Creole and Dialect Continua: Standard Acquisition Processes in Belize and China (Prc) (Creole Language Library)

CREOLE AND DIALECT CONTINUA CREOLE LANGUAGE LIBRARY (CLL) A companion series to the "JOURNAL OF PIDGIN & CREOLE LANGU...

Author: Genevieve Escure

27 downloads 1196 Views 15MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

CREOLE AND DIALECT CONTINUA

CREOLE LANGUAGE LIBRARY (CLL) A companion series to the "JOURNAL OF PIDGIN & CREOLE LANGUAGES"

Editors Pieter Muysken (Amsterdam) John Victor Singler (New York)

Editorial Advisory Board Mervyn Alleyne (Kingston, Jamaica) Norbert Boretzky (Bochum) Lawrence Carrington (Trinidad) Chris Corne (Auckland) Glenn Gilbert (Carbondale, Illinois) John Holm (New York)

George Huttar (Dallas) Salikoko Mufwene (Chicago) Peter Mühlhäusler (Adelaide) Pieter Seuren (Nijmegen) Norval Smith (Amsterdam)

Volume 18

Geneviève Escure Creole and Dialect Continua Standard acquisition processes in Belize and China (PRC)

CREOLE AND DIALECT CONTINUA STANDARD ACQUISITION PROCESSES IN BELIZE AND CHINA (PRC)

GENEVIEVE ESCURE University of Minnesota

JOHN BENJAMINS PUBLISHING COMPANY AMSTERDAM/PHILADELPHIA

The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences — Permanence of Paper for Printed Library Materials, ANSI Z39.48-1984.

Library of Congress Cataloging-in-Publication Data Escure, Geneviève. Creole and dialect continua : standard acquisition processes in Belize and China (PRC) / Geneviève Escure. p. cm. - (Creole language library, ISSN 0920-9026 ; v. 18) Includes bibliographical references (p. ) and index. 1. Creole dialects. 2. Dialectology. 3. Language acquisition. 4. Sociolinguistics. 5. Gram mar, Comparative and general-Syntax. 6. Pragmatics. 7. English language-Social aspectsBelize. 8. Chinese language-Social aspects. I. Title. II. Series. PM7831.E74 1997 417'.22-dc21 97-4278 ISBN 90 272 5240 8 (Eur.) / 1-55619-173-1 (US) (alk. paper) CIP © Copyright 1997 - John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. John Benjamins Publishing Co. • P.O.Box 75577 • 1070 AN Amsterdam • The Netherlands John Benjamins North America • P.O.Box 27519 • Philadelphia PA 19118-0519 • USA

Acknowledgments This book is first a tribute to the people of Belize, and more particularly the Creoles of Placencia and the Garifuna of Seine Bight, who shared with me their joie de vivre, and their love of language. There is no space here to thank all the men and women who opened their doors and inner lives to make me feel wel come. I especially want to acknowledge the warm friendship and support pro vided in Placencia by the Lopez family (Miss Cordelia, Clarence, Therese and Errol), Miss Tila (Athelo) Cabrai, Herbert, Miss Doris, Miss Lucille, Miss Sonia, Miss Tensy, Miss Pearl, Miss Lizzie, Blink, Dudu, Bobo, Captain Morgan, Peter, Karim, Philip, and all the pikni (many of them are now adults) especially Verna, Karen, Dennis, Dean, Edward, Chrissie,and Grayson. Special thanks to Errol who assisted me with the fieldwork in Placencia. This is also a tribute to the Seine Bight Garifuna. It was thanks to Mr. Flores, who has now passed away, that I discovered Seine Bight and, later, Placencia. We became friends during the long bus ride from Punta-Gorda, and he invited me to his distant home across the lagoon from Mango Creek. Had he not extended such warm hospitality, I would probably have bypassed the Placencia peninsula, then totally isolated from the main road from Punta-Gorda to Belmopan and Belize-City. My old friend Roman Zuniga also passed away. He was my Garifuna teacher, and he shared with me his humor and his pride in his people's, traditions and language. Little Mark, who at fourteen spoke Garifuna, Creole, Maya, and Kekchi was my most reliable guide in the bush of Toledo District. I cannot possibly mention the names of all the individuals who befriended me, showed me how to dive for conch and lobster, gut a fish, bake cassava bread, or invited me for fish and fufu. They were Creole, Garifuna, Maya, Kekchi, or Mestizo, in Orange Walk, Dangriga, Hopkins, Georgetown, Punta-Gorda, Barranco, San Antonio, Sartineha, Benque Viejo, Caye Caulker, and Belize-City. They instilled in me a profound love of Belize and its diverse people. In China too, I met warm and friendly people in the rural areas as well as

vi


in the crowded cities. I gratefully acknowledge the cooperation and open-mindedness of the officials and professors at Huazhong Central University of Science and Technology in Wuhan, especially Cheng En Hong, Lei Guo Pu, and Mu Li Ya, who invited me to teach a stimulating group of teachers for an entire acad emic year, and at subsequent times, and gave me the opportunity to travel freely and get to know China better than if I had been a mere tourist. I owe a lot to my students, many of whom became my friends. They taught me a great deal about the sociolinguistic situation in China. Particular thanks to Chen Kairong, Cheng Kola, Deng Shan, Guan Xiangyang, Liu Lichun, Peng Yuanchu, Tan Xiangyun, Wan Youzhong, Wang Qunying, Zou Hong, and final ly to Zhang Xianping who gave me the opportunity to give lectures in his dis tant Hubei hometown of Xianfan. I fondly remember the enthusiastic welcome I received at Xianfan University. I also benefitted from a summer teaching at Tsinghua University in Beijing. Thanks to Lu Ci and Wu Guohua for their hos pitality, and mostly to Chun-Jo Liu for her initiative in organizing the summer program. I gained a deep appreciation of the people and linguistic diversity of China mostly through chance encounters. Strangers went out of their way to help me when I got desperately lost in Chengdu—unable to get the right bus because of my difficulties with tonal differences; they bought tickets for me in the crowd ed train stations of Wuhan, Shanghai or Kunming; and they offered me tea on dusty Tibetan roadsides. Those strangers were kind to me; they saw me as a per son, not as an alien (the distrusted waiguoren), and they contributed to human izing the anonymous crowds of the Middle Kingdom. During long train rides across the People's Republic, I pored over my Chinese notebooks, trying to figure out a language survival kit that would enable me to achieve basic communication in diverse linguistic areas. A novice learner relating new knowledge to old knowledge, and looking for shortcuts, it occurred to me that Chinese syntax is very much like Creole syntax. Thus ger minated the idea of a comparison of Chinese and Creole structures. However, bringing the project to completion was as lengthy, hazy, and rocky as a train ride in a smoky "hard seat" compartment. None of this work would have been possible without the steady support of the University of Minnesota Graduate School, that provided leaves, research assistants, and limited but steady financial support. I also acknowledge occa sional contributions from the English Department. I am particularly grateful to J. Lawrence (Larry) Mitchell, for his encouragements to pursue my interest in nonstandard varieties, as well as to Sandy MacLeish for his positive attitude and moral support. All my thanks also to the various students and research assistants who discussed issues with me, and helped me transcribe data: Gail Ostrow, Dennis Dougherty, Chao Shoushing, Mao Luming, Ding Ersu, Guan

vii

Xiangyang, Angela Karstadt, and most of all Anna Fellegy, who provided invaluable scholarly comments and editing help in the final stages of the mansucript. Ï hold a deep appreciation of Pieter Muysken's encouragements and patience. Chris Corne has my special thanks because he gave me the confi dence to push onwards by providing helpful critical comments on the first part of the manuscript. I also acknowledge the help of Chun-Jo Liu in reading and commenting on parts of the Chinese sections. In the last stages, the detailed contributions of the reviewers, Stephen Matthews and Virginia Yip were extremely helpful in sharpening issues in the book and relating my findings to current developments in the field of Chinese sociolinguistics. Finally, many friends and colleagues helped in indirect ways, in particular as contributors to the meetings of the Society for Caribbean Linguistics, the Society for Pidgin and Creole Languages, and the Institut d'Etudes Créoles.

Contents Acknowledgments

v

I.

Introduction: Developmental continua 1. General issues in linguistic change 2. Aspects of acquisitional studies 3. The linguistic analysis of nonstandard dialects 4. Creoles and noncreoles

1 1 3 10 18

II.

The Belizean speech community and the use of English 1. Introduction to the history of Belize 2. Ethnolinguistic composition and census data 3. The language situation 4. Focus on the Stann Creek district 5. Conclusion

25 25 28 36 43 55

III. Creole acroiects as innovations 1. Radical creoles, postcreoles, and decreolization 2. Acroiects and standard dialects: Social aspects 3. Acroiects and standards: Linguistic aspects 4. Conclusion

57 57 65 73 86

IV. The interaction of syntax and pragmatics in acroiects: Topic marking 89 1. Syntax and discourse features 89 2. Types of topic strategies 94 3. Distribution of topic strategies 103 4. Comparison with American English 117 5. Pragmatic aspects of syntax 120 6. Conclusion 122 V.

Sociolinguistic perspectives on Chinese 1. Chinese 2. Aspects of the linguistic history of China 3. Main subgroups of "Chinese" 4. The reform of Mandarin: Guanhua, Guoyu, Putonghua

125 125 130 135 139

CREOLE AND DIALECT CONTINUA 5. Varieties of Mandarin outside Beijing: Wuhan 6. Wu varieties and Suzhou Wu 7. Conclusions

ix

144 146 148

VI. Topic mechanisms in Chinese: An overview 1. Introduction 2. Topic fronting 3. Topic repetition 4. Topic presentation 5. Summary of topic strategies

151 151 156 163 166 182

VII. Literary and colloquial Putonghua 1. The Putonghua corpus: Methodology 2. Literature and the Cultural Revolution: Two New Friends 3. Literature after the Cultural Revolution: The Matchmaker 4. Beijing colloquial Putonghua 5. Overview of topic strategies in Beijing Putonghua

185 185 188 202 209 219

VIII.Topic strategies in varieties of Putonghua as second dialects 1. Introduction 2. Putonghua spoken in Wuhan 3. Comparison of Beijing and Wuhan Putonghua 4. Wu Chinese 5. Comparison of native and non-native varieties of Chinese 6. Conclusions: Toward a universal view of topic processes

223 223 224 239 243 258 260

IX. Conclusions: Pragmatic universals in second dialect acquisition 1. Patterns of topic marking: Summary 2. Topic marking in the Belizean continuum 3. Patterns of topic marking in Chinese 4. Dialect versus standard: Sociolinguistic universals 5. Conclusions 6. Postscript: Directions for education

263 263 265 271 273 285 286

Bibliography

289

Index

303

Chapter 1 Introduction: Developmental Continua Abstract The issue of dialect acquisition is compared to other types of linguistic developmental continua, including second language acquisition. The impact of social attitudes on language development is examined in various communities, as such attitudes may affect the acquisition of standard varieties by speakers of nonstandard varieties. Methods and models for the analysis of language in spontaneous discourse are dis cussed, and the linguistic units selected are illustrated. A combination of sociolinguistic methodology and functional grammar appears to be the best suited to this study. Finally, I present a justification of the proposed comparative study of a creole context (Belize) and a noncreole context (China). 1. General issues in linguistic change Language is an essential element of practically every human activity, yet its flexible and adaptive nature in social communication is largely ignored, and even denied, by its users. Despite the advances made in linguistics over the last fifty years, most speech communities cling to general normative attitudes toward language, believing that it is (or at least should be) static, monolithic, homogeneous, and primarily preserved, perhaps further refined, by scholarly experts, academicians, writers, and educators. While linguistic change is rec ognized, it is deplored as contributing to the deterioration of "pure" linguistic norms. Pop grammarians contribute to this prevailing view, claiming to correct the process of disintegration by denouncing the solecisms or barbarisms com monly attributed to the poorly educated and the younger generations. In fact, the popular view that blames change on specific social factors and age groups effectively signals a general awareness that linguistic change is highly depen dent on such factors. Correlating social and linguistic factors for explanations of language

2


change is a relatively recent target in linguistics. The search for factors that motivate linguistic change has been applied to many areas of language study. Language change is most evident in such areas as historical linguistics, native language acquisition, second language acquisition, pidginization, and creolization which are all characterized by developmental continua—dynamic levels or stages which are not separated from each other by clear boundaries. In contrast to the early structuralist view that synchronic and diachronic processes were not comparable, there has been increasing awareness of the potential structural sim ilarities existing between different types of developmental continua. Such lin guists as Meillet (1921) and Martinet (1955) already emphasized the regularity of change, commonly referred to by structural linguists as the Uniformitarian Principle. According to that view, the same mechanisms which operated in the past can be observed in contemporary variation,' mostly because integrated rule systems are considered to be stable and to change or borrow only in the direc tion of "linguistic drift" or native tendencies. Sociolinguistic methodologies seem to reflect and endorse this approach in their use of apparent time (gener ational studies) and real time (longitudinal studies) to document linguistic change (Labov 1972a; 1972b). Recently, attempts have been made (Andersen 1983; Muysken & Smith 1986) to compare second language acquisition (SLA) to pidginization, creolization (Valdman 1983), and decreolization (Rickford 1983; Schumann 1978; Stauble 1978; Andersen 1983); first language acquisition, as well, has been compared to creolization/pidginization (Bickerton 1981). Nevertheless, there is still no clear consensus as to the nature of the structural similarities linking those developmental continua and the reasons motivating the use of operating principles in language development. According to Bickerton (1981:238) . . .no real connection exists between SLA and creolization: they differ in almost every particular. SLA is done alone, creolization is done in groups; SLA has a tar get, creolization hasn't; SLA is done mainly by adults, creolization mainly by kids; SLA gives you a second language, creolization gives you a first There is, however, general agreement that the adoption of formal princi ples does not operate in a vacuum but that it is linked to practical aspects of the learning situation, such as available linguistic input, psychosocial motivation, and historical factors. For some linguists, "the history of a language is a func tion of the history of its speakers" (Thomason & Kaufman 1991:4). Proponents of the imperfect second language learning hypothesis as applied to pidgins and creoles say that they display features of "interlanguage systems," such as invari ant verb forms, a lack of determiners, the use of demonstratives as determiners, the invariable placement of the negator in preverbal position, the use of adverbs to express modality, a fixed single word order, a lack of inversion in questions,

INTRODUCTION: DEVELOPMENTAL CONTINUA

3

and a reduced or absent nominal marking (den Besten, Muysken, & Smith 1995:97-98). In the search for a better understanding of the principles instrumental in language development, I will focus on two types of linguistic situations which can be characterized as subtypes of second dialect acquisition. Specifically, I will examine the acquisition of standard varieties of a language by speakers of the nonstandard varieties of that same language, with special attention to 1) a creole continuum (Caribbean) and 2) a non-creole situation involving extensive dialect variation (Chinese). The advantages of comparing similar phenomena in obviously distinct contexts are multifold and can be summarized as presenting the following possibilities: 1. 2.

3.

4.

the formulation of cross-linguistic generalizations in studies of acquisition, particularly as they apply to oral discourse; an inquiry into general social attitudes toward marginalized groups, and how this behavior affects the development and use of language varieties produced by those groups; the observation of the interaction of three primary language components— semantics, syntax, and pragmatics—in spontaneous oral communication; and an evaluation of the putative uniqueness of creole languages.

2. Aspects of acquisitional studies Although a substantial amount of linguistic research has been devoted to code switching phenomena between languages, and to second language acquisition (SLA) primarily by adults, relatively little attention has been given to the mech anisms underlying second dialect acquisition (SDA); yet, this type of acquisi tion is undoubtedly more widespread for the simple reason that dialect variation is universal. Whereas not all individuals find themselves in social situations which require the acquisition of a second language, there is no single human being whose repertoire is limited to only one language variety, style, or dialect. Furthermore, unlike second languages, second dialects are typically acquired earlier in life and continue to be acquired throughout adult life.1 Second dialect (usually standard) acquisition is often perceived to be an ambiguous, undefined linguistic process. In West Indian contexts, such learning situations have been characterized as those in which the standard to be learned is neither a native lan guage nor a foreign language (Craig 1971:376; Stewart 1964). SLA research has primarily focused on the analysis of acquisitional steps, particularly on errors occurring in the course of that process. Various mecha nisms have been hypothesized to account for "imperfect learning" and the mis-

4


interpretation of linguistic input, which results in the learner's unstable interlanguage continuum; the mechanisms include language transfer due to native language interference, overgeneralization, simplification, and the operation of universal strategies put into motion when linguistic input is deficient (McLaughlin 1987). Errors are usually traced to conflicts between the phono logical or morphosyntactic systems of the two codes in contact (the native ver sus the target model). The learnability of certain features is accounted for by markedness relations, a theory asserting that a language is easier to learn if it contains more unmarked features than the learner's native language. Jakobson (1940) first introduced this mode of interpretation which points to correlations between the development of phonological features in first language acquisition, aphasic loss, and the patterning of phonological inventories across languages (see also Ellis 1994). Recent attempts to identify types of contact-induced linguistic change distinguish between two main mechanisms of interference, obviously relat ed to acquisitional patterns: "interference that results from imperfect group learning during a process of language shift, and interference that results from borrowing, by native speakers of a language, of features of some other language with whose speak ers they are in contact" (Thomason & Kaufman 1991:212). A more focused group of studies have investigated non-native varieties of English (the New Englishes) examining the relationship between the acquisition of such varieties and claims made by theories of language acquisition and change, such as error analysis, mother tongue interference, and markedness theory (Williams 1987). For example, Platt (1991:376) considers English in Singapore as "a kind of fossilized interlanguage [which] became a lingua franca in the English-medium schools among students whose home language might be one of the Chinese dialects, an Indian language or Malay." He shows that many features of Colloquial Singapore English (articles, copula use, tense and number marking) suggest Chinese influence (Chinese constitutes 76% of the population of Singapore), especially from Hokkien, the dominant Chinese dialect. However, a recent investigation of reported language use and identity in Singapore docu ments the effects of the 1987 compulsory bilingual education policy, which made English the medium of education and required the study of a second "eth nic" language (Mandarin, Malay, or Tamil). At the same time, the Singaporean government discouraged the use of Chinese "dialects," such as Hokkien, enforcing a "Speak Mandarin" policy. Results indicate that English is increas ingly used and that, among the majority Chinese group, there is a generational shift in reported identity from Chinese (especially specific Chinese group iden tity such as Hokkien, Cantonese, or Hakka) to Singaporean (Hvitfeldt & Poedjosoedarmo 1995). One may wonder to what extent these attitudes are affecting Singlish, and more specifically, whether Chinese substratal compo-


5

nents are yielding to other features. Another New English acquired as a second language, Kenyan English is displaying mother tongue interference which reflects the learners' ethnic group. The Kikuyu show evidence of a liquid (1/r) phonological merger in their use of Kenyan English, whereas the Luo use affricates but no alveolar fricatives in their version of English. Both features echo native African features of Kikuyu or Luo (Schmied 1991:429). A similar distinctiveness linked to native language is found in Liberian English, a New English which happens to be integrated into a recognized creole continuum and similar to situations existing in other parts of Africa and in the Caribbean. Singler (1991:558), in his investigation of plur al marking in Liberian English, indicates that plural marking in Kru Pidgin English, whose speakers have Kru as a first language, is neither identical to plural marking in Liberian Interior English, whose speakers have Mande as a first language, nor to plural marking in Settler English, spoken by the descen dants of Black American immigrants to Liberia in the nineteenth century. The traces left by the native or ancestral language in the newly developed, typically contact-induced, variety are commonly referred to as substratai influ ences, which reflect the lower status of the speakers of the new language. On the other hand, any influence of the linguistic component traceable to the social ly dominant group is called superstrata!. Theories of genesis will be briefly dis cussed in chapter two, and a comprehensive presentation of existing approaches appears in Alleyne (1980), Arends et al. (1995:9-10; 99-109), Holm (1988), and Muysken & Smith (1986). Finally, pedagogically-driven research has focused on the identification of strata, or stages in the West Indian learning situation (and more particularly in Jamaica and Trinidad), and the implications for English-teaching methods (Bryan 1996; Christie 1983; Craig 1966; 1980; 1996; Pollard 1996; Roberts 1983; 1988; Shields-Brodber 1989). The important issue of the practical applications of lin guistic research to social change will be discussed in chapter nine. 2.1 Universalist explanations of acquisition One of the major explanatory approaches to issues of acquisition involves uni versalist explanations of the human language learning capacity. Chomsky's Universal Grammar (UG) provides innateness as an explanation for cross-lin guistic similarities. This theory postulates that all humans are biologically equipped with a language faculty that permits the generation of general lin guistic principles. The language faculty is available at birth and becomes acti vated when input is insufficient ¿o provide a full model of the target grammar, which is typically the case in first language acquisition. Grammar development consists of the setting of parameters for the specific language being learned. For

6


UG-proponents, the study of a child's language development is a sufficient basis for the research and testing of language universals, and external factors, such as communicative needs and social interaction requirements, are consid ered irrelevant. The "Bioprogram Hypothesis," a version of UG introduced by Bickerton (1981), incorporates special reference to the emergence of language universals in pidgin and creole languages, a case of first language acquisition. The latter develop rapidly in contact situations in which speakers evolve a new system without the benefit of an already existing model, thanks to the resources of their innate linguistic ability. Bickerton's genetic view allows for some social input, since demographic, historical, and sociopolitical elements are necessarily involved in the genesis of pidgins, but he fails to include systematic reference to daily interactive needs. Assuming that Chomsky and Bickerton are right about the existence of a language faculty in the brain, there may be other types of universals dependent on external factors, in addition to the commonalities which depend on the inter nal properties of the language faculty (Butterworth, Comrie, & Dahl 1984). Internal and external explanations can be distinguished as follows: Cognitive and psychological explanations involve formal operations that the human mind can vs. cannot accommodate. . .while pragmatic or sociolinguistic explanations involve (formal?) operations that a human society or individual within a society can vs. cannot accommodate. . . (Hyman 1984:68). Muysken & Veenstra (1995:121-134) summarize universalist models applied to pidgins and creoles, identifying two types of universals: procedural universals (referring to universal properties of processes such as L2 learning, grammaticalization) and constitutive universals which designate the properties resulting from those processes (e.g.,TMA systems or word order). These prin ciples and their applications in acrolects will be further discussed in subsequent chapters. In view of the previous attempts at understanding language acquisition and change in various contexts, the question is open as to whether linguistic com monalities, internal or external, also emerge in the acquisition of second lan guages, such as pidgins and creoles, and second dialects, such as acrolects or standard varieties acquired by native speakers of nonstandard varieties. Clearly, the issue must be differentiated from first language acquisition, as presented by Chomsky and Bickerton, since learners of a second dialect or language already possess a native linguistic system which interacts with the target system (either causing errors, transfers, interferences, markedness patterns, or substratai and superstratal influences, according to existing theories). If there are contactinduced linguistic universals, how can they be differentiated from language-


7

specific influences? In which language components are they more likely to be identified? These questions and their applications in acrolects will be further discussed in subsequent chapters. 2.2 The acquisition of second dialects The interpretations outlined above could conceivably be applied to other types of acquisitional situations of the "dialectal" type, especially in cases of second dialect learning—cases when there is no distinct mother tongue present in the competence of the standard learner. The general lack of interest in second dialect acquisition may be due to the difficulty involved in identifying second dialect situations as linguis tically distinct from first dialect contexts. In spite of the paucity of research on dialect acquisition, or perhaps because of it, there is a popular consensus that stan dard dialect acquisition is less traumatic than SLA. This attitude is obviously reflect ed in the educational establishment: a standard variety is never taught in a formal, organized manner and, in contrast to foreign language learning, is not supported by any language manual or the "bidialectal" equivalent of bilingual programs. The learner is assumed to know the standard before going to school. We can infer from such attitudes that there is a deeply entrenched conviction that the acquisition of dialects involves less distance between the native and the target varieties (D1 and D2) than is the case between L1 and L2. Of course, the nonstandard speaker may have previously acquired a passive or partial competence of the standard model, depending on the degree of exposure and motivation involved in a specific situa tion. Public opinion as well as educational systems assume that a speaker of Geordie, a variety of Newcastle English (Graham 1980), will find it easier to learn RP than to learn French, or that a speaker of American Black English should learn standard English faster than Spanglish, and Spanglish faster than Spanish. On the other hand, there is apparently no clear opinion concerning the relative learnability of nonstandard dialects by standard speakers. Would it be easier for a speaker of standard English to learn Black English2 or Geordie than Spanish? It is unusual, however, for speakers of prestigious dialects to strive to learn stigmatized forms, since there are neither the economic nor educational motivations to do so. Although the possibility is not excluded, SDA is typically a one-way social process. There are no empirical studies of dialect versus language acquisition that prove or disprove the validity of the above intuitive statements. Although some nonstandard varieties such as Black English have become the object of schol arly study (Labov 1972a), the actual process of the acquisition of standard American English by speakers of Black English still remains to be explored. The spontaneous acquisition of a standard code as second variety—whether we call it dialect or language—is highly constrained by social factors, subjective attitudes, and the psychosocial context of learning. It is, therefore, essential to

8


distinguish between guided learning, the artificial classroom acquisition of a second language, and unguided learning which includes the acquisition of a standard dialect or language in a real life context. Unguided learning situations create the ideal circumstances for the study of the interaction between syntax and pragmatics, as will be discussed below. 2.3 Social attitudes and language/dialect acquisition An important component of language development lies in the existence of social attitudes toward languages and their speakers; yet, SLA studies general ly do not look at the potential significance of the relative social status of the tar get language and of the learner's native language. They also do not take into account the related but distinct issue of the social status of language learners and of native speakers of the linguistic target. For example, there is obviously a different social dynamics involved when comparing the acquisition of English by a Mexican migrant worker in Minnesota, a Mexican migrant worker in California, a Chinese graduate student in an American university, and a Chinese resident of Singapore. There is more of a group support system for Mexicans, or Chícanos, in California than in Minnesota, which has predominantly German and Scandinavian ethnicity and only a seasonal influx of Chícanos. Whether or not a large Mexican population in California, and the universal presence of Spanish or Spanglish in the streets and on the airwaves, facilitates or impedes the acquisition of English by Mexicans remains to be studied. In the other case, a Chinese graduate student on an American campus has a focused, academic motivation to learn English, and the process of SLA acquisition does not inter fere with his or her sense of Chinese identity, unless of course there are personal motives for rejecting it to embrace an American identity. For the Chinese in Singapore, the local English variety is a lingua franca which now symbolizes their identity as Singaporeans, as seen above. Since Mandarin Chinese is offi cially supported by the Singapore government through the "Speak Mandarin" policy, there is good motivation to preserve a Chinese ethnic identity. Social factors are essential to examine because they determine motivation, the availability of learning tools, and other factors instrumental in the develop mental process of language acquisition. As pointed out by LePage & TabouretKeller (1985:247), "neither 'race' nor 'ethnic group' nor 'language' turns out to be a clearly-definable external object," and linguistic choices are "projections" of identity. The learning of non-native varieties, whether they are better defin able as dialects or languages, is at least partially triggered by psychosocial and economic motives. Language is perceived as a marketable asset, a tool for upward mobility. An individual expects to reap benefits, such as status, "class," distinction, recognition and related economic bonuses from the acquisition of a

