The Crosslinguistic Study of Language Acquisition, Volume 5: Expanding the Contexts 0805824219, 0805823115, 9780805824216

Citation preview

THE CROSSLINGUISTIC STUDY OF LANGUAGE ACQUISITION Volume 5: Expanding the Contexts

This page intentionally left blank

THE CROSSLINGUISTIC STUDY OF LANGUAGE ACQUISITION Volume 5: Expanding the Contexts Edited by DAN ISAAC SLOBIN

University of California at Berkeley

First Published 1997 by Lawrence Erlbaum Associates, Inc" Publishers Published 2014 by Psychology Press 711 Third Avenue. New Vork, NY 10017 and by Psychology Press 27 Church Road. I·lovc, EIlSt Sussex. BN3 2FA

Psych%gy Press i$ an imprinl ofl},e Taylor & Francis Group. an informa business Copyright CI 1997 by Lawrence Erlbaum Associmes, 1m;. Al l rights rcserved. No part oft his book may bc reprinted or reprodueed or utilised in any fonn or by any elcctronic. meehanieal. or other means. now known or hereafler invented. including photocopying and recording. or in any infonnation storage or retrieval system, without pcnnission in writing from the publishers. Trademark notiee: ProdUCl or corporate names may bc trademarks or registered trademarks. and are used only for identifieation and exp lanation wilhout intcnt tO infringc. Library of CongNss Cataloging in Publkation Data (Revised for vol. 5) The erosslingu islic study of language acquisilion. Includcs bibliographies and indexes. Conlent,: v. I. 11Je data- v. 2.11Jeoretical issues-v. 3. [withoul special tillej- v. 4. [without special ti llc ) - v. 5. Expanding lhe contexts. I . Language acquisition. I. S iobin. Dan Isaac, 193940 1',93 Pll S.C69 1985 85·274 11 ISBN 0·8058-2421 -9 (Vol. 5) ISBN 0·8058·2311·5 (Sct of Volulncs L 2, 3, 4, and 5) ISBN 978·0-805-82421-6 (hbk) Publisher's Nole The publisher has gone 10 great lengths to ensure the quality ofthis reprint but points out thaI some impcrfections in Ihe original may be apparenr.

Contents

Format and Abbreviations for Glosses

vii

Preface

xi

Contributors

xv

1

2

3

The Universal, the Typologieal, and the Particular in Acquisition Dan I. Slobin Language-Specific Input and Early Semantic Development: Evidence From Children Learning Korean Soonja Choi Language Typology, Prosody, and the Acquisition of Grammatical Morphemes Ann M. Peters

4

Variation in a Crosslinguistic Context Elena V. M. Lieven

5

The Origins of Grammaticizable Notions: Beyond the Individual Mind Dan I. Slobin

1

41

135

199

265

Author Index

325

Subject Index

331

v

This page intentionally left blank

Format and Abbreviations for Glosses'

All foreign language examples are given in Italics. (Small caps are used for emphasis and other usual functions of Italics.) In running text, English glosses and grammatical codes are given in single quotes, and optional free translations follow in parentheses, indicated by an equal sign. Grammatical codes are always given in capital letters (see list, below). For example: gel-me-di-n 'come-NEG-PAST-2SG' (= you didn't come).

In interlinear format, translation equivalents appear below the foreign language example and the free translation is placed below in single quotes: gel-me-di-n come-NEG-PAST-2SG 'you didn't come'

Hyphens in a morphological gloss always correspond to hyphens in the foreign example. If part of a foreign example corresponds to more than one grammatical code, the collection of codes is joined by colons; e.g., gel-medin 'comeNEG:PAST:2SG', or even gelmedin 'come:NEG:PAST:2SG'. If it is relevant to indicate the possibility of segmentation, plus signs can be used in place of colons. The preceding example consists of segmentable morphemes, and could also be glossed, for example, as gel-medin 'come-NEG+PAST+2SG'. Use of colons is neutral with regard to the possibility of segmentation, and in most instances either colons or hyphens are used. (The degree of precision of segmentation and glossing of an example, of course, depends on the rale it plays in the exposition.) 'The abbreviations are adapted from a list used by Bemard Comrie (The languages ot the Soviet Union, Cambridge University Press, 1981, p. xv). The format is based on useful suggestions offered by Christian Lehmann in "Guidelines for interlinear morphemic translations: A proposal for a standardization" (Institut für Sprachwissenschaft, Universität Köln, Arbeitspapier Nr. 37, 1980). The system presented here is offered as a proposal for standardization in child language studies.

vii

viii

Format and Abbreviations for Glosses

If a single lexical item in the original is expressed by several lexical items in a gloss, those items are separated by aperiod; e.g., hipil 'made.fall', kalk 'get.up'. Aperiod is also used when the name of a grammatical element consists of more than one item; e.g., DEF.ART = definite article. Combining the principies for use of colons and periods in grammatical codes, consider the gloss for the German definite article in its masculine singular accusative form: den 'DEF.ART:MASC:SG:ACC' .

LIST OF GRAMMATICAL CODES 1 First Person 2 Second Person 3 Third Person ABESS Abessive ('without X') ABL Ablative ('from X') ABS Absolutive ACC Accusative ACT Active ADESS Adessive ('towards X') ADJ Adjective, Adjectival ADMON Admonitive ADV Adverb(ial) AFFIRM Affirmative AGR Agreement AGENT Agent ALLAT Allative ('to(wards) X') AN Animate ANTI Antipassive AORIST Aorist APL Applicative ART Article ASP Aspect AUG Augmentative AUX Auxiliary BEN Benefactive BT Baby Talk C Consonant CAUS Causative CL Clitic CLASS Classifier CMPLR Complementizer CNTR Contrastive COMIT Comitative (,(together) with X') COMM Common COMPAR Comparative

COMPL Completive CONC Concessive COND Conditional CONJ Conjunction CONN Connective CONSEC Consecutive CONT Continuous, Continuative CONTEMP Contemporative COP Copula DAT Dative DECL Declarative DEF Definite DEICT Deictic DEM Demonstrative DER Derived, Derivation DESID Desiderative DIM Diminutive DIREC Directional DO Direct Object DU Dual DYN Dynamic (Nonstative) ELAT Elative ('out of X') EMPH Emphatic EQU Equative ERG Ergative ESS Essive ('as X') EVID Evidential EXCL Exclusive EXCLAM Exclamatory EXIS Existential EXP Experiential EXT Extension FACT Factive FEM Feminine FIN Finite

Format and Abbreviations for Glosses FOC Focus FUT Future GEN Genitive HAB Habitual HABITA Habitative HON Honorific HUM Human ILL Illative ('into X') IMP Imperative INAN Inanimate INCH Inchoative INCL Inclusive INCOMPL Incompletive INDEF Indefinite INDIC Indicative INESS Inessive ('in X') INF Infinitive INFER Inferential INSTR Instrumental INT Interrogative INTENT Intentive INTER] Interjection INTRANS Intransitive 10 Indirect Object IPFV Imperfective IRR Irrealis lTER Iterative LOC Locative MASC Masculine MKR Marker MOD Modal N Noun NEG Negative NEUT Neuter NEUTRAL Neutral NOM Nominative NOML Nominal NONPAST Non-past NONVIR Non-virile NUM NumeraI, Numeric OB] Object OBL Oblique OBLIG Obligatory OPT Optative PART Participie PARTIT Partitive PASS Passive

PAST Past PAT Patient PERF Perfect PERS Personal PFV Perfective PL Plural POL Polite POSS Possessive POST Postposition POT Potential PP Past Participle PRE Prefix PREP Preposition PRES Present PRESUM Presumptive PRET Preterite PRO Pronoun PROG Progressive PROL Prolative ('along X') PROLOC Prolocative PTL Particle PURP Purposive PVB Preverb Q Question QUANT Quantifier QUOT Quotative RC Relative Clause RECENT Recent RECIP Reciprocal REFL Reflexive REL Relative REM Remote REPET Repetition REPORT Reportative RES Resultative SG Singular SIMUL Simultaneous STAT Stative SUB] Subject SUBJV Subjunctive SUBL Sublative Conto X') SUFF Suffix SUPER Superessive ('on X') SUPERL Superlative T AGQ Tag Question TAX Taxis TEMP Temporal

ix

x

Format and Abbreviations for Glosses

TNS Tense TOP Topic TRANS Transitive TRANSL Translative ('becoming X') V Verb

VBLR Verbalizer VIR Virile VN Verbal Noun VOC Vocative VOL Volitional

Preface

In this fifth volume, we attempt to "expand the contexts" in which child language has been examined crosslinguistically. The chapters open themes that have been touched on, anticipated, and promised in earlier volumes of this series. The study of child language has been situated in the disciplines of psychology and linguistics and has been most responsive to dominant issues in those fields, such as nativism and learning, comprehension and production, eITors, input, and universals of morphology and syntax. The context has been, primarily, that of the individual child, interacting with a parent, deciphering the linguistic code. In these volumes, the code has been generally treated as a system of morphology and syntax, with little attention to phonology and prosody. Attention has been paid, occasionally, to the facts that the child is acquiring language in a sociocultural setting and that language is used in contexts of semantic and pragmatic communication. And there has been a degree of attention paid to the interactions between language and cognition in the processes of development. As for individual differences between children, they have been discussed in those studies where they could not be avoided, but such variation has rarely been the focus of systematic attention. Differences between individual languages have been of great interest, of course, but these differences have not often been placed in a framework of systematic typological variation. And although languages, and their grammars, change over time, the focus of attention on the individual child learner has generally led to neglect of explanatory principles that are best found on the level of linguistic diachrony, rather than the level of innate ideas or patterns of learning and cognition in the individual child. Here we seek to explore some of these neglected contexts in more depth. In my introductory chapter, "The Universal, the Typological, and the Particular in Acquisition," I propose that hinguages-and the processes by which they are acquired and used-fall into typological patterns. That is, each language is not totally different from every other, and the tasks of acquisition do not vary without limit. I suggest a mission for the crosslinguistic study of child language: "THE CHARTING OF CROSSLINGUISTIC DIVERSITY IN LANGUAGE ACQUISITION AND THE

xi

xii

Preface

EMPIRICAL DETERMINATION OF ITS PRINCIPLED LIMITATIONS." The expanded context in that chapter is typological analysis: "One cannot make claims about the acquisition or use of a grammatical form without situating it typologically, in a network of interactive psycholinguistic fa grammatical word > clitic > int1ectional affix Their discussion continues on the diachronie plane CHopper & Traugott, 1993,

p.7): Each item to the right is more clearly grammatical and less lexical than its partner to the left. Presented with such a cline, linguists would tend to agree that generally the points (labels) on the cline could not be arranged in a different order. ... It is often difficult to establish firm boundaries between the categories represented on clines, and indeed the study of grammaticalization has emerged in part out of a recognition of the general fluidity of so-called categories. It has also emerged out of recognition that a given form typically moves from a point on the left of the cline to a point further on the right ... 12Two overviews, both with the title Grammaticalization, have been provided by Heine, Claudi, and Hünnemeyer (1991) and Hopper and Traugott (1993). An early and insightful approach was developed by Bybee (1985), elaborated in successive papers with various collaborators, and most recently presented as The evolution (f grammar (Bybee, Perkins, & Pagliuca, 1994). Two volumes of conference papers, Approaches to grammaticalization, have been edited by Traugott and Heine (1991) and published in the John Benjamins Series, "Typological Studies in Language," which includes many books dealing with diachronie linguistic issues. The journal Language Variation and Change is a forum for diachronie research using statistical methods. The closely-related field of "typology and uni versals" places diachronie issues in a synchronie framework; see textbooks by Comrie (1981) and Croft (1990), and Linguistic Typology, the new journal of the recently established Association for Linguistie Typology.

5.

Origins of Grammaticizable Nations

299

The literature is full of examples of the lexical origins of grammatical items. Familiar English examples are the development of the verb go from a full verb of motion to a reduced future marker gonna, and the development of modals from verbs of cognition and ability-e.g., can originally meant 'know how to', may and might developed from a verb meaning 'have the (physical) power to'. The English contracted negative, n't, began in Old English as an emphatic form, ne-a-wiht 'not-ever-anything', used to reinforce another negative form, neo By the time of Middle English it had contracted to nat and eventually replaced the nonemphatic ne, becoming the new nonemphatic negative and finally contracting (Trau gott, 1972, pp. 146f). This is the typical progress along the cline-from full, stressed form with a more specific meaning, to reduced, unstressed form with a more general meaning. When such processes are traced out in full, the nature of grammatical morphemes-unstressed and general in meaning-is no longer mysterious. Here I will explore only one set of diachronic patterns in some detail, because it is relevant to several of the basic synchronic problems discussed in the previous section. 4.1. Origins and Extensions of Accusative Markers In the long history of Mandarin Chinese it is possible to see the entire developmental path from a lexical item to a grammatical morpheme (Lord, 1982, with examples from Li & Thompson, 1974, 1976). In the fifth century B.C. the verb ba was a full lexical verb meaning 'take hold of'. Much later, in the time of the Tang Dynasty (seventh-ninth centuries A.D.), it appears in serial-verb constructions, opening the way to reanalysis and eventual grammaticization. For example, the following sentence is open to two different interpretations: (8) Zui M zhü-gen-zi xi kim drunk M dogwood careful look

In the expected serial-verb interpretation the sentence means: (a) 'While drunk, (I) took the dogwood and carefully looked at it.'

However, a verb meaning 'take' can also be interpreted, in this context, as simply reinforcing the act of examining something that has been taken or held: (b) 'While drunk, (I) carefully looked at the dogwood.'

On this interpretation, ba has become a sort of object marker. Such possibilities of alternate interpretations open the way to the reanalyses that result in grammaticization. Sentence (8) "invites" a hearer to consider a single act-looking carefully, rather than two acts-taking and then looking carefully. This kind of "conversa-

300

Siobin

tional implicature" (Grice, 1975) or "pragmatic inferencing" (Hopper & Traugott, 1993) can set a full verb like ba off on the course toward a grammatical marker. And, indeed, that is what has happened in this case. In modern Mandarin, ba no longer has all of the syntactic properties of a fu\l verb: it can' t take an aspect marker and can no longer occur as the predicate of a simple sentence meaning 'take' (Li & Thompson, 1981, pp. 464ff). Now it functions as an objective casemarker in the frame, SUBJECT ba DIRECT.OBJECT VERB, as in: M jiu man-m~m-de he M wine slowly drink 'YOti drink the wine slowly!'

(9) nl

YOti

However, the ba-construction is still not a full objective or ACCUSATIVE casemarker. The process of grammaticization is long, and traces of the original meaning of a lexical item linger on to influence or restrict its grammatical function. The construction can only be appropriately used with aDEFINITE direct object, that is, to indicate a referent that the speaker believes the hearer knows about. And, most interestingly, it is further restricted to situations in which something happens to the object-in Li and Thompsons' terms (1981, p. 468), "how an entity is handled or dealt with." It cannot mark objects of verbs of emotion, like 'love' and 'miss', or verbs of cognition, like 'understand' or 'see', because these verbs do not imply manipulation or handling of the object. It may come to mark such objects at some future time, like accusative casemarkers in languages like German and Russian, but at present it still retains traces of its semantic origin. Lord (1982) describes almost identical grammaticization processes in several West African languages of the Benue-Kwa group. In Akan and Ga, a verb that meant 'take, hold, possess, use' no longer occurs as a verb in simple sentences and does not inflect for tense/aspect. It is now an invariant, noninflecting morpheme that functions as a casemarking preposition-but only when referring to physical manipulation and only in affirmative sentences (that is, manipulation that is actually realized). However, in a related language, Idoma, the corresponding morpheme has become aprefix and, although still restricted to affirmative clauses, can also mark the objects of experience, such as 'she PREFIx-tree saw'. Lord cites a parallel development in the Native American language, Chickasaw, which seems to be at an earlier stage in the process of using a verb meaning 'take' to mark instruments and objects that are moved by an agent. This is a brief summary of a long and interesting story, which also includes the development of transitive casemarking in ergative languages (e.g., Chung, 1977). But this is sufficient to raise several important points about learnability and nativism. It should now be evident why it is not possible to draw a line between lexical and grammatical items (Sections 3.1, 3.5.1). These continua do

5.

