Languages in Contact: The Partial Restructuring of Vernaculars

This page intentionally left blank

Languages in Contact The Partial Restructuring of Vernaculars

There is widespread agreement that certain non-creole vernaculars are structurally quite different from the languages out of which they grew: African American English, Afrikaans, Brazilian Vernacular Portuguese, Nonstandard Caribbean Spanish, and the Vernacular Lects of R´eunion French. Until now, however, these languages have proved remarkably resistant to the attempts of linguists to provide a plausible theory to account for either their genesis or their synchronic structure. Informed by the first systematic comparison of the social and linguistic facts in the development of these languages, this book argues that the transmission of their source languages from native to non-native speakers led to partial restructuring, resulting in the retention of a substantial amount of the source languages’ morphosyntax, but also the introduction of a significant number of substrate and interlanguage features. This study concludes with the proposal of a formal theoretical model identifying the linguistic processes that lead to partial restructuring, throwing into focus a key span on the continuum of contact-induced language change which has not been coherently analyzed up to now. It demonstrates how the insights gained from the comparative study of such vernaculars cast much-needed light on the relationship between the diachronic development and synchronic structure of this important group of languages, with some 200 million speakers. John Holm is Chair, English Linguistics and Director, Graduate Program in Descriptive Linguistics at the University of Coimbra, Portugal. He has written widely on the issues surrounding contact languages, and his previous publications include Introduction to Pidgins and Creoles (Cambridge University Press, 2000) and Pidgins and Creoles (Cambridge University Press, Volume 1, 1988; Volume 2, 1989).

Languages in Contact The Partial Restructuring of Vernaculars John Holm

   Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo Cambridge University Press The Edinburgh Building, Cambridge  , United Kingdom Published in the United States by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521430517 © John Holm 2004 This book is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published in print format 2003 ISBN-13 978-0-511-06755-6 eBook (EBL) ISBN-10 0-511-06755-0 eBook (EBL) ISBN-13 978-0-521-43051-7 hardback ISBN-10 0-521-43051-8 hardback

Cambridge University Press has no responsibility for the persistence or accuracy of s for external or third-party internet websites referred to in this book, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

For Mary, with love

Contents

List of maps List of tables Preface Acknowledgments List of abbreviations 1. The study of partially restructured vernaculars 1.1 1.2 1.3 1.4 1.5 1.6 1.7

Introduction Partial restructuring versus decreolization The study of African American English (AAE) The study of Afrikaans The study of Brazilian Vernacular Portuguese (BVP) The study of Nonstandard Caribbean Spanish (NSCS) The study of Vernacular Lects of R´eunionnais French (VLRF) The comparison of partially restructured vernaculars

2. Social factors in partial restructuring 2.1 2.2 2.3 2.4 2.5 2.6

Introduction AAE: the sociolinguistic setting of its development Afrikaans: the sociolinguistic setting of its development BVP: the sociolinguistic setting of its development NSCS: the sociolinguistic setting of its development VLRF: the sociolinguistic setting of its development Common sociolinguistic factors

3. The verb phrase 3.1 3.2 3.3 3.4 3.5 3.6

Introduction The AAE verb phrase The Afrikaans verb phrase The BVP verb phrase The NSCS verb phrase The VLRF verb phrase A comparison of the verb phrase

page ix x xi xvii xix 1 1 4 10 13 15 17 19 21

24 24 29 41 47 60 65 70

72 72 73 77 80 83 85 90

vii

viii

Languages in Contact

4. The noun phrase 4.1 4.2 4.3 4.4 4.5 4.6

Introduction The AAE noun phrase The Afrikaans noun phrase The BVP noun phrase The NSCS noun phrase The VLRF noun phrase A comparison of the noun phrase

5. The structure of clauses 5.1 5.2 5.3 5.4 5.5 5.6

Introduction The structure of AAE clauses The structure of Afrikaans clauses The structure of BVP clauses The structure of NSCS clauses The structure of VLRF clauses A comparison of clause structure

6. Conclusions 6.1 6.2 6.3 6.4

Introduction Social factors in partial restructuring Linguistic factors in partial restructuring Linguistic processes in partial restructuring Comparative studies in restructuring