INTRÓDUCTION: DEVELOPMENTAL CONTINUA

9

new linguistic code. Thus, subjective attitudes toward languages, accents, dialects, and other perceived linguistic entities are directly derived from social factors, and the speakers' histories (Thomason & Kaufman 1199:4). 2.4 Linguistic bias Although the popular (not the linguistic) view defines a dialect as deviant from the norm, the difference between a language and a dialect may be impossible to determine unambiguously, either linguistically or socially. Bias plays an essential role in our perceptions of language varieties and in their acquisition. Linguistic bias reflects the social stratification present in every community, creating a cir cular pattern of cause and effect: language is a social mirror and, thus, perpetu ates social and ethnic bias because linguistic behavior shapes attitudes and opin ions. The lack of power of a stigmatized group is compounded by the negative values associated with the group's linguistic forms (typically described as incorrect, inadequate, inappropriate, uneducated, illogical, and politically dan gerous). A nonstandard variety is often the subject of denial by its own speak ers and by the society harboring it (Ferguson 1959). If at all recognized, it is assumed to be an immature linguistic habit which will be shed in the course of normal intellectual development. If such "bad" linguistic habits persist, users find it harder to establish their social credibility; they may even be denied access to the powerful world of standard speakers. Such attitudes often apply to foreign accents as well and, therefore, to the SLA process. Native speakers often associate—perhaps unconsciously—unflattering intellectual connotations to non-native renderings of their language. A related attitude is the feeling of embarrassment all language learners experience in the first stages of conscious adult acquisition. Clearly, we are strongly motivated to assess an individual's value as a social human being on the basis of that person's verbal ability and conformity to conventional local linguistic standards in addition to other super ficial factors, such as physical appearance and behavior. A prime example of language-based bias is reflected in the recent cam paign to identify English as the sole official language of the United States. The English Only movement has as its goal to deny official status to Spanish, as well as to other immigrant languages, a position which may partly be derived from the common misconception that language unity promotes political unity and economic development.3 The threatening power of linguistic and ethnic diver sity has long been reflected in the stigma attached to Black English varieties in the United States, as well as to creoles in the Caribbean and elsewhere. As reflected in the case of Spanish in the United States, varieties which are commonly referred to as "languages," and that have official and prestigious sta tus in a given country, may hold low status elsewhere. Castilian Spanish, for

10


example, is the official language in Spain and constitutes the prestige form in most of Spain except in Catalonia, the Northeastern province, where it has, for political and historical reasons, a much lower status than the local language Catalan. The use of Catalan was made illegal during General Franco's twentyyear dictatorship. In contrast, Catalan was mostly tolerated in Catalogne, the southern province of France adjoining Spain, and it has fallen out of use there despite the fact that it is actively promoted by cultural organizations. Thus, lan guage differences sometimes disappear in cases of voluntary acculturation but are reinforced in cases of oppression. In Latin America, Spanish has a much higher official status than Indian languages (e.g., Quechua, Aymara) whose speakers are at the bottom of the social scale. In the United States, any variety of Latin American or Caribbean Spanish has low overt social status (which does not exclude high covert prestige) due to widespread negative social attitudes toward minorities, migrant workers, and recent immigrants. The circular pattern inherent in linguistic bias is particularly damaging to speakers of nonstandard or unofficial varieties who need to find creative ways to break the pattern of behavior and judgment. For example, the revalorization of Black English and African-American values through the medium of rap music illustrates the attempt to break the pattern of bias through a linguistic medium. Rap clearly has gone well beyond its original linguistic medium since it has been actively adopted in France with musical lyrics in standard Parisian French by Arab and African "rappeur" groups, as well as in Hong Kong as Canto-rap—rap with Cantonese lyrics by analogy with Canto-pop.4 Studies using the matched guise technique have documented the claim that social judgments are based on linguistic behavior, especially on the vague concept of accent (Giles, Bourhis & Taylor 1977). In England, Standard English (RP) is linked to traditional norms and the "public school" education reserved for the elite. A corpus of listeners found that RP evokes impressions of success, intelligence, and elegance but a certain lack of warmth, especially when used by women. Northern English varieties are associated with warmer but less educated and less successful individuals, which confirms that an official norm derives its prestige from the power of the speech community using it (Elyan et al. 1978). Such normative attitudes pro moting the supremacy of standard dialects or official languages are universally shared, and elementary and secondary education teachers are often encouraged to convey this view to their students. The two case studies to be analyzed in this book will illuminate such attitudes. 3. The linguistic analysis of nonstandard dialects Paradoxically, some of the most striking and productive developments in the


11

field of formal linguistics have been concomitant with a denial of the existence and importance of nonstandard varieties, which indeed runs parallel to public ideology. Transformational Generative Grammar (both in its original form and in its later version as Government and Binding), because of the priority it assigns to the search for an abstract linguistic competence and the identification of the universal linguistic capacity shared by all human beings, has intentionally ignored all references to linguistic variability and to the extralinguistic fac tors—social, cultural, psychological and political—that determine linguistic variability. For formal linguistics, the database consists of the linguists' intu itions. Apparently, those intuitions have not included nonstandard verbal struc tures; thus, omissions of large segments of speakers' repertoires are likely to have contributed to the rejection of nonstandard dialects as a valid field of study and a valuable component of education, and may even have reinforced the non standard bias and the low status assigned to their speakers. Schiffrin (1987b:392) claims that one of the major differences between formal linguistics and sociolinguistics is "a difference in data," adding that "the selection of a data type has a profound influence on the range of phenomena which a model aims to represent and a theory aims to explain." 3.1 Labovian sociolinguistics Until twenty years ago, there had been no large scale study of the dialects of groups with low social status. Labov pioneered the field with his seminal stud ies of the social stratification of English in New York City (Labov 1966) and of Black English (Labov 1972a). Although the study of nonstandard dialects has since greatly expanded in the direction of increased observational adequacy, there is still a glaring absence of data documenting the degree of deviance exist ing between standard dialects as acquired by nonstandard speakers and the ver nacular (native) forms of those standard dialects. Labovian sociolinguistics attempts to deal with speech variability per se, and its relevance to commu nicative competence rather than with the Chomskyan internalized competence of the ideal speaker/hearer. For the first time, questions are raised about the validity of our judgments of standard and nonstandard speech, and full scale scientific descriptions of nonstandard speech varieties are conducted. As issues of descriptive accuracy and observational adequacy are brought to the forefront, it becomes necessary to develop sound methodologies for the collection of lin guistic data. What are the observations that will reliably reveal the mechanisms opera tional in language development? How can we ascertain the difference between linguistic facts and the abstracted interpretation of those facts? Indeed, what theory will provide an explanation of the significant linguistic facts? The intu-

12


itions of generativists and the mailed questionnaires of dialect geographers obviously tapped limited, and manifestly selective, segments of the linguistic options open to speakers and are now no longer considered appropriate to sat isfy the sociolinguists' interests in establishing the range and extent of linguis tic variability. Matched guise techniques and laboratory experiments are com monly used to isolate linguistic features or identify subjective judgments in con trolled situations and produce interesting results, as in the case of the evaluation of RP outlined above; however, they have the disadvantage of creating artificial situations which can only remotely match genuine language use and, thus, pri marily define formal situations or tap conscious aspects of linguistic behavior. One of Labov's methodological principles—the Principle of Formality— states appropriately that "any systematic observation defines a formal context in which more than the minimal attention is paid to speech" (Labov 1972c). Labov advocates methods whose goals are to capture an individual's full per forming repertoire, from formal to casual, through elicitation of word lists and text reading (for formal speech) as well as through observations of spontaneous conversations (for careful to casual speech). The latter can only be successful if the fieldworker is a member of the speech community under investigation or has become fully accepted as a participant in the social activities of that com munity. Participation, therefore, helps solve the "Observer's Paradox," as one observes "how people speak when they are not being observed" (Labov 1972c). It has been claimed that sociolinguistics is merely a methodology and has no theoretical import. This is clearly inaccurate since all socially-oriented stud ies of language variation derive their premises from the basic theoretical view point that the social context determines human behavior, language being just one aspect of human behavior. Variation studies emphasize the necessity of quantitative analysis as a research procedure with the hypothesis that variabili ty in performance is predictable in terms of the correlated social and linguistic features of each natural interaction. Such approaches are traceable to a Marxist view of social class based on conflict and power. The issue of nonstandard dialects stands within this perspective because the social evaluation of language is based on power, or the perception of power and conflict, and on the human desire to change the power structure to one's advantage. Guy (1988:41) claims that Western social theories present an alternative definition of class based on social unity and status as opposed to the Marxist notions of power and conflict. He says that sociolinguists such as Labov have primarily been functioning from this "soft" perspective, with resulting gradient linguistic stratification, rather than the sharp stratification which is expected to evolve from violent conflict; however, sharp stratification in linguistic feature distribution or subjective atti tudes toward language types has been found to occur in issues of minority lin-


13

guistic behavior—whether in Black English, creoles, or in politically-motivated language choice (such as the case of Catalan mentioned above). Consequently, it seems pointless to separate or choose between unity-sta tus and conflict-power dichotomies. Both types of social behavior obviously cooccur and contribute to the fashioning and development of linguistic behavior. The approach adopted here assumes that a social dynamic is essential in non standard dialect learning but does not presume to predict the type of trigger involved in the acquisition process. It is, however, hypothesized that nonstan dard native speakers will only effectively learn the standard if they have strong incentives to do so—if they envision a practical positive outcome to the acqui sition of the linguistic code. If they assess the achievement of power as unreal istic, it can be predicted that they will not learn much of the standard. The popular definition of the term dialect—a nonstandard variety which is somewhat "deviant" from the local norm—is a useful heuristic sociolinguistic concept because it is based on social attitudes regarding language varieties and their speakers. The notion will naturally be essential in this investigation of the process of acquisition of standard varieties by speakers of dialects or nonstan dard varieties. 3.2 Language in oral discourse and pragmatics The data base underlying this investigation of second dialect acquisition pri marily consists of spontaneous conversations. It is the most common type of ordinary communication, however, which poses the most serious problems of analysis. The principles underlying oral discourse lie within the overlapping ranges of pragmatics and sociolinguistics, Yet, there is currently no compre hensive theoretical framework which adequately predicts the organization of discourse, "no theory of paragraphs and its parts which is nearly as elaborate as a theory of sentences" (Linde 1981:85). The field of pragmatics encompasses a broad range of perspectives on lan guage in context. Green (1989:2) places it at the intersection of a number of fields within and outside of cognitive science: not only linguistics, cognitive psychology, cultural anthropology, and philosophy (logic, semantics, action theory), but also sociology (interpersonal dynamics and social convention) and rhetoric contribute to its domain. The fuzzy boundaries that pragmatics shares in particular with sociolinguistics and semantics, as well as with other disciplines, make pragmatics hard to define clearly, although vague definitions are not lacking. It is defined broadly as "the study of understanding intentional human action," requiring reference to the central notions of belief, intention, plan and act (Green 1989:3), and narrowly

14


as "the interpretation of indexical expressions" (Green 1989:2). Levinson (1983) provides an interesting evaluation of various definitions of pragmatics, from its philosophical foundation to a more empirical data-driven approach, implying the growing importance of observational adequacy. Some of those defi nitions include "meaning minus truth conditions," "the study of the relation between language context that are basic to an account of language understanding," "the study of the ability of language users to pair sentences with the contexts in which they are appropriate." Levinson concludes that "conceptual analyses using introspective data [are] replaced by careful inductive work based on observation" (Levinson 1983:285). The functional approach endorsed by Dik (1980) and Givón (1979a; 1990), among others, also regards pragmatics as the primary framework within which "syntactic and semantic principles are explained in terms of the pragmatic pur poses and requirements of verbal interaction" (Dik 1980:2). The functional view of natural language claims to consider language primarily as an instrument of social interaction but, paradoxically, does not discuss real life contexts and pri marily aims to produce principles on the basis of isolated sentences. The advan tage of this approach is its search for "typological adequacy"; as Givón (1990:vii) notes, "surface diversity of cross-language typological facts masks behind it a great measure of commonality of human languages. Part of that com monality is due to semantic and pragmatic universals." Other linguistic research on language universals recognizes the combination of both formal and function al explanations for the commonalities found in human languages (Comrie 1984). 3.3 Theoretical framework for oral discourse The question remains of what theoretical model would most adequately account for the type of speech data to be considered here: namely, spontaneous dis course in the context of standard acquisition by native speakers of nonstandard varieties. The concept of communicative competence and, furthermore, polylectal competence, must be central to this putative model, since I will have to account for the competence of speakers who shift easily between different language varieties depending on the social context. The principles and methods of quantitative sociolinguistics, as briefly outlined above, are well-suited to the detailed analysis of specific linguistic variables. One major aspect of sociolinguistic analysis is empirical, as it involves close attention to social context and to the collection of reliable speech data, an essential element of any attempt at explanations of linguistic facts. In addition, an emphasis on typological generalizations will be essential in the projected comparison of two apparently unrelated linguistic contexts (Belizean and Chinese). Functional grammar provides that angle, with its spe-


15

cial emphasis on universal pragmatic aspects of human communication, espe cially in areas such as topic and focus. Since I elected to exclusively investigate topic marking in spontaneous verbal interaction, the functional approach pro vides a valuable framework within which to place the Belizean and Chinese facts. The processes involved in topic presentation and topic organization pre sent particularly interesting insights into the ambiguous overlap of syntactic and pragmatic processes. Thus, without committing myself fully to either model, I find it convenient and eminently feasible to combine the sociolinguistic framework with the func tional-typological perspective. Indeed, I find the two perspectives to be comple mentary, in the sense that what is lacking in one is supplied by the other. Whereas sociolinguistics is generally restricted to individual linguistic situations, func tional grammar provides options for cross-linguistic generalizations; while func tional grammar has little to say about specific contexts and speakers, sociolin guistics has a well-developed methodology and theory to identify and control the extralinguistic variables which have significant impact on language choice. Finally, the overlap of semantics, syntax, and pragmatics is an important aspect of the study of discourse, and the fuzziness of boundaries is often invoked in functional approaches. Although an account of the absence of neat ly separated components presents a serious challenge, the issue cannot be skirt ed. The notion of overlap is particularly relevant to the study of marginalized varieties. Since nonstandard varieties are usually highly stigmatized, there is no formal or canonical description of their structure, and they are not subjected to a norm. They are, therefore, relatively free to evolve. It is the intent of this investigation to observe whether and how spontaneous contexts favor the clos er association of semantic, syntactic, and pragmatic processes, as noted in cre le situations. A hypothesis to be considered here is that unguided learners, as well as learners deprived of a stable model, are forced to rely on general prag matic or functional principles rather than on the formal syntactic rules of the tar get language because of the continued inaccessibility of the standard grammar. In other words, speakers may pragmatically rearrange the target syntax. The hypothesis will be advanced that such pragmatic reorganization may derive from universally shared principles, regardless of the specific linguistic situa tion. The issue of the putative overlap of pragmatic and syntactic devices will be investigated here more particularly in the acquisition of Belizean acrolects by native speakers of Creole, and of Chinese standard dialects by native speak ers of nonstandard Chinese varieties. 3.4 Units of analysis An essential prerequisite in the linguistic analysis of discourse involves delim-

16


iting the domain of the study. The Chomskyan notion that the domain of lin guistic analysis can and should be the sentence unit constitutes the standard basis for most analyses—but it is easy to demonstrate that investigations limit ed to isolated sentences fail to provide essential information about the broader psychosocial context that generates the parameters within which a sentence occurs. Furthermore, it is often difficult to identify sentence boundaries and, thus, to break discourse into significant sentential units. This structural uncer tainty is represented in the following example, in which a Belizean speaker dis cusses "modernity":5 (1) This thing of modernity sometime, I am skeptical of it myself; I think when you talk of the economics of it you want to bring in some sophisticated system, make them become dependent on it, and, you know, look pan de land, look pan de people, and you just see them as factors of production rather than seeing them for what they really are, what the land can produce and how it will—de people who till it, rather than going into this sophisti cated system, say small is beautiful (Co, 42, 1980, Placencia) The definition of linguistic units in discourse, however, does require some reference to its boundaries and to its internal organization. Wald (1983:104) looks at narratives as "discourse units that are extended, potentially syntactical ly elaborate, commonly spoken, and easy to elicit." Such descriptions capture the complexity of discourse but do not contribute much empirical help to the analyst intent on defining descriptively adequate measurements. A more pro ductive attempt to delimit the "discourse origin" and "the end point" of a nar rative leads Wald to refer to a variety of cues, for example: "a coda such as that's it....or a falling intonation and slowing of the speech tempo finally laps ing into silence. . .[and] the following audience (or addressee) reaction" (Wald 1983:108). Labov and Fanshel (1977), in their study of the language of therapy, iden tify "units of talk" which function as speech actions, and their goal is to formu late the set of principles organizing those units of talk, either as concatenation rules sequéncing information, or mapping procedures which determine the pro duction and interpretation of verbal elements. According to Linde (1981:113), "discourse units have an internal structure that is as regular and accessible as the study of the structure of sentences." She identifies certain internal organizational principles which provide coherence in discourse, such as temporal ordering, relations of dominance and inclusion, and a wide set of social and cultural assumptions and presupposition. Linde defines major linguistic units in terms of genres, such as stories, jokes, recipes, narra tives of activities, and the like, all functioning as social units as well. One


17

advantage of this approach is that it makes sense to speakers, who are far more aware of undertaking to tell a story than they are of undertaking to produce a sentence. A disadvantage of the "genre as unit" interpretation is that it may be too broad. A story is likely to be subdivided into various components, inter rupted by interlocutors' comments, then continued and led to completion; thus, a topic can be maintained over alternating turns of conversation. It may be practically more effective to define linguistic units shorter than the genre, while leaving open the potential of expanding beyond conventional idealized sentential boundaries. Sample (2) shown below illustrates the problem of discourse division. It is excerpted from a conversation taped in Placencia, between Co and his friend Ro. Co and Ro used to be coworkers at the income tax department and the customs office in Belize-City. They recall a common friend Xo. The topic of Xo is introduced by Ro's basilectal question [We he do now?], but Co immediately shifts to an acrolect (more appropriate when dis cussing aspects of their official functions as civil servants in Belize-City) and sometimes to a mesolect (turn 5), although Ro occasionally shifts back to a basilect (turn 4). 6 Beside the issue of lectal shifting, a point relevant to the def inition of discourse units is represented in the fact that the topic is interrupted on turn 6 by Ro's digression into recollections of office relationships, to which Co responds (turn 7); Ro then returns to the topic of Xo (turn 8). Thus, the prob lem is how to break down the following chunk of discourse. Should there be eight discourse units (each turn constituting one discourse unit)? Could it be three, the first discourse unit including turns 1-5 (discussion of Xo), the second (6-7) about office relationships, then the third resuming the topic of Xo (turn 8)? Or could the entire paragraph be considered as one single discourse unit: (2) Ro: We he do now? (what does he do now?) 1. Co: I think he wanted to get into merchandising and maybe in bond stores, duty free shops. 2. Ro: I used to like work with he you know, they gave he lee [little] bad name and so but. 3. Co: I found him alright, boy, I got on pretty well with him, ahm, we were from way back, you and I know him even before being a civil ser vant, and ahm, then I worked with him the first time at income tax, then I worked with him at customs, but, I'm, telling you, I think he was a. . .he had ideas, right. 4. Ro: Dat da we I like with a [that's what I like about him]; and he got good, to me, right, we got good staff relationship, right. 5. Co: He used to fine [He used to be fine] 6. Ro: I know one big place like customs funny for you, because the staff too big, right, but to me, like, I use to like ya [there], they make the staff

18

CREOLE AND DIALECT CONTINUA there feel at home, right, for instance da you lee birthday today and things like that. 7. Co: Yah, and I think it was good, right, kinda, you work together you must also fun together, you socialize, you find it makes for better working relationship. 8. Ro: Dat's right, I never find no fault with the man, man. If you m ga wan man de but da Xo de [If you had to pick a man, it would be Xo]; yah, he da something else, he da dead scatterbrains, man. (Placencia 1980)

For the purpose of this investigation, I will opt to take the entire paragraph as a discourse unit of analysis. My definition of "topic units" is centered around the topic: all structures related to the topic make up the topic unit, as will be further illustrated in chapters four and six. In many languages, including English, the basic syntactic structure of the sentence involves an SVO order in which topics do not have an especially high priority; yet, topic mechanisms occur more extensively in contextualized casu al and nonstandard varieties related to English, which then appear to assign topic structures to the pragmatic component. The incidence of topic mecha nisms in creole varieties, especially in acrolects, is of special interest here and is related to the claim that the pragmatic component takes over in spontaneous contexts. On the other hand, some languages clearly assign syntactic status to what is pragmatically constrained elsewhere, a phenomenon akin to grammaticalization. This may be the case in Chinese which displays consistent topiccomment discourse structures. The concept of topic chain (Shi 1989; Tsao 1979), to be also discussed in chapter six, may provide an appropriate framework for the study of creole topic mechanisms because of its flexibility in allowing a definition of the topic unit which combines syntactic and pragmatic elements. According to the theory of topic chains, the topic extends its domain to a sequence of several propositions. The topic chain, or topic unit is thus a dis course level phenomenon that raises the issue of the interaction of syntax and pragmatics, as illustrated in the Belizean examples above. And it thus appears that cross-linguistic observations provide a productive perspective on our understanding of topic mechanisms.

4. Creoles and noncreoles Parallel to the growing interest in Black English evidenced in the last thirty years, and perhaps even more extensive, has been the development of the field of creolistics over the last twenty years. This creole scholarship has endeavored to rehabilitate stigmatized creole languages in their own communities, and else-


19

where, and to delineate a new field where social and linguistic factors are inter locked in language development. As indicated above, however, the attention devoted to the development of pidgin and creole-related continua has been somewhat limited. Several important books providing overviews of creoles, or case studies, have appeared since 1980 (Arends et al. 1995; Alleyne 1980; Bickerton 1975; 1981; Chaudenson 1992; Gilbert 1987; Holm 1988; LePage & Tabouret-Keller 1985; Muysken & Smith 1986; Rickford 1987; Romaine 1988; Thomason & Kaufman 1991). One of the current challenges in the field of creole studies involves pro viding satisfactory accounts and explanations of the linguistic variability inher ent in creole situations. There are apparently conflicting interpretations of vari ability, namely, the "continuum" and the "co-existent systems" perspectives (to be further discussed below). The varieties spanning a creole continuum are often characterized as unidimensional and displaying implicational patterns (Bickerton 1975; Rickford 1987). However, analyses leading to these interpre tations are usually based on phonological or morphological variables, rarely on observations of linguistic features beyond the sentence level. A different view holds that only a multidimensional model can accommodate creole situations. In particular, LePage & Tabouret-Keller (1985:180-185) claim that the linguis tic strategies used in Cayo District, a Western district of Belize with predomi nantly Spanish and Creole ethnicity, are necessarily multidimensional because each linguistic choice is an "act of identity" that implies solidarity with or dis tance from others. This perspective is in keeping with the speech accommoda tion theory, which involves the principles of "convergence" and "divergence" in human psychosocial interaction (Giles, Bourhis, & Taylor 1977). The multidimensional interpretation of linguistic variability need not be restricted to multilingual, multiethnic communities, as will be discussed later. The issue of decreolization is often linked to the continuum phenomenon. Decreolization has been interpreted as a development of the continuum in the direction of the standard model, leading from basilect to mesolect to acrolect, with concomitant loss of the "earlier" forms. The process has often been assumed to be a unidirectional substitution of the standard for the creole.7 Having embraced the decreolization view at early stages of my research (Escure 1981), I have come to the conclusion that the reality of creole continua, even in a relatively homogeneous speech community like Placencia, Belize, (see chap ter two) is more complex than can be accounted for by a unidimensional model. The learning of acrolects does not necessarily imply the loss of native basilectal vernaculars. On the contrary, the acquisition of acrolects involves an exten sion of the repertoire available to speakers rather than a systematic sliding of the continuum toward a more standard set of linguistic segments (Escure 1981;

20


1982; Rickford 1987:31). There is a reasonable possibility that the process of change through extension and remodeling may be typical of any acquisitional situation, implying the addition of a second code to a first variety. This is like ly to apply to SLA as well. I firmly believe that the acquisition of a second lan guage does not systematically and irreversibly move forward but is subject to backsliding and the preservation of socially meaningful intermediate forms, not unlike the acquisition of varieties along a creole continuum.8 If this is true, SDA may not be drastically different from SLA, and creole situations may be struc turally similar to other language variability situations, although they may still differ in other respects, perhaps historical or political. For Muysken & Smith (1995:4-5), "creole languages are not in the slightest qualitatively distinguish able from other spoken languages." A comparison of second dialect acquisition in creole and noncreole con texts will help clarify the definition of the concept of "creole." In particular, the following problems will be addressed: are creole languages and their continua structurally different from others? Is the pattern of standard acquisition differ ent for a native creole speaker as opposed to a native speaker of a noncreole, nonstandard dialect? Those questions will be examined in the very limited con text of the discourse variables selected for analysis. There is a strong indication that, even though the term creole is widely used in various types of contact-induced situations, it is not clear that a creole is typologically different from other language types which develop through some type of sociolinguistic contact (Corne 1995a; Escure 1993a). It will be hypothesized here that second dialect acquisition is somewhat analogous, though occurring under apparently less traumatic conditions, to the situations which permitted the genesis of pidgins and creoles. Even though a model exists (the standard is usu ally codified in written form and is also subject to informal standardization), it is never systematically related to the learner's native system; thus, the learner is forced into the same decision making situations—what to select from a variety of options. Furthermore, the same social pressures exist: both the incipient pid gin speaker and the nonstandard learner are in positions to be affected by vary ing degrees of social oppression and made to feel that they cannot meaningfully participate in social life unless they control, among other things, the "proper" code—which nobody is attempting to teach them properly or at all! In the case of nonstandard learners, language universals may have already been actively used in the natural development of their uncodified vernacular. 4.1 Two case studies It is hoped that a detailed analysis of selected aspects of two ethnically differ ent types of second dialect acquisition will contribute to our understanding of


21

cross-linguistic acquisitional mechanisms and of linguistic bias. First, the case of Belize, Central America: in that former British colony, the official language is English, directly inherited from the colonial past of what was previously identified as "British Honduras." English is the medium claimed to be used in all governmental and educational functions, but the reality is far from meeting official expectations. The Creoles (Afro-Europeans) overtly rec ognize the prestige of English, but their vernacular language is an Englishbased creole which has in-group prestige. Other groups manifest equally ambiguous attitudes and behaviors toward the standard acrolect. The acquisi tion of Belizean acrolects occurs in the context of a multilectal creole situation, which is likely to have always been variable, and definitely still is. It is a case of relatively recent language formation through pidginization and creolization. Second, the case of the acquisition of Standard Mandarin by speakers of other varieties of Chinese having local, but not official, status occurs in the more traditional context (noncreole) of dialect acquisition, so far undocumented. The People's Republic of China (PRC) has been actively promoting—through mass media and educational channels—the increased use of the new standard of Mandarin Chinese called Putonghua (the "Common Language"). Beijing natives claim that they cannot understand the provincial dialects spoken all over China, but most of the recent leading political figures of China are not natives of the Beijing province and sport remarkable accents: Mao Tse Tung was born in Changsha, Hunan, in the southern part of PRC, where a distinctive lan guage—Xiang—is spoken; Deng-Xiao-Ping is a native of Chengdu, Sichuan, whose Southwestern Mandarin variety is also very distinctive. Obviously, Belizean English-based creoles and Chinese dialects are genet ically and historically unrelated, and the countries in which those varieties are spoken are maximally different in size and ethnic composition. This apparent lack of connection is one reason why I elected to analyze the linguistic situa tions in those two geographically distant sites. Comparing the two types of acquisition should highlight universal aspects of acquisition, as well as the issue of whether there is any substance to the notion that creoles develop in any dis tinctive way (Escure 1993a; 1994). In addition, it is recognized that creoles constitute the linguistic product of relatively recent colonization and enforced transplantation and subjugation of one group to another region and culture, whereas a general assumption of homogeneity is typically associated with Chinese cultures and languages—a manifestly erroneous assumption. Both situations offer similarities in terms of the general social background underlying their respective developments. Chinese populations have been forever subjected to the constant oppression of various emperors and overlords invading each others' territories, and there has

22


been a plethora of successive waves of colonization from the north to the south of China. The most recent invasion was the Manchu takeover which resulted in the domination of one nation by an outsider group. Linguistically, the dominant Manchu are said to have adopted the language of the subjugated nation and, more specifically, the Mandarin variety spoken in Northern China (Beijing vari ety). The Manchu invasion occurred at about the same time as the main slave trade from Africa, which changed the ethnolinguistic map of the New World. It is obvious that, historically, similar types of linguistic contact and ethnic mixing phenomena underlie China and Belize. 4.2 Outline of chapters It is expected that close observations of the standard versions produced by native speakers of nonstandard dialects will provide interesting insights into the universal processes underlying the development of language in society. This volume investigates the development of second dialects in the context of the two distinct sociolinguistic situations outlined above: The first part of this volume, chapters two through four, examines linguis tic variation in the context of English-based creoles and the English acrolectal varieties acquired by native speakers of creole basilects, as spoken in Belize (previously British Honduras) where I have conducted fieldwork from 1978 to 1994. Chapter two provides the demographic, geographical, social, and histori cal background of the speech community of Belize, highlighting some of the conditions under which the creole continuum is currently developing English acrolects and focusing more particularly on the Stann Creek District which is the locus of this research. Chapter three is concerned with the acrolectal seg ment, the product of the acquisition of the standard as second (non-native) dialect. The intent is to identify potential differences between the acrolect and the standard variety, the apparent target in acrolectal development. Chapter four specifically investigates the use of topic strategies in Belizean acrolects, com paring syntactic and pragmatic aspects of standard acquisition to basilectal cre le patterns and casual American English varieties. The second part of this volume, chapters five through eight, investigates similar problems in Chinese dialects spoken in the People's Republic of China, including the native vernacular Mandarin/Putonghua spoken in Beijing, the non-native Putonghua acquired by speakers of the Wuhan varieties (Wuhanhua) in Wuhan (Hubei), and the narrative variety of Wu used for traditional tales in Suzhou (Jiangsu) which is also acquired as a second dialect. Speech data were collected since 1985 (Escure 1987), and vernacular data spanning the period from 1963-1985 are derived from contemporary written texts. As a counterpart to chapter two in the Belizean section, chapter five provides parallel background


23

information on the Chinese dialect situation with special emphasis on Beijing, Wuhan, and Suzhou. Chapter six outlines crucial aspects of Chinese syntax in discourse. Chapter seven looks into the use of various types of topic strategies in several varieties representing Standard Chinese/Putonghua. Chapter eight evaluates similar mechanisms in the context of two acquired dialects and leads to a comparison of first and second dialects in some representative Chinese var ieties. Finally, chapter nine establishes a comparison of the patterns observed in Belizean and Chinese standard acquisition and, more particularly, discusses the interaction of syntax and pragmatics, and the issue of the universality of dialect acquisition. From the above remarks, it follows that, should any structural sim ilarities be observed in the linguistic patterns adopted by learners of second dialects in Belize and in China, language-specific substrate and superstate mechanisms could be ruled out. Such potential findings would warrant careful scrutiny, with the ultimate goal of getting closer to an understanding of univer sal linguistic processes. Endnotes 1. Except in relatively restricted cases of simultaneous bilingual acquisition which does not involve diglossia-some sort of social discrepancy between the two languages. In fact, few, if any, such situations exist. 2. For example, Black English in the United States has become extremely popular among some white teenagers; however, I know of no full-scale study documenting the extent of its acquisi tion by groups other than African-Americans. 3. It has been demonstrated that there is no correlation between multilingualism and low eco nomic development or political instability, as discussed in Fasold (1984:5-9). 4. Thanks to Stephen Matthews and Virginia Yip for the information about Canto-rap. 5. In this excerpt, any representation of the phonological idiosyncracies of Belizean acrolects has been omitted since the point of this example is to discuss the challenge of defining discourse units. The only lexical unit that may be different from English is pan 'on' [look pan the land, pan the people]. 6. Linguistic features characteristic of basilects, mesolects, and acrolects will be discussed in later chapters. 7. A parallel interpretation of the acquisition of second languages assumes a unilateral shift from basilang to mesolang to acrolang (Stauble 1980:47; Schumann & Stauble 1983:263). 8. Although I don't know of any study testing in detail the hypothesis of the nonuniformity of L2 learning, all adult speakers of a second language experience the unreliability of L2 perfor mance which varies in terms of multiple factors in the context of interaction.

