Sei sulla pagina 1di 5

Access Provided by University of Zagreb, Faculty of Philosophy at 01/06/11 4:00PM GMT

498

LANGUAGE, VOLUME 81, NUMBER 2 (2005)

MARKMAN, ELLEN. 1992. Constraints on word learning: Speculations about their nature, origins, and domain specificity. Modularity and constraints in language acquisition (The Minnesota symposia on child psychology 25), ed. by Megan Gunnar and Michael Maratsos, 59102. Hillsdale, NJ: Erlbaum. OSHIMA-TAKANE, YURIKO. 1988. Children learn from speech not addressed to them: The case of personal pronouns. Journal of Child Language 15.94108. ROSCH, ELEANOR; CAROLE B. MERVIS; WILLIAM D. GRAY; DIANE M. JOHNSON; and PATRICIA BOYES-BRAEM. 1976. Basic objects in natural categories. Cognitive Psychology 8.382439. SMITH, LINDA B.; SUSAN S. JONES; and BARBARA LANDAU. 1996. Naming in young children: A dumb attentional mechanism? Cognition 60.14371. Institute of Child Development 51 East River Rd. University of Minnesota Minneapolis, MN 55455 [marat@umn.edu]

Frequency and the emergence of linguistic structure. Ed. by JOAN BYBEE and PAUL HOPPER. (Typological studies in language 45.) Amsterdam: John Benjamins, 2001. Pp. vii, 492. ISBN 1588110281. $55.95. Reviewed by JEAN-CHRISTOPHE VERSTRAETE, University of Leuven
What can frequency dataand, more generally, usage datatell us about linguistic structure? This classic question in linguistic theory has received answers ranging from distinctions like langue/ parole or competence/performance that underline the distinctness and independence of usage and structure (see Newmeyer 2003 for a recent statement), to psycholinguistic and diachronic work that shows how usage feeds back into structure (see Zipf 1935 for an early statement). Frequency and the emergence of linguistic structure takes up the fundamental problem of the relation between structure and usage, focusing on the role of frequency. The twenty contributions in this volume, subdivided into four sections, provide a convincing demonstration of the importance of frequency effects in phonology, lexicon, and syntax, using diachronic data, psycholinguistic experiments, and corpus-based methods. After an introduction in which Bybee and Hopper give a brief overview of the history and various lines of research in usage-based work, the first section of this volume brings together three papers that deal with the frequency of grammatical features in particular genres, specifically face-to-face conversation. In a study following up on their classic article about the scalar definition of transitivity (1980), PAUL J. HOPPER and SANDRA A. THOMPSON investigate the features of scalar transitivity in a small English conversational corpus. They come to the surprising conclusion that the large majority of clauses are either one-participant clauses or two-participant clauses with low transitivity, showing that the central clause type in the basic genre of conversation is rather different from the prototypical high transitive clause that is the focus of most grammatical studies. They also investigate the consequences of their findings for the traditional definition of argument structure. JOANNE SCHEIBMAN uses a larger corpus of English conversational data to investigate patterns of subjectivity in conversation. In a study of the distribution of pronouns, tenses, and verb classes, she shows that 1st and 2nd person pronouns predominantly occur with verbs of cognition and verbalization in the present tense (with high-frequency combinations like I guess or do you think), and that 3rd person pronouns are mainly used in impersonal constructions with relational verbs (such as it is good). Scheibman argues that these patterns of pronoun-verb-tense combination reflect the fundamentally subjective nature of conversation, which focuses on the expression and exchange of opinions in the here-and-now of the speech event rather than on the description of past events. Although many linguists would agree with Scheibman that subjectivity is a fundamental aspect of language structure (see Wright & Stein 1995), I think it is something of an overstatement to conclude that the prototypical structures of English clauses do not seem