Origins of Grammaticizable Nations

301

not have natural cutting points between the two syntactic categories because of the successive types of change that go on, in language use, over time. The Idoma prefix is "more grammaticized" than the Akan and Ga prepositions in that it is phonologically reduced and has a wider semantic range of application. The Chickasaw prefix is "less grammaticized" than any of the West African examples in that its semantic range is narrower, but it is "more grammaticized" than Akan and Ga prepositions, in that it is aprefix. All of these, including modern Mandarin bb. are "less gramrnaticized" than Gerrnan and Russian casemarkers, in that they don't mark the direct objects of all possible transitive verbs. Turkish is like Mandarin in that the accusative inflection is restricted to definite direct objects, but it is like German and Russian and Idoma in that it is not restricted to objects of manipulation. Russian, however, is like Mandarin in that its accusative inflection is limited to affirmative clauses. It is unlikely that one of these language-specific variants of the "accusative" or "direct object" or "patient" category corresponds to children's initial assumptions about the "grammaticizable notion" underlying the object marker. Clearly, the child must be guided by the patterns of the exposure language. To be sure, all of these examples are consistent with a COLLECTION of "grammatically relevant notions"-definiteness, negation, manipulability, agent vs. experiencer-but there are too many different packagings of such semantic and pragmatic characteristics to build in all of the possible packages in advance or rank them in terms of "naturalness" or "accessibility." At the same time, there is something intriguing about the fact that a verb like 'take' can repeatedly develop into an object marker in languages that have nothing in common geographically or typologically. In earlier work (Slobin, 1981, 1985) I suggested that children might begin to relate accusative or ergative casemarkers to a general notion of "prototypical direct manipulation," that is, "the experiential gestalt of abasie causal event in which an agent carries out a physical and perceptible change of state in a patient by means of direct body contact or with an instrument under the agent's control" (1985, p. 1175). Verbs like 'take' clearly fit this definition. Can we conclude, then, that this is a privileged grammaticizable notion (Seetions 3.3, 3.5.3)? It is important here to distinguish between what is salient to the cognition and life experience of a 2-year-old and the processes that drive grammaticization in the discourse of adult speakers of a language. The "manipulative activity scene" is central to a 2-year-old's interaction with the world, and grammatical markers that regularly occur in conjunction with such events may weil come to be associated with the notion of manipulation or direct effect on an object. But adult speakers of Tang dynasty Chinese, Akan, Ga, or Chickasaw do not set out to grammaticize manipulation or effect. Rather, they use a verb like 'take' in constructions that allow it to be interpreted as a marker of manipulation, and, over time, such verbs follow the familiar cline described by Hopper and Traugott. The processes, then, are quite different, though

302

Siobin

superficially similar. 13 In order to fill out the picture, it is necessary to understand the psycholinguistic forces that move a linguistic element along that cline. Before turning to those forces, though, let us consider one more diachronic example, which reveals a different type of gramrnaticization path for accusative markers. Quite another history of an accusative marker can be found in Persian (Hopper & Traugott, 1993, pp. 158ff, based on Bossong, 1985). Here the origin is a noun, radiy, meaning 'goal' or 'purpose'. Given the semantics ofthis starting point, the path does not begin with manipulation. As a result, much of the path is different, yet there are similarities to the examples we have just reviewed. In Old Persian (c. 600 B.C.) this noun is used as a postposition. By Middle Persian it is reduced to a suffix, -rao, used as a casemarker for dative-benefactive objects. In New Persian (beginning in the ninth century A.D.) it is reduced further, to the suffix -ra, and its use includes direct objects-but only if definite. At this point, then, its meaning is partly similar to the grammaticization paths reviewed above, but, because of its origin, the Persian suffix also marks dative-benefactive objects. In Classical Persian (twelfth-fourteenth centuries A.D.) the dative-benefactive uses expand to include marking of possessors and experiencers. At the same time, the accusative uses expand from marking individual human objects that are affected by an action to include inanimates and, eventually, indefinites. Note that these extensions are as much pragmatic as semantic. Hopper and Traugott point out that this follows a cline of discourse topicality, from animate, human participants to inanimate objects, and from referential to indefinite topics. In Modern Persian the suffix has become restricted to direct objects only, losing all of its capacity to mark dative-like indirect objects. It has thus become a "standard" sort of general accusative marker. These several paths of accusative development raise critical questions for the Conditions and Assumptions of Section 2. In Persian, as in the earlier examples, each step in the long evolution can be motivated by semantic and discourse factors-but at which point does the form mark a "true grammaticizable notion," and at which point is it a "true grammatical morpheme"? Which of the many "accusatives" in all of these language histories is the one to put on Pinker's innate chart of form- function correspondences? Which of these many historically attested grammatical morphemes corresponds to "core notions" like manipulation or purpose or goal? And, to return to the issue of 1anguage sampling, if the earlier versions of Persian were lost from the written record, would we be able to accurately assess the frequency of conflations of certain types of accusative with certain types of dative-benefactive casemarking? Or if the restricted Mandarin and West African object markers evolve into a Persian-type accusative by the twenty-second century, thereby reducing the number of such object markers in the sample of the world's languages, should linguists propose at that point in history that such object markers are "less accessible" to children? BFor a similar argument against equating processes of ontogeny and grammaticization, see my discussion of the development of the English PERFECT (Slobin, 1994).

5.

Origins of Grammaticizable Notions

303

4.2. Psycholinguistic Forces Responsible for Restrietions on Grammaticizable Nations Grammaticization paths such as those just sketched out take place, to begin with, in the processes of communication. Therefore they are shaped by the onIine demands on the speaker to be maximally c1ear within pragmatic constraints and maximally efficient within economy constraints, and by onIine capacities of the Iistener to segment, analyze, and interpret the message. Experimental and theoretical psycholinguists have learned much about these processes, in a Iiterature far too large to cite or review here. It is c1ear that pressures of expressivity, economy, and c1arity are always in competition, keeping language always changing in shifting states of balancing equilibrium (e.g., Slobin, 1977; Bates & MacWhinney, 1987; Bybee, 1985; Hawkins, 1983, 1995; MacWhinney & Bates, 1989). This is not the place for a detailed exposition of the psycholinguistic bases of language change. What I want to do is point to some psychoIinguistic processes that seem to account for the pecuIiar semantic Iimitations on grammaticizable notions that Talmy and others have discussed.

4.2.1. Frequency of Use and Genera/ity of Meaning Bybee (1985; Bybee, Perkins, & Pagliuca, 1994) has explained much of grammaticization in terms of the fact that lexical items that are used with high frequency also have general meanings. For example, motion verbs such as crawl, limp, hobble, creep, slither, wriggle are appIicable to describing a small number of situations and are, accordingly, not very frequent. By contrast, generalized motion verbs Iike come and go do not have such restrictions; they are applicable to a wide range of contexts and are used frequently. GeneraIity of meaning and frequency of use go hand in hand-both in the "open" and "c1osed" c1asses. For example, compare highly frequent EngIish prepositions Iike in and on with less frequent and more specialized propositions such as alongside, underneath, in back oj, throughout. The latter require more detailed attention to the geometry of the objects involved, and are therefore applicable to more Iimited contexts.

4.2.2. Frequency of Use and Reduction of Form It is also a commonplace that any motor pro gram that is called upon frequently is reduced and automatized. Zipf (1935) demonstrated the strong tendency for the length of a word to be negatively correlated with its frequency. Note that the more specialized prepositions just listed are also much longer than in and on. They are also more etymologically transparent-that is, they still have recognizable lexical components, inc1uding nouns (side, back) and more frequent prepositions (in, of, etc.). It is no mystery that as lexical items move along the grammaticization cIine they become phonologically reduced and bound to associated content words.

304

Siobin

4.2.3. Frequency of Use and Decidability In order for a speaker to express any notion in language, it is necessary to make a rapid decision with regard to the appropriate means of expression of that notion. Elements that are highly frequent and general-both content words and grammatical forms-must be easily accessible to online processing for both speaker and listener. Again, the same processing demands apply to content words and grammatical forms alike. Eve Clark (1978) has observed that early in English child language development the most frequent verbs are go, put, get, do, and make. She reports similar findings for Finnish, French, Japanese, and Korean. This pattern probably reflects the high frequency of such verbs in adult speech, but the fact of their early frequency also bears on the issue of decidability. In Clark' s examples, when a child says Do it! it might apply to unrolling some tape, taking out a toy, or building a tower. Make + NOUN can mean write, draw, move, cut out, and so forth. Do and make place low demands on decidability. In order to say write or draw or cut out the speaker must decide what kind of act of construction is involved, and determine the distinctions that are lexicalized in the language (for example, in some languages a single verb means both 'write' and 'draw'). These same sorts of "light" verbs appear as the sources of grammatical morphemes, as the examples of 'go' and 'take' discussed earlier. Hopper and Traugott (1993, p. 87) present this as a general fact of grammaticization: As we have noted in previous chapters, the lexical meanings subject to grammaticalization are usually quite general. For example, verbs which grammaticalize, whether to case markers or to complementizers, tend to be superordinate terms (also known as "hyponyms") in lexical fields, for example, say, move, go. They are typically not selected from more specialized terms such as whisper, chortle, assert, squirm, writhe. Likewise, if a nominal from a taxonomic field grammaticalizes into a numeral ciassifier, it is likely to be selected from the following taxonomic levels: beginner (e.g., creature, plant), life form (e.g., mammal, bush), and generic (e.g., dog, rose), but not from specific (e.g., spaniel, hybrid tea), or varietal (e.g., Cocker, Peace) (Adams & Conklin, 1973). In other words, the lexical items that grammaticalize are typically what are known as "basic words." Again, it is an illusion that child language development and grammaticization are due to the same sorts of processes. Children use basic verbs early on because they are easy to learn: They do not place high demands on decidability; they are frequent; they are used across a wide range of situations; they are short. But basic verbs appear at the beginnings of grammaticization clines because, when they are used in a conversational context, they contrast with the more specific verbs that COULD be used in that context, thereby allowing the hearer to infer that a more specific meaning may not have been intended. This opens the way for the kinds of pragmatic inferencing and reanalysis that lie at the heart of grammaticization.

5.

Origins of Grammaticizable Notions

305

Given these facts, it is evident that the special character of grammaticizable notions has its origin, in part, in the lexical items from which grammatical markers are pro ne to develop. That is, the "open c1ass" is already organized into general and specialized terms-and this division can be accounted for by quite ordinary psycholinguistic and communicative processes. There is no need to postulate a special "grammar module" as responsible for these facts about the meanings of frequent lexical items. Why are such words prone to grammaticize? Because of their generality they are both highly frequent and likely to be used in situations in which the speaker either does not intend a more specialized meaning or assumes that such a meaning can be reliably inferred by the listener as background information in context. If I say, for example, "While drunk, I GRASPED the dogwood" or "SEIZED the dogwood," the choice of a specialized verb of taking or holding suggests to the hearer that I wish to focus on the manner of taking or holding. This is simply an application of Grice's second maxim of Quantity: "Do not make your contribution more informative than is required" (Grice, 1975). The hearer may weIl assume that I used a specialized verb because I intended to draw attention to the manner of acting. If, however, I use a more general verb like 'take', the hearer is likely to assume that I have followed the first maxim of Quantity, "Make your contribution as informative as is required (for the current purposes of the exchange)," and will not attend to the manner of taking the dogwood. In fact, following this maxim, the hearer might arrive at the interpretation given earlier in example (8b )-that is, backgrounding the fact of taking entirely and focusing on the act of looking, which is, after all, what may be relevant in this communicative situation. In such situations, the way has been opened for the grammaticization of 'take' as an object marker.

4.2.4. Frequency of Use and Schematicization ofa Domain If a small set of linguistic items ends up being used frequently to reference divisions within a semantic domain, pressures toward easy decidability will inevitably move the system toward a schematic representation of that domain, selecting a set of parameters or features for sorting instances. The most familiar example of schematicization is an inflectional paradigm, in which slots are filled in for such features as person and number, or case and gender, and so forth. But schematicization is also evident in linguistic systems which might appear to be more lexical than grammatical. A good example is Levinson's (1994) analysis of the Tzeltal use ofbody-part terminology to locate an object in relation to a ground. In English we have suggestions of such a system in grammaticized expressions such as in back of the house and lexicalized descriptions such as the foot of the mountain. In Tzeltal, as in many Mesoamerican languages, body-part terms are used systematically to specify the grounds involved in locative relations. One says, for example, that an object is at the 'ear' (= corner) of a table or at the 'butt' (= bottom) of a bottle (P. Brown, 1994, p. 750). Levinson (1994) shows that choice

306

Siobin

of body-part term is based on apreeise geometrie schematicization of objects. For example, the base of an are defines the 'butt' of an object, including the large end of a pear, the bottom of a bowl, and the point where astern is attached to a leaf. If an object has two surfaces, the flatter, less-featured surface will be labeled 'back' and the opposing surface will be 'belly' if concave or convex and 'face' if flat. In order to use the system, in Levinson's analysis, the speaker must carry out aseries of algorithms, such as finding the orthogonal axis, finding the direction of the subsidiary arc, and finding junctures between surfaces. He proposes that by applying such calculations of the intrinsic shape of an object, speakers know how to use the body-part terms with regard to any particular object. Thus, Tzeltal words like 'butt', 'ear', 'belly', and 'neck' are as fully grammaticized as English prepositions. They constitute a smalI, closed set, with schematized representations of those features of their spatial characteristics that are used in the language to specify locative relations of particular types. Because body-part terms must be used to designate parts of any object in the world-doors, tables, computers, chile beans-speakers must be able to decide easily which term to apply to which part or surface of an object. Such a system cannot simply leave the speaker to pick a body-part term and search for a possible metaphorical extension; nor can it leave the speaker to use all possible body parts. Out of several hundred such terms, the language uses about 20 to label parts of inanimate objects. In order to apply this small set to all possible objects, there is no choice but to develop a way of schematizing their meanings within a structured semantic domain. Again, the system is a compromise between possible conceptual differentiation and the demands of online production and comprehension. The results of compromises of this sort cannot be built into the language module, but arise in processes of language use. 14 4.3. A Functionalist Account of the Classes of Grammaticizable and Non-Grammaticizable Notions If a domain is to be divided up such that each of the subcategories can be rapidly accessed online, by speaker and hearer, there cannot be too many divisions in the domain, nor can the deciding factors be infrequent or idiosyncratic. Typically, as forms become highly grammaticized, they divide up a domain exhaustively into a very small number of options: SINGULAR VS. PLURAL (with possible additions of DUAL), PERFECTIVE VS. IMPERFECTIVE, the six cases and three genders of Russian. Markers such as these are obligatory, which means they must be accessed in almost every utterance. There can be no ambiguities of online decidability. The notions that evolve into such very small and obligatory sets must (1) unambiguously divide 14Levinson proposes that the sehematieization is given by the visual system, thus raising a problem for modularity theories: "Aeeording to modularity arguments, linguistie proeesses should have no aeeess to strietly visual proeesses. Although the present faets are not deeisive, together with other observations they favor models where there is shared linguistie and visual aeeess to the underlying proeesses of volumetrie shape analysis" (1994, p. 791).

5.