Chapter 2 The Belizean Speech Community and the Use of English Abstract This chapter provides general demographic and socioeconomic information on the Belizean community, with special focus on Belize's ethnic and linguistic pluralism. The language situation is briefly placed within the specific context of the Englishbased creole continuum. The locus of the research is Stann Creek district and, in par ticular, the Creole village of Placencia and the neighboring Garifuma community of Seine Bight. The linguistic range, from basilect to acrolect, is illustrated in sample texts, with a brief outline of idiosyncratic features defining each lect. 1. Introduction to the history of Belize Aldous Huxley (1934:32) said: "British Honduras is not on the way from any where to anywhere else, has no strategic value, is all but inhabited; if the world has any ends, British Honduras would be one of them." With due respect to Huxley, nothing could be further from the truth! Far from being the dead end implied by Huxley's quote, Belize is a vibrant land of diversity and paradox, with city slums and unspoiled beaches, lush jungles and deforested land turned into sugar plantations only a few miles apart from each other. In multicultural Belize, ethnic diversity and miscegenation are taken for granted, yet Belizean society has inherited the color-coding system of colonization. Literacy and edu cation are high priorities but educated Belizeans are offered few opportunities at home. Thus, many of them have emigrated to the United States which has a higher Belizean population than Belize itself. Thc young average age of the population confirms the continuing scope of emigration patterns. According to the 1991 population census, 75% of the population is under twenty-one, about the same as in 1980. Belize, which originated in a lively brew of cultures, lan guages, and dialects is still developing in multiple dynamic directions. Steeped in a violent past of invasions, clashes, migrations, exploitation, piracy and slav-

26


ery, Belizean society offers fascinating insights into patterns of human adapta tion in the face of adversity and into the creativity of linguistic processes in complex communicative situations. The area was the scene of conflicts which almost eliminated Maya civilization, of clashes between Spanish soldiers and British buccaneers, of pirate landings, slave logging camps, Indian caste wars, Black Carib deportation, and finally Salvadoran refugees' immigration. In order to understand the range and overlap of the linguistic codes used in Belize, their social functions and iconic values, and eventually the nature of the second dialects spoken in the context of the traumatic history of Belize, it is essential to delve briefly into some historical aspects of the country. The name Belize was first applied to the main waterway and later to the principal city which became the capital of the independent territory of British Honduras. The origin of the lexical item Belize is somewhat unclear. It may be derived from the name of the assumed founder of the original settlement which eventually became Belize—a Scottish pirate named Peter Wallace, or Wallis, who was possibly a lieutenant of Sir Walter Raleigh.1 As reported in the 1827 Honduras Almanack,and as Bancroft (1883) and Asturias (1941) claim, the first settlement of British Honduras was founded either in 1617 or in 1638 by some eighty shipwrecked sailors led by Wallace/Wallis. The area settled would then have been named "Wallis" or "Ballis" by the Spaniards, due to some interfer ence of the Spanish bilabial fricative.2 The former British colony of British Honduras, self-governed since 1973, became independent on September 21, 1981, after a long period of negotiations with Guatemala which had long-standing claims to what is still referred to on Guatemalan maps as the Belice province. This 13,000 square kilometer territo ry (260 kms at its longest and 100 kms at its widest) is wedged between Spanish-speaking Central America (Mexico, Guatemala, and Honduras) and the Caribbean Sea. It is bounded on the north by Mexico's Yucatan state of Quintana Roo and on the south and west by Guatemala. Belize has the lowest population density of any Central American country, with an overall population of 189,392 according to the most recent (1991) pop ulation census. More than a quarter of the population is clustered in the capital Belize-City (population 43,710). Because of limited economic opportunities, Belize has been traditionally a land of emigration, and although it is difficult to have accurate figures, it can be assumed that there are at least as many Belizeans living in the United States as there are in Belize. The 1991 census finds that, overall, 3,000 Belizeans emigrated since 1980—over 85% of them to the United States—with a steady flow over the years (although no such data are available to document emigration before 1980). This figure is almost certainly an underrepresentation of the facts of emigration for two primary reasons stated in the

THE BELIZEAN SPEECH COMMUNITY

27

population census report: first, entire households who emigrated are not repre sented, and second, Belizeans residing in Belize may have been reluctant to report family members living abroad who have not yet regularized their resi dence status in their land of emigration. Considering those factors, I believe that a reasonable estimate of emigration since 1980 may be closer to 5,000 and, over thirty years may have reached as much as 150,000, and thus may have matched the current population officially residing in Belize. Such high emigration patterns appear to be confirmed by the relative youth of the Belizean population: 65% is under age twenty-four whereas the most pro ductive segment of the population (ages twenty-five to fifty-four) amounts to only 28%. Individuals over fifty-four constitute only 8% of the population. The 1980 census showed a similar trend in age distribution for the ten previous years. This overall pattern suggests that most of the breadwinners live abroad. Emigration is obviously an important factor likely to influence the lan guage situation, either because this young population is likely to have a power ful impact on language choice or because connections with United States resi dents may affect language development in Belize. Considering the pervading emigration, the population could be expected to show a slow increase or even a decrease. In fact, between 1980 and 1991, the population increased by 44,000, an increase of almost 30% and almost twice as much in real numbers as in the previous intercensal period (1970-1980).3 In comparison, the intercensal growth was less between 1970 and 1980 than in the previous ten years as shown in table 2.1: Table 2.1: Population growth in Belize (Population Census, 1991) Year

Population

1960

90,565

1970

119,645

1980

145,353

1991

189,392

Growth during decade 29,000 25,000 44,000

Why this apparent growth? In spite of regular emigration, increased childbirth is not the explanation. It appears that the traditional, or at least thirty-yearold, emigration channels to the United States have been offset by a strong incoming immigration trend. Over 25,000 people living in Belize in the last ten years reported being born abroad, which neatly accounts for the doubled inter censal growth noted above. Most of those immigrants are from Guatemala and El Salvador (both neighboring countries which suffered high political instabil-


28

ity) and, to a lesser extent, Mexico and Honduras. Those four countries, all Spanish-speaking, constitute 85% of the native lands of recent immigrants. This fact is an important factor related to language use. Spanish, one of the languages of Belize, has now gained majority status over English-based varieties. The eth nic composition of the country is also affected. 2. Ethnolinguistic composition and census data The diverse Belizean population results from repeated waves of immigrants usual ly forced upon the land. Settling at different times but adding extraordinary diversi ty to the indigenous Amerindian population have been a range of peoples: shipwrecked British seamen, French and British pirates, African maroons and slaves, Miskito Indians brought from the Miskito Coast (now Nicaragua), Black Caribs (Garifuna) deported from St. Vincent, then Honduras, Spanish and Indian refugees from Mexico, Mennonites fleeing religious persecution, and more recently Salvadorans and Guatemalans fleeing political oppression. The contemporary pop ulation includes four major groups: Amerindians, Creoles, Mestizos, and Black Caribs (Garifuna), and other smaller groups. Escure (1983a:32) attempted to inter pret the 1980 census data (Population Census 1980); the distribution of ethnic groups and the languages spoken is represented in table 2.2: Table 2.2: Ethnic groups, population, and languages of Belize in 1980 ETHNIC GROUP

Creoles Mestizos Garifuna Maya Kekchi Whites East Indians Chinese Other: Lebanese Mennonites Salvadorans

%

SECOND LG.

THIRD LG.

39.7 33.1 7.6 6.8 2.7 4.2 2.1 0.1

Creole Spanish Garifuna Maya Mopan Kekchi English English Creole Cantonese

English Creole Creole English Creole** Creole English Creole

Spanish* English English Creole**

3.6

Creole Low German Spanish

English Creole**

POP.

FIRST LG.

Spanish* English

* in some areas only **rarely

Although each Belizean is quite clear about his or her ethnic and cultural identity, the official classification of ethnic groups offers particular challenges


29

in Belize. The 1980 census used nine categories, including the labels Negro/Black and Mixed, which drew most votes with 76% and 13% respective ly. There was no reference, however, to the locally used terms of identification: Creole, Black Carib or Garifuna, and Mestizo.4 This classification overlooked the peculiar ethnic configuration of Belize and local attitudes toward identity. At least three of the major groups are mixed (Creoles, Black Caribs, and Mestizos), and the first two can be considered "black." Specifically, the Black Caribs, who are now officially referred to as Garifuna (see below for a more extensive discussion of this term and other alternatives) are indeed mixed (Afro-Indians), although they usu ally have dark skin. Creoles are also by definition mixed (Afro-European), but they range over a broad spectrum of coloring and morphological features. Yet, Creoles and Garifuna are culturally and linguistically very different groups. As to the Mestizos, they are also mixed (Hispanic/Indian), although many members of the group identify themselves as "Spanish," and there was no label indicating a Spanish identity in the 1980 census. Furthermore, no distinction was established between the Maya and the Kekchi, two distinct groups of Amerindians who speak different, though related, varieties of Mayan. The 1991 census is a significant improvement over the previous ethnic census because it used labels popularly used in Belize, twelve altogether as optional responses to the question: "To what ethnic, racial, or national group do you think you belong?" The overwhelming majority fits into the Mestizo (43.6%) and the Creole (29.8%) groups. The foreign-born (14% of the population) account for the shift in the majority groups. This figure, however, does not clearly distinguish between recent arrivals and traditional Belizean Mestizos. It might have been help ful to include national origin such as Guatemalan, or Salvadoran, or the use of the label Spanish/Spaniard/Hispanic besides Mestizo, which would have shed light on the ethnic shift apparently ongoing in Belize. Table 2.3: Ethnic Groups in Belize in 1991 ETHNIC GROUP

Creole Indian Garifuna Maya Mopan Kekchi Maya Other Maya German/Dutch Mennonite Mestizo Chinese Syrian/Lebanese White Other

POPULATION (%)

29.8 3.5 6.6 3.7 4.3 3.1 3.1 43.6 0.4 0.1 0.8 1.0


30

The most serious inadequacy of the 1991 census lies in its investigation of languages spoken and the discussion of the results. The census dealt with the language situation through two simple questions (Do you speak English?/Do you speak Spanish?) and three answer options in each case (very well/not so well/not at all). One might argue that subjective responses to fluency are generally unreli able because speakers tend to either overrepresent or underrepresent their lan guage skills, depending on the local or national status of those languages. This is a serious issue in Belize where English has the status of "official language" and serves as the language of education. However, the most serious problem may be that no attention is given to the use of the creole language. Furthermore, there is no attempt to assess the language acquisition and shifting strategies of the population. Indeed, the glaring omission of any reference to Creole confirms the continued presence of bias toward native vernaculars. Such an omission can only produce an inaccurate interpretation of the census data. Census figures indicate that 54.3% of the population believe that they speak English "very well," 22.5% "not so well," and 23.2% "not at all," sug gesting: "that just over half of our population speak the official language very well points to the need for some serious considerations, especially in view of the fact that fluency in an established language is an indispensable tool for the acquisition of any other form of knowledge" (Population Census 1991:8). On the other hand, the stated frequency of good Spanish fluency corre sponds to the percentage of Belizeans who, in 1991, identify themselves as Mestizo (43.6%), as shown in table 2.4: Table 2.4: Stated language fluency (1991 Population Census)

Speak English Speak Spanish

Very well 54% 43.8%

Not so well 22.5% 11.1%

Not at all 23.2% 45.1%

This table reflects a clear and complementary assessment of Spanish flu ency as "very well" and "not at all." The growing impact of Spanish from var ious proveniences suggests that other languages, such as the Mayan varieties spoken in the western and southern districts, may be affected by the develop ment of Spanish.5 In contrast to clear assessments of Spanish fluency, table 2.4 reflects the uncertainty related to what constitutes English fluency. The three major groups (Creole, East-Indian, Garifuna), who natively speak some form of English (assuming the creole is some form of English) constitute about 68% of the overall population, yet only 54% estimate that they speak English very well.


31

The uncertain piace and status of Creole vis-à-vis English is obviously part of the problem, an issue not addressed by the census. The issue of bilingualism or multilingualism is also not addressed. This official dismissal of Creole points to the necessity of including in census reports findings about the extent of sec ond language acquisition. Finally, it would have been interesting, if not crucial, to investigate which portions of the population estimate that they are fluent in Garifuna and in Mayan/Kekchi. Such figures may have helped assess the vital ity of the languages spoken by these two Belizean minority groups. The differential in language use and language fluency in English and Spanish is quite important and is obviously linked to the differential distribu tion of ethnic groups across the six districts of Belize, as represented in table 2.5. Creoles predominate in the Belize district (around the urban area of BelizeCity)6 and are also widespread in Cayo (west, near Guatemala) and Stann Creek (southern coastal area). Mestizos constitute the majority in Corozal and Orange Walk (both north near Mexico) and Cayo. The Garifuna only have a significant presence in Stann Creek. Finally, the southernmost Toledo district is clearly Maya territory, at least in rural areas, since the main town of Punta Gorda is in majority inhabited by Garifuna. Table 2.5: Ethnic groups in the six districts of Belize (1991) (arranged from northern to southern geographical locations; figures indicate percentages)

Districts COROZAL ORANGE WALK BELIZE CAYO STANN CREEK TOLEDO

Creole

Garifuna

Mestizo

Maya

7.6 7.4 67.9 23.0 25.1 5.7

1.3 1.2 5.3 1.7 36.2 10.0

74.1 71.7 18.7 58.0 23.7 11.9

5.0 9.1 1.2 8.7 8.0 62.8

A brief overview of Belizean ethnicity is outlined below, group by group, and in terms of their chronological presence or arrival on the scene of the territory which now constitutes Belize. 2.1 Amerindians The most ancient ethnic group in Belize is the Amerindian group. Maya civiliza tion flourished primarily in the Yucatan peninsula, extending into Belize and neighboring areas of Honduras and Guatemala, but abruptly declined in the ninth century A.D., possibly because of a violent popular revolt against the religious rulers. There is, however, archaeological evidence of continued occupation of the

32


Maya sites (presumably by farmers) well into the fourteenth century and proba bly much later, up to the time when the British began settling in Belize in the early seventeenth century. Ruins of cities and pyramids are scattered and half-buried under the tropical jungles, a silent testimony to the existence of a lost civilization forgotten even by its descendants. Some of the sites that have been excavated or surveyed include Altun Ha, San Jose, La Milpa, Mun Diego and others in northern Belize; Xunantunich, Actun Balam, Caracol, Cahal Pech and Tzimin Kax in the West (Cayo District near the Guatemalan border); and Lubantum and Pusilha in the southern Toledo District. The Mayas in Belize now consist of three groups, each speaking a differ ent language variety and living in separate areas. The Mopan and Kekchi Mayas are the descendants of the original inhabitants of Belize who reentered the area from Guatemala during the end of the nineteenth century; they live pri marily in the Toledo district, in southern and western Belize near the Guatemalan border. On the other hand, Yucatecan Mayas came from Mexico after the 1847-1853 Indian Caste War, and they now live in the northern Corozal and Orange Walk districts, close to the Mexican border. Most of the Mopan and Kekchi-speaking Mayas still live in isolated villages and, until recently, have had little contact with other groups. There is an increasing admixture of the Mopan and Kekchi, but Mopan speakers have a somewhat higher status locally, perhaps due to the fact that the largest village of San Antonio (pop. about 200) is primarily Mopan, with some assimilated Kekchi. The southern Mayas consider themselves to be different from the northern Yucatec Mayas who primarily speak Spanish and have been largely assimilated to Mestizos. Yucatec Mayan is still spoken in some rural areas, though Spanish is now becoming the first language as it is associated with progress and social mobility for members of this group. Northern Mayas and Mestizos, thus, close ly identify with Mexico, and Spanish is the primary marker of ethnicity for them. Creole, however, competes with Spanish as an indicator of youth soli darity and interethnic friendship. Even though Spanish-speaking people demographically dominate the northern Belizean community, the size of the Creole group has increased in the city of Corozal, and the limited Creole/Mestizo inter action allows for some Creole expansion in the community. In the western Cayo District, the Mestizo and Creole populations mostly balance out; thus, Spanish and Creole have roughly equal status in the commu nity at large. Interaction between the two groups is frequent, and intermarriage is more common than in the northern part of the country. Consequently, chil dren in Cayo have a choice of identities, whether Spanish or Creole, as attest ed in Le Page and Tabouret-Keller (1985).


33

2.2 Creoles (Europeans, Africans, Miskitos) The Spanish colonization of Central America in the sixteenth century was con centrated primarily in the Gulf of Campeche and barely affected the area of the Bay of Honduras; however, French, Dutch, and British buccaneers had occa sional contacts with the Belizean coastline. According to sparse records pre sented by Asturias (1941), Bancroft (1883), and Burdon (1935), British settlers had brought Jamaican lumbermen as early as 1655 for the purpose of logging the swamps along the Belize River. As buccaneers increasingly converted to the lumber business after privateering was outlawed, a larger labor force was need ed and African slaves were imported to the area from other parts of the Caribbean. In a 1724 report, a Spanish missionary referred to "300 English [liv ing in Belize], besides Mosquito Indians and negro slaves, these latter having been introduced but a short time before from Jamaica and Bermuda" (Bancroft 1883, vol.7:626). The Mosquito (now called Miskito) Indians mentioned in the above statement came from the British settlement located on the Pacific Coast area of present-day Nicaragua. The Miskitos had become the allies of the British against the Spaniards and, in exchange, had acquired certain privileges, for example the right to crown their kings in the Cathedral of Belize. British lumbermen moved freely between the two settlements until 1787, when the British had to evacuate the Miskito Coast. At that time, some 2,000 British, their African slaves, and Miskito allies moved to the Bay of Honduras to join the "Old Baymen" as the British Bay settlers called themselves (Escure 1983a:30). Miskito Indians merged with the rest of the population (they have also become Afro-Indians in Nicaragua), although many Belizeans still refer to them by the term Waika (generally as a derogatory term). On September 10, 1798, the Spaniards were defeated by the Old Baymen at the Battle of St. George's Cay, which is commemorated as Belize's National Day. Yet, it was not until 1862 that British Honduras became a colony; it became a Crown colony in 1871. Slavery was abolished in 1838, but freedmen were unable to acquire land because land ownership was monopolized by a very small num ber of absentee corporate landlords. The forests had been practically exhausted in northern Belize, and the country's economy could not be adequately converted to agriculture. Today, agriculture-related occupations involve only 5% of the popu lation now, with 42.4% stating that they have never worked. As indicated in tables 2.2, 2.3, and 2.5, the Creole population, which was a long-time majority group, has now slipped to second place after the Spanish-speaking Mestizo group, great ly expanded by recent Central American immigration.

34


2.3 Garifuna (Black Caribs) The ethnolinguistic composition of the Belizean population was further modi fied in the first part of the nineteenth century. The first significant event was the arrival, starting in 1803, of the "Black Caribs," a group of Afro-Indians whose native name is Garinagu, though they are now officially called Garifuna (the Arawakan term for their language). However, the anglicized terms Carib for the language, and Carib or Black Carib for the people, are still used freely and without negative connotations by the Garifuna-speaking Garinagu.7 The Garifuna group arose from the miscegenation of Arawak and Island Carib Indians with some African maroons who had been shipwrecked on the island of St. Vincent in 1635. After a failed uprising, supportėd by the French, the Garifuna were deported by the British in 1797 to the island of Roatan off the coast of Honduras. Most of them then moved to the coastal areas of Honduras (near La Ceiba and Tela), south to Nicaragua (Bluefields area), or north to Guatemala (Livingston) and to the Toledo and Stann Creek districts of Belize. The largest Garifuna town in Belize is Dangriga, the population of which was 6,435 in 1991 out of 18,085 for the entire Stann Creek district. Dangriga (previously called Stann Creek, like the district) was founded in 1823 by a group who had sided with the losing royalist cause against the republican gov ernment of Honduras. In 1939, other Garifuna fled the Honduran village of San Juan, again for political reasons, and founded the village of Hopkins, just south of Dangriga in Belize. There are now six Garifuna communities in Belize: Dangriga, Punta Gorda, Barranco, Seine Bight,8 Hopkins, and Georgetown— the only inland village, founded by Hopkins inhabitants after the destructive hurricane Hattie (Hopkins now has a sparse population). African phenotypical characteristics predominate in the Garifuna, who are physically indistinguish able from the darkest Creoles. Their African heritage is mostly represented in the folklore, whereas the Amerindian component has been preserved primarily in their native language, Garifuna, which is mostly Arawakan in structure (Taylor 1977), and will be briefly outlined at the end of this chapter. 2.4 Mestizos Another significant event leading to greater ethnolinguistic diversity was the Indian Caste War (1847-1853) in Mexico which led to a heavy migration of Yucatecan Mayas and Mestizos (Spanish/Indians) from Yucatan into the north ern (Corozal and Orange Walk) and western (Cayo) districts of Belize. Mexican Mestizos have continuously trickled in, many settling as fishermen on some of the islands off the northern coast of Belize in Ambergris Caye and Caye Caulker, for example.


35

The most recent wave of Mestizo/Hispanic immigrants consists of Salvadoran and Guatemalan refugees who started moving to Belize in the 1980s in order to escape their country's civil wars and political persecution. They have been greeted by the Belizean government as a welcome substitute for a consis tently depleted Belizean population due to massive economic emigration to the United States. They especially fulfill a dire need for agricultural workers in the banana and citrus plantations of the Stann Creek District, where they have settled in hastily built shantytowns. If not yet into a torrent, the Pania9 trickle has at least turned into a regular ly flowing stream, which is currently changing the ethnic and linguistic composi tion of Belize, turning at least half of the country into a Spanish-speaking nation. The other half—part of which I will be investigating—is still predominantly the domain of English and Creole. 2.5 Other groups Other ethnic groups arrived at different times to fill the need for indentured labor. The Chinese came from the southern Guangdong province of China in the 1860s to find labor; now, most own small businesses in Belize-City (dry goods stores or restaurants). They have preserved close ties with their ancestral coun try, especially with Hong Kong and the Guangdong province of the Peoples' Republic of China and, thus, speak Cantonese even now. Many still bring their wives from China through the age-old tradition of prearranged marriage.10 East Indians came shortly thereafter, in the 1880s, though not by choice. They were Sepoy soldiers, deported to Belize by the British after an aborted revolt against British rule in 1857. The British showed consistency in sending unwanted rebels twice to the backwaters of Belize (first, the Black Caribs, then the Sepoy), an attitude toward the forlorn colony obviously reflected in Huxley's quote which introduced this chapter. The East Indians mostly eked out a meagre living on the southern sugar plantations of the Toledo District which had been offered as land grants by the Belizean government to a small number of Americans planters fleeing the United States after the Civil War. The East Indians are now largely assimilated to the Creoles, both linguistically and eth nically, and are concentrated near the city of Punta Gorda. American plantations generally have been abandoned. Some merchants of Middle-Eastern descent, mostly Lebanese (usually referred to as "Syrians") have settled in Belize, as elsewhere in the Caribbean, at various times during the nineteenth and twentieth centuries and have started small businesses which compete with the Chinese. A more recent wave of immigrants arrived in the 1950s, as German Mennonites from Canada and the United States established farming communities in the western part of Belize.