References Index

92 92 93 96 101 105 108 112

116 116 116 120 123 127 129 133

135 135 135 137 142 144

147 166

Maps

1 United States: ‘Coloured Population, 1900.’ Originally published in Darby and Fuller (1978), The New Cambridge Modern History Atlas (Cambridge University Press). page 31 2 Southern Africa. Originally published in Holm (1989), Pidgins and Creoles Volume 2 (Cambridge University Press). 43 3 ‘South America, 1830–1956.’ Originally published in Darby and Fuller (1978), The New Cambridge Modern History Atlas (Cambridge University Press). 49 4 ‘Localizacion ´ aproximada de a´ reas afro-hispanas en Am´erica.’ Originally published in Perl and Schwegler (eds.) (1998), Am´erica Negra: panor´amica actual de los estudios ling¨u´ısticos sobre variedades hispanas, portuguesas y criollas (Iberoamericana). Reproduced by kind permission of Iberoamericana. 61 5 (a) ‘Carte de l’ˆıle de la R´eunion.’ R´eunion Ethnic groups. Originally appeared in the Chaudenson (1998), Cr´eolisation du fran¸cais et francisation du cr´eole: Le cas de Saint-Barth´elemy et de la R´eunion (previously unpublished paper presented at the Conference on Degrees of Restructuring in Creole Languages, University of Regensburg, June 24–27, 1998). Reproduced by kind permission of Robert Chaudenson. (b) ‘R´eseau ALR´e.’ R´eunion Lects. Originally appeared in the Chaudenson (1998), Cr´eolisation du fran¸cais et francisation du cr´eole: Le cas de Saint-Barth´elemy et de la R´eunion (previously unpublished paper presented at the Conference on Degrees of Restructuring in Creole Languages, University of Regensburg, June 24–27, 1998). Reproduced by kind permission of Robert Chaudenson. 66

ix

Tables

1 Degree of restructuring and social factors 2 Estimated population of (East) Virginia, 1685–1790 3 Estimated proportion of blacks in American colonies in 1750 4 Estimated population of South Carolina, 1685–1775 5 Estimated population of the Cape Colony, 1658–1798 6 Estimated population of Brazil, 1538–1890 7 Estimated proportion of whites in the Greater Antilles in the late eighteenth century 8 Estimated proportion of whites in various societies in the late eighteenth century 9 Key morphosyntactic features in partial restructuring

x

page 26 32 33 35 45 50 64 71 138

Preface

The naming of books is a tricky business, and it has consequences. In the case of this book – a comparison of five languages with a surprising commonality in structure and social history, and an account of the linguistic processes that formed them – the first thing that may strike anyone who knows my work is not so much the title, as what the title is not. Since the late 1980s, I have used the term “semi-creolization” for the process I discuss here. This book, however, is not called “semicreolization”. The change is in part strategic: I wish to make an argument which is, I think, original about the nature of these language varieties, which all raise current and important issues of politics and culture. I am not interested in exercising an imperial right to label, especially if it obscures discussion about the issues raised by this book. Creole language studies was one of the first post-colonial disciplines: Reinecke was surely far ahead of his times in seeing creoles from the perspective of their speakers rather than those who sought to be their speakers’ imperial masters. I am very aware that the social sciences’ long demand to label at will is always problematical. I also know that the use of a new or unfamiliar term in an established field is an irritant to others who, perhaps, have a word they like better or even a certain resistance to thinking again about issues of taxonomy. “Semi-creolization” has been the traditional term for what I am about to discuss since Reinecke first referred to Afrikaans as “semi-creolized” (1937:559). As used among linguists, this term has the advantage of a certain transparency: its first morpheme clearly means “half ” or “partly” in reference to the process of creolization. The problems that have emerged in connection with this term have entirely to do with the second morpheme. Creole linguistics is part of sociolinguistics, which is the most politically charged area of language study. At the very least, creoles involve the contact of different cultures and thus cultural frontiers and cultural politics. Usually the social setting of creolization has been that of enslavement, xi

xii

Languages in Contact

with all the racism and brute violation of human dignity on which slavery depended. Nobody who studies creole languages can be indifferent to all the resonances of their work, particularly among the people who speak the languages we study. Mufwene (2000c:67) has written that linguists’ self-licence to go around the world baptizing some vernaculars “creoles”, when in some cases their speakers do not even know the word creole, let alone how it is used in linguistics, is . . . the disenfranchising act by which some vernaculars are marginalized . . .