Origins 01 Grammaticizable Nations

307

the domain, and must (2) use criteria that are generally relevant to that domain. Thus it is no mystery that grammatical inflections do not indicate color or rate or ambient temperature: these are not generally salient aspects of experience; they are not universally decidable, applicable, or memorable with regard to all of the event types that we talk about. That is, they are not aspects that are relevant to how we interpret and store events IN GENERAL. In order, for example, to grammaticize a temperature marker or a color marker, it would be necessary, first, to have a speech community in which lexical items of temperature or color occurred frequently in discourse, and in which there were a few general terms that marked readily agreed-upon distinctions, such as cold - cool - warm - hot, or black - white - red yellow - blue/green. Such scenarios are unlikely for several reasons. For one, these distinctions are not relevant to most of human discourse. The things that we care to communicate about, by and large, are true on cool and warm days; the things we act upon are important regardless of their colorY Because we don't tend to store such information in memory, such a language would place terrible burdens on decidability. For example, if Iwanted to tell you a juicy bit of gossip, I would have to remember whether the reported event (or the time of my hearing about it) occurred on a warm or cool day. Or when a newscaster reports a bomb explosion in the Paris Metro, he would have to know the color of the bomb, or the Metro, or the explosion. We do not grammaticize such notions because we do not think or talk in such terms. "Lower" on grammaticization clines there are relatively small sets that provide options. For example, as discussed in Section 3.1.1, English has a small set of modal auxiliaries, supplemented by quasi-modals and some less clearly grammaticized terms. Most of the time modality can be simply left unmarked. The "zero option" means that it is not necessary to decide about the modality of every utterance. A similar function is provided by the general classifier ge in Mandarin. When a speaker does choose to mark modality in English or to classify a class in Mandarin, a small set of terms is provided, with more flexibility in their applicability. Erbaugh (1986) finds about 22 classifiers in ordinary speech in Mandarin, and she reports that the same object occurs with different classifiers in discourse; e.g., different speakers viewing the same film referred to a goat with the classifiers Yl-zhl 'one.anima!', Yl-to 'one.head', and Yl-tfao 'one.long.thing'. The choices in a set like the Mandarin classifiers do not unambiguously divide up a domain, but they are still semantically relevant to the nouns that are marked. Relevance, however, does not have to be part of a UNIVERSAL or INNATE human "semantic space." There is nothing in the nature of our cognitive and linguistic systems that precludes grammaticization of idiosyncratic information if it assurnes sufficient social or cultural relevance to be regularly communicable. For example, social structure is repeatedly grammaticized in choices of personal pronouns and 15Although, as mentioned in footnote 4, presence or absence of color may be more salient than distinguishing between colors.

308

Siobin

verb inflections. Although a European speaker would find it hard to decide whether each person addressed is older or younger than the speaker, this is obligatory in Korean, and children learn to pay attention to this feature. English speakers in France might find it hard to decide if an interlocutor falls into the tu or the vous category, and, as Roger Brown and Albert Gilman (1960) have shown, the criteria for choosing one of the two pronoun types has changed historically and varies between European countries. The languages of the world grammaticize an array of social categories of rank, status, relative age, servitude, and the like. These are sociocultural facts, and could not possibly be part of the child's innate linguistic categories or prelinguistic sensorimotor concepts. Yet they are grammaticized in those societies where they are relevant, and are marked with a small number of forms that are frequent and decidable online. The reason why languages have no grammatical markers for quantified categories of "fixed distance, size, contour, and angle" (Talmy, 1988, p. 171) is simply becausehuman beings do notregularly code, store, and report their experience in these terms-not because these categories are apriori excluded from the grammar module. I would suggest, then, that anything that is important and salient enough for people to want to refer to it routinely and automatically most of the time, and across a wide range of situations, CAN come to be grammatically marked, within the constraints of online processing briefly alluded to earlier. 16 I believe that similar arguments could be made with regard to each of the "conceptual domains NOT accessible to grammaticization," such as those listed in (l b). These arguments would draw on the factors of across-the-board relevance

lfiThe discussion of grammaticizable notions in the literature has focused almost entirely on "synthetic" language types ("inflecting" and "agglutinating"), and primarilyon inflectional, rather than derivational categories in those languages. "Isolating" and "polysynthetic" languages pose problems for the division between "grammatical" and "Iexical." In Talmy's analyses, serial verbs are treated as "satellites" in isolating languages such as Chinese, on the assumption that a "main verb" can be identified in a verb series. Similarly, grammatical morphemes in polysynthetic languages such as Atsugewi are also treated as "satellites" in relation to a central verb stern. Verbs in serial-verb constructions, and affixes in polysynthetic languages, are often more "contentful" or "Iexical" in their meanings, while still having a degree of generality and abstractness that characterizes the meanings of grammatical morphemes in synthetic languages. As such, they gualify as notions that speakers can refer to routinely and automatically across a wide range of situations. For example, Marianne Mithun (personal communication, 1996; also see 1989) has commented on cuiturally specific notions that can be marked by affixes in Amerindian languages: In Central Porno, the verb root ley 'exhaust, use up', when prefixed with c"- 'by gambling', results in c"ey 'lose all in gambling'; when prefixed with cu- 'by sitting', the resuit is cu!ey, which Mithun glosses as 'to wear a hole in the seat of your pants' or perhaps 'wear out the upholstery in the driver's seat of your car'. She also notes that Yup'ik Eskimo has a suffix that means 'have a cold NOUN', which combines with body-part noun roots such as 'nose' or 'toe' to yield 'have a cold nose', and so on. Thus, when one extends the definition of "grammatically marked," it seems that anything that is culturally significant can be expressed by grammatical morphemes associated with "content words." (The high incidence of "scare guotes" in this footnote indicates the degree to which all of the guoted terms are inadeguate to the task of sorting out the relations between grammar and cognition!)

5.

Origins of Grammaticizable Nations

309

to human experience and communication, online decidability, and the availability of high-frequency and general lexical items that could start off paths of grammaticization in those domains. I leave it to the reader to try to find examples or counterexamples.

5. CHALLENGES TO LEARNING THEORY 5.1. The Conditions and Assumptions of the Learning Task In Section 2 I listed three linguistic Conditions and three psychological Assumptions underlying standard definitions of the task of learning to use grammatical morphemes. It is time to return to that starting point. 5.1.1. The Conditions Condition 1: There is a distinct and identifiable collection of grammatical morphemes, arranged in small, closed classes. It turns out that there is a eh ne of linguistic elements, arising naturaUy over time, and that the "distinct and identifiable collection of grammatical morphemes" only defines the endpoint of that cline. However, looking at an entire language, one can only rank elements on various dimensions, both formal and functional. There are, to be sure, many smalI, closed and semi-closed sets of items-but they are not all grammatical morphemes. Thus the language does not present itself to the learner as a neat set of little packages labeled as "grammatical" and "lexical." Condition 2: These morphemes map onto a universal, limited set of semantic entities (grammaticizable notions). The further one moves to the right on the cline, the more true is this condition. And there are regular diachronie progressions of particular types of meanings toward the highly grammaticized pole of the cline. Condition 3: Grammaticizable notions are arranged in a universal accessibility hierarchy. If "accessible" means either "learnable" or "more frequent in human languages," we lack the data to evaluate this condition. If "accessible" means that some notions are more likely to grammaticize than others, the claim can be filled out with more and more data, and the patterns are amenable to explanation in terms of such interacting factors as online processing, pragmatic inference, and syntactic reanalysis. 5.1.2. The Assumptions Assumption 1: Conditions 1, 2, and 3 exist because of the structure of the mindJbrain (in modules for aspects of language, perhaps in conjunction with other modules). There is a great deal of evidence that the Conditions exist because of conditions on the processing, social use, and learning of form-function rela-

310

Siobin

tions. Such evidence greatly reduces the role of apriori specification of grammatical structures and their specialized meanings. Assumption 2: The role of linguistic input is to allow the relevant mental capacities to organize themselves in terms of the exposure language. This, of course, remains true-but relativized to the definition of "relevant mental capacities." Linguistic diversity in the domains considered here precludes a simple selection between prespecified alignments of formal and semantic categories. The role of linguistic input is to guide the child toward discovery and construction of the form-function relations inherent in the exposure language. That is, input is not a "trigger" but a "nutrient." Assumption 3: The child leams the meaning of a grammaticalform by isolating and identifying a particular stretch of speech as instantiating a grammatical form and attempting to map it onto a relevant grammaticizable notion. This formulation is built upon apriori definitions of "grammatical form" and "relevant grammaticizable notion"-the very concepts that demand reanalysis. The result of that reanalysis is, of course, the challenge to learning theory. 5.2. Toward a Solution It is not (and cannot be) the goal of this chapter to answer these challenges by

presenting The Adequate Learning Theory. At best, a reorientation might serve to head us toward different kinds of solutions. Once we have established a social-historical, rather than an individual-mind source of grammaticized notions and their means of expression, we can abandon the search for an innate formfunction module and follow Annette Karmiloff-Smith (1992) "beyond modularity." That is, we can take a DEVELOPMENTAL approach to the structuring 01' grammaticizable notions in the child. A major theme that emerges from the reanalysis is the proposal that the same learning mechanisms apply across the lexicon, including "content words" and "functors." To be sure, the child requires specialized mechanisms of perception (auditory for speech, visual for sign), storage, and analysis 01' linguistic material. And the ARCHITECTURE 01' syntax is certainly determined by quite different processes than those involved in learning the kinds of form-function mappings considered here. The reanalysis of the learning task places "grammaticizable notions" in the more general domain of concept formation. 5.2.1. The Problem of Constraints on Hypotheses: What Is "Economy"? Regardless of the revision of the task definition, the child will always be faced with a large set of possible form-function mappings. My very brief overview 01' a few problems of grammatical marking 01' semantic and pragmatic categories makes it evident that the child could be prey to many false starts and dead-end

5.

Origins of Grammaticizable Nations

311

attempts. This fact alone has led to the proliferation of "constraints," "predispositions," "parameter settings," "operating principles," and the like in the theoretical literature of recent decades. But there are no obvious constraints on the constraints, because we have no plausible metric of what makes a task "too hard" for a child learner. We know that children do acquire the manifold and subtle complexities of language. And we realize that this is a hard task for conscious, problem-solving adults (even linguists). Therefore we try to make the task "easier" for children by providing bootstraps that they can use to pul! themselves up with (an unclear metaphor at best). The list of grammaticizable notions was intended to provide an aid-intended to prevent the child from making too many false hypotheses. But, I would propose, we real!y have no way of knowing how many false hypotheses it takes to overburden the vastly complex human brain, or how quickly and efficiently they can be revised or dismissed. It is unsettling to realize how many of our theories are aimed at the simplistic criterion of "economy," when we have no rational measure of that economy. We have been cal!ed upon, in our training, to apply Occam's Razor to our theories. This is entirely reasonable: "What can be done with fewer [terms] is done in vain with more." But remember that William of Occam was concerned with choosing between theories that are equivalent in accounting for the data and differ only with respect to their complexity as theories. This is an aesthetic criterion, and has nothing to do with the functioning of organisms or machines or planetary systems. The technology of producing and marketing goods has given us the criterion of economy or efficiency-a standard based on time, cost, and value, and not elegance of theoretical formulation. These two norms, parsimony in explanation and economy in the marketplace, have become confused in our evaluations of psychological explanations. We are not in a position to apply Occam's Razor, because we do not have learning theories which are equal in accounting for the data, differing only in formal complexity. (In fact, we are at the stage in which all of our theories are vastly underdetermined by the available data.) If we, accordingly, abandon Occam and try to establish a metric of the efficiency or economy of a complex biological system, functioning with a complex language and culture, it should be evident that we have neither ground rules nor plausible criteria, and that therefore arguments based on "economy" are illusory.

5.2.2. What Is "Reasonable"? Having voiced these qualms about the soundness of our endeavor, I return to the attempt to give the child some guidelines for the task. Our theories are haunted by the risk that children might think that everything might be relevant to everything. Our data, however, suggest that children are more "reasonable" than that. Frank Keil' s (1994) discussion of the general problem of concept formation is helpful in placing our problem in a larger framework:

312

Siobin

People do not simply note feature frequencies and feature carrelations; they have strong intuitions about which frequencies and correlations are reasonable ones to link together in larger structures and which are not. Without these intuitions, people would make no progress in learning and talking about common categories given the indefinitely large number of possible correlations and frequencies that can be tabulated from any natural scene. These intuitions seem much like intuitive theories of how things in a domain wark and why they have the structure they do ... (p. 173)

Keil summarizes results of experiments on children's understanding of word meanings (Keil, 1989), giving the following example of children's "reasonableness": Even the youngest children were never simply tabulating up all salient feature frequencies and correlations. No child thinks that uncles must have glasses even if all the uncles they happen to have seen wear them. The features selected by even the earliest word learners were always constrained by some notions of reasonableness 1'ar the kind of thing in question (Keil, 1994, p. 177). Where might such "reasonableness" come from in learning the specialized meanings of linguistic items? How does the child know the reasonable factors to consider when encountering, say, verbs of motion or locative particles or casemarkers? Recall the diachronic pro ces ses of grammaticization (and, I would add, the processes of forming small sets of specialized verbs). The only available items are those which occur again and again in talking about a great range of experiences. They occur so frequently because they are applicable so generally . Therefore it should be no surprise that children find these notions salient. For example, the factors that apply to many instances of moving and placing objects include the force-dynamic and motoric aspects of picking up an object, moving it, and placing it in another location. It is "rcasonable" for grammatical items and small verb sets dealing with these actions to be sensitive to such factors as characteristics of figure and ground objects, direction of movement, and relation of the two objects at the endpoint of the action (e.g., tight fit, located near the bottom of another object, etc.).17 The color of the objects or the amount of daylight are not RELEVANT to this type of scene, just as eyeglasses are not relevant to the social and kinship status of uncles. In part, children are reasonable because languages are reasonable. It has been assumed in the literature that it is odd that systems of grammatical meaning, and children acquiring such systems, seem to be indifferent to "non-grammaticizable" notions such as those listed by Talmy (in Ib). However, if we look carefully at the communicative contexts in which 17What was, in retrospect, not "reasonable" was my Platonic hope that all children would start with the same semantic notions-the "Basic Child Grammar" of Slobin (1985). This issue must be explored in detail, domain by domain, before we can make any claims about the range and types of particular starting points in particular domains across languages and children.

5.

Origins of Grammaticizable Notions

313

language is used-both on the diachronic and ontogenetic planes-the situation seems much less odd. I suggest that the same factors that keep certain notions from becoming grammaticized also keep children from postulating them as the meanings of grammatical forms. There are at least two parts to this argument: frequency and relevance. For example, particular colors do not occur frequently in association with the linguistic encoding of particular event types. The child is not likely to encounter one set of object placement events that consistently occur with red objects and another with black ones. Even a simple model of statistical sampling, not to mention a connectionist network, would quickly drop color as a determining factor in choice of linguistic form-even if it is salient to the child in particular situations. Other features, however, do occur frequently. For example, every event takes place in an ambient temperature, and many events occur on hot days that do not occur on cold days. But, given the ways in which humans interact with each other and with the world, ambient temperature is not relevant to most of our behaviors. We put objects in containers, experience the visual world, maintain social status relations, and so forth, on days which are hot or cold, in daylight or at night. There is no reason to build into a language module factors that are basic to human life and action in general. 5.2.3. Why Ontogeny Does Not Recapitulate Diachrony

Paralleis between development of language in the individual and changes in languages over time have intrigued linguists for a very long time. With regard to the phenomena under investigation in this chapter, the parallel that has attracted attention is the fact that some of the notions that are salient to small children are also salient in the process of grammaticization. But it is important to underline, yet again, an important difference between aspects of experience that are salient to the child and those that end up as grammaticized notions. Consider features of temperature-which are NOT grammaticized in the languages that we know of. Children certainly do orient to temperature-both to temperature of objects and temperature of the surrounding air. And a child may very weIl expect, on first hearing a new linguistic item (either lexicalor grammatical), that it relates to a salient temperature experience. How else could children !earn adjectives like hot and cold? However, given the nature of human culture and interaction, temperature does not occur regularly and frequently with regard to the GRAMMATICAL structuring of utterances. In the presence of a hot object, mommy might say, "Don't touch it, it's hot." The lexical item hot DOES cooccur repeatedly with the salient experience of temperature-but touch, don't, and the syntactic constructions of subject-object relations and negative imperative do not. This is because most of the time temperature is irrelevant to human speech and action. Note, however, that this is not because children think that temperature is an unimportant feature of the world. (Nor do we know whether children might,

314

Siobin

fleetingly, postulate temperature as a possible grammaticizable notion.) To repeat my theme, the child does not recapitulate the history of the language, nor is the history of the language based on what is salient to the child. Rather, the only meanings that can end up as grammaticizable notions are those that are general, applicable, decidable, and so forth, as discussed above. Speech communities are at work (very slowly) in modifying grammars. Children are at work (quite quickly) in mastering ALREADY EXISTING grammars.