36


They rarely participate in the life of the country, restricting their interactions to market activities and to small businesses such as carpentry; many move on to other communities (in Central or South America) after a few years. As is the case elsewhere, they have preserved their Low German dialect. 3. The language situation Belize is certainly a multiethnic nation and can also be identified, overall, as a multilingual nation; yet, individual Belizeans are at most bilingual and usually monolingual, with superficial, if any, knowledge of some other language, as indicated by the census surveys reported above. Language use and the reasons for selection and code-switching vary from one region to the next and depend on the ethnic composition, the degree of interaction with other groups, the amount of education, and the relative status of each individual. As is usually the case in multiethnic societies, ethnocentricity and ethnic stereotypes exist and reflect the relative socioeconomic status of co-existing groups. Language choice is an accurate indicator of those attitudes. Since Creole is not officially recognized as distinct from English and is still overtly assigned a low value, one may wonder why it has so vigorously survived and is gaining prestige and a wider speakership. This vitality may be partly due to a "double standard" that has always prevailed in Creole societies. Even though the vernacular may be overtly denigrated by its very speakers and avoided for certain outgroup con tacts (see 4.2.3 below), it remains covertly—and sometimes overtly—presti gious and highly valued as a marker of ethnic identity and increasingly as a marker of Belizean national identity. Such a situation entails an almost schizo phrenic mechanism that requires subtle code-shifting based on a precise assess ment of the social situation (Hellinger 1972). 3.1 Language policy and education The official language of Belize is English. English may be unanimously recog nized as the language that one must acquire to participate in official government activities, but it is not commonly used in its external (American or British) stan dard form. This fact is largely ignored or unidentified by both language users and language planners. Although there is a general, informal agreement as far as the creole end of the continuum is concerned, what constitutes the "correct" local norm called English—the acrolect—is realized differently by different people, and they are not necessarily aware of such variations. The official (and unoffi cial) position is that any "non-English" variety is brokop—broken English— which often includes the creole varieties as well as the mesolectal area of the


37

continuum. Such is the persistent legacy of colonialism. Thus, in spite of the fact that it is gaining speakers and prestige, Belizean Creole is still officially ignored. A typical list of the languages spoken in Belize, as seen on a sign posted at the entrance of Punta Gorda, includes: English, Spanish, Garifuna, Maya, and Kekchi, with the glaring omission of Creole. This is disturbing in view of the facts presented above, and in particular when considering that English is not the Belizeans' first language but a second language or dialect for practically all. The paradoxical coexistence of two varieties, one carrying overt prestige and the other covert prestige, is not an unusual phenomenon. Indeed, this situation occurs wherever a "dialect" competes with a "standard," an issue to be exten sively discussed in the following chapters. Official directives to instruct in the overtly prestigious model are naturally passed on to educators; yet, the overlap of the model and the vernacular predictably accounts for the linguistic complexity encountered by children in the classroom. Teachers are expected to teach English, and are believed to be teaching English, but they are often unsure of the exact nature of English as distinct from Creole; more commonly, they are unaware that there is any difference at all. Furthermore, they frequently control a limited segment of the acrolectal end of the continuum which, ironically, is the only segment that they are instructed to teach. Because of the scarcity of teachers, young teachers and teachers' aides are put to work imme diately after primary school, with little or no teacher training. They admit that they frequently switch to Creole in the classroom for more fluent explanations. Similar problems are involved in teaching basic subjects to non-Creole Belizean children. Because teachers are often recruited from among the Garinagu, the Maya-speaking children of a remote village are likely to be intro duced to reading and writing in English by a native speaker of Garifuna who does not know a word of Maya and who has modeled his or her English on the fluctuating varieties she or he has heard in school, on radio reports which pre sent BBC news and local news in Belizean standard, and on Billy Graham's homilies which reach the most distant villages. The resulting effect can only lead to the maintenance, or even the expansion, of the bewildering array of options which make up the creole continuum. For example, Juan, a sixteenyear-old Maya teacher from the Mopan village of San Antonio who had just started teaching in the Garinagu village of Seine Bight told me that he could not speak either Garifuna or Creole, although he could understand the latter, but felt comfortable with English: "Mos'ly, I fin' it difficult wid der language because dey only speak Carib" (3.2: 3). Dennis, a forty-five-year-old Garifuna teacher in the same village, made the following comments, which illustrate, first, the linguistic confusion that must result from the contradictory messages received by the children, and sec-

38


ond, the socio-psychological factors that determine language choice in the com munity at large: You start speaking English to dem [the children], then you give dem de definition in Creole which is much easier to dem because dats what dey hear around dem. Outside you find dat de majority of people, der parents speak Creole, dey speak Carib and Creole. When I first started teaching, while I spoke Creole intermittent, and finally I found out dat I spoke too much Creole, and den I put a stop to it because I had learned English. So I practice my English more. I practice it more even when I speak to my friends. So you would find out when I would stand in front of a class and I speak to dem, I try to speak as fluent as I can. When dey [my friends] speak Carib to me, I speak Carib. When dey speak Creole, I speak English to dem, dats de way I treat dem, right? (3.1: 3). But, later in the conversation, he says: Let's say I go among some friends and, like, we going to have a fine time. Okay, de boys start talking and all everybody just discussing Creole. So instead of trying to make myself feel, or try to show off myself dat I am a better man or I am different dan all o' dem, I associate myself wid dem. So I speak Creole, understan'? So nobody can just say he talk funny, he must be from a different land. Ya understan'? I play when you are in Rome you do like the Romans do. (3.1: 10). Of course, the cost of training teachers to achieve proficiency in English and to acquire a sound methodology for teaching English as a second language or dialect is no small problem, and Belize is still a long way from solving it. It seems, however, that a first significant step would be to recognize the separateness of English and Creole, their distinct linguistic and social values, and in par ticular to recognize the intrinsic value of Creole as a linguistic system in its own right. In the recent past, government policies have acknowledged the existence and cultural heritage of the Maya, Garinagu, and Mestizo communities, thus encouraging the use of Maya, Garifuna, and Spanish in school and in local pub lic events. The traditional November 19 celebration of the first Carib settlement in Belize has become another Belizean national holiday earning full coverage in the local mass media. In 1985, Radio Belize broadcast an average of three to four hours in Spanish out of a total seventeen hours; however, earlier attempts at using Creole on the mass media (short programs, storytelling on Radio

THEBELIZEAN SPEECH COMMUNITY

39

Belize, and brief articles in local press) have been discontinued. 3.2 English and the Creole continuum In spite of the official designation of English as the language of choice, English is not anyone's native language in Belize. Since the goal of this study is to exam ine how English is acquired by speakers of related dialects, in particular by native speakers of the English based creole, an essential prerequisite is to define the patterns which characterize Belizean Creole. Only then will it be possible to interpret potential differences between the target model and the version achieved by its learners. 3.2.1 Putative origins of Creole The variety which is overall the most widely spoken in Belize is, undoubtedly, Belizean Creole, the English-based creole which is the vernacular of the Creole (Afro-European) group. Like other Caribbean creoles, Belizean Creole proba bly developed from contacts between Africans and Europeans who emigrated to the American continent. It also ftmctions as the lingua franca for most Belizeans of various ethnic origins. The sociohistorical outline presented in pre vious sections suggests that the African slaves were imported via the Miskito Coast, Jamaica, or other Caribbean islands such as Barbados, which served as a stepping-stone for British colonization (Alleyne 1980:23). Once in Belize, those slaves either created ab ovo the language which was to become Belizean Creole, or they may have brought with them some form of English-based pid gin or creole. It is also possible that the pidgin developed through earlier con tacts between Africans and Europeans along the West African coast (Hancock 1980 assumes a Proto-Creole originating in the Upper Guinea Coast). However, it is practically impossible to ascertain with any degree of accuracy the under lying components of Belizean Creole, both because of the complex population movements which occurred in the area and because of the absence of historical linguistic and demographic documentation. Such paucity of data is largely due to the traditionally low status assigned to creoles as the vernacular languages of a long oppressed group. It is one of the tantalizing mysteries presented by cre le languages that, even though they are very recent languages, so little is known of their precise origins, and thus, practically any interpretation is specu lative.11 In this respect, Belizean Creole is no different from other creoles; in fact, its origins may be more difficult to trace because so many ethnic groups were involved in its development. Although the focus of this study is on the cur rent status of some selected segments of creole continua rather than on their ori gin, it is useful to briefly review current theories of creole genesis, as they illu-

40


minate possibilities in developmental processes (a comprehensive critical overview appears in Arenas et al. 1995:87-134). Proponents of the monogenesis of pidgin and creole formation hold that all creoles have a common origin. The earlier interpretation assumed predomi nantly European (superstrate) linguistic influences, dating back to fifteenth- or sixteenth-century European arrivals. According to this perspective, the ancestor of creoles may have been either a nautical lingua franca or a Portuguese-based version of the Medieval Mediterranean trade language (Sabir), later relexified. The nautical element may have been acquired by slaves in Africa or during their sea voyage between Africa and America, or it may even have been borrowed through contacts with white colonists after their arrival in the Caribbean (den Besten, Muysken & Smith 1995:92-93). Some nautical lexical items are evident in most Belizean lects. Terms now obsolete in Standard English are in common use; in particular, frequent function words are likely to have originated in Scottish or other Northern English dialects, such as yonder 'over there,' pan ; deictic introducers follow closely (21.3%), but verbal prepositions which are highly favored in the short stories examined have a restricted distribution in Wuhan speech (table 8.). In conclusion, a comparison of Wuhan and Beijing PH varieties reveals more similarities than differences in topicalization strate-

242


gies. Differences occur, but primarily in the detail of lexical idiosyncrasies, and category distribution. 1) Educated speakers of non-native PH use more topic particles overall than a similar group of Beijing speakers. In Wuhan, the three women used an average of more than two particles (2.4) in each utterance but Beijing speakers of a similar education level used less than two on average. 2) In terms of specific particle selection, Wuhan speakers prefer to use adverbial particles. The AP category is over-represented in the Wuhan version: 52.7% of all particles are pre-comment adverbials, as opposed to 30.6% for the officials, and 38.2% overall in native Putonghua. One single particle jiu is espe cially common in non-native PH. 3) As far as TIs are concerned, only a moderate amount of deictics (21.3%) is produced in Wuhan (compared to 44.1%, for native PH), but the use of the clefting particle shi prevails. EPs constitute the only group showing the infre quent distribution previously observed in native Putonghua. As far as topic mechanisms are concerned, non-native varieties match most closely in appearance the speech produced by the characters in the two vernacular stories. This is not overly surprising, considering certain facts of Chinese history. Those stories were generally commissioned by Maoist directives issued in the 1960s, and the intent was to glorify peasant working classes, and extol a work ethic that would boost production. It is thus natural to find commune leaders producing the official PH with non-native accents. On the other hand, the three Wuhan women grew up during the period of the Cultural Revolution, and were exten sively exposed to the overwhelming propaganda of the time. Furthermore, they all spent several years in the countryside for "re-education," and became further attuned to peasant language. They are very much aware of the importance of the popular standard and of the status it holds, even though they are not themselves officials or "cadres"—the highest professional level to be attained in China.8 But even the egalitarian ideology (that led to the ubiquitous use of tongzhi 'comrade') would seem to encourage a leveling of all language varieties as is indeed the intent behind the term Common Language. This could at least partially explain the stan dardization of informal topic marking strategies. A cultural explanation partly jus tifies the unified pattern of topic marking strategies found in all varieties of Beijing Chinese and Wuhan PH, offsetting the substratai effect of native dialects in acquired dialects. On the other hand, the possibility of universal pragmatic strate gies in presenting topics cannot be discounted, especially when considering that there is independent evidence (suprasegmental, phonological, lexical) of first dialect interference. Observations of another second dialect variety will help in this investigation of the universal nature of topic strategies.

TOPIC STRATEGIES IN VARIETIES OF PUTONGHUA

243

4. Wu Chinese The second variety of non standard Chinese to be examined is a type of Suzhou Wu produced in traditional story telling. It was selected for several reasons. First, it is extremely popular in the province of Jiangsu, and the story-telling genre uses speech forms believed to be representative of earlier forms of Wu; second, it is a vernacular, informal speech form; third, it is not directly related, linguistically, to northeastern Mandarin (Beijing Putonghua), nor to southwestern Mandarin (Wuhan); and finally, it is an acquired dialect. 4.1 The Suzhou oral narrative The Suzhou story is a classical and traditional form of oral literature which is spoken in the Wu dialect used in the Suzhou area, Jiangsu province (south of Shanghai). It is an ancient tradition of story-telling called pingtan which was already well-established as a professional art at the time of the Yuan Dynasty, and was at its peak by the late Ming and early Qing periods in the sixteenth and the seventeenth centuries (Bender 1988:57). The tales can be performed with musical accompaniment, and are present ed in serial fashion over weeks or even months, as they recount the exploits of heroes, princes, scholars and beautiful ladies. Episodes are heard regularly on the Suzhou radio station, and are probably as popular, at least among older peo ple, as Chinese Opera (such as Beijing Opera and other local forms). The lan guage used in these stories is typically identified by Suzhou natives as a formal speech style mostly because it contains many literary words which do not occur in colloquial forms of contemporary Suzhou dialect. Not surprisingly, consider ing universal conservative attitudes toward linguistic change, the Suzhou story style is said to be "the real Suzhou dialect", still spoken only by older people, and preserved in the traditional stories. However, what makes those stories par ticularly intriguing is that they contain a great deal of style-shifting, at least in the segment studied here. In spite of some formal or literary features (including occasional sentences in Suzhou Putonghua), they are for the most part spoken in a vernacular style, and contain colloquial passages, which create a realistic and often facetious presentation of the characters, in the narration itself, as well as in short dialogues. The sample on which this analysis is based constitutes the twenty-second episode of the tale of Yang Nanwu and Xiao Baicai, told in about two hours by two narrators, one male and one female, and recorded from Suzhou radio in spring 1988. The two narrators' voices alternate, the male voice presenting the hero Yang Nanwu's actions, thoughts, emotions and words, but also referring to other male characters, whereas the female narrator deals with the events and

244


emotions affecting the female character Xiao Baicai. The latter has a limited participation (as befits a shy heroine in traditional Chinese tales), thus most of the speech is produced by the male narrator (150 out of 188 topic-units), and there is no apparent difference related to sex. The story recounts the misfortunes of Yang Nanwu, scholar and former dignitary who was thrown in prison because of the accusations of Xiao Baicai, the woman who was his former lover. The specific chapter examined here pre sents the encounter in Yang Nanwu's cell of the two protagonists, which takes place the night before Yang Nanwu's scheduled execution. The unfortunate scholar attempts to persuade Xiao Baicai to admit that she framed him, which is especially important since he is well aware that high dignitaries are spying on their encounter through a hole in the wall. 4.2 Text 2: Suzhou Wu (narrative style) The following excerpt from chapter 22 of the tale of Yang Nanwu and Xiao Baicai depicts the observations and feelings of one of the high dignitaries, Prince Chen, who has been pushing the other (lesser) dignitary out of the way in order to get a better view of Yang Nanwu through a discrete opening in the wall of his cell. The story humorously dwells on the spying mandarin's physical discomfort in his cramped position, and on the lesser dignitary's resentment at being nudged away. An interesting discourse feature represented here as well as throughout the tale is the narrator's use of several omniscient voices, switching from observations of each dignitary's physical behavior, to revelations of their inner feelings, as well as to the main character Yang Nanwu's inner thoughts. This text also exhibits linguistic characteristics of Chinese oral discourse observed in the context of other samples, in particular revealing the existence of stacked predicates relating to one single topic. For example, the first topic utterance (1) consists of one topic—the prince's reactions after watching through the hole, and nine subsequent comments (la-li). Some of the most obvious characteristics of Suzhou Wu (henceforth SU) are suprasegmental and phonological. The tone system in SU is fundamentally different from the native PH system (see chapter five). However, since the focus of this study is on morphosyntactic features, tonal details are not included in the following transcrip tions. An approximate transliteration of the Suzhou phonetic representation has been attempted in pinyin, an orthographic representation of Chinese loosely based on Wu pronunciation, although normally associated to PH. Each sentence is accompanied by its PH version, when possible (when no direct translation is available, the approximation is parenthesized, as well as by a literal English translation. A regular English translation following the text. Many lexical and morphological items found in the Suzhou narrative do


245

not have exact counterparts in PH, or are assigned different semantic or stylis tic values, some of them to be discussed below: (la) Ge PH (zhègè

Yang Nanwu jinze Yang Nanvu j ī n t i a n

gongfu gōngfu

[This(TI) Yang Nanwu today (lb) wang ya PH (váng yé

skill

se jiù AP

gang he, zheyang hao) AP

good

sezei qiy fe sie, s h í z à i chî bu x i ǎ o )

[prince master AP (lc)genji PH(èrqiě)

eat

wang ya vàng yé

not digest neisi biyjia, nàixîn b i i j i a o )

[besides(TI) prince master patience poor (ld) nin me li zi sang yue, PH (rén m e ( z h a n ) - l e chang yuan) [body TE stand-ASP long (Ie) dongdong me [hole hole TE ye yào

9

leide e

P H (dòngdòng me ( f ē i c h a n g )

(If) ge PH (zhègè

time

(bù shūfū)

AP

awkward

me wei qi zi, me vān__zhe)

[this (TI) waist TE bend—ASP (lg) dei sen de, PH (tóu shun dao) [head hang down, (lh) sue

z

sue

de lei,

P H (sāan

shì

sūan

)

[stiff

TI

stiff

EP]

yibi te tài

feel AP

di, dî)

low

246


(li) a sei fe si ze PH (yě (rěn) bù zhù le) [AP endure not stay EP] (2a) dei wei gulei dei Wang Xin mangmang: PH (tóu hui guòlái dùi wcng xīn wangwang) [Head turn back

TI Wang Xin look look]

(2b)nei jinze siang kue, PH(nï j ī n t ï a n xiǎng kan, [you today

want

nei kue nǐ kan

fa, gou gue ba, vǒ gān

nue niang nei. yuan rang nï

watch, you watch TE, I willing let

(3a) bilang Wang Xin a xiang fe lo: PH(yìpang) wáng xín (mèihuà kě shōu) [at side, Wang Xin AP make sound not say:] (3b) "yizei me niang gou, gou zen e a fe ye kue". wǒ, wǒ bù yào kàn) PH (xìanzài me rang ["now

TE let

me, I

AP AP not want watch]

(4a) name wang ya ciy sei yiyqie yiyqie qie PH (name vang yéqian suì yíqiao yíqiao qiào [TI

prince 1000

years limp

(4b) de

le

lang

(PH) dào.

lâo wèizhi shàng zuo

weizi

arrive old seat (5a) "Ai" yi PH (ai

la

tei

you zài tan

on

sou din. )

sit down]

qi qì

ze le)

["Ah" again ASP exhale breath EP. (5b) gerne Yang Nanwu nahang? PH(zhege Yang Nánwu zenyàng ) [TI

limp

Yang Nanwu what?]

limp

you]

TOPIC STRATEGIES IN VARIETIES OF PUTONGHUA (5c) siang siang ndo jinze fe ye tei PH ( nïmén j ī n t i ā n bù yào tàn [ TI

TI

you

today

247

qi. qi)

not must exhale breath]

(5d)"ηou Yang Nanwu li wo lei" PH (wo Yáng Nanwu zì

si yu beifa." yǒu banfa)

["I Yang Nanwu link-down-come self have method", (5e) yinci PH (yîncï

siang de geilang Yang Nanwu dang din zi ji, xiang dào zhèshàng Yang Nanwu dǎ dìng zhu j iàn)

[consequently(TI) think

AP that

(5f) jinze yiy fangmi P H ( j ī n t ï a n yì fangmiàn [today

qu quàn

Yang Nanwu make up own mind li qiy jiu ta (hē) j i u )

one direction-side persuade her drink wine

(5g) lin yiy fangmi PH (lìng yì fāngmiān

ye yaò

keisi dei quezi * ze, kāishï dōu quānzǐ le)

[other one direction-side must begin

go around

EP

(5h) yindang zenben keisang men li zenzin: "sue mei. . ." PH (yïndāng zhèngběn kāishï wen tā zhënqíng:"shuān mèi..." [should seriously begin ask her truth: "Shuan sister."] (Chapter 22:108-110) (1a) The skill Yang Nanwu displayed today was so great that (1b) the prince was overwhelmed (1c) Furthermore, the prince had little patience (1d) He had been standing for so long (le) bending in front of this /(1f) awkwardly low hole, (1g) his head inclined (1h) that he was getting very stiff, (1i) and he could not stand it any more (2a) Turning his head toward Wang Xin, he said: (2b) "You wanted to watch today, so watch! I am willing to let you do that" (3a) Wang Xin did not know what to say: (3b) "You finally let me watch now, but I don't really want to any more". (4a) The noble prince (of thousand years) limped(and limped and limped)

248


(4b) back to his previous seat. (5a) [Prince sighing]: "Ah". [Narrator, talking about the prince]: There he is, sighing again! (5b) Now, what is Yang Nanwu up to? (5c) [Talking of Yang Nanwu's thoughts] You should not sigh today. (5d) [Yang Nanwu thinking: "I, Yang Nanwu know what to do next" (5e) His reflections had led him to decide this: (5f) today, he would first make her drink; (5g) and then, he would skillfully persuade her (5h) to tell the truth: "Shuan sister. . . ." 4.3 Topic strategies in Suzhou Wu SU exhibits the same three-way system of topic mechanisms that was found in other varieties: a limited amount of topic fronting, repetition, and the use of presentative elements, including a type of clefting. The distribution of the 40 topic marking strategies exhibited in text 2 (comprising a total of five topic units) is rep resented in table 8.14. Once again the presentative strategy is clearly dominant: Table 8.14: Topic strategies in Suzhou Wu (text 2) (N^total number of topic strategies) N [40]

Fronting 10% (4)

Repetition 13% (5)

Presentation 77.5% (31)

It was observed before that short samples (that is, the two literary Putonghua excerpts, as well as excerpts from Beijing interviews and Wuhan conversation) are representative of the complete texts. This is also the case in the story of Yang Nanwu, which includes 188 topic units altogether. Table 8.15 presents the results of the analysis of topic strategies in the story, showing sep arately the male narrator's and the female narrator's productions. As indicated before, there is a discrepancy between the amount of speech produced by the two narrators: the female character—Xiao Baicai—has a subdued role in this chapter, and only occasionally responds (mostly through the indirect musings of the female narrator) to the male character—Yan Nanwu—now in jail and await ing execution because he was betrayed by his love. It is clear, however, that the small female sample matches very closely the more substantial male sample:


249

Table 8.15: Topic strategies in the Wu story

Male narrator Female narrator Both

N [679] [154] [833]

Fronting 55(8.1%) 16 (10.4%) 71(8.5%)

Repetition 184(27.1%) 35 (22.7%) 219(26.3%)

Presentation 440(64.8%) 103 (66.9%) 543(65.2%)

In both, presentative strategies consistently rank within the 60% to 70% range, whereas topic repetition ranks moderately high (around 26%), and fronting has a minimal incidence (8.5%). Those two minor strategies are briefly illustrated below, then the dominant presentative mechanism is discussed. 4.3.1 Fronting The fronting of complements, either to the beginning of a topic unit, or internally, is rare in Suzhou Wu, as it is elsewhere, but when it happens, it often co-occurs with topic repetition, typically combining preverbal and postverbal complements, as illustrated in (15) in which objects are italicized: a postverbal structure occurs in (15a) following the verb mang 'forget', whereas the preverbal (fronting) strategy appears in (15b), before the same verb fronting similar types of objects (referring to previously felt emotions such as affection, anger, and others), thus indicating that colloquial Suzhou Wu, like other forms of Chinese, does not, at least in this case, take into account general/specific or known/new information. (15a) jinze be mang siy (PH)[jïnzhāo bù wàng qián (today

siyn, qíng ];

not forget previous affection)

(15b)rjousei ni lixiang-ne hen yu (PH) [wǒ soin n ián l i - d e hèn yuan (I three years — G

qi qì

siybu cei mangji te quánbù wàngjì dào]

hatred, complaint, anger, all

forget-arrive)

'Today (I) still remember my previous affection, and I have forgotten all the hatred, complaint, and anger of the last three years' (Ch 22:11) Finally, left dislocation is even scarcer in the Suzhou text than in native Putonghua, or in Wuhan varieties. Only two cases were found out of 71 cases of fronting (less than 3% for 188 topic units), as in (16):


250

(16) Ge Sen cinwang li le kue ya (PH) [z.hè chén qīnwang, tā zài kàm a] (TI Chen Prince,

he ASP see TE)

'Prince Cheng, he watched.' (Ch.2: 36). In most cases, even with a long initial topic, there is no resumptive pronoun, as represented in (17): (17) name wang ya ciy

sei yiyqie yiyqie qie

(PH) (name vang yeqiän suì yíqiao yíqiao qiào

[TI de

prince 1000 years limp limp le weizi lang sou din.

(PH) dào läo wèizhi shàng

TI

old seat

on

zuò

limp

)

sit down]

'The noble prince (of thousand years) limped painfully back to his previous seat' (Text 2:4) 4.3.2 Repetition Onomatopoetic repetition or reduplication is a prominent feature of the SU oral narrative which is more widespread in this variety than in PH. A duplicated ele ment adds emphasis to the text, warmth, and often humor to the rapport between the narrators and their audience. Several examples occur in text 2, (e.g. mangmang 'looked intently' (2a); yiyqie yiyqie qie 'limped (very stiffly), as shown in (17) above. The following sentence—the complete version of (16) shown above—includes several repeated topics, especially in reference to the shed ding/dripping of tears {dada di and do ŋeili a di celei): (16a) Ge Sen cinwang, li le kue ya,  -de PH ( chén qînvàng, tā zài kàn a, kàn-dào [TI Chen prince 16b)

gabiy gébì

he ASP see TE, see-arrive next door two persons]

kuo de lei rjeili dada di, li de mei sue Hang, PH [kū-de yànlèi dî, tā dào man shàn Hang [cry-G AP tear

lianggadei liànggèrén)

drip drop, he AP AP kind

good,

TOPIC STRATEGIES IN VARIETIES OF PUTONGHUA 16c)

251

wei donghuo zin qi, longde e wangya PH(huì tónghùa j ìn qù, shï'dé zhè wángyé [ AP same

enter go, make-G TI

prince ]

16d) ge Hang do qeili a di celei, ŋe PH (zhè Hang dî yanlèi ye dî chū-lai, áo [TI

two

drop tears

fe le ze. bù liǎo le)

AP drop come out, suffer not end

EP]

'When Prince Chen saw those two persons shedding tears next door, and because he had a kind and generous heart, he could not contain his tears'. (SU 22: 36) Similarly, the four-way reduplication yeye hehe nienie lolo of yehe nielo 'be close friends' in (18) evokes a warm friendship which stands in ironical contrast to the lovers' anticipated demise due to one lover's betrayal of the other. Both expressions have no exact counterparts in Putonghua, as indicated in the cognate approximations, and furthermore, the English translation is unable to capture the humorous or rhetorical effects of the repetitive strategy. Note also the repetition of sei 'hand' in a serialized verbal structure: (18a) Menze PH(Mingtian

ni yeye hehe women yàohao

[To-morrow we (18b) sei cei PH shou chān

nienie lolo rènao

be close friends hot and jolly

sei de shou dào

yingei yīnj iān)

[ hand support hand arrive netherworld] 'Tomorrow we should be very close friends and hold hands to enter the nether world' (SU 22: 22) Topic repetition is also evident in Suzhou Wu classifiers (measure words) which in Chinese grammar are usually systematic markers of certain semantic properties of a noun (i.e., length, bulk, height). In the following text, they func tion as reduplicated intensifiers, for examples zangzangjiji, buebue,yangyang, to indicate excellence, a device not much used in PH. The semantic link is that the adjective applies in every dimension—flat, long, thick, etc.—a notion which has no direct counterpart in English:


252

(19a) Be PH (Bié

ninga

di-de,

li la zonggui

rénjid

tí-dào,

t a ya z h ō g g ū i

[Other people

mention-arrive she TE always

se shūo

say

19b) si yangyang he, buebue he, jiji he, zangzang he, sese he, PH (ta (yàng) hǎo, (bîan) hǎo, (jǐ ) nao, (zhāng ) hdo, (shìshì)hao [she very good, 19c) jise

very good,

very good, very good, everything good,

mbe yiy yang fe he.

