Fo- 3. Government is a local relationship benveen elements (for example, a verb governs its object ris Publications. NP) that plays a central role in several of the CHOMSKY, N. (1986a). Knowled e of subtheories of grammar. More precisely, a cate- Lan~age: Its Origin gory governs another category if both these ca- New York: Praeger. tegories m-command each other, where an CHOMSKY, N. (1986b). Barriers. Camelement is said to m-command al1 the elements bridge, Ma.: MIT Press. contained inside its domain. CHOMSKY, N. (1988). "Some Notes on Economy of Derivation and Repre- 4. The topic phrase is generated in the topmost sentation". MIT Working Papers in position and it is claimed to have referential properties which are derived from the discourse. Linguistics 10. 43-74. DE HAAN, G. and K. TUIJNMAN. (1988). 5. In the Principies and Parameters Theory it is "Missing subjects and objects in child suggested that the language system itself is mogrammar". Jordens and Lalleman, dular in the sense that ir consists of a number of Language Development 10 1 - 12 1. different subsystems, each of which constrains a Dordrecht: Foris Publications. particular type of grammatical entity and cons- FELM, S. W. (1984). "Maturational Astruct. pects of Universal Grammar". In Da- 132 Links & Letters 1, 1994 Reviews vies, A. and A. Howatt (eds.) Interhn- JOHN Cm and COLIN YALLOI! guage, 133-161. Edinburgh: Edin- Inmddn toPho&a~Phomb8 burgh University Press. 0xford:Basil Blackwell, 1990. GUILFOYLE, E and M. NOONAN. (1988). Functional Categories and Language 400 pages. Acquisition. Paper presented at the Thirteenth Annuai Boston University Conference on Language Develop- Phonetics and phonology are van discirnent. plines which include rnany different to- HYAMS,N. (1986). LnnguageArquirition pies- Thereforel 'Orne degree of and the Theory parameters. Darcornplexity, associated with any scientific drecht: Reidel. field, is unavoidable when studying HYMs, N. (198». rtThe nieory of Paspeech in general. Clark 81 Yallop's An rarneters and Syntactic Develointroduction to phonetics and phonology prnent". Roeper, T and E. Williarns achieves the airn of giving an introduc- (eds.) Parametter Setting, 1-22. Dortory account of speech sciences, and sucdrecht: Reidel. ceeds in presenting a clear and concise OUHALLA, J. (1 99 1). Functional Categodescription of both areas of study. In rnany aspects, this is an irnpressive book. riesandfarametric VnMion. London: Its rnain merit lies in its putting together Routledge. POLLOCK, J. Y. (1989). "Verb Movernent, in one book a vast range of different issues covered in ~honetic and phonological UG and the Structure of IP". Linguitheory. tic Inquiry 20. 365-424. The bookstarts with a short introduc- IUDFoRD> *. (1990). Syntactic Theo~ tory chapter one which presents and out- and the Acquisition ofEnglishSyntax lines the rnain topics dedt with in The ¡Vature 'f Ear!J' Child Grammars text. Chapter 2 deals with the organs of 'fEngLish. Oxford: Basil B1ackwell. speech and various functions. It be- TSIMPLI, 1. (1992). Functional Categories gins with a very exhaustive descnption of and Maturahont The Prefinctional the anatorny of speech production, pay- Sta@ ofLa%uage Ac9uisition. Doctoing attention to muscular acral dissertation. Universi'7 College tion and the respiratory cycle. Much of London. the detail present in this chapter is remi- WEISSENBOm, J. (1990). "Functiond niscent of Borden and Harris (1984). Categories and Verb MOvernent in This part may be too cornprehensive for Early Ckrrnan: The acqusition of Gerthe beginner student of phonetics, but rnan syntax reconsidered". RothweigriVes an excellent overview to any-one lec, M. (ed.) Spracherwerb with interests in speech anatorny. The Grarnrnatik. Linguistiche Untersuchuwhole chapter is centred on the gen m~m l3werb van S~ntm UndMortion of laryngeal excitation, particularly, phologie. Gottingen: Verlog. rnodai phonation, that is, normal vocal fold vibration. Chapters 3 and 4 are concerned with Montserrat Capdevila i Batet some basic issues within phonetic theory. Departarnent de Filologia Anglesa i Ger- These include phonation rnodes, vowel rnanística and consonant articulation, rnanner and Universitat Autbnorna de Barcelona place of articulation, voice onset time, secondary articulation and coarticulation, syllabification, etc. It also describes the diversity of speech sounds to be found in languages around the world. This is done less systematically than in other