5.2.4. Iconic Bootstrapping There is a different sort of "relevance" that also plays a role in form-function mappings. It has long been noted by linguists that grammatical morphemes are placed in association with the content words with which they have the most conceptual affinity-for instance, tense is marked on verbs rather than nouns, shape elassifiers are placed in relation to object nouns or verbs of handling, and so forth. A elassic formulation of this principle was offered by the German linguist Behaghel (1932, p. 4): "What belongs together mentally is placed elose together syntactically."18 Bybee (1985) has refined the principle, showing not only that particular notions are relevant to verb sterns, but that grammatical morphemes reflecting such notions are ordered in a reasonable way, with those meanings that are most relevant to the meaning of the stern occurring elosest to the stern. The details are not important here, but Bybee's conelusion gives the child another part of a "reasonableness bootstrap" (pp. I1f): Verbal inflections differ with respect to the extent to which they are RELEVANT to the verb, that is, the extent to which their meanings DIRECTLY AFFECT THE LEXICAL CONTENT OF THE VERB STEM. The different degrees of relevance of verbal categories that can be inflectional is reflected diagrammatically in three ways: (I) The more relevant a category is to the verb, the more likely it is to occur in a synthetic or bound construction with the verb; (2) The more relevant a morphological category is to the verb, the c10ser its marker will occur with respect to the verb stern; (3) The more relevant a morphological category is to the verb, the greater will be the rnorpho-phonological fusion of that category with the stern. Bybee's analysis is part of aseries of discoveries of the "inconicity" of formfunction mappings in language (e.g., Haiman, 1985a, 1985b). To the extent that the arrangement of linguistic items is a "diagram" or "icon" of the arrangement of mental items, the child may be aided by "iconic bootstrapping." There are many examples of iconicity in children's early grammars, across languages, summarized in Slobin (1985). 18"Das oberste Gesetz ist dieses, daß das geistig eng Zusammengehörige auch eng zusammengestellt wird."

5.

Origins of Grammaticizable Nations

315

5.2.5. Typological Bootstrapping

Keil's analysis of concept formation emphasizes that children are building up "explanatory systems" that are relevant to c1asses of phenomena (cf. Gopnik & Meltzoff, 1996, on children as "theorists"). He proposes (pp. 178f): There are shift, m understanding which explanatory system is most relevant to a class of phenomena. A child might realize that 'uncle' is better understood in terms of a set of biological relations that comprise kinship and not in terms of social relations that govern friendship; the social explanation is discovered to generate more serious mi stakes and is abandoned.

This example comes from the learning of word meanings, but the framework is applicable to learning the meanings of grammatical morphemes and constructions as weil. Wh at is important in this approach is the child's ability to learn "which explanatory system is most relevant to a c1ass of phenomena." I suggest that as the child develops a successful explanatory structure for part of the exposure language, other parts become more accessible-that is, a coherent theory of the language begins to emerge. This is true, in part, because the language really IS a fairly coherent system-as a result of constant balancing out of competing forces. Over time, each language acquires a typological character resulting from the particular interplay of forces in its history. (There is a small number of language types, but this is not because there is a small number of innate parameter settings; rather, there is a small number of solutions to the kinds of competing forces which shape language in use.) At the risk of overburdening the child's shoe-rack, I propose yet another kind of bootstrapping: TYPOLOGICAL BOOTSTRAPPING (for further discussion, see Siobin, this volume). Consider verbs of motion again. As the Korean-speaking child 1earns more linguistic constructions describing motion events, the lexicalization patterns and grammaticized notions of the language become an established pattern. She comes to expect that paths will be lexicalized in verb sterns, that caused-motion verbs attend to factors of tightness of fit, and so forth. The English-speaking child comes to expect verb partic1es to structure domains in terms of locative and temporal relations, and finds that certain locative and temporal notions occur again and again. That is, to some extent, the language structures itself as it is learned. Certain patterns of semantic and formal organization become more and more familiar, and, to use an old term, habits are established. This is possible because of the fact that languages naturally develop into coherent systems of various types. In the process of learning various pieces of the system, they come to interrelate because of inherent typological factors. In Karmiloff-Smith' s (1992) terms, "representational redescription" occurs-in this case aided by the systematicity inherent in the language that is being learned. (I am aware that this formulation still leaves open the mechanisms that a child might use to detect

316

Siobin

such systematicity and "representationally redescribe" it. The goal of this chapter is to reformulate the question.) An intriguing consequence of typological bootstrapping is that children come to formulate experience for linguistic expression in quite different ways, depending on the type of language they are learning. I have suggested that each type of language fosters its own modes of"thinking for speaking" (Slobin, 1991, 1996a). Because of the systematic crosslinguistic diversity in selection and patterning of grammaticizable notions, different patterns of online mental organization result. In crosslinguistic work on narrative development, Ruth Berman and I have identified a number of ways in which children come to structure dis course in terms of the typological characteristics ofthe particular language (Berman & Slobin, 1994). By school age, children have acquired typologically distinct ways of describing events and constructing connected texts. From this point ofview, grammaticizable notions have a role in structuring language-specific mental spaces, rather than being there at the beginning, waiting for an input language to turn them on. I am aware that this formulation stillleaves open the mechanisms that a child might use to detect and "representationally redescribe" the systematicity of the exposure language. Various sorts of "operating principles" and "procedures" will be needed in order to give substance to the formulation. However, the very fact that form-function relations become systematically patterned in the course of acquiring a particular language points to an importance learning mechanism. As suggested above, in the course of development the child comes to attend to particular types of meanings and to expect them to be expressed by particular types of forms. Such a combination of thinking for speaking and typological bootstrapping seems to guarantee that language-specific form-function patterns will be established and maintained by learners.

5.2.6. The Problem of Primitives The child learner must begin the task of form-function mapping with SOME semantic notions, and must continue to employ semantic hypotheses in the process of being "guided" by the language to the specific notions that are packaged into linguistic forms. In 1985, I assumed that the starting points were unanalyzed semantic configurations or prototypes, and that language-specific categories would emerge later (p. 1174): When functors are first acquired. they seem to map more readily onto a universal set of basic notions than onto the particular categories of the parental language. Later in development, of course, the language-specific use of particular functors will train the child to conceive of grammaticizable notions in conformity with the speech community ... At first, however, there is considerable evidence that children discover principles of grammatical marking according to their own categoriescategories that are not yet tuned to the distinctions that are grammaticized in the parental language.

5.

Origins of Grammaticizable Notions

317

It is c1ear from the argument of the present chapter, however, that the picture is not that simple. We have evidence both for language-specific categories from the start, and for idiosyncratic categories that do not directly reflect the exposure language. Eve Clark (in press) discusses the latter as "emergent categories." These are "categories that surface fleetingly during acquisition and then vanish again." She notes, for example, that a general notion of SOURCE appears in early child speech in various languages, such as the use of a locative preposition (English fram, Dutch van, Italian de) to mark agents of actions, possessors, and standards-of-comparison-although this array of meanings is not a unitary category in the exposure language (for details, see Clark & Carpenter, 1989a, 1989b). The "prototypical scenes" of Slobin (1985) can also be seen as emergent categories: manipulative activity, result, and the like. They are fleeting because they are at the wrong level of granularity for the language being acquired. For example, English uses the preposition by to express the SOURCE of AGENCY, uses than to express the SOURCE of STANDARD OF COMPARISON, and so forth. If the child beg ins with aglobai notion of SOURCE, it must later be analyzed on some dimensions, using relevant semantic features or components. Emergent categories are intriguing because they seem to reveal conceptual starting points for grammaticized notions. As Clark and Carpenter put it (l989a, p. 22): "Emergent categories offer evidence for the conceptual representations that underlie linguistic categories and that have linguistic consequences .... Emergent categories, we propose, reflect universals of conceptual representation." However, we do not yet know whether such patterns are truly universal, or whether there is a great range of individual variation, by child and by language. Another approach is to equip the child with fine-grained "primitives" and let linguistic experience sort them out into the appropriate categories. This is implicit in the various discussions of "semantic packaging" reviewed above: Something has to "get packaged." For example, in Bowerman's (l996a, 1996b) studies of the language of space, children learning different languages must attend to various features of locative relations. She lists many such features: for example, "spontaneous versus caused motion," "contact with and support by an external surface," "attachment by hooking," "put elongated object to base," "juxtapose surfaces that are flat," and so forth. Using such semantic characterizations, Bowerman succeeds in delineating the differences between English, Dutch, and Korean in the meanings of spatial terms and traces out the course of acquisition of these languages. Could the child be using distinctions at these levels of granularity to arrive at the proper definition of grammaticized notions in the language? As Bowerman puts it (1996a, p. 422): If semantic categories are constructed, they must be constructed out of something, and an important question is what this something iso Here we come squarely up against one of the oldest and most difficult problems for theorists interested in the structure of mind: identifying the ultimate stuff of which meaning is made.

318

Siobin

The facts of crosslinguistic diversity, and the inadequacy of current formulations, make Bowerman skeptical of accounts which attribute semantic primitives to the child; however, she DOES succeed in using various kinds of "primitives" to make her exposition comprehensible. It Seems to me that our task is to find the proper level of analysis to distinguish between languages and to adequately describe children' s uses of linguistic forms over the course of development. This is, indeed, "one ofthe oldest and most difficult problems." But it should ultimately be solvable on the basis of converging attempts at careful analysis and description of many languages and many child 1earners. In the end, the overlap between such descriptions should reveal psychologically plausible levels of primitives for the construction of grammaticizable notions and lexical forms in general. It may weil turn out that some developmental paths go from general to specific and others from specific to general; that some domains show great uniformity in underlying organization across children and languages while others are open to considerable variation; and so forth. For now, however, I believe that these questions must remain open. 5.2.7. In My End Is My Beginning

Allow me to end on a retrospective note. For more than thirty years our linguistic, psychological, and philosophical disciplines have sought to replicate themselves in the mind/brain of the child. The modules that are postulated often have names that evoke suspicion: they are the names of our own academic fields (linguistics, mathematics, physics, biology) or subfields (closed-class morphemes, grammaticizable notions). Could God or evolution have anticipated the academic and intellectual organization of late twentieth-century America? At the beginning of my career I was skeptical of building academia into the child. Later I found it attractive to "help" the child by removing some problems from the learning task. Now-partly to my surprise-I find myself thinking things that I said long ago (Slobin, 1966, pp. 87ff): [According to Chomsky] the reason that human languages utilize such strikingly uni versal grammatical relations and formal devices is ... due to the fact that these universal characteristics are themselves part of the innate structure of man .... I would rather think of the child as learning [a category such as the Russian animate accusative] through feedback than to have hirn waiting for confirmation of dozens of such categories from his mother' sexpansions. It seems to me more reasonable to suppose that it is language that plays a role in drawing the child' sattention to the possibility of dividing nouns on the basis of animacy; or verbs on the basis of duration, or determinacy, or validity; or pronouns on the basis of social status, and the like.

I propose that it is reasonable to return to this formulation and study how children use linguistic cues to discover the collections of semantic elements that are packaged in the lexical and grammatical items of the language.

5.

Origins of Grammaticizable Notions

319

AKCNOWLEDGMENTS

The ideas developed in this chapter were presented to a conference on "Language Acquisition and Conceptual Development," Max Planck Institute for Psycholinguistics, Nijmegen, The Netherlands, Nov. 13-17, 1995. A shorter version appears under the title "FormIFunction Relations: How Do Children Find Out What They Are?" in M. Bowerman and S. C. Levinson (Eds.) (in press), Language Acquisition and Conceptual Development (Cambridge University Press). I have benefited from many long discussions of these topics with Melissa Bowerman, and she will find her influence obvious in the revisions of my earlier position. I also owe much to Ioan Bybee, Alison Gopnik, Len Talmy, David Wilkins, the many colleagues in Nijmegen who have provided stimulation and (re-)education, and the 1995 conference participants. Thanks also to Paul Bloom, Marianne Mithun, Izchak Schlesinger, Elizabeth Traugott, and Tania Kuteva for valuable correspondence on topics raised here.

REFERENCES Adams, K. L., & Conklin, N. F. (1973). Toward a theory of natural c\assification. In C. Corum, T. C. Smith-Stark, & A. Weiser (Eds.), Chicago Linguistic Society 9 (pp. 1-10). Bates, E. (1979). On the evolution and development of symbols. In E. Bates, The emergence oj symbols: Cognition and communication in infancy (pp. 1-32). San Francisco/London: Academic Press. Bates, E., Frederici, A., & Wulfeck, B. (1987). Grammatical morphology in aphasia: Evidence from three languages. Cortex, 23, 545-574. Bates, E., & MacWhinney, B. (1987). Competition, variation, and language learning. In B. MacWhinney (Ed.), Mechanism~ of language acquisition (pp. 157-193). Hillsdale, NJ: Lawrence Erlbaum Associates. Bates, E., & Wulfeck, B. (1989). Crosslinguistic studies of aphasia. In E. Bates & B. MacWhinney (Eds.), The crosslinguistic study of sentence processing (pp. 328-374). Cambridge: Cambridge University Press. Behaghel, O. (1932). Deutsche syntax: Eine geschichtliche Darstellung: Vo/. 4. Wortstellung. Periodenbau. Heidelberg: Carl Winters. Berman, R. A., & Siobin, D. I. (1994). Relating events in narrative: A crosslinguistic developmental study. Hillsdale, NJ: Lawrence Erlbaum Associates. Bickerton, D. (1981). Roots of language. Ann Arbor, MI: Karoma Publishers. Bossong, G. (1985). Empirische Universalienjärschung: Differentielle Objekt-markierung in den neu iranischen Sprachen. Tübingen: Gunter Narr. Bowerman, M. (1985). What shapes children's grammar? In D. I. Slobin (Ed.), The crosslinguistic study ollanguage acquisition: Vo/. 2. The data (pp. 1257-1319). Hillsdale, NJ: Lawrence Erlbaum Associates. Bowerman, M. (1993). Typological perspectives in language acquisition: Do crosslinguistic patterns predict development? In E. V. Clark (Ed.), The proceedings of the Twenty~jijth Annual Child Language Research Forum (pp. 7-15). Stanford, CA: Center for the Study of Language and Information. Bowerman, M. (1994). From universal to language-specific in early grammatical development. Philosophical Transactions of the Royal Society of London B, 346, 37-45.

320

Siobin

Bowerman, M. (1996a). Learning how to structure space for language-A crosslinguistic perspective. In P. Bloom, M. A. Peterson, L. Nadel, & M. F. Garrett (Eds.), Language und space (pp. 385-436). Cambridge, MA: MIT Press. Bowerrnan, M. (1996b). The origins of children's spatial semantic eategories: Cognitive vs.linguistic deterrninants. In J. J. Gumperz & S. C. Levinson (Eds.), Rethinking linguistic relativity (pp. 145-176). Cambridge, England: Cambridge University Press. Bowerrnan, M., de Le6n, L., & Choi, S. (1995). Verbs, particles, and spatial semanties: Learning to talk about spatial actions in typologically different languages. In E. V. Clark (Ed.), The Proceedings of the Twenty-seventh Annual Child Language Research Forum (pp. 101-110). Stanford, CA: CSLI. Bowerman, M., & Pederson, E. (1992, December). Crosslinguistic perspectives on topological spatial relationships. Paper presented to Annual Meeting of the American Anthropological Association, San Francisco. Brown, P. (1994). The INs and aNs ofTzeltal locative expression: The semanties of statie deseriptions of loeation. Linguistics, 32,743-790. Brown, R., & Gilman, A. (1960). The pronouns of power and solidarity. In T. Sebeok (Ed.), Aspects (!t style in language (pp. 253-277). Cambridge, MA: MIT Press. Bybee, 1. L. (1985). Morphology: A study (!t the relation between meaning and form. AmsterdamlPhiladelphia: John Benjamins. Bybee, J., Perkins, R., & Pagliuea, W. (1994). The evolution (!/'grammar: Tense, aspect, und modality in the languages (!/' the world. Chieago: Chieago University Press. Caplan, D. (1992). Language: Structure, processing, and disorders. Cambridge, MA: MIT Press. Choi, S. (this volume). Language-speeifie input and early semantie development: Evidenee from ehildren learning Korean. In D. I. Siobin (Ed.), The crosslinguistic study of language acquisition: Vol. 5. Expunding the contexts. Mahwah, NJ: Lawrence Erlbaum Associates. Choi, S., & Bowerman, M. (1991). Leaming to express motion events in English and Korean: The influence of language-specific lexiealization patterns. Cognition, 41, 83-121. Chung, S. (1977). On the gradual nature of syntactic change. In C. N. Li (Ed.), Mechanisms oj syntactic change (pp. 3-56). Austin: University of Texas Press. Chomsky, N. (1965). Aspects of the theory (!/' syntax. Cambridge, MA: MIT Press. Clark, E. V. (1978). Discovering what words can do. In D. Farkas, W. M. Jacobsen, & K. W. Todrys (Eds.), Papers from the parasession on the lexicon (pp. 34-57). Chicago: Chicago Linguistic Society. Clark, E. V. (in press). Emergent categories in .first language acquisition. In M. Bowerman & S. Levinson (Eds.), Language acquisition und conceptual development. Cambridge: Cambridge University Press. Clark, E. V., & Carpenter, K. L. (1989a). The notion of source in language acquisition. Language, 65, 1-30. Clark, E. V., & Carpenter, K. L. (1989b). On children's uses of 'from', 'by', and 'with' in oblique noun phrases. Journal of Child Language, 16, 349-364. Comrie, B. (1981). Language universals and linguistic typology. Chicago: University of Chicago Press. Craig, C. (Ed.). (1986). Noun classes und categorization. AmsterdamlPhiladelphia: John Benjamins. Craig, C. G. (1993). Jakaltek directionals: Their meaning and discourse function. Languages (!lthe World, 7, 23-36. Croft, W. (1990). Typology and universals. Cambridge: Cambridge University Press. Erbaugh, M. S. (1986). Taking stock: The development of Chinese noun classifiers historically and in young children. In C. Craig (Ed.), Noun classes und categorization (pp. 399-436). AmsterdamlPhiladelphia: John Benjamins. Gleitrnan, L. R., Gleitrnan, H., Landau, B., & Wanner, E. (1988). Where learning begins: Initial representations for language learning. In F. 1. Newmeyer (Ed.), Linguistics: The Cambridge Survey:

5.