PH(shénmì

döu hao)

[simply not one kind not good] 'All are unanimous in saying that she is perfect (good in every possible way') (SU 22: 67) Finally, a recurrent expression in chapter 22 involves the presentative phrase se gang zong {cue jin/yangzi) 'such type of a miserable situation/atti tude', which is usually placed at the end of the utterance, thus often functioning like a variant of right dislocation, as in (19) or (20). It is also reminiscent of phrases such as nèi yang, or n èi diǎn shìr 'those kinds of things' used in  similar manner in Beijing PH (see chapter six: text 3: 4; 6b), and would have counterparts in English, such as whatever, this kind ofthing, stuff like that. (20a) Sei ni du lixiang hei-de Yang Nanwu, PH (san nián dūo lǐ hài-de Yang Nanvǔ, [Three years more — 20b) se

gang

PH j i ù

zhèyàng zhōng cāo

[AP this (TI)

zong

suffer-G Yang Nanwu,

kind

cue

jin,

yinci

jì ng, yîncï

mbe mikong ji

li.

méi

tā)

miserable state, so(AP) not

(liân)

j iàn

face

see him]

"In the past three years and more, she had caused so much suffering to Yang Nanwu that she was ashamed to face him" (SU 22: 11)

4.3.3 Presentative strategy: Topic particles Considering the privileged status of presentation, it is no surprise to find an average of almost three particles per utterance, the same frequency for both nar rators, despite discrepancies in sample size (table 8.16). Chapter 22 contains an


253

overall number of 457 particles for 188 topic-units, when combining the contributions of both the male and the female narrators: Table 8.16: Frequency of topic particles in Suzhou Wu TU 150 38 188

Male narrator Female narrator Total: CH. 22

Particles 440 103 543

Percentage 2.9 2.8 2.9

As attested by tables 8.17 (text 2) and 8.18 (complete chapter), the Suzhou narrative style has a predilection for adverbial particles, that amount to as much as 43.8%) of all particles in the whole story, but there is an almost equally high percentage of topic introducers (38.9%)), which is in part due to the common use of clefting with z/si (see below). Table 8.17: Topic particles in Suzhou Wu (text 2) (n=number of presentative particles) n [31]

TI 38.7% (12)

 16.1% (5)

 32.3% (10)

 12.9% (4)

Table 8.18: Topic particles in Wu

Male Female Both

n [440] [103] [543]

TI 169(38.4%) 42(40.8%) 211(38.9%)

 59(13.4%) 11(10.7%) 70(12.9%)

 188(42.7%) 50(48.5%) 238(43.8%)

 29(6.6%) 0(0.0%) 29(5.3%)

Although pre-comment particles (APs) play an important role in the topic unit, there is no evidence that any one single morpheme has a predominant dis tribution, as is the case for jiu in the Wuhan sample examined above. Suzhou Wu utilizes the j iù reflex se (see text 2: la) 43 times out of 238 APs (18.1%), thus much less than the 4 5 % found in Wuhan for the same adverbial (table 8.6). In contrast to the selective use of a few APs in Wuhan PH, the variety of adver bial particles is overwhelming in Suzhou: over forty different particles occur before verbs or adjectives. In addition to se, one finds a/ya/ye (=PH yě) 14.7%, wei (=PH hái)—7.9%, and da/de (=PH dào, but rare as AP in Putonghua 1 0 )—6.3%. In this sense, the variety of AP particles in SU is more like that found in native Putonghua than in Wuhan Putonghua. This particle diversity is represented in all four particle categories, as is

254


briefly illustrated in table 8.19. The varieties represented in the SU oral narra tive include a combination of literary (or conservative) items, and others which are currently used in colloquial forms of Suzhou Wu. 1. Literary items: Some particles found in Suzhou pingtan tales are no longer in use in contemporary Suzhou varieties, such as  'consequently' (PH b iàn); the variants ya/ye 'still, also' (PH ye), now replaced in Modern Suzhou Wu by a, as illustrated in Text 2: 1 i (16d); genji 'besides' (PH érqi e) in text 2:1 c; or da/de; (PH dào) which is used in conservative Suzhou Wu to indicate surprise in predicate-ini tial position, a usage atypical in PH, shown in text 2: 5e. (5e) yinci siang de geilang Yang Nanwu dang din ziji consequently(TI) think to(AP) this YN make up own mind 'These thoughts led YN to decide this' The more regular usage of de is represented in text 2:lg, or in (16a): (lg) dei sen de head hang-arrive (serial verb sequence) '(His) head hung down' (16a) kue de gabiy lianggadei see-arrive next door two people '(The prince) saw next door two peop]e(16a) 2. Colloquial items: Other morphemes are commonly in use in vernacular varieties of Suzhou Wu. Most common are probably the deictics e, ge, gerne, gezi, name (la; If; 4a; 5b) either before a noun or an adverb; and adverbial mitigators which also function as topic introducers (i.e., siang reduplicated in 5c), which has no PH reflex. Prepositional/verbal morphemes include the object introducer nuo (PH bǎ. ) and dei ( PH dùi ). There is a wide range of a SU adver bial predicates which have no lexically related equivalents in PH. Some com mon pre-adjectival particles functioning as intensifiers include leide. . .e (text 2,le), z. . .de lei (text 2, lh); other predicate particles occur before verbs, such as a (see text 2: li; 3a), or se (text 2,1a), to mention just a few. Those Suzhou items do not occur in Beijing PH or Wuhan PH, and are generally difficult to translate, though they are pragmatically equivalent to English discourse mark ers such as as far as (X) is concerned/regarding/kinda/like or even the increas ing usage of the suffix -wise in American English (meteorological forecasts con-


255

stantly include temperaturewise; weatherwise). All such particles function as intensifiers: they are almost completely bleached of their original semantic value, although their emphatic value is probably due to their original meaning (i.e.sezei 'really'). Table 8.19: Particle reflexes in Putonghua (PH) and Suzhou Wu (SU) [C=colloquial; L=conservative or literary] TOPIC INTRODUCERS

SU:

genji(L)

e/ge(C) geme(C)

PH:

érqie

zhègè

SU: de i (C) PH: du i

he(C) hǎo

gezi(C)

name siang(C)

segang(C) nuo(C)

te(C)

zhèyàng

gēng

bǒ

si/z shî

TOPIC ENDERS

SU: ba(L) fa/va/ve(C) va(C) la(C) ya(C) PH: ba ba a a a

na(C) na

niang(C) nie(C) me(C) ne ne me

ADVERBIAL PARTICLES

SU: da/de(L) ya/ye(L) bi(L) PH: dào ye biàn SU: PH:

zen(zen)(C) zhēn

mei(C) yu(C) man you

leide e(C) a(C) ye

sezei(C) shizài (quèshi)

sei/seibu(C) se/sezi(C) döu/quān(bu) j i ù / j i ù s h i

te(C) tài vei(C) hái

END PARTICLES

SU: ne/ni(C) la/ya,a(C) PH: ne na/ya/a

niang(C) ne

ze(C) le

me(C). me

The relatively high incidence of Tis shown in table 8.18 (38.9%) is partly due to the common use of the two SU reflexes of shi (si and z), which function as a clefting structure. This morpheme is the most common topic introducer (40.8%), as represented in table 8.20 which displays the use of topic introduc ers by both narrators in the Suzhou story. The use of this particle and its variant appears twice in (20a), where both si and z occur on the same line. There is no apparent constraint on the use of either variant (the variant z occurs slightly more often than si):


256

(20a) Ge jinze si nei fe he, dedi nei z hei ninzin, PH (zhègè jīnzhāo shí nǐ bù hǎo, dàodí nǐ shí hài rěnjīng [T]

today

AP you not good, AP

you AP betray-person-spirit]

(20b) heide Jou Yang Nanwu, se gang zong yangzi PH (hài-de wǎ Yáng Nanwu, jiù zhèyang zhōng yàngzi [frame-G me Yang Nanwu, AP this

kind of situation]

"Today, (it's) you (who) are to blame, after all, (it's) you (who) are the trai tor, (who) framed me Yang Nanwu (and put me in) this miserable situation" (SU 22: 43). Obviously, this clefting particle represented by the two variants si/z can be placed either before the topic si nei or after the topic to be given salience nei z, with the same value 'it's you', and this double strategy is reminiscent of the functioning of the equivalent cognate particle shi in Wuhan and Beijing, as well as its counterpart da in Belizean Creole (see chapter nine, section 4.2). Table 8.20: Topic introducers in Suzhou Wu

Male narrator Female narrator BOTH

N [169] [42] [211]

Deictics 56(33.1%) 15(35.7%) 71(33.6%)

Verb/Prep 12(7.1%) 4(9.5%) 16(7.6%)

Adverbs 32(18.9%) 6(14.3%) 38(18%)

Shi 69(40.8%) 17(40.5%) 86(40.8%)

As far as the other introducers are concerned, deictics are quite frequent as well (33.6% overall), whereas adverbial introducers are limited. Finally, verbal prepositional TIs, such as nuo, te, dei, have a relatively minor function in topic presentation, as compared to the other categories, contrary to PH. 4.4 Stylistic shifts A particularly interesting aspect of SU narrative lies in its unique mix of formal and informal forms, literary and colloquial items, involving lexical morphemes as well as particles, which provides insight into the stylistic constraints on lex ical choice. Lexically related items are associated with different stylistic levels (ranging from colloquial to literary) in SU and PH, as illustrated in table 8.21. Typically, some unmarked PH variants have formal counterparts in SU, and vice-versa, informal SU variants may have formal reflexes in PH:


257

Table 8.21: Stylistic variations in lexical items (SU=Suzhou Wu; PH=Putonghua) [< > indicates that the item ranges over formal and literary styles ] Informal mang kàm

Formal Literary kue <wàng, d ī ng>

Gloss 'see, watch'

mikong 1 idn

lian miàm (kŏng)

'face'

PH SU PH

minze; jinze minti; jinti 'tomorrow; today' m í n g t i ā n ; j í n t i ō n <mingzhāo;j īnzhāo>

SU PH

SU

SU PH

Nonstandard zang, za qiáo

qiy hē

ha yĭn

yong yòng

'drink'

Table 8.21 shows for example that kue 'see, watch' is very formal in SU (2b in text 2) and typically used by high dignitaries, whereas the corresponding cognate PH item kán 'see' is unmarked. SU also has a colloquial counterpart mang (reduplicated in 2a), which appears when the narrator jocularly describes some of the characters' frustrations. The PH reflex vàng has a formal or poetic connotation ('gaze'). Similarly SU mikong 'face' is stylistically equivalent to the colloquial use of PH 1 i an, whereas the etymologically related PH item miànkŏng is very formal. This crossover pattern is extremely common. There are, however, some SU items which do not have clear equivalents in PH. For example, some SU vernacular forms rarely used in formal style, such as  'awkward' (le), or liwolei 'to arrive (at a decision)' (5d) have no appropriate PH counterpart, according to my Chinese sources. There is ample use of colloquial metaphors idiosyncratic to Suzhou Wu, such as qiy fe sie (lb), literally 'eat not digest', meaning 'to be overwhelmed', or xiang fe lo (3a), literally 'make sound not say', meaning 'did not know what to say'. In addition to those stylistic discrepancies, there are interesting semantic shifts, which obviously indicate an early relationship between SU and PH. For example, whereas PH differentiates between chl 'eat' and he 'drink', SU uses qiy, a reflex of chī for both 'eat' and 'drink', and with identical tones ( l b ; 5f), and uses ha for 'drink' only in very formal contexts (table 8.22). On the other hand, Suzhou Wu differentiates distance in time and place, in contrast to PH (or English "far"), indicating that both varieties, though related, developed in dif ferent directions.

258

CREOLE AND DIALECT CONTINUA Table 8.22: Semantic differentiation in Suzhou Wu and Putonghua

Gloss SU PH

'stand';'stand up' li li zhàn lí

'eat';'drink' qiy qiy chī hē

'far'(time);'far'(place) yue yue jiu yǔan

4.5 Suzhou Wu: Conclusions This brief analysis of the dialect used in Suzhou oral narratives, as represented in one radio-broadcast chapter of the tale of Yang Nanwu and Xiao Baicai shows that the morphosyntactic correlates of topic prominence occur also in varieties not directly related to Northern varieties of Mandarin, though also acquired as second dialects. In view of the objective pursued here—namely, to identify the nature of the linguistic features attributable to second dialect acquisition or, as it has become increasingly obvious, to eliminate those features which cannot be attributed to second dialect acquisition—it is of significance to observe that the narrative variety of Suzhou Wu surveyed in this chapter displays high topic prominence. This is identified primarily through the structure of the tales and the incidence of topic strategies. In those tales, Suzhou Wu has a unique value in that it is simultaneously formal/literary and colloquial/popular. The language used in those radio-broadcast tales represents in a sense a highly stylized, and some what literary and conservative variety because it describes high dignitaries, princes and mandarins in the tradition of Chinese opera. Yet, on the other hand, it shows the human foibles of those privileged characters and is intended to serve as a popular form of entertainment for mass consumption. Consequently, there is constant shifting between formal and colloquial variants depending on the context and the characters portrayed. It is thus remarkable to find that, regardless of the stylistic level, topic strategies exhibit regular patterns, as observed above. This is clearly represented in the excerpt selected because of its vernacular, often jocular stylistic level, and in the distribution of topic strate gies in the complete chapter. The relevant conclusion is that topic prominence is a universal feature of all Chinese varieties, both literary and spoken, and, therefore, one that cuts across native and non-native varieties.

5. Comparison of native and non-native varieties of Chinese When comparing the Suzhou Wu narrative style, the acquired Putonghua of Wuhan, and varieties of native (Beijing) Putonghua, it is clear that all three var-

, TOPIC STRATEGIES IN VARIETIES OF PUTONGHUA

259

ieties highly favor topic strategies in a very similar manner, and more particu larly the use of topic particles which have been specifically investigated and quantified in the three corpora. Table 8.23 compares the overall use of topic mechanisms in all three types of varieties: Table 8.23: Overall topic strategies

Beijing PH Wuhan PH SuzhouWU

N [1933] [400] [833]

Fronting 11.2% 4.8% 8.5%

Repetition 23.3% 26% 26.3%

Presentation 65.5% (Table 9) 69.2% (Table 9) 65.2% (Table 8.15)

Although differences related to second dialect acquisition clearly exist at the suprasegmental, phonological, morphosyntactic, and lexical levels, there is a remarkable homogeneity of topic mechanisms observed throughout all vari eties of Chinese, across the diachronic, synchronic, and stylistic dimensions, as well as across small or large samples. The syntactic/pragmatic level, however appears to function uniformly. This supports the universal hypothesis that dis course requires special structures and morphemes to mark syntactic sequences of communicative units. Topic particles range between two and three per TU, and occur at specific locations in discourse. Table 8.24 presents the general dis tribution of the four categories of presentative morphemes investigated throughout this analysis of topicalization: Table 8.24: Particle-types in three varieties of Chinese

Beijing: Wuhan: Suzhou:

(N) [1266] [277] [543]

TI 35.8% 27.1.% 38.9%

 10.6% 14.8% 12.9%

 44.8% 52.7% 43.8%

 8.8% 5.4% 6.6%

There is also a striking homogeneity in particle use: all three varieties use predominantly the pre-comment particles (APs), with Wuhan Putonghua lead ing in this respect (52.7%). The TI category follows closely overall, though the gap is wider in Wuhan again, which complements its higher use of APs with a lesser frequency of TIs. The other categories of particles do not pattern differ entially in any significant way. However, the internal composition of each par ticle category may present certain differences, beside the actual lexical idiosyn crasies which are most obvious in Suzhou Wu, as discussed in the previous sec tion. The most obvious differences in this respect involve:


260

1) the specialized use of a few specific adverbial particles in Wuhan PH (jiu and jiushi in particular) as pre-comment particles, whereas other varieties, both native Putonghua, and especially Suzhou Wu, display much more diversi fication in the use of particles, and 2) different priorities in the choice of topic introducers from the four cate gories identified throughout this study. It is clear from table 8.25 that whereas native Beijing Putonghua as a whole assigns equivalent importance in the 30% range to deictics (zhe, zhege, zheme, zheyang), and to verbal prepositions (ba, dui, gen, bei), both non-native vari eties of Putonghua and Suzhou Wu show a clear preference for the use of the clefting particle shi, which amounts to about 40% of all topic introducers in each case. All three varieties, however, widely use deictics, which is obviously a frequent feature of vernacular discourse. The consistent frequencies of presentative deictics (ge, gerne, gang in Suzhou for example) reflect the sponta neous level of most varieties investigated here, as well as the importance that Chinese assigns to the popular language, as reflected in the two short stories analyzed here: Table 8.25: Topic introducers in Chinese ("All native PH"=combined varieties from chapter seven)

All native PH Wuhan PH Suzhou WU

N [449] [75] [211]

Deictics 35.6% 21.3% 33.6%

V/Prep 31.8% 8.0% 7.6%

Adverbs 12.7% 29.3% 18.0%

shi 19.8% 41.3% 40.8%

However, the combined frequencies of Beijing Putonghua varieties obscures the fact (pointed out in chapter seven) that a striking difference between Literary Putonghua (reported 'peasant Putonghua') and direct record ed Spoken Putonghua is represented in the use of verbal prepositions, which is high in literary PH (40.7%), but less valued in spoken PH (19.4%), and even less so in other varieties: only eight percent in Wuhan PH and 7.6 percent in the Pingtan style of Suzhou Wu 6. Conclusions: Toward a universal view of topic processes I have thus far evaluated the topic patterns represented in Belizean Creole (chapters two to four), and in various types of Putonghua (chapters six to eight). The Chinese varieties analyzed in the latter chapters included literary interpre-


261

tations of the vernacular in the context of a political system which assigned sta tus to lower class varieties; a longitudinal perspective on two types of literary Putonghua spoken over twenty to thirty years; spoken Beijing Putonghua in the 1980s (a native variety); spoken Wuhan Putonghua in the 1980s (an acquired variety); and the pingtan style of Suzhou Wu. In spite of some minor differences in particle choice, topic strategies are remarkably homogeneous across all dimensions investigated, whether in terms of diachronic change, stylistic variation, or acquisitional types. Although the observations conducted and discussed here are limited in their scope and time range, I believe that they are extensive enough to suggest that spoken varieties of native and non-native dialects use similar discourse strategies, and that such pragmatic consistency may be due to universal patterns in the use of topic mechanisms, a theory which is substantiated when Chinese patterns are com pared to Creole patterns (see chapter nine). The intensive use of presentative morphemes noted above applies in a remarkably consistent manner to all varieties examined, whether native (Beijing informal Putonghua) or acquired as second dialects (Wuhan Putonghua and Suzhou Wu), as well as in creole acrolects. The only potential for a differential occurs within the prominent presentative strategy. There are specific differences related to the choice of certain particles, such as the preferential use of a few par ticles, versus the diversified use of a broad set of elements fulfilling similar func tions in certain pivotal articulations of the topic utterance. However, it is not clear that those particle choices separate native and non-native varieties of Chinese. For example, although it is tempting to conclude from the observed frequency of the clefting particle shi in Wuhan and Suzhou that it is a characteristic of non-native patterns, it must be remembered that the group of eight officials commenting on the value of the science program for middle school students all consistently use that particle shi 38.8% of the time (table 7.21 in chapter seven), thus differing from students, parents and teachers. More data are needed to assess the putative effect of stylistic context on particle choice, but there is no doubt whatsoever that topic prominence is an essential part of discourse processes in all varieties of Chinese, and that this type of information presentation is primarily dependent on the intensive use of particles in a variety of positions at structural boundaries. Important boundaries are obviously located at the beginning of comments, or before topics anywhere in the utterance, whether at the very beginning of the utterance, or after a sub ject which is not the main topic, or even within comments when another topic is introduced. Particles clearly contribute to the grammaticalization of structur al units, and appear to fulfill a universal function essential to information pro cessing that is not dependent on the type of learning.

262

CREOLE AND DIALECT CONTINUA Endnotes

1. There are mutually unintelligible dialects or languages, even in the close proximity of Wuhan and the neighboring state of Hunan (the Shangsha area is well-known for a distinctive language Xiang) 2. They were also particularly aware of dialectal variation in Chinese, since they had been assigned various sociolinguistic research projects involving local variation. 3. Although jiushi is normally an AP, two instances in this text qualify as topic introducers (13; 10), because they provide the setting for the whole proposition, not just the topic. 4. This borrowing is not typical of Wuhan Chinese. It is due to the fact that the two women are English teachers. 5. My consultants did not identify shi as deviant here, perhaps because it does not immediate ly precede the verb dai. 6. As indicated in chapter seven, there is no observable difference between the two short sto ries analyzed. 7. Cantonese also makes extensive use of jauh, a cognate of jiu, as attested by Matthews and Yip (p.c. 1996) (see also Matthews & Yip, forthcoming). 8. However, PG, as party representative, probably holds a fairly high status relative to her classmates and is most likely to become a cadre. 9. According to my consultants,  and liwolei are Wu idiomatic expressions with no direct correlates in Putonghua (see section 4.4). 10. The same particle da/de functions like PH dao when it occurs as a verb ('arrive') or as a preverbal preposition ('to') typically in serial verb constructions (see example in "Literary items," section 4.3.3 of this chapter).

Chapter 9 Conclusions: Pragmatic Universals in Second Dialect Acquisition Abstract This final chapter draws conclusions from the comparison of the discourse features observed in the acquisition of standard versions of Belizean English and Chinese. Those genetically and areally unrelated distinct corpora will serve as a basis for a dis cussion of such issues as the putative distinctiveness of creole languages, the signifi cance of adult language acquisition in language change, the interaction of syntax and pragmatics, and universal aspects of second dialect acquisition in discourse contexts. 1. Patterns of topic marking: Summary It has been assumed that, unlike child language acquisition, adult acquisition, broadly defined as involving the acquisition of second dialects or formal styles, does not contribute much to language change: The ability to master a language like a native, which children possess to an extra ordinary degree, is almost completely lacking in the adult. . . .I conjecture that changes in later life are restricted to the addition of a few rules in the grammar and that the elimination of rules and hence a wholesale restructuring of his gram mar is beyond the capabilities of the average adult. (Halle 1964:344) Halle's view—originally applied to phonology—may appear to be sup ported by observations of discourse strategies which exhibit remarkably homo geneous mechanisms across first and second dialects. On the other hand, one might argue that adults, on the contrary, play a significant role in language development simply by ensuring the maintenance of topic mechanisms in acquired dialects (Hopper & Traugott 1993:212).

264


The survey of topic marking phenomena, as presented in the preceding chapters, suggests a striking homogeneity of topic strategies among two distinct sets of varieties-—the acrolects of Belizean Creole and two corpora of Chinese dialects (Wuhan Putonghua and Suzhou Wu), all acquired as second dialects. In both cases, learners give a high priority to topic-comment structures, in which topics are clearly promoted and highlighted according to at least three system atic strategies: fronting, repetition and presentation. The focus was kept intentionally and unavoidably narrow because of the methodological approach adopted in the context of this analysis. This perspec tive is predicated on the principle that observational adequacy can only be appropriately met if the spectrum of acquired varieties is spanned as thorough ly as possible; this requires a corpus including speakers' extensive repertoires, in which social context and individual characteristics cannot be overlooked. Consequently, the choice was made to forfeit an analysis of a greater number of linguistic features that would have necessarily entailed a corpus limited to an unacceptably small subset of the linguistic repertoire. Extensive, if not exhaus tive, repertoire investigations demand time and effort, but they constitute pre requisites to the successful evaluation of a given variety, and more specifically in the case of second dialect acquisition. Restricting this study to one single linguistic feature can only provide a par tial understanding of the overall process of the acquisition of standard varieties by speakers of second, or third "dialects," or nonstandard varieties. Nevertheless, the feature selected is a complex linguistic variable which anchors the communicative event and has far-ranging effects over the whole sentence, proposition, or topic unit. Topic mechanisms relate to crucial aspects of communication and are clearly linked to distinctive grammatical structures such as serialization or relativization. Topics are presented effectively and persuasively through the use of various devices, such as special placement in the utterance, repeated copies, and a wide array of presentative structures and grammatical elements underscoring those top ics. Thus, in effect, topic marking mechanisms collapse syntactic and pragmatic processes in a challenging fashion. Close observations of versions of the standard produced by speakers of nonstandard dialects in Belize and in China, as compared with their native pro duction of nonstandard varieties, provide a broad perspective on discourse mechanisms. Patterns of topic marking in acquired second dialects turn out to be similar, quantitatively as well as qualitatively, to patterns of topic marking in first dialects, whether standard or nonstandard—at least for those examined here (Belizean basilects, American standard English, and Chinese/Putonghua). Ordinary discourse does not consist of independent structural units, but of a concatenation of meaningful sequences which are likely to be goal-oriented in

PRAGMATIC UNIVERSALS IN SECOND DIALECT ACQUISITION

265

the sense that speakers and hearers intend to exchange messages. To this effect, they develop strategies that enable them to communicate productively. One such strategy is the general set of mechanisms presenting core information or topics. Although some language may assign topics a relatively stable syntactic position in the grammar (in the case of topic prominent languages), it is likely that all types of languages, whether the product of native acquisition or adult acquisition develop similar mechanisms. 2. Topic marking in the Belizean continuum As far as the creole situation is concerned, it was shown that acrolects can be differentiated both from other lects and from the standard lexifier in terms of a number of phonological and morphological features. However, such lectal dif ferentiation does not apply to discourse features. Topic strategies are found to be remarkably alike in all Belizean English-based varieties—ranging from basilects to acrolects, regardless of context, style, status, or gender. Another interesting, though still preliminary finding, provides insight into patterns of diachronic change in the creole continuum. Varieties produced twenty years apart do not evidence any major difference, either in terms of dis course topics. All speakers investigated use large amounts of topic enhancing strategies, with somewhat differential usage of the three identified mechanisms. The most favored strategy overall for highlighting topics is the one involv ing the use of presentative structures, occurring 53.6% of the time in acrolects and 45.5% in basilects (see tables 4.2 - 4.5). Thus, presentation is slightly more favored in acrolects, with a predominance of various introductory phrases, such as one of the things; dat s one ting wid him; de next problem; the point is. These strategies are innovative because their function is clearly one of grammaticalized focus, much beyond the literal semantic value of the lexical items includ ed in those strings. Several presentative elements can occur contiguously before or after a topic (including any combination of NP, V, VP or adverb), and the topic itself may be repeated, commonly producing a string, as shown in (1): (1)

# (PRESENTATIVE)11 + TOPIC (X) (PRESENTATIVE)11 + (TOPIC)11 (Y) #

This pattern is illustrated in (2) and discussed in chapter four, which includes as many as three separate presentative phrases even before the first occurrence of the verbal topic, a post-topic element (now), and three addition al introductory phrases preceding the second topic reference:

266


(2) #You see, the point is, if you goin' to think PRES

PRES

PRES

+

now,

ok,

TOPIC + PRES, PRES

let's say, PRES,

for instance, each one of us thinhin ' differently.1 # PRES

TOPIC

Topic repetition closely follows presentation as a favored topic marking strategy, with an average 38.7% repetition overall (36.6% in acrolects and 46.0%) in basilects). In my corpus, basilects and acrolects slightly differ in their choice of strategies, since basilects show an equivalent distribution of repetition and presentation strategies, 46.0% and 45.5% respectively, whereas acrolects show a preference for presentation. However, it should be noted that repetition is closely linked to presentation, since the most common strategy is a version of left dislocation with a full nominal copy. Focus is assigned to a topic by the use of introductory phrases, as indicated above; the topic is then repeated fully or partially later in the sentence, and additional topic copies can also be pre ceded by other presentative elements. This pattern is illustrated in (2), with the repetition of the topic think. Finally, fronting, the least favored of the strategies represented in my corpus, never amounts to more than 8% or 10% of all topic strategies in a given lect. The crosslectal homogeneity of topic mechanisms in the Belizean varieties examined suggests that a second dialect offers extensive similarities with a first dialect in the context of nonstandard native speakers, and in particular, that the discourse/pragmatics level is the language component least likely to show first dialect substratai effects. This is generally confirmed by the longitudinal study of Belizean varieties over a ten-year interval, including data collected in 1984 and 1994, which reveals no significant difference in topic marking, except a slight decrease that would need to be confirmed by further study. The surpris ing finding that there is crosslectal homogeneity in Belize runs counter to offi cial educational policy and popular belief that acrolects are regular "good" stan dard English, and basilects are rural, uneducated, and "inferior" dialects. This discovery was made possible through extensive comparison across lects and speakers. Clearly, these findings have important implications involving educa tional policy and putative curriculum changes, as will be discussed below. 2.1 Topic consistency in English-based continua When comparing Belizean data to American English, it was found that casual American English presents strong evidence of general trends in topic marking similar to those found in Belize. Comparing Belizean to an external standard English not necessarily in direct contact, but serving as a possible formal exter-