The problem with this honorable concern is that all social scientists, to some degree, work by setting the reality of other people’s lives within a framework and a vocabulary that is particular to a discipline; it is inevitably a question of power, but it is also how analysis becomes possible. To refuse that task is to refuse the science. Not every Englishman is glad to know he speaks a “Germanic” language; but the classification has its uses, even if it reflects the German origins of historical philology. Not everyone knows what a “clitic” and a “copula” are, but these are terms of art which linguists need to analyze language. And the disenfranchising of a language, and of its speakers, may have more to do with economic and political issues than the choices of labels made by social scientists far away. Still, Kautsch and Schneider are certainly correct in their assertion that “Terms such as ‘semi-creoles’ . . . suggest certain undesired associations and are found inappropriate by many” (2000:247). Linguists do indeed have a duty to be sensitive to such questions. But these questions may be exceedingly complex, and relative. Many creole-speaking areas are now independent countries where creole-speakers themselves control the media, and to that extent the local use of words; most of them seem to find the term creole to have positive connotations. Some do not, particularly in some areas where partially restructured languages are spoken. Few words are truly neutral. The term Geechee (sometimes used in reference to Gullah) can have undesirable connotations in African American English, i.e. “a Gullah or any Black person whose speech is peculiar or unintelligible to the hearer” (Folb 1980); “She was a big, burly-looking, dark type sort of girl, a real geechy-looking girl” (quoted by Labov 1972a:390). In fact, the American Dialect Dictionary defines Geechee as “a negro from the islands, as from the Bahamas” (Wentworth 1944), while for Bahamians, “A Geechee is what you could call a Merican who’s work field” (Holm with Shilling 1982). And therein lies the analogy: a descriptive word will carry the burden of attitudes to those it describes, and come in time to be identified with those attitudes. Whatever terms linguists may

Preface

xiii

use, many African Americans and Brazilians find semi-creole or semi-crioulo unacceptable in reference to their own speech – while many R´eunionnais find semi-cr´eole unacceptable because it is not creole enough. For all these reasons, I have decided to refer to semi-creoles as “partially restructured languages,” although I will continue to refer to creoles as such, rather than as “fully restructured languages.” My logic is that linguists still need to distinguish creoles from other fully restructured languages, such as those that are intertwined (partially intertwined languages having never, to my knowledge, been reported). There may be some loss of precision here, since I am not convinced that the processes which produced the languages discussed in this volume are in all points the same as those which produced partially restructured language varieties such as Irish English (Hickey forthcoming, pace Winford 2000:216). It is also true that the languages examined here share at least part of their substrate with the Atlantic creoles, which means that they also share to varying degrees a broader cultural heritage. The more neutral term does not convey this cultural difference. But it may be that this cultural and historical difference is why some find the term semi-creole unacceptable. I believe that my choice of terms is not euphemistic; that is, it is not an attempt to deny historical realities, like the social position of slaves, in order to make the world seem more tasteful or untroubling than it actually is. Thus I will continue to use one term with which linguists have developed the most surprising problems: the world “creole” itself. “Creole” is a term of art for linguists, as is “creolization”; we try to give it a precise meaning in order to analyze and understand a particular kind of language contact and change. It must clearly be something beyond the obviously circular definition: “the process which produces a creole language.” We think we mean something like McWhorter’s pithy definition: “radical reduction by non-natives followed by reconstitution into a natural language” (2000:115), but the problems begin as soon as we try to explain how and why this happens. It does sometimes seem that we are more interested in debating the general processes even before we have provided a proper account of the specifics involved, but this failing may be understandable if not forgivable: there are so many creole languages and they appear to have been produced by such a wide range of social and linguistic circumstances that it often seems unclear precisely what a creole language is, much less what the processes are that produce such a language. This has led to a singularly unhelpful confusion. Parkvall (2002:362) has written that “for most creolists, there is nothing at all that sets creoles apart from non-creoles.” If this is so, it has to do with the belief that