Origins of Grammaticizable Nations

321

Vol. Ill. Language: Psyeho[ogiea[ and biologieal aspeets (pp. 150-193). Cambridge: Cambridge University Press. Gleitman. L. R., & Wanner, E. (1982). Language acquisition: The state of the state of the art. In E. Wanner & L. R. Gleitman (Eds.), Language aequisition: The state of"the art (pp. 3-48). Cambridge: Cambridge University Press. Goldberg, A. E. (1995). Construetions: A Construetion Grammar approach to argument strueture. Chicago: University of Chicago Press. Gopnik, A., & Meltzoff, A. N. (1996). Words, thoughts, and theories. Cambridge, MA: MIT Press. Green, 1. N. (1982). The status of the Romance auxiliaries of voice. In N. Vincent & M. Harris (Eds.), Studies in the Romanee verb (pp. 97-138). LondoniCanberra: Croom Helm. Greenberg, J. H. (1963). Some universals of grammar with particular reference to the order of meaningful elements. In J. H. Greenberg (Ed.), Universals of" language (pp. 73-113). Cambridge, MA: MIT Press. Grice, H. P. (1975). Logic and conversation. In P. Cole & J. L. Morgan (Eds.), Syntax and semanties: Vo!. 3. Speech aets (pp. 41-58). New York: Academic Press. Haiman, J. (Ed.). (1985a). Jconicity in syntax. AmsterdarnlPhiladelphia: John Benjamins. Haiman, J. (l985b). Natural syntax: Jconicity and erosion. Cambridge: Cambridge University Press. Hawkins, J. A. (1983). Word order universals. New York: Academie Press. Hawkins, J. A. (1988). Explaining language universals. In J. A. Hawkins (Ed.), Explaining language universals (pp. 3-28). Oxford: B1ackwell. Hawkins, J. A. (1995). A performance theory ot order and constituency. Cambridge: Cambridge University Press. Heine, B., Claudi, U., & Hünnemeyer, F. (1991). Grammaticalization: A eoneeptual framework. Chicago: University of Chicago Press. Hopper, P. J., & Traugott, E. C. (1993). Grammaticalization. Cambridge: Cambridge University Press. Jackendoff, R. (1983). Semantics and cognition. Cambridge, MA: MIT Press. Jackendoff, R. (1987). Consciousness and the eomputational mind. Cambridge, MA: MIT Press. Jaekendoff, R. (1990). Semantic structures. Cambridge, MA: MIT Press. Karmiloff-Smith, A. (1992). Beyond modularity: A developmental perspeetive on cognitive science. Cambridge, MA: MIT Press. Keil, F. C. (1989). Coneepts, kinds and cognitive development. Cambridge, MA: Bradford Books. Keil, F. C. (1994). Explanation, association, and the aequisition of word meaning. Lingua, 92, 169-196. Landau, B., & Gleitman, L. R. (1985). Language and experience: Evidence from the blind ehild. Cambridge, MA: Harvard University Press. Landau, B., & Jackendoff, R. (1993). "What" and "where" in spatiallanguage and spatial cognition. Behavioral and Brain Scienees, 16,217-265. de Uon, L. (1994). Exploration in the acquisition of loeation and trajectory in Tzotzil. Linguisties, 32, 857-884. Levin, B. (1993). English verb classes and alternations: A preliminary investigation. Chicago: University of Chicago Press. Levinson, S. C. (1994). Vision, shape, and linguistic description: Tzeltal body-part terminology and objeet deseription. Linguisties, 32, 791-855. Li, c., & Thompson, S. (1974). Coverbs in Mandarin Chinese: Verbs or prepositions? Journalot Chinese Linguisties, 2, 257-278. Li, c., & Thompson, S. (1976). Development of the causative in Mandarin Chinese: Interaction of diachronie processes. In M. Shibatani (Ed.), Syntax and semantics: Vo[. 6. The grammar ot eausative eonstructions. New York: Academie Press. grammar of Mandarin Chinese. BerkeleylLos Li, C., & Thompson, S. (1981). A funetional r~terence Angeles: University of California Press.

322

Siobin

Lord, C. (1982). The development of object markers in serial verb languages. In P. J. Hopper & S. A. Thompson (Eds.), Syntax and semantics: Vol. 15. Studies in transitivity (pp. 277-299). New York: Academic Press. Lyons, J. (1968). Introduction to theoretical linguistics. Cambridge: Cambridge University Press. MacWhinney, B., & Bates, E. (Eds.). (1989). The crässlinguistic study of sentence processing. Cambridge: Cambridge University Press. Meillet, A. (1912). L'evolution des formes grammaticales. Scientia (Rivista di scienza), 12(26). [Reprinted in Meillet, A. (1982). Linguistique historique et linguistique generale (pp. 130--148). Geneva: Slatkine/Paris: Champion.] Mithun, M. (1989). The subtle significance of the locus of morphologization. International Journal or American Linguistics, 55, 265-282. Pick, A. (1913). Die agrammatischen Sprachstörungen. Berlin: Springer-Verlag. Pinker, S. (1984). Language learnability and language development. Cambridge, MA: Harvard University Press. Pinker, S. (1989). Learnability and cognition: The acquisition of argument structure. Cambridge, MA: MIT Press. Sapir, E. (1949). Language: An introduction to the study o{ speech. New York: Harcourt Brace. (Original work published 1921) Sapir, E. (1958). The grammarian and his language. American Mercury, 1, 149-155. [Reprinted in D. G. Mandelbaum (Ed.), (1958). Selected writings o{ Edward Sapir in language, culture and personality (pp. 150-159). Berkeley/Los Angeles: University of Califomia Press.] (Original work published 1924) Schlesinger, I. M. (1979). Cognitive and linguistic structures: The case of the instrumental. Journal ofLinguistics, 15,307-324. Schlesinger, I. M. (1982). Steps to language: Toward a theory or language acquisition. Hillsdale, NJ: Lawrence Erlbaum Associates. Schlesinger, I. M. (1988). The origin of relational categories. In Y. Levy, I. M. Schlesinger, & M. D. S. Braine (Eds.), Categories and processes in language acquisition (pp. 121-178). Hillsdale, NJ: Lawrence Erlbaum Associates. Schlesinger, I. M. (1995). Cognitive space and linguistic case. Cambridge: Cambridge University Press. Sinha, c., & Kuteva, T. (1995). Distributed spatial semantics. Nordic Journal of Linguistics, 18. Siobin, D. I. (1966). Comments on "Developmental psycholinguistics": A discussion of McNeill's presentation. In F. Smith & G. A. Miller (Eds.), The genesis o{ language: A psycholinguistic approach (pp. 85-91). Cambridge, MA: MIT Press. Siobin, D. I. (1977). Language change in childhood and in history. In l. Macnamara (Ed.), Language learning and thought (pp. 185-214). New York: Academic Press. Siobin, D. I. (1979, April). The role o{ language in language acquisition. Invited Address, Fiftieth Annual Meeting of the Eastem Psychological Association, Philadelphia. Slobin, D. I. (1981). The origins of grammatical encoding of events. In W. Deutsch (Ed.), The child's construction o{ language (pp. 185-200). London: Academic Press. Siobin, D. I. (1985). Crosslinguistic evidence for the Language-Making Capacity. In D. I. Slobin (Ed.), The crosslinguistic study o{language acquisition: Vol. 2. Theoretical issues (pp. 1157-1256). Hillsdale, NJ: Lawrence Erlbaum Associates. Siobin, D. I. (1991a). Aphasia in Turkish: Speech production in Broca's and Wemicke's patients. Brain and Language, 41, 149-164. Siobin, D. I. (199Ib). Learning to think for speaking: Native language, cognition, and rhetorical style. Pragmatics, 1,7-26. Slobin, D. I. (1994). Talking perfectly: Discourse origins of the present perfect. In W. Pagliuca (Ed.), Perspectives on grammaticalization (pp. 119-133). AmsterdamlPhiladelphia: lohn Benjamins.

5.

Origins of Grammaticizable Nations

323

Siobin, D. 1. (1996a). From 'thought and language' to 'thinking for speaking'. In J. J. Gumperz & S. C. Levinson (Eds.), Rethinking linguistic relativity (pp. 70-96). Cambridge: Cambridge University Press. Siobin, D. 1. (l996b). Two ways to travel: Verbs of motion in English and Spanish. In M. Shibatani & S. A. Thompson (Eds.), Grammatical constructions: Theirform and meaning (pp. 195-219). Oxford: Clarendon Press. Slobin, D. 1. (1997). The universal, the typologieal, and the particular in acquisition. In D. 1. Siobin (Ed.), The crosslinguistic study of language acquisition: Vol. 5. Expanding the contexts. Mahwah, NJ: Lawrence Erlbaum Associates. Siobin, D. 1. (in press). Formlfunction relations: How do children find out what they are? In M. Bowerman & S. C. Levinson (Eds.), Language acquisition and conceptual development. Cambridge: Cambridge University Press. Siobin, D. I., & Hoiting, N. (1994). Reference to movement in spoken and signed languages: Typological considerations. Proceedings 01' the Twentieth Annual Meeting 01' the Berkeley Linguistics Society, 487-505. Talmy, L. (1978). The relation of grammar to cognition-A synopsis. In D. Waltz (Ed.), Proceedings 0I'TINLAP-2 (Theoretical Issues in Language Processing). New York: Association for Computing Machinery. Talmy, L. (1983). How language structures space. In H. Pick & L. Acredolo (Eds.), Spatial orientation: Theory, research, and application. New York: Plenum Press. Talmy, L. (1985). Lexicalization patterns: Semantic structure in lexical forms. In T. Shopen (Ed.), Language typology and semantic description: Vol. 3. Grammatical categories and the lexicon (pp. 36-149). Cambridge: Cambridge University Press. Talmy, L. (1988). The relation of grammar to cognition. In B. Rudzka-Ostyn (Ed.), Topics in cognitive linguistics (pp. 166-205). AmsterdamlPhiladelphia: lohn Benjamins. Talmy, L. (1991). Path to realization: A typology of event conflation. Proceedings 01' the Seventeenth Annual Meeting 01' the Berkeley Linguistics Society, 480-519. Thompson, W. D. (1917). On growth and.f{mn. Cambridge: Cambridge University Press. Traugott, E. C. (1972). The history 01' English syntax. New York: Holt, Rinehart & Winston. Traugott, E. c., & Heine, B. (Eds.). (1991). Approaches to grammaticalization: Vol. 1. Focus on theoretical and methodological issues: Vol. 2. Focus on types 01' grammatical markers. AmsterdamlPhiladelphia: John Benjamins. Zipf, G. K. (1935). The psycho-biology 01' language. Boston: Houghton Mifflin.

This page intentionally left blank

Author Index

A Aarssen, J ., 27 Abney, S. P., 179,192 Adams, K. L., 304, 319 Akhutina, T.V., 237, 256 Aksu-Ko9, A. A., 28, 165, 175, 182, 184, 192, 206,225,234,246,256 Allen, G. D., 186,192 Anilovich, Y., 28 Argoff, H. D., 218, 243,256 Aske,J., 17,35 Au, T. K., 51, 128 Aviezer, 0., 42, 131

B Baiiiargeon, R., 122, 128 Baldwin, G., 220, 226, 233, 260 Bamberg, M., 28, 29, 30, 31,35 Barnes, H. D., 210,215,260 Bates, E., 32, 34, 35, 43, 48, 51, 57, 128, 129, 130,202,204,208,209,212,215,216, 225,236,237,238,239,240,245,247, 248,253,256,25~261,281,282,297,

303,319,322 Bavin, E., 28, 46, 129, 205, 228, 256 Bayles, K., 240, 261 Beckman, M., 166 Behaghel, 0., 314,319 Bellugi, U., 223, 230, 256, 257 Bender, B., 170 Benedict, H., 243, 258 Berman, R. A., 3, 4, IS, 17, 28, 29, 35, 36, 45, 46,129, ISO, 151,168,179,192,204, 213,230,232,234,248,256,287,292,

316,319

Bernstein-Ratner, N., 153, !54, 192 Bickel, B., 270 Bickerton, D., 93, 99, 126, 129,266, 273,319 Bishop, D., 247, 256 Bloom, L. M., 49, 57, 63, 94, 108, 113, 120, 122,129,164,177,192,205,209,229, 233,234,248,256,257

Bloom, P., 281 Bocaz, A., 22, 28, 38 Bolinger, D., 161, 192 Borer, H., 237, 252, 257 Bortolini, U., 174, 195 Bos, P., 27, 28 Bossong, G., 302,319 Bottari, P., 137, 179, 186, 192, 206, 207, 257 Bowerman, M., 9, 10, 29, 36, 45, 46, 64, 72, 73, 74,80,83,93,95, 100,101,123,125, 127, 128,129,130,210,218,239,25~ 265,274,276,285,287,289,290,291, 296, 317,319,320 Braine, M.D., 210,257 Branigan, G., 163, 192 Braunwald, S., 210, 247, 257 Bretherton, 1., 202, 204, 208, 209, 212, 236, 237,238,239,245,247,256,257 Brice Heath, S., 212, 241, 257 Bridges, A., 229, 257 Brinkmann, U., 100, 102, 129 Brody, G., 108, 120, 132 Brown, P., 128, 130, 305,319 Brown, R., 126, 130, 177, 192,220,223, 257,

308,319 Bryk, A., 214,259 Buczowska, E., 248, 263 Budwig, N., 250, 257 Butterfield, S., 145, 192 Bybee, J., 108,130, 190,192,275,276,298,

303,314,319 Byrnes, J.P., 119, 130

c Camaioni, L., 206, 245, 257 Caplin, D., 282,319 Carnevale, G. F., 248, 256 Carpenter, K. L., 276,317,320 Carroll, M., 28 Carter, D. M., 145, 192 Caselli, M. C., 174, 195, 196 Castillo Pintado, J., 127, 128, 132 Cazden, C., 223, 257

325

Author Index

326 Charney, R., 73, 131 Chilosi, A.M., 137, 179, 186, 192,206,207, 257 Chimombo, II Cho, S. W., 219,240,257 Choi, S., 9, 36, 50, 51, 53, 54, 55, 57, 60, 61, 62,63,64,67, 72, 73, 74,93,94,95, 103, 105, 110, 113, 121, 123, 127, 129, 130,131,2!4,2!5,238,239,248,25~