267

nal model and educational tool, permits an assessment of the universal status of the pragmatic strategies implemented in topic mechanisms. Although American English was selected as a matter of convenience, any other standard variety (British, Canadian, Australian, etc.) would have achieved the same goal, name ly, to assess whether non-native standard dialects differ from native standard varieties. Surprisingly, and contrary to popular belief, a semi-casual, profes sional American English sample (see chapter four, and table 4.9) exhibits an even more polarized use of topic strategies, as summarized in table 9.1: Table 9.1: Topic strategies in Belizean and America

(semi-casual)

N

Fronting

Repetition

Presentation

Basilects Acrolects

[189] [717]

8.4% 9.7%

46.0% 36.6%

45.5% 53.5%

ALL BELIZEAN

[906

9.4%

38.6%

51.8%

SEMI-CASUAL AMERICAN

[252]

5.0%

21.4%

76.4%

Presentative structures rank high—76.4%—on the list of topicalizing devices used in casual American English (as compared with an average 52% for Belizean), whereas repetition averages only 21.4%, and fronting shows mini mal incidence. This sample is only a brief control because the issue here is not to achieve a full comparison of Creole English or English acrolects with stan dard varieties such as American English but simply to shed light on the putative distinctiveness of second dialect acquisition processes. The conclusion is that some styles of English do use large quantities of topic marking strategies, just like creole acrolects. In this case, casual, as well as semi-casual, American English would seem to definitely favor presentation over repetition, even more so than Belizean varieties. Considering the wide range of adverbials found to introduce topics or sepa rate topics from comments, an adequate description of topics versus nontopic ele ments must be accomplished in relational terms because human language is always relational and cannot exist out of natural contexts. Comrie (1989:64) argues that English does not grammaticalize topics because it would be inappropriate to intro duce a chapter with a cleft construction, e.g. in this chapter, it's relative clauses that we are going to talk about. However, one can argue that a formal written style is not a necessary and sufficient test of the structure of English. From a more realis tic viewpoint, spoken informal varieties are statistically much more widely repre sentative of human communicative activities than such formal tasks as introducing chapters, and therefore, it seems that an assessment of linguistic properties such as the grammaticalization of topics (to be discussed in section 4.2 below) should be more appropriately based on casual speech, or at least should refer to comparably

268


relaxed spoken or written contexts. As indicated above and amply represented in the lawyers' conversations analyzed in chapter four, and summarized in table 9.1 above, there is ample evi dence that English varieties—at least casual varieties2—have multiple ways of assigning grammatical functions to lexical items in an effort to distinguish top ics from their comments, and that they do so with a high frequency of topicstrategies, especially topic particles per topic unit (over four). 2.2 Topic marking in formal and written styles Formal and written styles also use creative devices to present topics. To cite only one example of academic prose, a linguistic discussion displays plentiful usage of a variety of presentative (though semantically empty) verbal structures which function exactly like the traditional (informal) cleft construction. In each of the following sentences, the message is in essence: "this chapter/section is about. . ." or "it is X that we will talk about." (3) In this section, we will contrast two major methodological approaches to language universals that have been adopted in recent linguistic work (Comrie 1989:1). (4) In this section, we will establish some more practical reasons why the study of language universals must operate with data from a wide range of lan guages (Comrie 1989:3). (5) In the present chapter, we will look in considerably more detail at one par ticular aspect of this problem, namely the definition of subject cross-lin guistically (Comrie 1989:104). The above introductory sentences demonstrate unambiguously that devices such as paraphrases, verbal introducers (look at one particular aspect of this problem, consider, contrast) or adverbial introducers {namely and even in considerably more detail) truly function like the cleft (it is X. . .that Y. . .), which occurs primarily in spoken varieties. Even semi-formal argumentative styles do not escape the elaborate setup provided by presentative strategies. To illustrate this situation, I adduce a few excerpts from a conversation conducted in an American university context. The following sentences are drawn from an evaluation of junior faculty performance. Participants are all senior faculty members in the English department of the University of Minnesota. It is clear that academic contexts partake in the widespread reference to topic marking devices. The samples shown below include four introductory phrases, and an


269

instance of threefold repetition in (6); the recurrent use of the discourse mark er you know, which may function as topic introducer sentence internally, as well as as topic ender (both in (7) and also in (8 and 9); the use of deictics as in this thing .. .talking to people, and this taking too many people in (8), which is also related to a type of left dislocation combined with clefting in (8) this taking too many people, that's something; there is also a type of verbal topic presentation in (9) which is a variant of a pseudo-cleft what she has done, she has accept ed; there is even a type of end particle in (10) so, which occurs at the end of one person's sentence; finally, there is evidence of prepositions functioning as topic introducers with fronted topics—with A in (II)—thus, evidencing a behavior not unlike that of verbal prepositions/coverbs in Chinese, such as bǎ, duí, gĕi, or cόng, as illustrated in (12) and (13) below: (6)

In other words, maybe, I think, it's also worth saying that A is in an unusu al position. There is no- she has no faculty member, there really is no one else who has established networks in the arts [A, Minneapolis, 11/2/94]

(7)

So she talked about, you know, different categories of people, you know [B, Minneapolis, 11/2/94]

(8)

And so, what she has done, she has accepted too many of those invita tions. So, I think, this thing, you know, talking to people in different ways, this taking too many people on [committees] that's something we should talk about [A, Minneapolis, 11/2/94]

(9)

I find it, it s, you know, someone has to do something for her, you know. But I think, one of the things that we could do, you know what I mean [B, Minneapolis, 11/2/94].

(10) This is really rich, so [B, Minneapolis, 11/2/94] (11) I think with A, it's not because she can't do it [A, Minneapolis, 11/2/94] (12) Xiànzài, gēn jiāzhǎng, [Now (TI) with (TI) parents, ASP]

dōu shuō bú shàng shenme l e = all (AP) talk not result s o m e t h i n g -

=jiùshí yī shuō ba, j iāzhǎng, dōu bǎ nǐmén dāng= [so (TI)—{AY) speak (TE), parents, all (AP) (TI) you take ] =háizi [child

kán. see]

270


'Now there is almost nothing you can talk about with parents, that is, when you talk, your parents treat you as children' (MSS.1985, 39:20). (13) duí zíjǐde nan ren yào j ī n g c h á n g j iàoyù ne 'Regarding (your) own men, (you) must always educate (them), don't you' These short excerpts suggest that contemporary American English, even in professional contexts, is as richly subjected to the ubiquity of topic particles as any of the Belizean or Chinese varieties examined. Deictics are also used as introducers, as well as prepositions (11). Particles introducing topics, such as so in (8) can also be used to close an utterance, as in (10). This may be a relative ly new development, but it is nevertheless noticeable in the United States, and it is definitely not restricted to the routinely lamented "sloppy" speech of teenagers. Structural units are generally linked into tighter amalgamations by the use of such phrases as in terms of, on the basis of, or given that, as also observed by Hopper & Traugott 1993:177). There is clearly a shift from the tra ditional meanings of adverbials or phrases (so, you know, I think, really, actu ally, etc.) to a grammatico-pragmatic function, which pervades all stylistic con texts. In brief, there is ample evidence that even semi-formal English in the con text of those educators whose mission is to instruct in the use of elegant English makes extensive use of topic presentation in particular, and repetition as well. 2.3 Topic marking by listeners There are other creative ways to mark off syntactic boundaries in discourse, and they are also represented in listeners' utterances to signal interest in the speak ers and punctuate semantic/pragmatic units. For example, minimal responses, such as mmhmm, yeah, uh huh, right appear at strategic places at phrase bound aries in predicates in 95% of all cases (Fellegy 1995), as shown in (14-16): (14) Oh urn g[V is [ N p velocity] [pp of the apparatus]] in the-I think it's just a ya ya beam of light sorta (15) s[The alleged victim may remove [NP any member] [pp of the committee] mmhmm [pp for good cause]] and an alternative. . . mmhmm


271

(16) s [she had this kind of mystical experience [pp with this woman]] ya s [coming [pp down]] and s [resting [pp by her]] mhmm mmhmm (Fellegy 1995:190) This pattern is found to be consistent across various groups of speakers in Minneapolis and not related to gender or style variables. The minimal response thus effectively functions as "a hinge between linguistic and communicative competence" (Fellegy 1995:196), which demonstrates that even paralinguistic elements can fulfill needed functions in the syntax of discourse. In conclusion of this brief summary, the interesting finding emerges that all varieties of English make active use of topic marking devices, either in the form of straightforward presentative structures or under the guise of adverbs, verbs, or more complex structures which in effect illustrate the ongoing grammaticalization of topic marking. In view of the findings presented in the previ ous chapters and summarized here, there is justification to doubt the validity of the general belief that "English does not have any grammaticalization of topic versus nontopic (comment) status" (Comrie 1989:64). In fact, Comrie and oth ers somehow mitigate such statements when admitting that there are "restricted kinds of topic or focus that can be grammaticalized," such as contrastive topic and focus, which require reference to a context, and in particular to pragmatic relations between noun phrase arguments and their predicates. 3. Patterns of topic marking in Chinese The examination of native and non-native varieties of contemporary Chinese demonstrates the importance of topic mechanisms in all forms of Chinese as well, not surprisingly in light of the general knowledge that Chinese is a topic prominent language, in contrast to other languages such as English. What is unexpected is the extent of the topic strategies used in Chinese and the equal importance assigned to topic strategies in native as well as non-native varieties, an homogeneity parallel to that found in Belizean and American texts. All var ieties give a high priority to presentation, but use fronting minimally. Yet, Chinese differs from all English-based samples in as much as left dislocation is practically absent from the Chinese corpus examined here, whether in the reported peasant speech of the 1960s and 1980s literature, in spontaneous Beijing and Wuhan conversations, and in pingtan Suzhou story-telling style. Each of these corpora includes no more than one or two examples of left dislo cation, in the context of an already restricted usage of the fronting strategy. Chapter six (section 2.2) also adduces evidence from literature that confirms the


272

extreme rarity of resumptive pronouns with fronted elements. However, this is not tantamount to saying that left dislocation is ungrammatical in Chinese, as it may occur in varieties not examined here. For example, Matthews and Yip (forthcoming) say that Cantonese structures with topicalized sentence subjects are grammatical with or without the resumptive pronouns. The question is how frequently the resumptive pronoun is realized. Topic strategies are remarkably alike in native and acquired varieties of Chinese. There is an overwhelming frequency of the presentative strategy, which prevails over the other two strategies in the 65% to 69% range in all three corpora: Beijing (native) Putonghua, Wuhan Putonghua and Suzhou Wu (nonnative varieties) as shown in chapter eight (table 8.23). Table 9.2: Topic strategies in Chinese (table 8.23 in chapter eight)

Beijing PH Wuhan PH Suzhou WU

N [1933] [400] [833]

Fronting 11.2% 4.8% 8.5%

Repetition 23.3% 26% 26.3%

Presentation 65.5% 69.2% 65.2%

No major difference in this respect emerges either from the longitudinal comparison of the peasant vernacular represented in a sample of 1964 literature, a similar type of vernacular represented in 1983 literature and, in 1985-1986, semi-casual television interviews and spontaneous conversations. Topic presentation is so pervasive in Chinese that four categories of topic particles were identified to sort out the complex mechanics of topic patterning in Chinese: 1) pre-topic morphemes (topic introducers) include, besides regular adver bials and deictics found in other languages, a category of verbal prepositions which have no equivalent in English, as they require the fronting of the topic they modify; 2) post-topic particles (topic enders) which closely follow a topic, and also occasionally occur in English, such as the now illustrated in (1) above, or the ubiquitous you know; 3) pre-comment particles, dubbed "adverbial particles" for Chinese because this class is primarily composed of adverbials or intensifiers, which structurally separate the topic from a following verb-centered commentary on the topic; and 4) post-comment particles (end particles) which signal the finish of an utterance (in English they can appear as disclaimers). All of those topic parti cles were found to have similar parsing functions, namely, to separate topics


273

from comments and, in other words, to grammaticalize the structure of topicoriented utterances. As indicated in previous chapters, APs and TIs—the precomment and the pretopic particles—are most frequently represented, which indicates that the initial position, or the position at the beginning of any syntactic unit, is the most efficient in presenting information. This pattern applies particularly to the nonnative Putonghua spoken in Wuhan, where the AP category amounts to 52.7% of all topic particles. Furthermore the Wuhan variety is characterized by the use of very few recurring elements (such as jiù, and yě). Such extreme specializa tion suggests that some adverbs previously restricted in their structural func tions are now grammaticalized into complementizers or topic-comment links. A similar specialization is represented in the dominant choice of the socalled Chinese copula shî or its cognate si/z in Suzhou Wu as topic introduc er or focus marker, which constitutes 41.3% of all introducers in Wuhan, and 40.8%) in Suzhou (but only 19%> in Beijing Putonghua), as summarized in chap ter eight (table 8.25). Thus, although all varieties of Chinese show that topic strategies and more particularly topic particles are essential in discourse structure, non-native vari eties appear to exhibit a more generalized version of the systematic ongoing grammaticalization of topic marking through the selection of a few representa tive elements. 4. Dialect versus standard: Sociolinguistic universale There are obviously differences between Belizean acrolects and Chinese acquired dialects; for example, the Belizean varieties investigated here make ample use of left dislocation, whereas that type of structure rarely occurs in Chinese. Further research is needed to verify those putative differences, but they appear to be relatively minor compared to the similarities observed in the ways Belizean acrolects and Chinese second dialects in Wuhan and Suzhou assign prominence to topics. There is ample evidence that the generally presumed distinction between topic-prominent and non-topic-prominent languages is a matter of degree rather than a clear-cut distinction. All varieties of natural discourse, including the peasant speech reflected in vernacular Chinese story writing of the Cultural Revolution include some form of topic prominence. Presentation is consistent ly favored—and even more so in the short casual American sample than in the Chinese corpus. On the other hand, Belizean acrolects display more internal variability in terms of the use of presentative elements; in some contexts, there is an accumulation of particles, whereas in other situations, there is more

274


emphasis on paratactic types of structure and repetition. In English-based situ ations, the use of particles may be more commonly triggered by a highly charged emotional context than in the Chinese context in which focus strategies constitute a regular part of the communicative mechanisms. This suggests that such mechanisms are not yet as deeply grounded in the syntax of English-based varieties as they are in Chinese, in which the systematic use of particles at syn tactic boundaries indicates a more advanced, or established, stage of the grammaticalization of topic morphology, and extensive bleaching of adverbials functioning as topic markers. The patterns shared by the Belizean and Chinese continua suggest the operation of sociolinguistic universals that develop in the context of dialect encounters, or putative conflicts between standard and nonstandard varieties. As indicated earlier, conflicts between "dialect" and "standard" appear to lead universally to the co-occurrence of overt and covert prestigious varieties. To that extent, the educational endorsement of an official standard, and concomi tant de facto exclusion of other dialects, do not seem to succeed in uprooting stigmatized varieties because such varieties preserve strong grassroot values. This may be a factor in achieving the retention of certain endangered minority languages,3 as well as the survival of ethnicity-linked features in groups long removed from their original native lands. For example, elderly Swedish-Americans in Minneapolis, Minnesota, and Lindsborg, Kansas, exhibit syntactic Swedish features in discourse, such as rel ative markers and other particles, which help maintain their Swedish-American mixed identity (Karstadt 1996; forthcoming). Based on oral history interviews, Karstadt found that the vernaculars produced by elderly Swedish-Americans consistently differ from those of socially comparable native speakers of Minnesota English in their predominant use of nonstandard variants, such as dat and a zero relativizer in subject position, as also discussed in chapter four, and illustrated in (17-18): (17) Ja, and, and, and den dey sold it to people[ø] came in dar to da farmers (18) So in 1947 I went to Bloomin'ton [ø] was all farm out dar by Minnesota wiver and bought me tree acres and made my home dar (Karstadt, forth coming) Although a substratai explanation can account for the invariable use of dat, which parallels its Swedish counterpart som, the absence of a relativizer in sub ject clauses cannot be traced to Swedish, or to standard English for that matter. Thus, this innovation may be typical of contact-derived varieties, and point to


275

universal explanations of acquisition. Research in New-Ulm, Minnesota, a German-American community also suggests the implementation of linguistic devices, such as deictics, that differentiate the speech of German-Americans, though long established in the United States, from that of other ethnic groups (Fellegy forthcoming). Fossilization has been used to refer to the persistence of external influ ences in a system, but the term fails to capture the dynamic, innovative, and— at least subconsciously—intentional use of old features to preserve a sociolinguistic identity distinct from the majority (usually dominant) group identity. Decreolization has been viewed from a similar perspective—as an innovative acquisition of strategies linking the past to the present rather than as the result of an approximation of a new target implying the intended abandonment of an earlier system or identity. It is clear that the creole in Belize, as well as local vernaculars in China remain strongly established in their respective communi ties, in spite of increased access to the educational system. Attachment to native vernaculars remains strong, in spite of the fact that most vernacular speakers are persuaded that their native nonstandard varieties have "no rules," "no gram mar," a belief that seems to be universally shared as well, in view of the wide spread purist views of language discussed in chapter one. It is possible to pro vide a sociolinguistic definition of the vernacular in terms of its emotional value, in contrast to the standard which may be defined in terms of its intellec tual value. The acquired second dialect or acrolect may have a more complex mixed identity in the sense that linguistic behavior matches ethnic identity. Contact-induced change may be characterized, at first, by the co-existence of two or more socially-conditioned variants, followed by hybridization producing new structures. For example, spoken Cantonese requires obligatory agent phrases in passives, but agentless passives (Mandarin-influenced) occur in for mal or literary registers (Matthews 1996:1275-1276). This phenomenon is a form of diglossia applied to syntax (named ditaxia by Matthews), which even tually may result in hybridization, with the possible outcomes of redundancy, double marking, or hypercorrection (Matthews 1996:1277-1278). Such linguis tic consequences of languge contact are widely represented in the study of topic structures, whether in Belizean acrolects or in Chinese acquired dialects, all showing a high incidence of repetition, extensive particle use, and the extra assignment of salience to various lexical items. 4.1 Causes of pattern consistency The crucial issue that must ultimately be raised is that of the formal cause(s) of the observed acquisitional consistency of topic marking. What is the relative importance of the substrate on the uniform development of second dialects

276


(whether Belizean Creole or Wuhan Putonghua, for example)? What of univer sal patterns of dialect acquisition? What is the function of dialect leveling as interpreted by historical linguistics? Societal and educational attitudes and century-old biases are partly respon sible for the generalized acquisition of second dialects, and the confusion and conflict thus created is certainly causally involved in the pattern consistencies observed. From a more strictly linguistic perspective, it has been claimed in prior discussions of linguistic change that clashes between different systems produce universal structures, sometimes referred to as the operation of natural ness conditions, or the production of unmarked features. The two approaches are not necessarily equivalent, and the issue of what constitutes "natural," or "unmarked" features is far from being settled. For many linguists a "natural" structure is defined as "unmarked" or more frequently "distributed." Speakers may revert to Universal Grammar (Givón 1979c), or they may produce forms which reflect the reemergence of bioprogram features originating in their innate system (Bickerton 1981). There is general disagreement as to the relative effects of substratai, superstratal, and universal influences on contact-induced change, perhaps because widely different sets of data, and often severely restricted data, have constitut ed the data base underlying interpretations. It has been said that "the universal ist and substrate hypotheses complement one another" (Arends 1986; Mufwene 1986). On the other hand, some superstratal influence is typically taken for granted, since the bulk of the lexical component is derived from that superstrate which is assumed to be the target language. Yet little has been said about the potentially continuing effect of the superstrate, thus leading to an extension of the linguistic repertoire without necessarily implying decreolization in the tra ditional sense, as discussed in chapter three, or the total loss of an earlier sys tem. In this sense, the acrolect constitutes an alternative to the standard (exter nal target), and perhaps an end in itself. This is attested by "a matched-guise experiment conducted in Hong Kong by Kingsley Bolton and Helen Kwok in which they found that Honk Kong speakers saw as their target fluent English with a Hong Kong accent, rather than standard American or British English." (Matthews and Yip pc 10/96). In fact, the more qualitative 4 linguistic data are adduced, the more com plex the situation appears to be, and this is really not an unexpected finding, since it is likely that ongoing linguistic change happens no matter where and when, and that "creoles do not enter a situation of stasis once nativized" (Muysken & Smith 1986:9). Therefore, traditional explanatory notions used in historical linguistics to explain language development, such as "borrowing," "dialect interference," or "dialect leveling" turn out to be vague and rarely rel evant when dealing with extensive corpora of spontaneous speech in the full


277

context of discourse. Some linguists have attempted to differentiate more clearly between bor rowing and substratum interference: .. .in borrowing proper, many words will be borrowed before any structural inter ference at all occurs; but in substratum interference . . .structural interference comes first. . . . In fact, the distinction between borrowing and substratum inter ference is so important that.. .we will hereafter use the term "borrowing" to refer only to "the incorporation of foreign elements into the speakers' native lan guage," not to interference in general. (Thomason & Kaufman 1991:21) Attempts at differentiating borrowing and substratum interference, however, often fail to account for the phenomenon of grammaticalization which widely occurs in the development of the creole continuum, and especially the acquisi tion of acrolects, Chinese non-native standard varieties, and the development of second dialects. It is doubtful that borrowing may occur without any kind of structural change or some restructuring, including the assignment of "old" (per haps substratai or native) functions to a newly borrowed lexical item. A "for eign" term can apparently be borrowed and still be assigned, without delay, a function that may be a projection of the substratai system, frequently because the borrowed item has some phonological resemblance to a substratai item. I have identified three general patterns occurring in cases of language mix ing. They serve to identify multiple, complex, or ambiguous identities. Frequency variability quantitatively alternates old and new variants, yet in a nonerratic fashion. If marked, new variants—such as interdentals—occur first in highly visible strong positions, such as word-initial placement; past marking occurs first in auxiliaries, though often with a pragmatic topic marking function (see chapter four; also Escure, forthcoming). The second strategy, structural hypercorrection, is characterized by multifunctionality. It occurs when a value is marked twice or more frequently; or a new item acquires a second function while still preserving its old function. This strategy can combine with frequen cy distribution. The new function, or variant, occurs less frequently than the old function and may gradually overlap. The third strategy, grammaticalization, relates to structural hypercorrection as well, as new items acquire new func tions, and those functions become part of the system. I am thus moving toward a causal interpretation of topic marking consis tency in discourse, according to which substratai and superstratal elements are manipulated to convey old (substratai) functions through new (superstratal) items. The selection of such forms may reflect an essential universal of lan guage contact.

278


4.2 Grammaticalization in topic marking strategies Meillet was apparently the first to identify, though very succinctly, a phenome non akin to grammaticalization. He referred to the grammatical system as the most stable part of language, though progressively undergoing change: "La part de la langue qui se maintient de génération en génération en se transformant progressivement est le système grammatical" (Meillet 1924:6). Martinet also mentions the mechanism leading to the development of grammatical forms out of earlier lexical items: Les nouveaux indicateurs de fonction (prépositions, conjonctions, locutions pré positives ou conjonctives) sont formés à partir d'éléments autonomes. Ceux-ci peuvent être des monèmes autonomes (l'anglais up dans he went up the hill, employé comme fonctionnel dans up the hill) ou des syngtames autonomes (fr. sans égard [pour]). (Martinet 1961:180). Grammaticalization, less commonly called grammaticization is thus a linguis tic process involving structural change in the system, perhaps the most dynam ic and fundamental vehicle operational in second dialect development. It has been variously defined as "the linguistic process, both through time and synchronically, of organization and coding" (Traugott & Heine 1991, vol. 1:1), as "the harnessing of pragmatics by a grammar" (Hyman 1984:73), and more extensively as: the process whereby lexical items and constructions come in certain linguistic contexts to serve grammatical functions, and once grammaticalized, continue to develop new grammatical functions. . .it refers not only to processes observable in language, but also to an approach to language study, one that highlights the interaction of use with structure, and the non-discreteness of many properties of language (Hopper & Traugott 1993:xv). Any definition of this dynamic process must account for the interaction of structure and use, or of syntax and pragmatics. Thus it is a crucial issue to deter mine whether grammaticalization is the result of discourse strategies, or is due to other influences. Although it is impossible to identify and separate all the fac tors, cognitive, distributional, social, substratai or superstratal, which may con tribute to language change in general, and to grammaticalization in particular, crosslinguistic observations are likely to shed light on the operation of gram maticalization in the context of topic processing. The overlapping of the prag matic properties of topic marking, and of the syntactic mechanisms contributing to the delineation of propositional structure has been amply demonstrated in all the corpora examined here. What is involved is the transformation of various lex-


279

ical items (deictics, adverbials, verbs or prepositions) into items that fulfill gram matical functions necessary to the communicative process. The items selected for those functions are usually frequently distributed, and probably selected because of their visibility. In some varieties, topic marking functions rely on a few selected items which recur with high frequency (da in Belize; j i u in Wuhan), whereas other varieties, such as Beijing Putonghua, assign those same functions to a wide range of lexical items. This grammaticalization process is well illustrated in the extremely wide spread use of da which can alternatively function as topic introducer (19) or as post-topic marker (20) in Belizean basilects, in parallel structures and with equivalent meanings: (19) Da

Habat#

mέ

di kapten

TOP Herbert PAST ZERO COPULA the captain

[It's/that's H. (who) was the captain] 'H. was the captain/The captain was H.' (Escure 1983c: 194) (20) Habat# Herbert

da

me —

di kapten

TOP PAST ZERO COPULA the captain

[Herbert, he/that (was) the captain] 'H. was the captain/The captain was H.' (Escure 1983c: 196) The occurrence of the morpheme da can be interpreted as the appropria tion of the superstratal deictic/demonstrative that which because of its frequen cy and iconicity became easily extended to fulfill pragmatic functions, and at the same time remained fixed in syntactic positions in the close proximity of topics. It is likely that the shifting from restricted morphological categories to broader syntactic-pragmatic values was facilitated by the influence of analo gous substratal mechanisms (Escure 1983c). An issue to be raised is the reason why certain items are systematically and selectively subjected to grammaticalization, whereas others are not. The logical response to this puzzle lies in the universal priorities found in communicative processes. Highlighting the topic appears to be one of these universal priorities. Thus, the morpheme da becomes so important in topic marking that it can also occur after the topic as well, in comment-initial position, thus functioning as a left-dislocated structure (20), as well as in topic initial position, thus function ing as a cleft (19). A consequence of this ubiquitous position of da in Belizean basilects is that this topic marker has often been mistaken for a copula because of its postnominal position.5 It is manifestly impossible to identify this da as a copula

280


because the same highlighting value is associated to the element contiguous to da. Both sentences are cleft/dislocated types which provide salience to the fact that "Habat" is new information in regard to the identity of the captain of the boat.6 In addition, in both cases da is followed by typically preverbal morphol ogy (here the past or anterior marker me; elsewhere the progressive/iterative marker ds), which means that da cannot be a copular verbal element; thus, m s must precede a "zero-copula." The universal aspect of pseudo-copula/topic marker ambiguity is supported by data from Chinese. As indicated previously (see chapters seven and eight), the same situation obtains in Chinese under identical conditions: shí (si/z in Suzhou Wu) can occur both in topic-initial and in comment-initial positions, with the same topicalizing function in what can be considered to be a copula-free language, con trary to some assumptions that sh i is an occasional copula: (21) Fǎnzhèng shí tĭng =nèige —dòng zhī xīn-de [Really TI AP —TI —move one heart-G] '(It) truly moves me deeply' [Beijing Saga B: 3; see chapter six (28)] (22) yīge shi, Tzhǔyào, shi wŏ àirén, ta gōngzuò yòu máng a, you shi yào shàng kè, yào dài háizī (Wuhan:Text l: 5a-b) 'One problem is really about my husband, his work keeps him very busy, yet he not only must teach classes, but he must also look after the child' [chapter eight (1)] (23) G e j i n z e si nei fe he, dedi nei z hei ninzin, heide Yang Neiwu se gang zong yangzi (Suzhou Wu 22:43)