REVIEWS

499

geared to objective relating of events (86). Subjective types of structure like tense, mood, or person deixis do not replace more objective types of structure like transitivity or case systems; they simply coexist and interact with them, a standard assumption in many functional models of grammar (see e.g. Halliday 1994, Dik 1997). In the final paper in this section, NAOMI HALLAN compares the frequency of nonspatial uses of the English path morphemes on and over (e.g. in phrasal verbs like put on) with their spatial uses (e.g. on the sofa), the latter having traditionally been considered the prototypical use in semantic analyses. Using adult and child language corpora, Hallan shows convincingly that the apparent conceptual basicness of spatial uses is contradicted by the fact that nonspatial uses such as in phrasal verbs are more frequent in adult and child corpora and are acquired earlier by children. Unlike the other two papers in this section, which merely use frequency as a heuristic to explore the distribution of specific grammatical features in a particular genre, Hallans paper shows how frequency data can provide insight into structure, in this case the internal semantic structure of path morphemes. In addition to the traditional concept-based definition of prototypical organization, prototypicality can also be defined in terms of frequency in adult use and in the process of acquisition, and Hallan clearly shows that the two definitions do not necessarily match. The second section in this volume contains papers that deal with frequency effects in phonology and morphology. In the only diachronic contribution of this section, BETTY S. PHILLIPS summarizes and refines earlier work on the role of frequency in the lexical diffusion of sound changes. Using data from the history of English, she shows that some sound changes affect the least frequent words first, whereas others affect the most frequent words first. She argues that the difference relates to whether the sound change in question requires some kind of analysis of the word to which it applies in terms of part of speech or morphological structure (least frequent words first) or whether it has a purely articulatory basis (most frequent words first). The next two papers in this section deal with frequency effects in phonology. JANET B. PIERREHUMBERT addresses the fundamental question of how to deal with lexically specific phonetic knowledge (like t/d-deletion in English, which correlates with lexical frequency) in phonological theory, in which rules are traditionally taken to apply to any lexical item that has the right structure, irrespective of usage factors such as frequency. In a rather programmatic discussion, Pierrehumbert uses statistical simulations to show how the findings of exemplar theory, originally developed in cognitive psychology as a model of perception, can be extended to speech perception and production in order to model specific phonetic knowledge in phonology. STEFAN A. FRISCH, NATHAN R. LARGE, BUSHRA ZAWAYDEH, and DAVID B. PISONI examine the nature of speakers knowledge of phonotactic constraints. Using experiments with native speakers of English and Jordanian Arabic, they demonstrate that phonotactic knowledge is not an absolute set of rules, but emerges instead from the speakers mental lexicon. They show, for instance, that variation in speakers judgments of wordlikeness correlates with statistical tendencies in the lexicon, and that judgments may also be affected by the size of an individual speakers lexicon. The last two papers in this section deal with the issue of regularity and irregularity in morphology. MARY L. HARE, MICHAEL FORD, and WILLIAM D. MARSLEN-WILSON investigate frequency effects for regular past tense inflection in English. Although such effects are well known for irregular forms, and while most theories accept some kind of lexical storage of irregular forms, the status of regularly formed verbs is less clear, with some theories assuming lexical storage and others positing derivation by rule. The authors argue for a model in which highly frequent regular verb forms can be stored just like irregular ones, and they support this with evidence from dictation and priming experiments which show that more frequent regular past tense verb forms behave just like irregular ones in terms of accessibility. The relation between irregularity and frequency is further investigated by GREVILLE CORBETT, ANDREW HIPPISLEY, DUNSTAN BROWN, and PAUL MARRIOTT, who go beyond the traditional dictum that irregularity correlates with frequency and use a statistical corpus study of Russian noun inflection to pinpoint the precise locus of frequency effects in the structure of the lexeme. Distinguishing between different types and degrees of irregularity, they show that frequency effects can be located both on the level of the lexeme as such (i.e. as an effect of frequency of the irregular form relative to the

500

LANGUAGE, VOLUME 81, NUMBER 2 (2005)

corresponding form of other lexemes) and to a lesser extent also on the level of the subparadigm of the lexeme (i.e. as an effect of frequency of the irregular form relative to regular forms of the same lexeme). The papers in the third section deal with frequency effects in word combinations and constructions. DANIEL JURAFSKY, ALAN BELL, MICHELLE GREGORY, and WILLIAM D. RAYMOND set the scene with a paper about their probabilistic reduction hypothesis, which states that words tend to be pronounced in reduced form when they are more predictable, using a measure of probability that covers both predictability of words as such (in terms of relative frequency in a corpus) and in a particular context (in terms of frequency of cooccurrence with its neighbors, with or without control of individual frequency). Specifically, they use data from an English corpus to show that function words tend to undergo more vowel reduction and shortening in contexts where they are predictable from their neighbors, whereas content words tend to undergo more t/d-deletion in contexts where they are predictable in terms of their relative frequency. The same principle of probability is further explored in two papers on word-boundary processes. NATHAN BUSH shows that the predictability of a word on the basis of the preceding word can explain the distribution of glide palatalization on word boundaries in English (e.g. would you realized as [wdU] rather than [wdjU], 256), and JOAN BYBEE uses the same principle to provide a more adequate account of the classic problem of liaison on word boundaries in French (e.g. les in les enfants realized as [lz] rather than [l], 339). Both authors also discuss implications of this finding for the nature and development of constituency. The next three papers turn to diachronic developments in English and their synchronic reflections. CATIE BERKENFIELD investigates the phonetic properties of that in a corpus of spoken English, comparing the demonstrative pronoun function with other functions like complementizing and relativizing that, which are known to have developed out of demonstratives. She shows that mean vowel duration and the frequency of the first formant are higher for demonstratives than for complementizers and relativizers and argues that such features could be regarded as emergent phonological features in the differentiation of these diachronically related forms. MANFRED G. KRUG identifies a similar type of differentiation of diachronically related forms in the progressive univerbation of expressions like want to and going to to wanna and gonna in auxiliary constructions. K. AARON SMITH investigates the competition between be and have as auxiliaries with past participles in the history of English, which has traditionally been related to semantic features of the main verb (like mutativity). On the basis of a study of the two constructions in an extensive diachronic corpus, Smith argues that the distribution in the older phases of English does not seem to be semantically determined; instead, it reflects a development that is more akin to leveling processes (e.g. regularization of plural marking), with the have-construction extending its domain on the basis of higher type frequency (i.e. the higher number of verbs in the lexicon that use have) and first affecting the be-construction in contexts of lower token frequency (i.e. verbs with a lower number of instances in a corpus). The last two papers in this section turn to sociolinguistics and variation theory, the classic domains of frequency studies in linguistics. JOYCE TANG BOYLAND looks at sociolinguistic, syntactic, and frequency effects in the use of hypercorrect pronouns in English (e.g. between you and I), and SHANA POPLACK investigates the relative significance of syntactic, semantic, and frequency-related factors in a number of French constructions that allow variation, such as variation between indicative and subjunctive in complement clauses or between conditional and imperfect verb forms in conditional clauses. The final section brings together three papers that discuss frequency and other usage-related effects in more general terms. BRIAN MACWHINNEY gives an overview of the concept of emergence as it is used in different disciplines, distinguishes between different levels of emergence, and introduces a number of psycholinguistic models that have been used to model emergent phenomena in language. GERTRAUD FENK-OCZLON summarizes earlier work on frequency effects in the framework of a more general theory of processing effects on linguistic structure, using examples from work on word order, word length, and phonetic reduction. OSTEN DAHL rounds off the volume with a nice paper about different types of inflationary effects in language, such