259,276,285,287,290,319,320 Chomsky, N., 13, 14, 36,273,297,320 Chung, S., 300, 320 Cipriani, P., 137, 179, 186, 192,206,207,257 Clancy, P.M., 50 94, 115, 120, 130, 181, 183, 184,186,192,212,213,215,218,219, 222,224,232,248,249,257 Clark, E. V., I, 36, 42, 122, 130,204,213,218, 221,258,276,304,316,317,320 Claudi, U., 298,321 Comrie, B.,l2, 36,298,320, 167, 169, 170,192 Conklin, N.'F., 304, 319 Connelly,M. J., 165, 192 Corrigan, R., 42, 131,247,258 Craig, C., 283, 284, 320 Croft, W., 2, 15, 36, 252, 258, 298, 320 Cutler, A., 145, 154, 192, 194 Cziko, G. A., 93, 130

D Dale, P., 48, 51, 57, 129, 130, 202, 209,212, 215,216,256 Dapretto, M., 51, 128 Dasinger, L., 2, 4, 5, 32, 35, 36, 218,232,242,

250,258 de Boysson-Bardies, B., 148, 194, 202,263 DeFries, J. C., 236, 261 Dehnhardt, A., 64, 67, 70, 133 DeLancey, S., 64, 130 deLeon, J., 123, 129 de Leon, L., 287,319, 321 Della-Corte, M., 243, 258 Demuth, K., 11, 32, 36, 141, 153, 156, !59, 165, 179,183,193,195,207,220,233,258 Devescovi, A., 32, 34, 35 de Villiers, J. S., 230, 232, 233, 235, 248, 262,

263 de Villiers, P. A., 230, 232, 233, 235, 248, 263 Donegan, P., 146, 174, 184 Dore, J., 120, 130,203,258 Dresner-Barnes, H., 208, 210, 215 Dromi, E., 174, 182,193 Duff, M. A., 119, 130 Dupoux, E., 154, 195 Durand, C., 202, 263 Durova, N. V., 28

E Echols, C. H., 137, 152, 155, 160, 162, 193 Eisenberg, A. R., 240, 258 El'konin, D. B., 6, 36 Elbers, L., 225,258 Elman, J. L., 253,256 Engstrand,0.,8,38, 156, !58,193 Erbaugh,M.S., 150,181,193,228,233,243, 258,283,307,320 Eriksson, A., 143, 193

F Faria, I. H., 28 Farrar, M., 223, 224, 243, 258 Fensan,L.,202,209,212,215,216,256 Penson, L., 48, 129 Fernald, A., 152, 153, !54, 193, 241,258 Fiess, K., 234, 257 Fisher, C., !53, 176, 193 Foley, W. A., 252, 258 Fortescue, M., 168, 170, 171, 173, 182, 193, 218,226,240,246,258 Francis, W., 214,259 Friederici, A., 153, 194,281,319 FromeLoeb, D., 239, 261 Fujiwara, 222 Fukui, N., 179, 193 Furrow, D., 119, 132, 241,258

G Garding, E., 146, 148, 174, 187,193 Gentner, D., 9, 36, 47, 48, 49, 51, 61, 130,214,

258 Gerken, L., 148, 149, 156, 162, 163, 177, 191,

193 Gillis, S., 214,242,258 Gilman, A., 308,319 Giv6n, T., 252, 258 Gleitman, H., 152, 156, 194, 280, 320 Gleitman, L. R., 152, !56, 193, 194, 280, 281,

282,320,321 Goldberg, A. E., 280, 321 Goldberg, R., 100, 131 Goldfield, B. A., 48, 52, 57, 130,239,241,258 Goldowsky, B., 178,194 Goldsmith, J. A., 143, 160, 194 Gopnik, A., 44, 49, 50, 51, 53, 54, 57, 60, 61, 62,63, 73, 74,88, 110,124,127,130, 131,215,238,247,25~258,259,315,

321 Green, J. N., 279, 321 Greenberg, J. H., 12, 36,276,321

Author Index

327

Greenlee, M., 202, 263 Gregoire, A., 147, 175, 194 Grice, H. P., 300, 305, 321 Gropen, J., 100, 131 Gruendel, J., 122, 131 Guillaume, P., 221,259 Gvozdev, A. N., 6, 36

Jaeggli, 0., 15, 37 Jisa, H., 28 Johnston, B., 148, 184, 194 Johnston, J. R., 9, 37, 42, 44, 47, 128 131 248

259

H Hafitz, J ., 248, 257 Haight, W., 214, 259 Haiman, J., 314, 321 Hakuta, K., 233, 262 Halle, M., 143, 194 Halle, P., 148, 194 Halliday, M.A. K., 48, 131 Halpern, E., 42, 131 Hampson, J., 49, 132, 241, 259 Hankamer, J., 168, 186, 194 Hardy-Brown, K., 235, 259 Harris, A., 183 Harris, J. W., 146, 194 Hart, E., 214,259 Hartung, J., 48, 129,202,209,212,215, 216

'

'

256

K Kaiser, G., 169, 194 Karmiloff-Smith, A., 3, 4, 37,237,259, 310

315,321

'

Haugen, E., I 48, 194 Hawkins, J. A., 4, 12, 14, 35, 36, 37, 275, 276

303,321

'

Hawkins, S., 186, 192 Hayashi, M., 213, 259 Hayes, B., 143, 144, 194 Heine, B., 15, 38,298,321,323 Hickmann, M., 4, 37 Hirsh-Pasek, K., 153, 194 Hirst, W., 119, 131 Hoff-Ginsberg, E., 108, 127,131,132 Hogg, R., 141, 143, 145, 194 Hoiting, N., 17, 38,292,323 Hollander, M., 100, 131 Hood, L., 63, 94, 108, 113, 120, 129, 205,209, 234,257 Hopper, P. J., 15, 37, 294, 298, 300, 302 304

321

'

Jonsd6ttir, H., 38 Joos, M., 148, 194 Jusczyk, A.M., 153, 194 Jusczyk, P. W., 153, 154, 194

'

Hiinnemeyer, 298, 321 Huttenlocher, J., 73,131,214,259 Hyams, N., 14, 37,251,259

I lmedadze, N., 183, 194,205,207,222,224, 225,248,259 Indefrey, P., 227, 259 Izutani, M., 183, 194

J Jackendoff, R., 272, 289, 293, 296, 297, 321 Jacobsen, T., 248, 258

'

'

Karttunen, F., 146, 172 Kay, E., 202, 263 Keil, F. C., 311 312,315,321 Kelly, M., 153, 195 Kemler Nelson, D. G., 153, 194 Kennedy, L., 153, 194 Kern, S., 28 Kessler Shaw, L., 49, 132 Kim, C., 64, 67, 70, 133 Kim, Y.-J., 60, 92, 103, 131, 218,219,230,233, 234,243,249,252,259 Kita, S., 67, 130 Klein, D., 241,243,258,259 Klein, H. B., 137, !55, 159, 160, 163, 194 Knudson, S. L., 191, 196 Koda, K., 93, 130 Konieczna, E., 248, 263 Kucera, H., 214,259 Kuczaj, S. A., 230, 240, 259, 260 Kunene, E., 165, 195 Kiintay, A., 28 Kuteva, T., 291,322

L Labov, W.,230,259 Lahey, M., 233, 234, 257 Landau, B., 152,156,177,193,194,272,280,

281,293,320,321

Lanza, E., 8, 38 Lederer, A., I 53, 195 Lee, H. B., 51, 131 Lee, H. S., 105, 106, 108, 113, 117, 131 Lee, K., 105, 131 Lee, K.-0., 252, 263 Leiwo, M., 8, 38 Lennert Olsen, L., 168, 171, 173, 182, 193 Leonard, L. B., 174, 182, 193, 195 Levin, B., 295, 321 Levinson, S.C., 128, 132, 265, 305, 306, 321 Levy, Y., 15, 37, 224, 260

328

Author Index

Li, c., 283, 299, 321 Li, P., 94, 126,132 Lieven, E. V. M., 137, 195,205,208,209,210, 212,213,215,220,225,226,233,239,

241,246,248,259,260,261

Lifter, K., 234, 248, 257 Lightbown, P., 63, 94, 129, 205, 209,257 Lloyd, E., 240, 260 Locke, J. L., 237, 260 Longobardi, E., 206, 245, 257 Lord, C., 299, 300, 322 Lust, 8., 13, 37, 252, 263 Lyons, J., 266, 295, 322 Lyons, T., 214, 259 Lyytinen, P., 127,133

M Maciver, D., 108, 132 Macken, M. A., 137,195 MacWhinney, B., 172, 179, 195, 204, 218, 224,

226,232,246,253,256,260,303,319,

322 Major, R. C., 142, 187, 195 Mandler, J., 47, 122, 123,132 Maratsos, M. P., 55, 61, 132, 224, 230, 240,

259,260 Marchman, V., 27, 48, 129,190,191,196,202, 209,212,215,216,256 MarguIis, C., 49,129 Markman, E., 47, 49, 132, 214, 260 Mayer, M., 17,37 McCuIIey, C. 8., 141, 143, 145, 194 McCune-NicoIich, L., 238,260 McGregor, K. K., 174,195 McIntosh, B. J., 156, 177,193 McNew, S., 239, 257 McRoberts, G., 153, 193 McShane, 1., 48, 132 Mehler, 1.,154,195 Meier, R. P., 152,195,244,260 MeiIIet, A., 298, 322 Meisel, J., 169,194 Meltzoff, A. N., 44, 50, 53, 57,62,63,73,74, 88,110,124,127,131,238,259,315, 321 Menn, L., 137, 152, 156, 160, 162, 163, 164, 177,183,185,189,195,202,206,261 Merriman, W. E., 47, 133 Mikes, M., 6, 37, 225, 262 MiIIs, A. E., 182, 186, 195, 207,218,221,232,

234,260

Mithun, M., 155, 187,195,308,322 Moore, C., 119, 132 Morgan, 1. L., 152, 153,195 Morikawa, H., 241, 258 Morisset, C., 51, 57, 130

N Nagano-Madsen, Y., 142,195 Naigles, L., 128,132 Neeman, Y., 28 NelfeIt, K., 156, 195 Nelson, K., 48, 49, 52, 57, 122,132,208,209, 211,213,214,215,239,240,241,258

260

'

Newport, E. L., 152, 155, 160, 178,193,194,

195,244,260

Nordqvist,

A., 8, 28, 38

o Ochs, E., 212, 218, 221, 244, 260 Ohsiek, D., 173, 195 Okubo, A., 215, 216, 260 Olsen, L. L., 218, 226, 240, 246, 258 OIson, S. L., 240, 261 OrsoIini, M., 28

p PagIiuca, w., 108,130,275,276,298,303,319 Pao, Y., 239, 261 Pea, R. D., 229, 261 Pederson,E.,9, 10,36, 128,129,289,319 PeIIigrini, A., 108, 120,132 Perkins, R., 275, 276, 298, 303, 319 Pemer, J., 249, 263 Peters, A. M., 1,3,8,38,136, 137, 152, 156, 158, 160, 162, 163, 164, 172, 177, 183, 185,189,195,196,202,203,204,205, 206,207,208,209,210,211,223,224, 246,261,263,281 Piaget, 1., 43, 118, 132 Pick, A., 281, 322 Pine, J. M., 208, 209, 210, 212, 213, 215, 220,

225,226,233,241,242,246,248,260,

261 Pinker, S., 7, 99,100, 126,131,132,226,261, 266,273,274,282,289,296322 Piwoz, J., 153,194 ' Pizzuto, E., 174,196 Plank, F., 2 Plornin, R., 236, 261 Plunkett, K., 1, 8,37,38, 137, 149, 156, 158, 186,187,189,190,191,196,204,230, 231,234,246,249,261 PoIi, P., 137,192 PuIIeyblank, D., 141, 196 Pure, K., 119, 132 Pye,C., 148, 150, 155, 157, 158, 162, 166, 178, 196,239,245,261

Author Index R Radford, A., 224,246,251,261 Radulovic, L., 6, 37,218,261 Ragnarsd6ttir, H., 8, 28, 38, 190 Raisanen, A., 242, 261 Ramer, A., 203, 230, 240, 248, 261 Rankin, 1., 225, 256 Redanz, N. 1., 154, 194 Reilly, 1., 48, 129, 202, 209,212,215,216,256 Reittinger, E., 127, 128, 132 Remez, R. E., 177, 193 Renner, T., 27, 240, 258 Reznick, 1. S., 48, 51, 52, 57, 129,130,202, 209,212,215,216,256 Richtoff, U., 8, 38 Rispoli, M., 93, 95, 97, 132 Rocissano, L., 108, 113, 120, 129 Roeper, T. W., 13, 37, 221, 262 Rossi, F., 28 Rowland, C., 241, 261

s Sabbadini, L., 174, 195 Safir, K., 15, 37 Sapir, E., 266, 267, 268, 322 Savic, S., 28, 225, 262 Scarpa, E., 137, 186, 196 Schieffelin, B. B., 181, 186, 196, 212, 213, 218, 219,232,242,262 Schlesinger, I. M., 102, 132, 211, 262, 288, 289, 321,322 Scollon, R., 163, 177, 196 Sebastian, E., 22, 28, 33, 37 Segui, J., 154, 195 Selkirk, E. 0., 175, 196 Seltzer, M., 214, 259 Sera, M., 127, 128, 132 Serra Raventos, M., 245, 262 Shatz, M., 108, 120, 127, 128, 132 Shore, C. M., 236, 262 Shteiman, M., 174, 182,193 Siewierska, A., 180, 196 Silva, M., 248, 258 Simonetti, M. Z., 206, 262 Simonsen, H. G., 8, 38, 137, 163, 186, 190, 196 Sinha, C., 291,322 Slobin, D. I., I, 3, 4, 6, 7, 9, 15, 17, 21, 22, 23, 29,32,33,36,3~38,42,44,45,46,93,

101,108,110,120,124,129,131,133, 136, 149, 157, 165, 171, 172, 175, 177, 181,182,184,190,192,196,205,206, 211,223,224,225,230,231,234,235, 244,246,248,255,256,262,266,273, 276,281,287,292,296,301,302,303, 312,314,315,316,317,318,319,322, 323

329 Smiley, P., 73, 131 Smoczyriska, M., 6, 7, 28, 38, 96, 133, 182, 184, 196,207,218,220,224,227,228,234, 240,249,262 Snow, C. E., 242, 262 Snyder, L., 202,204,208,209, 212, 236, 237, 238,239,245,247,256,257 Sole Planas, R. M., 245, 261 Song, Y., 51, 128 Speas, M., 179, 193 Spelke, E., 122, 133 Starr, S., 209, 262 Stephany, U., 110, 119,133,220,231,234,262 Stem, C., 137, 164, 186, 196 Stem, W., 137, 164, 186, 196 Stoneman, Z., 108, 120, 132 Striimqvist, S., I, 8, 28, 37, 38, 149, 156, 158, 186,187,189,193,196,230,231,234, 246,249,261 Sundberg, U., 202, 263 Sufier, M., 13, 37 Suzman, S., 11,38 Svenkerud, V. Y., 153, 194

T Tager-Fiusberg, H., 233, 262 Takeff, J., 233, 257 Talmy, L., 5, 16, 17, 38, 64, 65, 72, 133,269, 270,271,272,274,276,285,286,292, 293,296,303,308,312,323 Tang Boyland, J., 154 Tanouye,E.K.,215,262 Tardif, T. Z., 51, 54, 57, 127, 133, 165, 166, 181,19~214,215,262

Thai, D., 48, 129,202,209, 212, 215, 216, 256 Thompson, S., 283, 299, 300, 302, 321 Thompson, W. D., 297,323 Tinker, E., 49, 129 Toivainen, 1., 8, 38, 172, 189, 197 Toivainen, K., 8, 38 Tokura, H., 153, 176, 193 Tomasello, M., 46, 49, 57, 63, 128, 133, 215, 219,224,246,262 Topping, D. M., 148, 197 Torrens Garcia, V., 245, 261 Torrens, V., 251,262 Toupin, C., 32, 35, 36 Traugott, E. C., 15, 37, 38,294, 298,299, 300, 304,321,323 Tseng, C., 168, 188, 189, 197 Tuite, K., 183, 194,205,207,222, 224, 225, 248,259

v Vainikka, A., 15, 37 van der Auwera, J., 2

Author Index

330 van Geert, P., 248, 263 Van Valin, R. D., Jr., 2, 5, 38, 231, 246, 252, 253, 258, 263 Vassiere, J., 148,197 Vergnaud, J.-R., 143,194 Verhoeven, J., 27, 28, 242, 258 Vihman, M. M., 148,194,202,204,263 Vlahovic, P., 6, 37 von Stutterheim, c., 28

w Wanner, E., 152, 156,193,194,280,282,321 Waterson, N., 162, 197 Weil, J., 119, 131 Weist, R. M., 220, 248, 263 Weist, R., 127, 133 Wesseis, J., 153,194 Wexler, K., 237, 252, 257 Whitman, J., 13,37,252,263 Whorf, B. L., 43, 127,133 Wienold, G., 64, 67, 70, 133 Wiley, A., 250, 257