ηou

'Today, (it's) you (who) are to blame, after all, (it's) you (who) are the traitor, (who) framed me Yang Nanwu (and put me in) this miserable situation' [chap ter 8 (20)] The universal interpretation of the grammaticalization of deictic elements receives strong support from classical Chinese data, as represented in texts such as the Confucian Analects and Mencius (dating from 551 to 221 b.c.). In those early texts, the deictic-demonstrative shi is "usually anaphoric, with no par ticular implication of closeness or remoteness. . .and functions to recapitulate a phrase or a series of phrases which is the subject of a noun predicate out of which its later use as a copula develop" (Pulleyblank 1995:85). However, the interpretation of shí as "modern copula" does not appear to be substantiated


281

by the acquired data adduced in the corpus presented here. It is clear that shî still has a primarily salient value. Thus, the striking similarity of the cleft (pseu do-copula) in Chinese and Belizean creole-related varieties at all lectal levels, including acquired dialects, reinforces the universal hypothesis that some spe cific particles serve to focus a topic in pre- or post-topic position. If those par ticles do not exist, they will be created through grammaticalization. A second example illustrating the topic marking universal comes from the use of other items such as adverbials. For example, it was shown in chapter four that many adverbials, such as still, suddenly, again in acrolectal varieties have assumed topic marking or topic-comment delimiting functions. In the case of acrolect acquisition, this process is interpreted as grammaticalization of simple adverbials which function as syntactic boundaries and partly lose their original meaning. The bleaching of adverbials is even more advanced in all varieties of Chinese, to the extent that practically every topic chain requires its chain links. The particles which operate the smooth articulation of topics and comments as substitutes for the most common basilectal da. In Chinese also, numerous adverbials function as topic markers. This can be viewed as a case of substrate or first dialect interference through borrowing, and also involving restructur ing of the system in the sense that the grammaticalized adverbs are added in positions and in quantities not used in formal standard versions of the same sen tences. Finally, there is the case of the development of formal styles which use stilted verbal introducers in lieu of simple clefts or pseudo-clefts, as illustrated above in samples of chapter introductions (see 3-5 above). Thus, similar types of essential but subtle grammaticalization occur throughout the developmental continuum—whether in acrolect acquisition or in stages leading to the devel opment of the creole basilect, and in the stylistic continuum of standard lan guages as well. Tracing this phenomenon to the universality of topic presenta tion in natural discourse does not conflict with the substratai interpretation, but only confirms why certain features are selected over others. Those features which are common to substrate and superstrate are eventually preferred. These examples support the claim that reanalysis is a major factor in change (Hopper & Traugott 1993:48), and therefore that change may be accomplished by adult learners of second dialects. 4.3 The syntax of discourse The consistency of topic marking observed in the previous chapters yields insights into the relation between syntax and discourse. Perhaps the question to be posed is: Is there syntax in discourse? However, this question only arises if syntax is defined narrowly as a subset of formal properties defining an

282


autonomous level of structural organization, as it is in traditional linguistic the ory. If so, the answer must necessarily be negative, considering the shifting val ues of grammatical elements, both in Belizean and in Chinese. An autonomous level would exclude the possibility of interaction with the pragmatic level, of concrete communication. Perhaps, a better question to ask would be: Is syntax independent from pragmatics? Based on the evidence provided in the previous chapters, the answer must also be negative. Clearly, syntax is highly dependent on discourse strategies—even centered around the communicative events of normal dis course communication. My data strongly support Givόn's definition of syntax as a "dependent functionally motivated entity" (Givón 1979b:208). The Belizean and Chinese corpora—all collected in natural discourse situ ations as documented above—illustrate what I called a stacked syntax, that is high incidence of paratactic structures with a low frequency of embedding and avoidance of subordination and complementation. This phenomenon has also been referred to as clause chaining by DeLancey (1991) who describes the ori gin of verb serialization in Tibetan through the grammaticalization of verbs. Avoiding complementizers, including relative pronouns, naturally leads to rep etition as shown in some of the examples of relative clauses, and others, given in chapter four. However, repetition is not a redundant, ponderous vehicle for communication but rather a necessary enlightening device—Ariana's thread running through the complex convolutions of natural discourse. The fact that there is a predominance of parataxis, as found in Belizean and Chinese discourse, does not mean that there is a lack of structural organization. Discourse is highly structured around topic-comment constituents: an initial phrase (topic) sets the stage for what is said about it (comment). Thus, although in formal contexts syntax avoids redundancy by the use of pro-forms or para phrases, creole and Chinese discourse—spanning the whole continuum from native varieties to acrolects/acquired dialects—shuns proforms to favor repeti tion and presentation as topic-marking devices. This results in a fluid type of structural organization which cannot be neatly broken into delineated NP-VP sequences, thus forcing an analysis of discourse units as spanning semantic units beyond sentence boundaries. But even formal contexts may use clitics which will follow the initial topic, as in "the judge, however, took a different view," implying a contrast between "the judge" and some other participant, and thus clearly marking "the judge" as the topic of this sentence (Hopper & Traugott 1993:134). Such strategies allow structural boundaries to separate effectively topics from their comments. Givón's discussion of the diachronic process of syntacticization con tributes some stimulating ideas which are highly relevant to the present discus sion. Second dialect acquisition can indeed be considered diachronic since it


283

involves the sequential addition of one system to another. Givón assumes that it is typical for "loose paratactic pragmatic discourse stmctures [to] develop— over time—into tight grammaticalized syntactic structures" (Givón 1979a:208). In fact, he posits two extreme poles of communication: the pragmatic mode and the syntactic mode. His discussion implies that the pragmatic mode is a stage prior to the syntactic mode, with the following associations: Pragmatic

Syntactic

Pidgins → Creoles Child language → Adult language Informal language → Formal language (Givón 1979a:223) However, such a characterization seems to imply that there is a unilateral development from a "loose" pragmatic organization to a "tight" syntax, and more importantly, that the syntactic stage is more evolved than the pragmatic stage. The situation is manifestly more complex than what seems to be implied by this claim of directionality toward syntacticization.7 It appears more likely that the situation is such that both "modes" coexist, interact, and are simultaneously available to speakers in different degrees, and subject to the effects of context. Indeed, formal versions of standard languages are learned later than informal varieties under normative conditions, and typi cally through the strict constraints dictated by educational systems. Yet, infor mal varieties are never abandoned, and the acquisition of formal styles merely implies an extension of the stylistic repertoire available to individuals. Similarly, the acquisition of non-native acrolects in Belize and in Wuhan, Hubei (PRC) expands the stylistic/lectal range available to native speakers of Belizean Creole and of nonstandard Wuhanhua. Observations of production data in SLA also support the claim that dis course functions and syntactic positions are simultaneously elaborated. Japanese and Chinese learners of English produce a high frequency of interlanguage structures representing topic prominence, especially periphrastic topic constructions withfor and pseudo-passives in which the fronted topic is an argu ment of the verb (Yip & Matthews 1995), as illustrated below: (24)

For first language acquisition, the acquiring process is unconscious...while for L2 learners, they also imitate and follow the pro nunciation of their target language (Yip &Matthews 1995:26)


284

(25) These ways can classify two types (Yip &Matthews 1995:22) This structure is parallel to its Chinese version (26), in which these ways is the object of the verb: (26) Zhexie fangfa keyi fen These way

liang

zhong

can classify two types

'These methods can be classified into two types' (Yip & Matthews 1995:22) In conclusion, a close look at second dialects in natural discourse indicates that the syntactic structures of acquired dialects partake of the same pragmatic strategies. The organization of discourse in terms of clearly marked topic chains (or series of topic/comment sequences) is a natural development in non-native dialects, regardless of the degree of topic prominence of the corresponding native systems of second dialect learners. What is remarkable is that informal (spontaneous) varieties of a language also exhibit highly topicalized structures (as represented in the casual American sample presented in chapter four). The major finding emerging from this study is that discourse is a great equalizer, the common denominator to all languages and all varieties. Regardless of language type, and irrespective of whether the language is a cre le or not, natural discourse exhibits similar types of pragmatic structures in topic presentation, or pragmatic inferencing (Hopper & Traugott 1993:63). In other words, the role of speakers negotiating meaning in communicative situa tions cannot be discounted and may in fact be a strong motivator in language acquisition. This is a strong claim which is testable, and indeed, should be test ed in the context of more languages and more extensively documented speak ers' repertoires. The finding that pragmatic principles are universally pre-eminent in spon taneous narratives is not tantamount to saying that such discourse is incoherent, disorganized, and redundant (which is the popular belief). On the contrary, top ics and their related propositions are tightly coded with high frequencies of par ticles—or various lexical items—grammaticalized to function as topic particles, as well as as by minimal responses produced by listeners, which clearly sepa rate the major semantic units of discourse for easier processing. If a language does not have pre-existing presentative particles, other elements will become grammaticalized units, thus creating a true syntax of discourse ruled by prag matic principles. Finally, the issue of the social context of second dialect acquisition must


285

ultimately explain a striking feature of second dialects. Formal (written) forms of a standard language like American English (which can be viewed as second dialects acquired after the native vernacular) do exhibit only one type of topic marking (presentation through formal verbal structures), whereas the informal end of continua evolving from nonstandard dialects exhibit richly varied topic and multipropositional structures, with extensive repetition as well as presenta tion. One of the reasons for this style-linked differential may be that the acqui sition of formal dialects in a standard context is highly constrained and codified within the educational system. Repetition is a strong taboo, whereas elaborate introductory devices are prestigious. Such attitudes are primarily determined by social constraints acquired in educational and professional situations. In contrast, the acquisition of standard dialects by nonstandard speakers is not directly addressed within an educational system, since nonstandard varieties are rarely given overt recognition, whether in school or in any official context. Speakers of Belizean Creole are "assumed" to speak English, and speakers of Wuhanhua are "assumed" to speak Putonghua. It is generally "assumed" that the acquisition of the standard is not problematic. In China, the problem is com pounded by the fact that the written logographic system, which signals formal literacy, does not reflect extensive spoken differences between different forms of Chinese. Even the use of the pinyin written system in the early years of pri mary education does not resolve the issue of dialect discrepancies because the pinyin orthography is based on Beijing Putonghua. Thus, the acquisition of acrolects and other standard forms is totally spontaneous and obviously func tions like a vernacular. In this case, language acquisition (and change) is goaloriented, or even teleological, a perspective that fits with a ftmctional view of language. Language change is determined by strategies used by speakers/hear ers in producing an understandable flow of speech, and this requires a strong anchoring of topics. 5. Conclusions A thorough examination of unrelated vernacular languages demonstrates that the discourse component, and specifically topic marking, functions in a univer sally manner consistent with such psychosocial motivations in communication as "economy, efficiency, clarity, expressivity or routinization" (Hopper & Traugott 1993:67). Many different dimensions were investigated in this survey of second dialect acquisition, namely, wide geographical distribution (Central America, China, and to a limited extent, the United States); genetically unrelat ed languages (African/Indo-European for Belizean; Sino-Tibetan for Chinese); acquisitional diversity (native versus acquired varieties); longitudinal compar-

286


isons (1960s and 1980s for Chinese; 1980s and 1990s for Belizean); and final ly stylistic discrepancies (from written to casual styles). Throughout those diverse dimensions, the varieties examined inescapably demonstrate that the syntax of discourse has universal procedural priorities, all varieties displaying the systematic marking of syntactic boundaries through a regular three-way sys tem of topic mechanisms. When there is no formal pre-existing topic morphol ogy, the grammaticalization process assigns required syntactic values to avail able items such as adverbs, deictics or pseudo-copulas, and even minimal responses. Furthermore, in languages like Chinese that are already programmed for topic orientation, the language develops supplementary items to increase the syntactic definition of topics and their predicated propositions. In this sense, the increasing use of lexical items, adverbs, coverbs, deictics, etc., may be viewed as the process by which lexical elements, partially or totally bleached of their original semantic value, come to mark grammatical relations. In dialect acquisition, it is clear that the incorporation of substratai features is not merely a calquing process but that it is, in part, a dynamic reinterpretation through superstratal influences, and also affected by contact universals that appear to operate most effectively at the pragmatic level. Universal elements may be primarily identifiable in terms of the common features which are those transferrable into the new code. This implies that a new system pragmatically evolves, in priority, a discourse-oriented syntax. Acrolects thus have great potential for modifying and remodeling an available standard target in innova tive ways that match pragmatic requirements. 6. Postscript: Directions for education—-from language to social change The ultimate purpose underlying any study of socially marginalized groups in developed or developing nations is to eventually achieve the demarginalization of those groups. One of the major stumbling blocks interfering with a balanced incorporation of standard learners into the world of native speakers of the stan dard resides in the uncertainties of those learners who are regularly faced with language conflicts. Linguistic insecurity is naturally a reflection of the lack of recognition of nonstandard dialects and related social fallout. This entails igno rance of the linguistic social and linguistic distinctiveness between standard and nonstandard varieties. This gaping absence has an immediate practical conse quence in the sense that learners have no clear starting point, or point of reference, to help them develop productively throughout the required standard acquisition process. The assumption that all Belizeans speak English, or that all Chinese speak Mandarin/Putonghua, is counterproductive because it is manifestly wrong. The artificial sense of security resulting from the fallacy of linguistic commonality


287

results in confusion in the minds of teachers and students alike. I noted once the equation here-hair-hear figuring prominently on the blackboard of a Belizean classroom. This clearly represented the instructor's intention to teach distinctive orthographic representations of what he perceived as a homophonic set pronounced as [her]. On another continent, my Chinese graduate students had never realized that their acquired Putonghua omitted Beijing-type retroflexes (with concomitant vowel and tone changes) until they started sociolinguistic fieldwork in their hometown of Wuhan. Dangerous ped agogical traditions are illustrated in both cases, namely, the habitual priority given to writing and reading, and the general omission of any systematic refer ence to spoken forms, as well as to the linguistic variability linking basilects and acrolects. Education is potentially the most effective equalizer because it can reach children early enough in life. The issue of language acquisition and the associated sense of social and ethnic identity in relation to others is proba bly the most crucial, yet the most downplayed, element in educational systems. The main findings of these observations of Belizean and Chinese dialect continua include the general distinctiveness of first and second dialects, espe cially on the suprasegmental and phonological levels, the strong covert prestige assigned to nonstandard vernaculars, the special psychosocial value assigned by their speakers to non-native acquired standards, the universal spontaneous strategies used to process information, and the general overlap of syntax and pragmatics. All of these observations can easily be applied to develop progres sive, nonprescriptive, and more flexible systems of education. Some possible innovations might include: 1) providing equal educational time and respect to native and non-native varieties co-occurring in a given society. Thus varieties no longer need conflict, but can co-exist peacefully; 2) training teachers to incorporate locally-appropriate discussions of the social functions and linguistic aspects of the varieties in contact, and; 3) recognizing that spontaneous oral communication is the easiest medium through which to initiate the acquisition of second dialects (rather than through formal and written forms), since they are identical to primarily oral vernaculars, at least in the area of topic strategies. Such commonalities would provide a sta ble and positive base to the acquisition of the standard. The task of creating unity in plurality is not as paradoxal as it appears, and leveling or exclusion is not the solution. The Roman phrase e pluribus unum has been interpreted in Belize as Belize da all a wi. Perhaps not surprisingly, a stan dard English translation "We are Belize" fails to capture the salience of the Belizean structure, more appropriately reflected in its casual left dislocated ver sion: "Belize, that's all of us."

288


Endnotes 1. This sentence is discussed as (11) in chapter four. 2. It is reminded here that the lawyers' conversation could be qualified as "semi-formal" rather than casual because of the professional context. 3. Thanks to Stephen Matthews and Virginia Yip for pointing out this important implication of "dialect" versus "standard" conflicts. 4. "More" is used here in a qualitative, not a quantitative sense, as stated above. It is felt that a greater range of the community's or the speaker's repertoire should serve to constitute the data base, as opposed to using more data of a single type. 5. An unconscious bias to compare any language to familiar Indo-European grammars cannot be discounted. 6. This is a case where a topic provides new information, as opposed to the traditional claim that the topic introduces old information, followed by new information in the comment. In both Belizean and Chinese data, it is common to find new information topicalized, as occurs in English in such cases as in "As far as X is concerned. . ." or "Now about X. . ." where X is a new topic. 7. Givón later states that "syntactic structure in time erodes via processes of morphologization and lexicalization" (Givón 1979b:208), thus leading apparently to cyclical uses of pragmatic and syntactic modes.

Bibliography Alleyne, Mervyn. 1971. Acculturation and the cultural matrix of creolization. Pidginization and creolization of languages ed.by D. Hymes, 169-186. Cambridge University Press. —. 1980. Comparative Afro-American. Ann Arbor, MI: Karoma Publishers. Andersen, Roger, ed. 1983. Pidginization and creolization as language acqui sition. Rowley, MA: Newbury House. Anonymous Editors. 1980. 1978 prize-winning short stories. Beijing: People's Literature Publishing House. Arends, Jacques, Pieter Muysken, & Norval Smith. 1995. Pidgins and creoles. Amsterdam: John Benjamins. Arends, Jacques, Silvia Kouwenberg, & Norval Smith, eds. 1995. Theories focusing on the non-European input. Pidgins and creoles: An introduction. ed. Jacques Arends, Pieter Muysken, & Norval Smith, 99-109. Amsterdam: John Benjamins. Arends, Jacques. 1986. Genesis and development of the equative copula in Sranan. Substrata versus universals in creole genesis ed. by P. Muysken & N. Smith, 103-128. Amsterdam: John Benjamins. Asturias, Francisco. 1941. Belice. Guatemala: Tipografía nacional de Guatemala. Bailey, Beryl Loftman. 1966. Jamaican Creole syntax. London: Cambridge University Press. Bancroft, H.H. 1883-87. History of Central America Vol 6-8. San Francisco: The History Company. Bender, Mark. 1988. The storyteller's aesthetics and 'Song Sigong' Journal of the Chinese Language Teachers Association, 23.2.55-67. Bickerton, Derek. 1975. Dynamics of a creole system. Cambridge University Press. —. 1981. The roots of language. Ann Arbor, MI: Karoma Press. —. 1984. The language bioprogram hypothesis. Behavioral and Brain Sciences 7.173-88. —. 1989. Seselwa serialization and its significance. Journal of Pidgin and Creole Languages 4.155-83.

290


—. 1993. Subject focus and pronouns. Focus and grammatical relations in cre óle languages ed. by Francis Byrne & Donald Winford, 189-121. Amsterdam: John Benjamins. —. 1994. The origins of Saramacan syntax: A reply to John McWhorter's 'Substratai influence in Saramaccan serial verb constructions'. Journal of Pidgin and Creole Languages 9.1.65-78. Biq, Yung-O. 1988. From focus in proposition to focus in speech situation: Cai and jiu in Mandarin Chinese. Journal of Chinese Linguistics 16.1.73-107. Blust, Robert. 1994. The Austronesian settlement of Mainland southeast Asia. Papers from the Second Annual Meeting of the Southeast Asian Linguistics Society ed. by by K. L. Adams & T. J. Hudak, 25-83. Tempe: Arizona State University. Bryan, Beverley. 1996. How much English does the Jamaican primary school child hear? Perceptions of a group of primary school teachers. Paper pre sented at the Eleventh meeting of the Society for Caribbean Linguistics. St Maarten, N.A. Burdon, Sir John Alder. 1935. Archives of British Honduras. London: Sifton Praed. Butterworth, Brian, Bernard Comrie, & Osten Dahl, eds. 1984. Explanations for language universals. Amsterdam: Mouton. Byrne, Francis. 1987. Grammatical relations in a radical creole. Amsterdam: John Benjamins. Cadiot, Pierre. 1992. Matching syntax and pragmatics: A typology of topic and topiclike constructions in spoken French. Linguistics 30: 57-88. Cassidy, Frederic G. & R. B. LePage. 1980. Dictionary of Jamaican English. Second Edition. Cambridge University Press. Chao, Yuen Ren. 1968. A grammar of spoken Chinese. Berkeley: University of California Press. Chaudenson, Robert. 1992. Des îles, des hommes, des langues: Essai sur la créolisation linguistique et culturelle. Paris: L'Harmattan. Chen, Matthew. 1976. From Middle Chinese to Modern Peking. Journal of Chinese Linguistics 4.2/3.113-277. Cheng Chin Chuan. 1973. A quantitative study of Chinese tones. Journal of Chinese Linguistics 1.1.93-110. Chinese Writers'Association. 1963. Duan Pian Xiaoshuo, Vol. 3. Selected Short Stories. Beijing: Writers' Press. Chomsky, Noam. 1973. Conditions on transformations. A festschrift for Morris Halle ed. by Stephen R. Anderson & Paul Kiparsky, 232-286. New York: Holt, Rinehart & Winston. —.1977. On WH-Movement. Formal syntax ed. by P. Culicover et al., 71-132. New York: Academic Press.

BIBLIOGRAPHY

291

—. 1981. Lectures on government and binding. Dordrecht: Foris. Christie, Pauline. 1983. In search of the boundaries of Caribbean creoles. Studies in Caribbean language ed. by L. Carrington, 13-22. Trinidad: Society for Caribbean Linguistics. Comrie, Bernard. 1984. Form and function in explaining language universals. Explanations for language universals ed. by B. Butterworth, B. Comrie,, & O. Dahl, 87-104. Amsterdam: Mouton. —. 1989. Language universals and linguistic typology. Chicago: The University of Chicago Press. Corne, Chris. 1989. Un créole à base lexicale française en Nouvelle-Calédonie: le tayo ou le patois de Saint-Louis. Etudes créoles 12.2.29-42. —. 1994. Relativization and thematization in Tayo and the implications for cre ole genesis. Journal of Pidgin and Creole Languages 9.2.283-304. —. 1995a. A contact-induced and vernacularized language: How Melanesian is Tayo? From contact to creole and beyond ed. by Philip Baker, 121-148. London: University of Westminster. —. 1995b. Nana  nana, nana  napa: The paratactic and hypotactic relative clause of Reunion Creole. Journal of Pidgin and Creole Languages, 10.11.57-76. Corne, Chris, Deirdre Colemann, & Simon Curnow. 1995. Clause reduction in asyndetic coordination in Isle de France Creole: The 'serial verbs' prob lem. From contact to creole and beyond ed. by Philip Baker. London: University of Westminster. Craig, Dennis. 1966. Teaching English to Jamaican Creole speakers. Language Learning 16: 1-2 —. 1971. Education and Creole English. Pidginization and creolization of lan guages ed by D. Hymes, 370-391. Cambridge University Press. —. 1980. Models for educational policy in creole-speaking communities. Theoretical orientations in creole studies ed. by A. Valdman & A. Highfield. New-York: Academic Press. —. 1996. Putative absurdity in student writing. Paper presented at the Eleventh meeting of the Society for Caribbean Linguistics. St. Maarten, N.A. Crystal, David. 1987. The Cambridge encyclopedia of language. Cambridge University Press. de Beaugrande, Robert. 1992. Topicality and emotion in the economy of dis course. Linguistics 30: 243-266. de Francis, John. 1950. Nationalism and language reform in China. Princeton: Princeton University Press. DeCamp, David. 1971. Toward a generative analysis of post-creole speech community. Pidginization and creolization of languages ed. by D. Hymes, 340-376. Cambridge University Press.

292


den Besten, Hans, Pieter Musken, & Norval Smith. 1995. Theories focusing on the European input. Pidgins and creoles ed. by J. Arends, P. Muysken, & N. Smith, 87-98. Amsterdam: John Benjamins. DeLancey, Scott. 1991. The origins of verb serialization in modern Tibetan. Studies in Language 15: 1-23. Dictionary of English Etymology. 1966. Oxford: Clarendon Press. Dik, Simon. 1980. Studies in functional grammar. London: Academic Press. Dittmar, Norbert. 1992. Introduction. Linguistics 30:1-3. Edwards, Walter. 1992. Sociolinguistic behavior in a Detroit inner-city black neighborhood. Language in Society 21.93-115. Ellis, Rod, ed. 1994. The study of second language acquisition. Oxford University Press. Elyan, O., P. Smith, H. Giles, & R. Bourhis. 1978. R. P.-accented female speech: The voice of perceived androgyny? Sociolinguistic patterns of British English ed. by P. Trudgill, 122-31. London: Edward Arnold. Escure, Geneviève. 1978. Vocalic change in the Belizean English-Creole con tinuum and markedness theory. Berkeley Linguistics Society IV, 158-167. Berkeley: University of California Press. —. 1979. Linguistic variation and ethnic interaction in Belize: Creole/Carib. Language and ethnic relations ed. by Howard Giles & Bernard SaintJacques, 101-116. Oxford: Pergamon Press. —. 1981. Decreolization in a creole continuum: Belize. Historicity and varia tion in creole studies ed. by A. Highfíeld & A. Valdman, 27-49. Ann Arbor: Karoma. —. 1982. Contrastive patterns of intragroup and intergroup interaction in the creole continuum of Belize. Language in Society 11.2.239-264. —. 1983a. Belizean Creole: Chapter Two. Central American English ed. by John Holm, 29-70. —. 1983b. The use of creole as interlanguage by the Black Caribs of Belize. Mid-America linguistics conference papers ed. by Frances Ingemann, 271282. Lawrence: University of Kansas Press. —. 1983c. The Belizean copula: A case of semantic shift. Studies in Caribbean language ed. L. Carrington, 190-202. Trinidad: Society for Caribbean Linguistics. —. 1984. The acquisition of creole by urban and rural Black Caribs in Belize. York papers in linguistics 11: Urban creoles ed. M. Sebba & L. Todd, 95106. York, England: University of York. —. 1987. The acquisition of Putonghua (Mandarin) by speakers of the Wuhan dialect. Variation in language: NWAV-XV at Stanford ed. by K. Dennings, S. Inkelas, F. McNair- Knox, & J. Rickford, 121-136. Stanford University: Department of Linguistics.

BIBLIOGRAPHY

293

—. 1988. Topic structures as language universals. Journal of Pidgin and Creole Languages: 3.2: 159-5176. —. 1991. Gender roles and linguistic variation in the Belizean Creole commu nity. English around the world: Sociolinguistic perspectives ed. by Jenny Cheshire, 595-608. Cambridge University Press. —. 1991. Serialization in creole oral discourse. Development and structures of creole languages ed. by F. Byrne & T. Huebner, 179-192. Amsterdam & Philadelphia: John Benjamins. —. 1993a. Anaphoric reference in creoles and noncreoles. Language Variation and Change 5.3.233-259. —. 1993b. Focus, topic particles and discourse markers in the Belizean Creole continuum. Focus and grammatical relations in creole languages ed. by F. Byrne & D. Winford, 233-247. Amsterdam & Philadelphia: John Benjamins. —. 1993c. Gender and linguistic change in the Belizean Creole community. Locating power Vol. I, ed. by K. Hall et al., 118-131. Berkeley Women and Language Group, University of California. —. 1994. Référence anaphorique dans l'espace créolophone. Etudes Créoles. Paris: Didier Editions 17.1.68-83. —. Forthcoming. Gender and tense/aspect in Belizean Creole. Contact languages ed. by F. Mugler & J. Tent. Canberra: Pacific Linguistics, The Australian National University. Farmer, A. K. & Hamish, R. M. 1987. Communicative reference with pronouns. The pragmatic perspective ed. by Verschueren & Bertuccelli-Papi, 547-65. Amsterdam: John Benjamins. Fasold, Ralph. 1984. The sociolinguistics of society. Oxford: Basil Blackwell. Fellegy, Anna. 1995. Patterns and functions of minimal responses. American Speech. 70.2: 186-199. —. Forthcoming. Locative Deictics as Discourse Markers. University of Minnesota PhD dissertation. Ferguson, Charles. 1959. Diglossia. Word 15: 325-340. Forrest, Robert A.D. 1973. The Chinese language. London: Faber & Faber. Fromkin, Victoria & Robert Rodman 1983. An introduction to language. Third edition. Holt, Rinehart and Winston. Fudan daxue fangyan Diaocha gongzuo zu. 1958. Jiangsu Shanghai Fangyan Diaocha. (Dialects survey group of Fudan University Jiangsu and Shanghai Dialect survey). Vol.1. Shanghai: New Knowledge Press. Gass, Susan. 1980. An investigation of syntactic transfer in adult second lan guage learners. Research in second language acquisition ed. by Robin Scarcella & Stephen Krashen, 132-141. Rowley: Newbury House Publishers.