xiv

Languages in Contact

the typological similarity that creole languages may or may not bear to one another is analogous to the typological similarity of languages whose structural affinities can be attributed to a family resemblance – that is, the similarities found between genetically related families of languages such as the Germanic or the Bantu languages. The confusion undermines the fundamental and useful meaning historical linguists give to the genetic relationship between languages when it comes to discussion of the relationship of creoles to their superstrates. Thomason and Kaufman (1988) maintain that because creoles (and the pidgins from which they developed) were not passed from one generation of speakers to the next by normal language transmission, they are not genetically related to the languages spoken by their creators – in the technical sense used in historical linguistics, which requires systematic correspondences between languages not only in lexicon but also in structure. This position has been attacked by Mufwene (2000a:9–11) as “disenfranchising” the creoles, and by DeGraff (2001) for implying that creoles are “abnormal.” Thomason (2002) responds that it would nonetheless be unwise to abandon the traditional definition of genetic relatedness: it is the central concept of the comparative method, which has, for well over a century, enabled historical linguistics to accumulate a vast and reliable body of knowledge about language change. Perhaps, before attempting to dismantle a science, we should have reasons more substantial than the possibility that others will misconstrue its technical terms. In any case, the category to which creole languages belong is sociolinguistic rather than typological in the genetic sense. To indulge again in the tautology, creoles are creoles because they have undergone a sociolinguistic process called creolization, not because they “attest the features that define creoles as creoles” (Markey 1982:170). Because the study of Atlantic creoles has dominated the field, and because these varieties happen to share genetically related Indo-European superstrate and Niger-Congo substrate languages as a historical result of the Atlantic slave trade, it has perhaps come to seem that certain syntactic features found in most of the basilectal Atlantic creoles are “creole features” in a sense compatible with Markey’s. But this is not the case. Creolization is a sociolinguistic process: there is widespread consensus that its defining characteristics must include social as well as linguistic phenomena (Holm 2000a:68–71). Because it is a sociolinguistic process, like language attrition or code switching, there is no logical reason to expect that languages which have undergone creolization necessarily bear any more typological similarity to one another than languages that have not. The nub of the problem seems to be confusion over what “typology”

Preface

xv

means in linguistics. Genetically dissimilar languages can, by coincidence, be typologically similar (i.e. isolating, synthetic, etc.). In respect to its paucity of inflectional endings, English is typologically closer to Chinese than it is to Latin (Crystal 1986:319) despite the fact that genetically it is closer to Latin. On the other hand, certain linguistic processes can also result in closer typological similarity: two genetically unrelated languages undergoing drift can eventually emerge with a similar degree of loss of inflectional morphology. There is, indeed, abundant evidence that genetically unrelated languages that are creolized undergo an extensive if not complete loss of inflectional morphology. Building on Chaudenson’s (1992) view of creoles as simply varieties of their superstrate languages – approximations of approximations of the earlier regional dialects spoken by European colonists – Mufwene (1994, 1996, 1997) concluded that creole is not a valid term for classifying languages, and therefore one language cannot be said to be more or less “creole” than another. McWhorter (1998) countered that creoles are indeed synchronically distinguishable from non-creole languages in that they combine all three of the following traits: they have little or no inflectional affixation; they make little or no use of tone lexically to contrast monosyllables or encode syntax; and they have semantically regular derivational affixation. This creole prototype, graded like most phenomena, is “the direct result of severely interrupted transmission of a lexifier, at too recent a date for the traits to have been undone by diachronic change” (1998:812). Later, McWhorter (2000) specified that the gradience in the proximity of creoles to the synchronic prototype that he had proposed could be accounted for by sociohistorical factors such as the continuing presence of the lexifier language. This reaffirmation of the validity of the concept of “creole” has nonetheless been challenged by the claim that the differences between creoles and other languages, such as their lexifiers, are in fact the result of differences in parameter settings, which determine various interrelated morphosyntactic configurations in current generative theory. For example, historical syntacticians such as Roberts (1999) have argued that in languages with highly inflected verbs, the verb tends to move to I, the locus of tense and agreement, whereas in languages with verbs having few or no inflections (such as creoles) verbs tend to remain in place (but note the strict verb-second rule of Afrikaans, a language with scarcely any verbal inflections, section 3.2.1). McWhorter (2002) further argues for the validity of creoles as a delineable syntactic class with evidence that “the syntax-internal model . . . is, by itself, inadequate to explain creole genesis” (2002:31). Distinguishing between features that are universal to all languages for cognitive reasons