Williams, E., 13, 37 Williams, K., 158, 193 Wilson, B., 246, 263 Wimmer, H., 249, 263 Witkowska-Stadnik, K., 248, 263 Wong'FiIImore, L., 204, 263 Woodward, A., 153,194 Wulfeck, B., 281, 282, 319 Wysocka,H., 127,133,248,263

y Yoshida, M., 64, 67, 70, 133 Youssef, v., 126, 133 Yurieva, N. M., 28

z Zakharova, A. v., 6, 38 Zijlenmaker, c., 27 Zipf, G. K., 303, 323

Subject Index

The abbreviations given below indicate discussion of an index category with regard to a particular language. Entries without a language code refer to a general discussion of the category in question. All references refer to discussions of acquisition or of generallinguistic issues in the chapters. ABBREVIATIONS FOR LANGUAGE CODES Ab ASL BEn Ba Ca Ch Cz Da Du En Ek Es Fi Fr Ge Gn Gk Ha He Hu Ie IE It

Jp Ka Ki

Arabic American Sign Language Black English Bantu Catalan Chinese Czech Danish Dutch English Eskimo Estonian Finnish French German Georgian Greek Hawaiian Hebrew Hungarian Icelandic Indo-Europeanlanguages Italian Japanese Kaluli K'iche'

A accent, Sw 158f, 188 accentgroup, 144, 163 accessibility, 166, 203 accessibility hierarchy, 274-277, 301, 309 accusative, 5,289,290, Chiekasaw 300f, Ge 301, Ma 299ff, Persian 302, P17, Rs 7, 301, Th 301, West African 300f acoustic salience (see salience, acoustic) adjective, It 225

Ko Ma Mo My Na No PI Pr Ro Rs

Sa SC Se Si SI Sp Ss Se Sw Ta Tg Th Tz Wa Zu

Korean Mandarin Mohawk Mayan languages Navaho Norwegian Polish Portuguese Romance languages Russian Samoan Serbo-Croatian Semitic languages SiSwato Slavic languages Spanish Sesotho Scandinavian Swedish Tamil Tagalog Turkish Tzeltal Warlpiri Zulu

agglutinating languages, 139f, 169, 281, 308, ASL 180, Fi 180, Gn 205, Ki 157, Ss 169, Tg 170, Th 169f, 180,207,213, Wa205 agrammatism, 28lf agreement, 190, 217, ASL 244, En 220, Gk 220, Si 165,Ss 165,220 Akan,300f allomorphy, 169-172, 188, 190, En 183 amalgam, 179f, 186, He 179f American Sign Language (ASL), 180,230,244

331

332 analytic languages, 167ff, Cantonese 168, En 180, Ge 180, Germanie 167, Ma 168, 180, Sw 180, Vietnamese 167f anaphora, 29 animacy, Jp 97, Ko 95ff, 100ff, PI 96, Rs 96 aphasia, 281f Arabie lraqui,289 Moroccon, 28 artieies (see also definiteness; determiners), Da 156,189, En 148ff, 281, Fi 4, Fr 4, Ge 182,281, It 281,206, No 156, 189, Sc 149f, 156, 189, Sw 149, 156, 189 aspect, SI3 assimilation, Ek 171, 173, En 171 Atsugewi, 286, 308 auxiliary, Sp 279

B baby talk (see infant -directed speech) Bantu, 3, 11, 146, 156, 165 Basie Child Grammar, 45 Benue-Kwa languages, 300f Belhare, 270 bilinguals, bilingualism, Da 213, Jp 213 bioprogram, 99,126,273 Black English, 230 Bontoc,151f bootstrapping, 237 iconic, 314 prosodie, 152-156, 281 semantie, 5, 126 typological, 315f

c Caddo, 286 Cantonese, 140, 168 case, case marking, 217-223, 254, En 217, Fi 218, Ge, 181, Gn, 222, Jp 222, Ka 181, 219, Ko 90-102, P16f, 93, Rs 6f, SC 6, 218, Tu 217 Catalan, 251 categorization, 62, 238 En vs. Ko, 63, 127f certainty (ofaproposition), En 119, Ko 107, 115f, 120 Chamorro, 148, 151 Chickasaw, 300f Chiehewa, 11 child-raising practiees, 128 Chinese, 136, 239, 268 Chippewa, 268 classifiers, En 283, 307, Ma 189,283,307, Ss 157, Thai 189

Subject Index cIause chaining, Sp 22 cline, grammatical, 294, 298-302, 309 semantie,288-291 cIitie, Fr 169, Ki 148 cogniiive development, 42-48, 62f, 73, 103, 110, 118f, 125, 127,230, 238f, 249f, 273,276,311 individual differences in, 247f, En 248, Gn 248, He 248, PI 248 universals of, 43f, 122, 125f comitative, 288f competition, competition model, 253, 303 complex sentences (see also coordination; negation; subordination; syntax), 232-235, En 233, 240, 248, Ko 233, Ma 233, Ss 233 complexity, 44, 90 comprehension methodology, En 156, 177 and production, 203, 247, 255 computationallinguistics, 190f concept formation, 31 Of connectionism, 3, 237, 253 consonant gradation, Ek 147, Es 172, Fi 147, 172, 189 consonant harmony, 160, 162 constraints acquisitional, 311f cognitive, 49, 56 lexieal,214f linguistie, 2 construction, grammatical, 280, 291f continuity (of strategy, style), 209-213, 249 continuum (see cline) coordination, 233f, En 248 core grammar, 14ff, 251 creole languages, 126 cue, acoustic, 148, 152ff, 185, En 153f reliability, 153, 156 culture, 308 Czech,281

D Danish, 8,137,140,156,173,176,186, 188f, 204, 206, 213 decidability, 304-309 deixis, Ge 30f, 285f and motion events, 64, Ko 67f, 75, 78, 80f definiteness (see also articIes; determiners), 29, Ge 4, Fi 4, Fr 4 determiners (see also articIes; definiteness), En 206, Sc 246

Subject Index development, eognitive (see eognitive development) diaehrony, linguistic (see also grammaticization), 15,227,294,296-309,312, En 299 and ontogeny, 313f sound ehanges, 169 discourse, discourse eontingeney, Ko 109, 113f, 115-120 diversity, linguistie, 2, 266, 275, 296, 310 typological, 5 dual-Iexicon hypothesis, 282 duration, 141ff, Fi 142, Hu 142, Jp 142 Duteh, 9f, 23, 27-29, 83, 87, 123,225, 290f, 317

E eeonomy, 310f egoeentrism, 118 elision, Gk 146f, 174f ellipsis (see also pro-drop) 213, 242f, En 33, He 15, Jp 180f, 212f, 215f, Ko 180,215 emergent eategory, 316f English, 9-12,16-25, 27f, 30, 32-34,42, 45-49,53t56t60-63,65,68,70-74, 78-90,119-122,124-127,136-144, 146, 148f, 153f, 162, 164f, 168f, 17lf, 173f, 176, 179f, 182f, 185f, 189f, 202-206,208t212-217,220,223t226, 230-233,235,238-243,245t248,250, 278,283,285t290t295,299,302,304, 307f, 315, 317 ergative, ergative languages, 3, 5, 300, Ka 219, Sa 46, 22lf, 244 erosion, 169, Fr 169, Pr 169 errors, 246, 250, 254, Ko 97-102,126 mispereeption, 145 and U-shaped learning, 101 Eskimo, 136f, 140, 147, 168, 171, 173, 180, 182,186,191,215,218,223,226,240, 246,308 Eskimo-Aleut languages, 170 Estonian, 146, 172 event paekaging, En 33, Sp 33 evidentials (see also modals, modality), Ko 103-108, 117f, 123 expeetation (see also knowledge, shared; newly aequired information; old vs. new information), Ko 105

F feedback morphemes, Sc 249 fietion, 21, 23, 29 Filipino languages, 151

333 fillers, prosodie, 160, 163-166, 177, 184ff, 191, 205-207,254, Da 206, En 137, 164f, 203-206, Fi 206, Ge 164, 206f, Gn 205, 207, It 179, 206, Jp 184,207, Ki 163, Ma 165f, 184, No 163,206, Pr 206, Si 165, Ss 165, 207, 10 141, 165, 182, 184, 206f, Wa205 Finnish, 3f, 136, 140, 142, 146f, 168f, 17lf, 176, 179t 186, 189,206, 218,242t 304 Finno-Ugrie languages, 17 first words, 48-51, 64, 160, En 53f, 63, Ko 51-58,63, Ma 54 foeus, Tg 180 foot metriea1, 143ff, 160, 163, Sw 149 troehaie, En 145, 148, 154, 162, Ss 156f, Sw 156 form-funetion relation, 4, 15f formalism, 15 formulae, low seope, 227, 247, 249, 254, Ek, 226, 246, En 226, 246, Hu 226, 246, PI 228 formulaie style (see style, phrasal) Freneh, 4, 23, 28f, 34,121, 136f, 140, 144, 147, 168f, 174f, 187f, 204, 213, 221, 230, 238,268,304,308 frequeney erosslinguistie, 274f, 302 inputJmatehing, 24lff, En 214, 250, Ge 227, Hu 250, Ko 214, Ma 214 ofuse, 303ff, 312f funetionalist grammars, 2, 15, 252f, 266,306 fusion, phonologieal, 170 semantie (see also portmanteau morpheme), 169f, 182f, Ge 182, PI 182, Rs 182 fusionallanguages, 169, Ge 169, He 170, PI 169, 180, Rs 169

G Ga, 300f gender, Ge 182, He 3, PI 7, Rs 7 Georgian, 137, 168, 183-186, 191,205,207, 223-225,248 German, 4, 9f, 23, 26-29, 45,137,140,146, 164,168-170, 172f, 176, 179f, 182, 186, 189f, 205-207, 218, 221, 227, 23lf, 268, 281,285,301 Germanic languages, 167,287, 292f grammaticizable notions, 265-318 grammaticization (see also diaehrony, linguistic), 269, 298-309, 312 Greek, 146f, 174f, 220, 231

334

H Hawaiian, 146 hearsay (see evidentials) Hebrew, 3,15,26--28,32,34,45, 140, 150f, 168-170,174,179, 181f, 186,204,213, 230,232,248 historicallanguage change (see diachrony, Iinguistic; grammaticization) homophony, 182f, 189f, En 183, 189, Ge 182, Jp 183, Ma 181, No 189f, Rs 6,182, Wa217 honorifics (see also politeness), Ko 249 Hungarian,3, 140, 142, 146, 168, 169, 171, 172,179,204,218,226,232,246,250

I iconicity, 314 Icelandic, 28, 190 Idoma, 300f IDS (see infant-directed speech) imitation, 120,211,242, En 137, 177,Jp 115, Ko 115, 118 implicature, conversational, 299f incorporating languages, 168 individual differences (see also style; strategy; variation, individual), 200-255 causes of, 235-244 cognitive explanations for, 238f, En 238, Fr 238, Ki 239, Ko 238f, Ma239 guidelines for research, 254f input explanations for, 241-244 methodology, 244--250 neuropsychological explanations for, 236ff in phonology, 139, 155, 159f, 162ff, 166, 176fL 186, 188, 191,202, Da 137, En 137ff, 202, Ge 137, It 137, Jp 181, Ka 181, No 137, Pr 137 pragmatic exp1anations for, 248f temperamental explanations for, 239f, Ek 240, En 240, Ko 240, PI 240 Indo-Europeanlanguages, 17, 170,275 infant -directed speech, 136, 152ff, 191, En 153f, Jp 153,176, Ma 154, PI 153 inference, pragmatic, 300, 304, 309 infinitival complement, En 233, Ss 233 infixation, 207, Bontoc 151f, Chamorro 151, Micronesian 170, Palauan 151, Tg 169f,180

Subject Index infleeting languages, 281, 308, Ek 137, Ge 137, Th 137 vs. isolating, 137, 139 vs. agglutinating 139f inflection, En 136, Fi 136, Fr 169, He 181f, Jp 181, P16f, Rs 6f, 136, 184, SC 6, Th 136,216 infleetional imperialism, 182-184, 226 innateness (see also nativism), 5, 13f, 48f, 93, 99f, 126f, 201, 235f, 25lf input, 42-49, 214, 277, 310, 313, 318, ASL 244, En 49, 57, 60-62, 243, Fi 242f, Ka 242, Ki 245, Ko 56f, 60-63, 100f, 109, 115, 118,243, Ma 54,57,243, Sa 244 and individual differenees, 201, 212, 241-244,250,255, En 241f, Fi 242, Jp 241 and semantic development, Ko 120-128,238 instrumental, 288f intensity, 141, 143 interdigitation, He 150f, 169f, 179 interrogatives (see questions) irregularity, En 190 isolating languages, 308, En 137,204, Fr 137, Ma 137 vs. inflecting languages 137, 139f Italian, 28, 34,137,143,170,173,174,179, 186,206,207,225,245,281

J lakaltek, 284 lapanese,3, 17,23,29,65,67,93-95,97, 115, 123,140-144,153, 168f, 174, 176, 180f, 186, 183f, 188f, 205, 207, 212f, 215f, 218L222,224,230,232,241, 249,281, 286,292,304

K Kaluli, 181, 186,213, 218f, 232f, 242 K'iehe', 140, 148, 150, 154, 155-158, 162f, 166,168,179,183,185,231,239,245 knowledge, shared (see also expectation; newly aequired information; old vs. new information), 120, Jp 115, 249, Ko 115-118, 120 Korean, 3, 17,41-128, 140, 169, 180, 188f, 214f, 218f, 230, 233, 238-240, 243, 248f, 252, 285f, 290f, 304, 308, 315, 317 Kwakiutl,268

L language aequisition theories, 2, 4f, 266f

Subject Index Latin, 12, 170 learnability (see learning theory) learning theory, 274, 277, 283, 287-290, 296f, 300,302,309-311 lexical development, early, 202f individual variation in, 214--217, 249f, En 214f, Jp 215, Ko 214, Ma 214 liaison, Fr 147, 175, Tu 175 linguistic change (see diachrony, linguistic; grammaticization) linguistic determinism, 127 linguistic relativity, 267 locative (see also motion event; motion verb), 9, 270,272,276,289,291,317, Du 9f, 291f, 317, En 9f, 16, 42, 46,82-90, 123-125, 290f, 293,317, Ge 9f, Jp 123, Ko 82-102, 121, 123f, 290f, 317, PI 93, Rs 93, SC 44, Sp 16, Tu 44, Tz 305f, Wa46

M MacArthur Communicative Developmental Inventory, 203, 209 Mandarin, 54, 57, 94, 126, 137, 140, !50, 154, 165, 168, 180f, 184, 188f, 214f, 228, 233,243 Manipulative Activity Scene, 44f, 124, 290, 301,317 maturation, functional categories, 224f, 237, 246f, 251f Mayanlanguages,284,286 means-end tasks, 238 En vs. Ko, 62f, 127 methodology comprehension, En 156, 177 crosslinguistic, 187-190 imitation, 137, 177 individual differences, 244--250 minimal pair studies, 187-190 productivity, measuring, 226f, 247 Micronesian languages, 151, 170 modals, modality (see also evidentials ), 108, En 278,307, Ko 102-120, 124, Sp 279f agent -oriented, 119 deontic, Ko 103f, 108 epistemic, En 119, 124, Ko 103-108, 119, 123 modularity, 235-237,251,266,272, 276f, 281, 295, 305f, 309f, 318 Mohawk, 140, 155, 163, 168, 185, 187 mood, 203, En 120, Ko 109f mora, mora-timed languages, 142f, 176, Jp 143, 174,207,212 morpheme, Ma 168, Vietnamese, 168 bound, 167f