294


Gilbert, Glenn (ed.) 1987. Pidgin and creole languages: Essays in memory of John Reinecke. Honolulu: University of Hawaii Press. Giles, H., R.Y Bourhis, & D. Taylor. 1977. Towards a theory of language in eth nic group relations. Language, ethnicity and intergroup relations ed. by H. Giles, 307-348. London: Academic Press. Givón 1979a. On understanding grammar New York and London: Academic Press. —.1979b. Discourse and syntax, syntax and semantics 12. New York: Academic Press. —. 1979c. Prolegomena to any sane creology. Readings in creole studies ed. by Ian F. Hancock, 3-35. Ghent: Story-Scientia. —. 1990. Syntax: A functional-typological introduction. Vol II. Amsterdam: John Benjamins. —. 1992. The grammar of referential coherence as mental processing instruc tions. Linguistics 30: 5-56. Görlach, Manfred. 1986. Middle English—a creole? Across historical and geo graphical boundaries ed. by Kratovsky & S., 329-344. Graham, Frank. 1980. The new Geordie dictionary. Newcastle upon Tyne: J. & P. Bealls Ltd. Green, Georgia. 1989. Pragmatics and natural language understanding. Hillsdale, N.J.: Lawrence Erlbaum Associates. Guy, Gregory 1988. Language and social class. Linguistics: The Cambridge survey. Vol IV: Language: The sociocultural context ed. by F. Newmeyer, 37-63. Cambridge University Press. Halle, Morris. 1964. Phonology in generative grammar. The structure of lan guage: Readings in the philosophy of language ed. by Jerry A.Fodor & Jerrold J.Katz. Englewood Cliffs, NJ, Prentice-Hall. 334-52. Han, Yang Saxena. 1995. A pragmatic analysis of the ba particle in Mandarin Chinese. Journal of Chinese Linguistics 23.2.99-127. Hancock, Ian. 1980. Gullah and Barbadian: Origins and relationship. American Speech 55.17-35. Hao Ran. 1984. Hao Ran Xuan Ji Vol. 2. (Hao Ran's Selected Works). Tianjin: Baihua Press. Hashimoto, Anne Yue. 1971. Mandarin syntactic structures. Unicorn 5. Princeton University. Hashimoto, Mantaro J. 1973. The Hakka dialect: A linguistic study of its phonology, syntax and lexicon. Cambridge: Cambridge University Press. Hellinger, Mariis. 1972. Aspects of Belizean Creole. Folia Linguistica (Acta Societatis Linguisticae Europaae) 6.Y2.118-35. Hengeveld, Kees. 1990. A functional analysis of copula constructions in Mandarin Chinese. Studies in Language 14-2.291-323.

BIBLIOGRAPHY

295

Holm, John A. (with A. Shilling) 1982. Dictionary of Bahamian English. Cold Spring, N.Y.: Lexik House Publishers. Holm, John. 1988. Pidgins and creoles. Vol. I. Cambridge University Press. Hopper, Paul & Sandra Thompson. 1980. Transitivity in grammar and dis course. Language 56: 251-99. Hopper, Paul J. & Elizabeth C. Traugott. 1993. Grammaticalization. Cambridge Textbooks in Linguistics. Hsieh-Hsin-L 1973. A new method of dialect subgrouping. Journal of Chinese Linguistics 1.1.64-92. Huang Chu-Ren.1990. A unification-based LFG analysis of lexical discontinu ity. Linguistics 28.263-307. Huang, C.-T. James. 1984. On the distribution and reference of empty pro nouns. Linguistic Inquiry 15.531-74 —. 1987. Remarks on empty categories in Chinese. Linguistic Inquiry 18.321-37. Huang, Yan. 1994. The syntax and pragmatics of anaphora. A study with spe cial reference to Chinese. Cambridge: Cambridge University Press. Huxley, Aldous. 1934. Beyond the Mexique Bay. New York and London: Harper & Brothers Publishers. Hvitfeldt, Christina & Gloria Poedjosoedarmo. 1995. Language use as identity across generations in Singapore. Paper presented at the Fifth Meeting of the Southeast Asian Linguistics Society, May, 1995, Tucson, AZ. Hyman, Larry. 1984. Form and substance in language universals. Explanations for language universals ed. by B. Butterworth, B. Comrie, O. Dahl, 67-86. Amsterdam: Mouton. Jakobson, Roman. 1940. Kindersprache, Aphasie und allgemeine Lautgesetze. Selected writings I ed. by R. Jakobson. The Hague: Mouton. Karlgren, Bernhard. 1915. Etudes sur la phonologie chinoise. Stockholm: Leyde Imprimerie Orientale. Karstadt, Angela. 1996. Relative markers in Swedish-American English: Evidence for a contact language phenomenon? American Speech71.1:27-43. —. Forthcoming. Swedish American English: A longitudinal study of linguis tic variation and identity. University of Minnesota PhD dissertation. Keenan, Edward & Bambi Schieffelin. 1976. Topic as a discourse notion. Subject and discourse topic ed. by Charles Li. New York: Academic Press. Keenan, Edward & Bernard Comrie. 1977. Noun phrase accessibility and uni versal grammar. Linguistic Inquiry 8: 63-100. Labov, William. 1966. The social stratification of English in New York City. Washington, D.C.: Center for Applied Linguistics. —. 1972a. Language in the inner city. Philadelphia: University of Pennsylvania Press.

296


—. 1972b. Sociolinguistic patterns. Philadelphia: University of Pennsylvania Press. —. 1972c. Some principles of linguistic methodology. Language in Society 1.97120. — & W. Fanshel. 1977. Therapeutic discourse. New York: Academic Press. Lalla, Barbara & Jean D'Costa. 1991. Language in exile: Three hundred years of Jamaican Creole. Tuscaloosa: University of Alabama Press. Le Page, Robert & Andrée Tabouret-Keller. 1985. Acts of identity. London: Cambridge University Press. Lefebvre, Claire. 1986. Relexification in creole genesis revisited: The case of Haitian Creole. Substrata versus universals in creole genesis ed. by Pieter Muysken & Norval Smith, 279-300. Amsterdam: John Benjamins. Leggett, Mead, & Kramer. 1988. Handbook for writers. 10th edition. Prentice Hall. Levinson, Stephen. 1983. Pragmatics. Cambridge University Press. Li, Charles N. and Sandra Thompson. 1975. The semantic function of wordorder: A case study in Mandarin. Word order and word order change ed. by Li & Thompson, 47-45. Austin: University of Texas Press. —. 1977. A mechanism for the development of copula morphemes. Mechanisms of syntactic change ed. by Li & Thompson, 419-443. Austin: University of Texas Press. —. 1981. Mandarin Chinese: A functional reference grammar. University of California Press. Linde, Charlotte. 1981. The organization of discourse. Style and variables in English ed. by Shopen & Williams, 84-113. Cambridge: Winthrop Publishers. Liu, Lanying (ed.) 1981. Grammar and rhetoric. Nanning: Guangxi Publishing House. Liu, William W. 1979. Dialect features and communication problems in Linxian. Journal of Chinese Linguistics. 7: 29-43. Lu, Shuxiang. 1980. On language. Beijing: San Lian Press. Lu, Zhongti. (with Celia Millward). 1989. Chinese given names since the Cultural Revolution. NAMES 37.3: 265-280. Manessy, Gabriel. 1989. De quelques notions imprécises (bioprogramme, sémantaxe, endogénéité). Etudes Créoles 12.2: 87-111. Martinet, André. 1955. Economie des changements phonétiques. Berne: Francke. —.1961. Eléments de linguistique générale. Paris: Armand-Colin. Matisoff, James. 1994. Following the marrow: Two parallel Sino-Tibetan et mologies. Papers from the Second Annual Meeting of the Southeast Asian Linguistics Society ed. by K. L. Adams & T. J. Hudak, 213-234. Tempe: Arizona State University

BIBLIOGRAPHY

297

Matthews, Stephen. 1996. Ditaxia and hybridization in Chinese dialect gram mar. Proceedings of the Fourth Conference on Pan-Asiatic Linguistics, Bangkok 1274-1283. — & Virginia Yip. 1994. Cantonese: A contemporary grammar London: Routledge. —. Forthcoming. Verb-copying constructions in Cantonese. Proceedings of the Third International Conference on Chinese Linguistics. City University of Hong Kong. McLaughlin, John. 1987. Theories of second language learning. London: Edward Arnold. McWhorter, John. 1992. Substratai influence in Saramaccan serial verb constructions. Journal of Pidgin and Creole Languages 7.1-53. Meillet, Antoine. 1921. Linguistique historique et linguistique générale. Paris: Champion. —. 1924. Les langue du monde. Paris: Champion. Milroy, Lesley. 1980. Language and social networks. Baltimore: University Park Press. Mufwene, Salikoko. 1986.Universalist and substrate theories complement each other. Substrata versus universals in creole genesis ed. by Pieter Muysken & Norval Smith, 129-162. Amsterdam: John Benjamins. Muysken, Pieter & Norval Smith. 1986. Substrata versus universals in creole genesis. Amsterdam: John Benjamins. —. 1995. Study of pidgin and creole languages. Pidgins and creoles ed. by J. Arends, P. Muysken, & N. Smith, 3 -14. Amsterdam: John Benjamins. Muysken, Pieter. 1994. Saramaccan and Haitian: A comparison. Journal of Pidgin and Creole Languages (Column). 9.2: 305-314. — & Tonjes Veenstra 1995. Universalist approaches. Pidgins and creoles ed. by J. Arends, P. Muysken, & N. Smith, 121-134. Amsterdam: John Benjamins. Nichols, Patricia 1983. Linguistic options and choices for black women in the rural south. Language, gender and society ed. by B. Thorne, C. Kramarae, & N. Henley. Rowley: Newbury House Publishers. Norman, Jerry. 1988. Chinese. Cambridge Language Surveys. Cambridge: Cambridge University Press. Oxford Dictionary of English Etymology. 1985. ed. by C. T. Onions. Oxford: Oxford University Press. Platt, John. 1991. Social and linguistic constraints on variation in the use of two grammatical variables in Singapore English. English around the world: Sociolinguistic perspectives ed. by J. Cheshire, 376-387. Cambridge University Press.

298


Pollard, Velma. 1996. A singular subject takes a singular verb and other causes (?) of hypercorrection in Jamaican speech. Paper presented at the Eleventh meeting of the Society for Caribbean Linguistics. St Maarten, N.A. Population Census of Belize. 1980. Belmopan. Statistical Office: Central Planning Unit. Population Census of Belize. 1991. Major findings: Belmopan. Central Statistical Office. Postma, Johannes M. 1990. The Dutch in the Atlantic slave trade, 1600-1815. New York: Cambridge University Press. Pulleyblank, E.G. 1984. Middle Chinese: A study in historical phonology. Vancouver. University of British Columbia Press. —. 1994. The old Chinese origin of type A and  syllables. Journal of Chinese Linguistics 22.1.73-99. —. 1995. Outline of classical Chinese grammar Vancouver: University of British Columbia Press. Ramsey, Robert. 1987. The languages of China. Princeton University Press. Rickford, John. 1983. What happens in decreolization. Pidginization and creolization as language acquisition ed. by R. Andersen, 298-319. Rowley, MA: Newbury House. —. 1987. Dimensions of a creole continuum: History, texts and linguistic analy sis of Guyanese Creole. Stanford: Stanford University Press. — & Jerome Handler. 1994. Textual evidence on the nature of early Barbadian speech: 1676-1835. Journal of Pidgin and Creole Languages 9.22: 1-55. Roberts, Peter. 1983. Linguistics and language teaching. Studies in Caribbean language ed. by L. Carrington, 230-244. Trinidad: Society for Caribbean Linguistics. —. 1988. Second language acquisition theory and second dialect acquisition. Paper presented at the Seventh meeting of the Society for Caribbean Linguistics, Nassau, Bahamas. Romaine, Suzanne. 1988. Pidgin and creole languages. London: Longman. Ross, Claudia. 1983. On the functions of Mandarin de. Journal of Chinese Linguistics 11.2: 217-246. Ross, John. 1967. Constraints on variables in syntax. MIT Dissertation. Ruhlen, Merritt. 1991. A guide to the worlds languages. Vol. 1. Stanford University Press. Sampson, Geoffrey. 1989. Review of J. Norman Chinese. Journal of Linguistics 25: 225-235. Scarcella, Robin. 1983. Developmental trends in the acquisition of conversa tional competence by adult second language learners. Sociolinguistics and language acquisition ed. by Nessa Wolfson & Elliott Judd, 175-183. Rowley: Newbury House Publishers, Inc.

BIBLIOGRAPHY

299

Schiffrin, Deborah. 1987a. Discourse markers. Cambridge University Press. —. 1987b. Toward an empirical base in pragmatics. (Review article). Language in Society 16.381-396. —. 1992. Conditionals as topics in discourse. Linguistics 30: 165-198. Schlobinski, Peter & Stephan Schütze-Coburn. 1992. On the topic of topic and topic continuity. Linguistics 30: 89-122. Schmidt, Richard. 1983. Interaction, acculturation, and the acquisition of com municative competence: A case study of an adult. Sociolinguistics and language acquisition ed. by Nessa Wolfson & Elliott Judd, 137-174. Rowley: Newbury House Publishers, Inc. Schmied, Josef. 1991. National and subnational features in Kenyan English, English around the world: Sociolinguistic perspectives ed. by J. Cheshire, 420-422. Cambridge University Press. Schumann, John. 1978. The pidginization process: A model for second lan guage acquisition. Rowley: Newbury House. — & Stauble, Ann-Marie. 1983. A discussion of second language acquisition and decreolization. Pidginization and creolization as language acquisi tion, ed. by R. Andersen, 260-274. Rowley, MA: Newbury House. Selinker, Larry 1972. Interlanguage. International Review of Applied Linguistics 10: 209-231. Shi, Dingxu. 1989. Topic chain as a syntactic category. Journal of Chinese Linguistics 17.2: 223-254. Shields-Brodber, Kathryn. 1989. Standard english in Jamaica: a use of compet ing models. English World-Wide 10 1: 41-53. Singler, John 1991. Social and linguistic constraints on plural marking in Liberian English. English around the world: Sociolinguistic perspectives ed. by J. Cheshire, 544-562. Cambridge University Press. Stauble, Ann-Marie. 1978. The process of decreolization: A model for second language development. Language Learning 28.29-54. Stauble, Ann-Marie. 1980. Acculturation and second language acquisition. Research in second language acquisition, ed. by R. Scarcella & S. Krashen, 43-50. Rowley, MA: Newbury House. Stewart, William. 1964. Foreign language teaching methods in quasi-foreign language situations. Nonstandard speech and the teaching of English ed. by W. Stewart. Washington, D.C. Center for Applied Linguistics. Stochl, John. 1975. A dictionary of Central American Carib (3 vols.). Revised by R. E. Hadel and R. Zuniga. Belize-City: BISRA. Sun, Chaofen. 1995. Transitivity: The ba construction and its history. Journal of Chinese Linguistics 23.1.159-195. Taylor, Douglas. 1951. The Black Carib of British Hondura. Viking Fund Publications in Anthropology, 17. New York: Wenner-Gren Foundation.

300


— 1955. Phonemes of the Hopkins (British Honduras) dialect of Island-Carib (Island-Carib, I).International Journal of American Linguistics 21:233-41. —. 1956a. Island-Carib, II: word classes, affixes, verbs, nouns. International Journal of American Linguistics 22:1-44. —. 1956b. Island-Carib, III: locators, particles. International Journal of American Linguistics 22:138-50. —. 1958. Island-Carib, IV: Syntactic notes, texts. International Journal of American Linguistics 24: 36-60. —. 1977. Languages of the West Indies. Baltimore: The Johns Hopkins University Press. Teng, Shou-Hsin. 1977. A semantic study of transitivity in Chinese. Taipei Student Book Company. —. 1979. Remarks on cleft sentences in Chinese. Journal of Chinese Linguistics 7.1: 101-113. Thomason, Sarah Grey & Terrence Kaufman. 1991. Language contact, creolization and genetic linguistics. University of California Press. Todd, Loreto. 1974. Pidgins and creoles. London: Routledge and Kegan Paul. Traugott, Elizabeth Closs & Bernd Heine. 1991. Approaches to grammaticalization (Vol. I & II). Amsterdam: John Benjamins. Tsao, Feng-Fu. 1979. A functional study of topic in Chinese. Taipei, Taiwan; Student Book Company. —. 1987. A topic-comment approach to the ba construction. Journal of Chinese Linguistics 15.1: 1-53. Valdman, Albert. 1983. Creolization and second language acquisition. Pidginization and creolization as language acquisition ed. by R. Andersen, 212-234. Rowley, MA: Newbury House. Wald, Benji. 1983. Referents and topic in discourse units. Discourse perspectives on syntax ed. by Flora Klein-Andreu, 91-116. New York: Academic Press. Wang, Li. 1954. A Chinese modern grammar. Beijing: Zhonghua. Wang, William S-Y. 1969. Competing changes as cause of residue. Language 45.9-25. —. 1989. Language in China: A chapter in the history of linguistics. Journal of Chinese Linguistics 17.2. 183-221. Williams, Jessica. 1987. Non-native varieties of English: A special case of lan guage acquisition. English World Wide 8.1.161-198. Winford, Donald 1993. Predication in Caribbean English creoles. Amsterdam: John Benjamins. Xu Liejong and D.T. Langendoen. 1985. Topic structures in Chinese. Language 61.1.1-27.

BIBLIOGRAPHY

301

Yip, Virginia & Stephen Matthews. 1995.I-interlanguage typology: The case of topic prominence. The current state of interlanguage: Studies in honor of William E. Rutherford ed. by L. Eubank, L. Selinker, & M. Sharwood Smith, 17-30. Amsterdam: John Benjamins. Yule, George & Terrie Mathis. 1992. The role of staging and constructed dia logue in establishing speaker's topic. Linguistics 30: 199-216. Zou, Ken. 1993. The syntax of the Chinese ba construction. Linguistics 31.4.715-736.

Index acrolect 6, 19, 22, 36-7,42, 51-2, 55, 57-123, 263-6, 273-5, 281 distinctiveness 111-2 longitudinal dimension 112-4 phonological features 77-9 morphological features 79-80 syntactic features 83-5 adverb(ial); 81, 99-103, 166182, 206, 209, 218, 233-7, 241-2,252-4,257,265,270, 281 see topic particles semantic change (of) 99-103 anaphora anaphoric binding 91-2, 1526, 161-3 nominal 163-5 pronominal 161-5 zero anaphora 153, 186 Arawak(an) 52, 34 aspect 49, 51-2, 61,79, 83, 103 ba 166, 182 see topic ender bǎ-construction 133, 158, 161, 166, 172-8, 198-200 see verbal preposition basilect 19, 22, 41, 48-9, 58-71,

77, 86-89, 101-2, 109, 122, 181-2,266-7,279-80 longitudinal study 112-4 Belize census 27-31 Chinese 28-9, 35 Creolesee creole, basilect emigration 26-37 ethnicity 28-36 history 25-36 languages 28-55 longitudinal dimension 1124,266 people 28-55 see Creoles, Garifuna, Maya, Mennonite, Mestizo Stann Creek 43-55 bèi -construction 161, 172, 1778, 192, 198,200 see verbal preposition bioprogram 6, 41, 59 Black Carib see Garifuna Black English (US) 7, 9, 11, 18, 70,79 bleaching 192, 274, 281, 283, 286 Cantonese(Yue) 28, 35, 130, 132, 136-8,262,275

304


Catalan 10 Chinese Cultural Revolution 128-30, 141, 162, 187-208 continuum 137 dialect differentiation 137-41, 145-8, 187-221,223-261 genetic classification 131-7 history 125-30, 141-2 language see Cantonese Mandarin, Putonghua, Wu, Wuhan longitudinal dimension 187, 208 naming 129-30 post Cultural Revolution 2028 topics 151-261 see topic code-shifting 63-4 cleft(ing); pseudo-cleft 51, 84, 90,97-8,231-2,267-8,280-1 co-indexing 91, 152-3 Common language see Putonghua coverb see verbal preposition copula; pseudo copula 51, 71-84, 166,234,280-1 see da; shǐ creole continuum 3, 19-20, 39-43, 61-4, 102-3,265-6 genesis 39-41, 135 people(Creoles) 28-34, 42-48 post-creole 61-4 radical 57-63 structure see basilect use 30-2, 35-8,41-4, 52-4 creolization 2, 59, 121, 131, 134 da 103, 170,279-81 decreohzation 2, 19, 57, 62-5, 70,

275-6 definition 64 deictic 103, 160, 171, 197-8,218, 235, 269-70, 279 cataphoric 160 diglossia 64, 224, 238, 275 empty category 153 English American 42, 68, 117-20, 266-71 Belize 30-1, 35, 38,41-4, 523, 67-9 English Only movement 9 German-American 275 Singapore English 4, 8 Swedish-American 115-6, 274 focus 49, 51,83-4, 93-4, 103, 160-1,265 fossilization 275 French 41,52, 60, 134 frequency variability 76-82, 89, 277 functional analysis 14-15, 153-4, 281-5 fronting see topic gapping 156-8 Garifuna Garinagu 34 history 26, 34 language 52-5 people 31, 34, 37-8,44-9, 52, 71-2 use of creole/English 38, 49, 50-4, 71-3 women's language 54-5 gender 54-5, 70-1, 193-7 Geordie 7

INDEX Government and Binding 11, 902, 152-3 grammaticalization 77, 83, 265, 271-3,277-81 Guanhua 139 guided learning 7 Gullah 70 Guoyu 139-40 Haitian 40, 60 Hakka 138 Island Carib see Garifuna Jamaican Creole 61 jiù '166, 181,233-4,237,241, 253 Kekchi 28 Kenyan English 5 language acquisition 2-10, 63-5, 90, 114-7, 145-8,263-4,275-85 attitudes 69-73, 129,139-41 276 bias 9-10, 30, 71-2 education 286-7 identity 8, 32, 70-5 policy 36-9, 266 universals 77, 274-5, 280-6 left dislocation 51, 90, 94, 161-2, 230,266,271-3 lexical diffusion 135-6 Liberian English 5 longitudinal study see creole, Chinese Manchu 22, 128, 133-4, 139 Mandarin 130-46 markedness 4, 276-7

305 matched guise 10, 12, 276 Maya 26-32, 37, 43 Melanesian 41 Mennonite 28-9, 35-6 mesolect 19, 36, 42, 49-51, 55, 63,67 Mestizo 28-32, 34-5 Miskito 28, 33 multifunctionality 277 nautical lingua franca 40 nonstandard variety 10-1,61, 122, 264-75 objects fronted 160-1, 192,230-1 non-fronted 160-1, 230 observational adequacy see sociolinguistic methodology oral discourse 13-8, 209-21, 223-61 parataxis 114-6, 122,282 pidginization 2, 39-41, 57, 131 pingtan 148,243-61,271 possessive 156 pragmatics 14-18,263-88 pragmatic mode 122 presentation see topic Putonghua 21, 140, 185-221 Beijing Colloquial 22, 125-7, 141-4, 161-2, 166,209-21, 272-3 Literay 162, 187-208 phonology 142-4 rural variants 197-200, 206-7 Wuhan 22, 125, 144-6 rap 10 relativization 83-4, 114-6, 156,

306

CIŒOLE AND DIALECT CONTINUA

264,274,281 relexifïcation 40, 76, 81, 83, 89 repetition see topic resumptive pronoun 84-5, 230, 272 reflexivization 154 Saramaccan 58-60 serialization 59-60, 85, 282 Seselwa 59 shí/si/z 167-70,197,201,206, 218,234-8,255-6,273,280-1 sociolinguistic methodology 11-2, 155,264 Spanish 26-37, 52 style minimal responses 270-1 shifting 109-11, 206-7, 194-7, 256-8 standard 10-3, 37, 61, 65-74, 79, 81, 123,264-75 definition 66 structural hypercorrection 76, 801,89,277 substrate 40-1, 58, 266, 275-81, 286 superstrate 5, 40, 83, 277-81 Suzhou Wu see Wu Tayo 41,60 topic chain 18, 154-5, 186-90,2024,209-11,215-6,225-9,2458, 264-5 definiteness 160-1,230 definition 92-4, 153 fronting 84, 94-6, 106-20, 155-63, 191-220,230-2,239, 248-50 given/new status 160-1, 230,

280 multiple 158-60 particles see topic particles presentation 84, 94-120, 155, 166-182, 191-220,232-42, 252-6, 265-6 repetition 96-7, 106-20, 155, 163-5, 191-220,232,239 theoretical interpretation 903, 151

unit 104-6, 155, 186-93,2448 topicalization 90, 93-5, 153-6, 161 topic particles AP(adverbial particles) 166, 179-81,193,201,213,217 EP(end particles) 166, 181-2, 193,201,213,217 TE(topic enders) 178-9, 193, 213-9 TI(topic introducers) 166-78, 192-3, 197-201,206,213, 217 unguided learning 8 Universal Grammar 5, 153, 276 universals see language univer sals verbal preposition 172-8, 192, 197-200,206,218 with fronted objects 206, 270, 280 see topic particles (TI) see bǎ , bèi women's language see Garifuna, gender word-order 136, 156-61, 230-1

INDEX Wu (Suzhou) area 146-7 language 136, 147-8,243-61, 272-3 Wuhan area 145 normative status 236-8 use of Putonghua 145-6, 22442, 272-3

Creole and Dialect Continua: Standard Acquisition Processes in Belize and China (Prc) (Creole Language Library)

Creole and dialect continua: standard acquisition processes in Belize and China

Mbugu (Creole Language Library)

Creolization and Contact (Creole Language Library)

Structure And Variation in Language Contact (Creole Language Library)

Creole Discourse: Exploring Prestige Formation and Change Across Caribbean - English-lexicon Creoles (Creole Language Library)

Roots of Creole Structures: Weighing the contribution of substrates and superstrates (Creole Language Library)

Reading Acquisition Processes (Language and Education Library)

Creole Blues

Creole Formation as Language Contact: The Case of the Suriname Creoles (Creole Language Library)

Creole Genesis and the Acquisition of Grammar: The Case of Haitian Creole (Cambridge Studies in Linguistics)

Defining Creole

Kriyol Syntax: The Portuguese-based Creole Language of Guinea-bissau (Creole Language Library)

Kriyol syntax: the Portuguese-based Creole language of Guinea-Bissau (Creole Language Library)

Creole Fires

Creole Fires

Deconstructing Creole (Typological Studies in Language)

Noun Phrases in Creole Languages: A multi-faceted approach (Creole Language Library)

Variation in the Caribbean: From creole continua to individual agency

Contact Languages: A Wider Perspective (Creole Language Library)

Settler and Creole Re-Enactment

Contact languages: a wider perspective (Creole Language Library)

Haitian Creole English Dictionary

Creoles, Contact, and Language Change: Linguistics and Social Implications (Creole Language Library)

From Creole to Standard: Shakespeare, Language, and Literature in a Postcolonial Context. (Cross Cultures)

Second Dialect Acquisition

History, Society And Variation: In Honor of Albert Valdman (Creole Language Library)

Language in exile: three hundred years of Jamaican Creole

Second Dialect Acquisition

An Arabic Creole in Africa: The Nubi Language of Uganda

Creole Identity in the French Caribbean Novel

Creole and Dialect Continua: Standard Acquisition Processes in Belize and China (Prc) (Creole Language Library)