REVIEWS

501

as cases where overuse of forms in contexts that do not strictly speaking require them (e.g. overuse of polite address terms) leads to loss of their specific value. In general, the data and analyses presented in this volume are original and often surprising, clearly demonstrating the need to take frequency effects seriously and making a convincing case for a usagebased approach to some aspects of linguistic structure. In a volume that demonstrates the role of frequency effects so effectively, however, one aspect that is sometimes missing in the discussion is the explanation of the effects that are observed: Frequency effects as such do not constitute an explanation but are themselves an effect of more general cognitive and processing-related principles. Although some authors include references to such principles, it would have been useful to include more frequent and more explicit discussion of links with work in cognitive psychology. In addition, it is also a bit surprising to find so few references to and papers on language acquisition and natural language processing, two domains in which frequency-related phenomena are becoming increasingly important (see Tomasello 2003 and Manning & Schutze 1999 for relevant overviews). These are minor defects, however, in a volume that so convincingly makes a case for the importance of frequency effects in linguistic structure, a point that has often been made in impressionistic terms but has not often been elaborated in such detail.

REFERENCES
DIK, SIMON. 1997. The theory of functional grammar. Part 1: The structure of the clause. 2nd edn. Part 2: Complex and derived constructions, ed. by Kees Hengeveld. Berlin: Mouton de Gruyter. HALLIDAY, M. A. K. 1994. An introduction to functional grammar. London: Arnold. HOPPER, PAUL, and SANDRA THOMPSON. 1980. Transitivity in grammar and discourse. Language 56.25199. MANNING, CHRIS, and HINRICH SCHUTZE. 1999. Foundations of statistical natural language processing. Cambridge, MA: MIT Press. NEWMEYER, FREDERICK J. 2003. Grammar is grammar and usage is usage. Language 79.682707. TOMASELLO, MICHAEL. 2003. Constructing a language: A usage-based theory of language acquisition. Cambridge, MA: Harvard University Press. WRIGHT, LAURA, and DIETER STEIN. 1995. Subjectivity and subjectivisation: Linguistic perspectives. Cambridge: Cambridge University Press. ZIPF, GEORGE. 1935. The psychobiology of language. New York: Houghton Mifflin. Department of Linguistics University of Leuven Blijde-Inkomststraat 21 3000-Leuven Belgium [jean-christophe.verstraete@arts.kuleuven.ac.be]

Linguistic borrowing in bilingual contexts. By FREDRIC W. FIELD. Foreword by BERNARD COMRIE. (Studies in language companion series 62.) Amsterdam: John Benjamins, 2002. Pp. xviii, 255. ISBN 1588112853. $110 (Hb). Reviewed by PETER BAKKER, Aarhus University
Fredric Field intends in this book to come to grips with universals of borrowing; he proposes a number of hierarchies and constraints and applies them to a corpus of recorded texts in which borrowings from Spanish are frequent. His book provides a detailed overview of earlier proposals for borrowing patterns and uses semantic and morphological typology to formulate constraints on borrowing. The introduction (124) outlines the background: Borrowing refers primarily to the integration of forms into a recipient language (2). F distinguishes between social factors relevant in the study of borrowing (cultural dominance, prestige, lexical gaps, affect) and linguistic factors such as word frequency and equivalence. Words are seen as form-meaning sets, and concrete nouns can be expected to be more prone to borrowing than verbs because of the latters more abstract

Potrebbero piacerti anche