335 closed-class, 265-318, En 56 defining, 277-282, 294 grammatical, 136-192, Bontoc 151, En 137, 139, 154, 156, 177, 185, He 150f, It 174, Ki 150, 157f, Ma 166, Sw 158f input frequency, En 223 open class, 265-318, Ko 56 morpheme order, 182, 187, 189,314, Da 8, Ek 168, Gn 168, Ki 168, No 8, Sw 8 morphological paradigm (see paradigm, morphological) morphology, Ek 223, 226, 246, En 224, Gn 224, He 3, 138, Hu 246, It 225, Jp 224, PI 224f, SC 225, Tu 225, 246, 280f and individual differences, 223-229 verb, Ic 190, Ko 55f, No 190 morphophonoiogy, 139, 147, 169, 171-173, 188, Ek 147, 171, 173, En 137, 183, Fi 147, 169, 172, Ge 169, 172, He 169, Hu 169, Jp 169, Ko 169, PI 169, Pr 169, Rs 16~~16~~16~Tgl~,Tu1~

morphosyntax, 137, 145-148, 150 motion event, 64--66, 285, En 72, 74, 79-82, 121, Ko 66-70, 75-82,93, 95f, 121 motion verb (see also locative), 303, En 16-27, 29f, 48, 65, 68-72, 79f, 121, 285f, 290-293,315, Ge 26f, 29-31, He 26f, Jp 93, 284, Ko 57-59, 63,65-73, 75-80, 121,285~290-292,315,My 284, Sp 16-27,48, 65f, 292f

N naming, En 60 vs. activity-oriented utterances, En 62, Ko62 naming explosion (see also verb spurt; vocabulary explosion), 48, 52, En 62, 127, 238, Fr 238, Ko 63, 238 narrative development, 17, 316, En 17-19, 21f, 24f, 27, 45, Ge 27, 45, He 27, 45, Sp 17-19, 21f, 24f, 27, 45, Tu 27 satellite- vs. verb-framed languages, 28f nativism (see also innateness), 13, 42, 99, 126, 237,266,272~297,300

Navaho, 186 negation, 229-231, ASL 230, BEn 230, En 121, 230f, 248, Fr 121,230, Ge 231, Gk, 231, He 230, 248, Jp 230, Ki 231, Ko 121,230, PI 230, Sc 230f, Tu 230 scope of, 231, 253 Nez Perce, 286 neutralization, 173, Ge 172, Rs 184 newly acquired information (see also expectation; knowledge, shared; old vs. new information), Ko 105-107, I !Of, 114,118

336 nominalization, Ek 168 Nootka, 268 Norwegian, 8, 28, 137, 140f, 148, 156, 163, 176,186,188-190,206 nouns, 46--49, En 49, 53f, 57, 60f, 214f, Ko 51-54,61, Ma 54 collective, 214 concrete, 48, 214 early semantics of, Ko 57 mass, 214 vs. verbs, 214-217, Ek 215, En 214-216, Jp 215f, Ko 214f, Ma 214f weak, Ge227 noun morphology, Ko 56 null pronoun (see ellipsis; pro-drop)

0 object permanence, 43f, 62, 238 old vs. new information (see also expectation; knowledge, shared; newly acquired information), Ko 103, 108, 110--114, 117f, 120 omission, 177, 182f, Ek 182, En 183, Jp 183, Pll82, Rs 182, Ss 183 one-unit utterance, 179f one-word stage, En 83, 12lf, Ko 83, 121 online processing (see processing) operating principle, 3, 7f, 101, 108, 149,224, 273f, 281, 311 order of acquisition, 119, 246, En 223f, 243, 248, Gn 224, Jp 120, Ki 245, Ko 120f and coordination, 233f, 248 and nativism, 251f and negation, 230 and questions, 232 order, morpheme (see morpheme order) word (see word order) overextension, 42, Du 9, Ko 82, 88f, 122, 124, 126, Ma 126 overgeneralization, 226, Ko 97-102, PI 227f, Rs 6, Sw 159 overmarking, Ma 228, PI 228, Wa 228 oversegmentation, Ka 213

p paradigm, morphological, Ge 182, Pl6-8, 182, 184f,Rs6--8, 182, 184,SC6 parameter setting, 2, 5, 12-16,201,216,251, 294,311,315 particle pragmatic, Jp 3, 115, 181, 248f, Ko 3, 51, SSf, 102-120, 123f, Ma 181

Subject Index verb, Da 8, 189, En 16--18,26, 74, 83, 315, Ge 30, No 8, 189, Sw 8, 189 passive, Chichewa 11, En 11, Ss 11, Zu 11 perception, acoustic, 136f, 152-160 perfect, En 302 Persian, 302 phonology autosegmentall60f, 191 individual differences in, 202 metrical, 143, 162, 191 and morphology, 136--139, 146f, ISO phrasal style (see style, phrasal) pitch, 141f, 148, 186, 188, Chamorro 148, Da 188, No 148, 188 pitch-accent languages, 141, Jp 141, Mo 155 pivot structure, 181,205,207, 209-213, Da 213, En 212, He 213, Jp 216 plurifunctionality, Ko 94 Polish, 6f, 28, 93, 96, 140, 146f, 153, 169f, 173, 176, 180, 182, 184, 187f, 207,218,220, 224,227f,230,240,248f politeness (see also honorifics), Ko 249 Polynesian languages, 65 polysynthetic languages, 139f, 168, 308, Ek 168, 180, Mo 168 Porno, 308 portmanteau morpheme (see also fusion, semantic), 182, Ge 170, IE 170, It 170, PI 170, Pr 170, Rs 170, Sp 170 Portuguese, 28, 137, 140, 142, 169f, 173f, 176, 186f, 206 possessive, possession, 211, SC 225 pragmatic inference (see inference, pragmatic) pragmatics, 203f, Jp 248f, Ko 249, Pl249, Sc 249 prefix, 186f, Ba 3, Gn 186, He 186, Mo 187, Navaho 186, Ss 186 preposition, 272, Du, 9f, En 9f, 16, 20, 22, 26, 293, 303, Ge 9f, Sp 16, 22, 293 locative, 293 primitives, conceptual, 123f, 316--318 processing, 186, 303f, 306, 308f, Th 186 pro-drop (see also ellipsis), En 216, Sp 33 production, early, 155f, 161-166, 177, 184, 188, En 186, Ki 155f, 158, Mo 155, Ss 159, Sw 158f productivity, measuring, 226f, 247 pronoun, 250, 307f, En 308, Fr 308, Ko 308 null (see ellipsis; pro-drop) prosodic fillers (see fillers, prosodic) prosody, 139-166, 188 definition of, 203 individual differences, 203-207 place holders (see fillers, prosodic) protomorpheme (see also fillers, prosodic), 163, 184f, En 177, Ge 164, It 179, No 163

Subject Index prototypical event, 124f, 273, 276, 290, 301, 316f pseudomorpheme (see finers, prosodic; protomorpheme)

Q questions, 231-233, En 205, 232, Ge 205, 232, He 232, Hu 232, Jp 205, 232, Ka 232f Quiche (see K'iche')

R rate of development, 202, 224, 208f, 287 reanalysis, 299, 304, 309 reduplication (see also fillers, prosodic), 160, 170, En 205, Ma 154, PI 207 referential style, 209, 212, 239, 24lf vs. expressive, 208f, En 208 register, Sa 221 relational words, En 56, 62, Ko 56 relative cIause, 29, En 32-34, 235, Ge, 234, He 32, 34, It 34, Ko 234, 252, PI 234, Sc 234, Sp 32-34, Tu 34f, 234f relevance, semantic, 307f, 312-314 reorganization, Ko 99f, 102 resyllabification (see syllable structure) rhetoric, typology of, 18-20,26 rhetorical style, 25, 28, 32, 34, Ge 31 rhythm, 141, 143-145, 154, 162-166, 173-176, 186, Da 206, Ki 166, No 206 Role and Reference Grammar (RRG), 231, 252f Romance languages, 17,29,32,34,65,187, 218,221,286,292 rote learning phrases, 208-210, 212, 214 Russian, 6f, 23, 28f, 93, 96,136,140,142,144, 169f, 173f, 176, 182, 184, 187f, 237, 268,301

s salience, acoustic, 139, 144, 155, 174, 178, 188, Da 8, En 49, Ko 50, 60,108,119, No 8, Sc 149, Ss 159, Sw 8, 158f Sarnoan, 46, 218, 221, 244 satellite-framed language, 19, eh 17, Du 23, En 17-20, 22, 65, Finno-Ugric 17, Ge 23, 26, JE 17, Rs 23 Scandinavian languages, 149f, 156, 168, 189f, 230f, 246, 249 scope of interrogation, 232f ofnegation, 231, 253 of operators, 254 secondlanguage,204

337 segmentation, 139-141, 143, 145-147, 150-166,169,171-176,179-181,185, 187, Ba 3, En 137, 145, 203f, Fr 136f, 147,175, Gn 183f, He 179f, Ka 213, Ki 179, Ma 136f, Tg 180 semantic development, 42-47, 239 sernantic domains, 287f semantic fusion (see fusion, semantic) semantic relations, 210f semantic relevance (see relevance, semantic) semantic transparency, Ko 108 Semitic languages, 3,17,29,32,34,65,150, 286,292 Serbo-Croatian, 6, 28, 44, 218, 225 Sesotho, 11, 140, 156f, 159, 165, 169, 180, 182f,186,205,207,220,233 Siswati, 165 Slavic languages, 3, 6, 8, 287, 292f Slovak,288 social-expressive style (see style, social-expressive) Spanish, 16-28,32-34,45,48,65, 140, 146f, 169L 174, 187L 245, 251, 279L 292f Specific Language Impairment (SLI), 174, He 18lf Spotlight Hypothesis (see also morpheme, grammatical; salience, acoustic) stern, 179, 186, Ss 157 strategy (see also individual differences; style; variation, individual), 159, 166, 176-185, En 212, Jp 212 continuity of, 211-213 morphologieal, 179, 181, 191 prosodie, 188,203-207,213, Da 204, En 203f, Fr 204, He 204, Hu 204, Tu 207 syntactic, 178f, 181, 191 stress, 160, 173f, Ba 146, Da 8, En 14lf, 145, 176, Fi 146, 179, Fr 188, Hu 142, 146, 179, Jp, 188, Ki 148, 157f, 162f, Ko 188, No 8, 141, PI 7, 146-148, Pr 142, 188, 188, Rs 7, 142, 188, Sp 188, Sw 8, Tu 146, 280f, Wa 146 cIosed vs. open dass words, 280f contrast, 139 stress-timed languages, 143-145, 173-176,186, Da 173, En 143f, 173f, 204, Ge 173, He 174, It 173f, PI 173, Pr 173, Rs 144f, 173f, Sw 143, Wa 173 style (see also individual differences; strategy; variation, individual), En 209-212 continuity of, 209-213, 249 nominal/pronominal, 211, Fr 213 phrasal, 209-213, 239, 24lf, 249, Da 213, Jp 213, Ka 213 referential (see referential style) rhetorical (see rhetorical style)

338 social-expressive, 208f, 239 "syllable-and-segment," 205-207 "tune," 205-207, 209, Da 206, Fi 206, Ge 206, He 213, It 206, No 206, Pr 206 subject-oriented language, eh 11, Eu 11, 32 subordination, 29, 234f, Ge 234, Gk 234, Ko 234, PI 234, Sc 234, Tu 234 suffix, Ek 186, Gu 186, He 186, Ss 186, Tu 186 suppletion, Eu 172 suprasegmentals, 16lf Swahili, 289 Swedish, 8,28,143,146,156, 158f, 176, 180, 182, 188f syllabification (see syllable structure) syllable, 144f, 154, 160, 162, 184f, Ki 155-158, 185, Mo 155, 185 "syllable-and-segment" children (see style, "syllable-and-segment") syllable structure, 139, Boutoc 15lf, Eu 146, Fr 147, 175, Ge 146, Gk 174f, He 150f, Ki 150, 154, 185, Ma 150, 154" Sp 146f, Sw 146 syllable-timed languages, 143, 174f, 186, Fr 144, 174, Jp 144, Sp 174, Thai 174, Tu 207 syllable weight, 142, 173f syntax (see also complex sentences; negation) 47,310, PI 249 synthetic languages, 168, Fi 168, Tu 168

T Taiwanese, 189 Tagalog, 140, 169f, 180 Tamil,288 telegraphic speech, 205, 209-213, He 213 Thai, 174, 189 timing (see rhythm) tone, 141, 186,204, Ma 166, 188f, Ss 159, 182, 205, Sw 159, 182, 188 topic-oriented languages, Ss 11, Sp 32, Zu 11 topological notions, 271, 273, 276 truth (of a proposition), Ko 106, 112, 115 "tune" children (see style, "tune") Turkic languages, 17,29,65,286,292 Turkish, 3, 23, 26-29, 34,44, 136f, 140f, 146, 165,168-172,175,180,182,184-186, 206f, 213, 216f, 225, 230, 246, 280-282, 301 two-unit utterance, 180-182, 186, Ek 191, Gn 191 two-word utterance, 209-214, En 79f, 82, Ko 8lf, 121 typology linguistic, 1-35,65,292,298, 315f

Subject Index morphological, 139f, 167-176, 178, 185,187,191, En 179, He 170, Tg 170 Tzeltal, 305f Tzotzil, 123

u underextension, 42 universal grammar (UG), 13f, 224f, 229, 247, 25lf universals, acquisitional, 5, 273, 277, 252 implicational, 11-17, 26, 32, 276 linguistic, 2, 5, 12, 269, 298, 318

v variation, individual (see also individual differences; strategy; style), 200-203, 206f, 209 and acquisition theories, 250-254 in lexical development, 214-217, 249f, En 214f, Jp 215, Ko 214, Ma214 Vepsian, 189 verbalization, Ek 168 verb, 42, 46-49, En 47, 49, 53f, 61, 63, 190, Gn 183, Ko 47,50-61,63, Ma 54, 57, Ss 159, Tu 165 early semantics of, Ko 57 irregular, En 190 light, En 304, Fi 304, Fr 304, Jp 304, Ko304 motion (see motion verb) vs. noun, 214-217, Ek 215, En 214-216, Jp 215f, Ko 214f, Ma 214f object destruction, En 295 serial, Ma 299,308 stative vs. process, Ko 126, Ma 126 verb argument structure, 217, Eu 246, Fi 218, Hu226 verb c1asses, 190, 294f, En 190, Ge 190 verb-framed languages, 16-29, Fr 23, He 26, Jp 17, 23, Ko 17, 65f, Ro 17, Se 17, Sp 17,20, 22f, 65, Tu 17, 23, 26 verb island, 246, Ka 219 verb morphology (see morphology, verb) verb spurt (see also naming explosion; vocabulary explosion), En 53, Ko 51-53,55, 121,127 vertical construction, 163 Vietnamese, 140, 167f vocabulary explosion, 247f (see also naming explosion; verb spurt)

Subject Index vowel, Pr 142 vowel harmony, 147, 17lf, Es 146, Fi 3,146, 171, Hu 3,146, 171f, Th 3,146, l7lf vowellength, Da 176, En 142, Ge 176, No 176, Pr 142, Rs 142, Sw 176 vowel quantity, 145, 174 vowel reduction, En 142,174,176,186, PI 188, Pr 142, 174, 188, Rs 142, 174, 176, 182, 184,188, Sp 188

w VVarlpiri,28, 46, 146, 173,205,217,228 VVest Greenlandic Eskimo (see Eskimo)

339 word order, 4,12-14,217-223, 242f, 254, He 34, Ka 219, Ko 60, 103, Sp 34, Th 34f grammatical, En 12, 217 pragmatic, 222, Latin 12, Gn 225, PI 220 pragmatic vs. syntactic, 220-222 variation, Ek 218, Fi 218, Fr 221, Ge 218,221, Hu 218, Jp 218f, Ka 218f, Ko 218f, PI 218, 220, Ro 218, Sa, 218, 22lf, SC 218, word stress (see stress)

z Zulu, 11