xvi

Languages in Contact

(e.g. definite/indefinite opposition) and those which are “ornamental” (e.g. grammatical gender), he points out a number of differences between creoles and their source languages that cannot be explained by the syntaxinternal hypothesis. This is “because much of what distinguishes creoles from older grammars is that, as recent descendants of pidgins, their grammars are less accreted with such ‘grammaticalizational overkill’ ” of highly particular, non-universal features (2002:9). Whether or not McWhorter’s characterization of creoles as a delineable synchronic class of languages is supported by future research, I must point out here it will still be the case that creoles differ from languages that did not result from contact: creoles develop much more rapidly. One of the earliest creolists, Van Name, recognized this: The changes which [creoles] have passed through are not essentially different in kind, and hardly greater in extent than those, for instance, which separate the French from the Latin, but from the greater violence of the forces at work they have been far more rapid . . . here two or three generations have sufficed for a complete transformation. (Van Name 1869–70:123)

For example, there could have been no Caribbean Creole French before the French brought their first slaves to St. Kitts in 1626 and then captured Martinique and Guadeloupe in 1635. Yet we find the existence of the language documented by a text dated 1671 containing linguistic features unequivocally identifying it with the modern creole of the latter islands (Carden et al. 1991). The structural gap between French and Creole French is at least as great as that between French and its source language, vulgar Latin: it is the difference in the speed of formation – a lifetime instead of a millennium – that is justification enough for distinguishing creoles from languages whose genesis was not induced by contact. All this confirms a simple point which has direct relevance here: the concept of a “creole” remains an essential point of reference in our understanding of language contact and its relation to language change. Without it, we have a less accurate, less truthful, view of language.

Acknowledgments

My greatest debt is to my former doctoral students at the City University of New York (CUNY) for the insights on language restructuring that I gained from their research. A number of talented young linguists participated in the seminars I organized at CUNY on partially restructured languages, some of whom were native speakers of the varieties they described or the source languages involved. These seminars were extraordinarily productive, leading to publications on American Indian English (Craig 1991), Dominican Spanish (Lorenzino 1993, 1998), Surinamese Dutch (Healy 1993), and Afrikaans (Slomanson 1993; de Kleine 1997). Others wrote relevant dissertations on Nagamese (Bhattacharjya 2000), Dominican Spanish (Green 1997), Brazilian Vernacular Portuguese (Mello 1997), and the Vernacular Lects of R´eunion French (Chapuis forthcoming). I am happy to have this opportunity to acknowledge my gratitude to them. I would also like to thank the following linguists for help and advice while I was working on this volume: Alan Baxter, Jeanette B´eraha, Hans den Besten, Ernesto d’Andrade, Raymond Hickey, Liliana Inverno, Alexander Kautzsch, Alain Kihm, Ernst Kotz´e, Dante Lucchesi, John McWhorter, John Rickford, Paul Roberge, Edgar Schneider, Armin Schwegler, Dominika Swolkien, Sarah Grey Thomason, Jean-Philippe Watbled, and Walt Wolfram. I am also grateful to my editors at Cambridge University Press, particularly Andrew Winnard and Helen Barton, for their excellent support. Of course the present volume builds on my earlier work, and I would like to thank once again those who helped me with that. I am particularly grateful to Michael Pye for unearthing the quilombo song (section 2.3.2), possibly one of the earliest examples of restructured Brazilian Portuguese. However, responsibility for the errors, omissions, and other shortcomings in this book is mine alone. Finally, it is a pleasure to thank the following institutions for travel grants that allowed me not only to exchange ideas at international meetings, but also to do fieldwork and library research relating to language re´ structuring: from the Comit´e International des Etudes Cr´eoles to present xvii

xviii

Languages in Contact

a paper at a conference on R´eunion (2002); from the University of Puerto Rico to give a mini-course in San Juan on restructuring (2001); from the Linguistic Society of Southern Africa to present a paper at a conference at the University of Cape Town (2000); from the University of Coimbra to present a paper at a round table discussion of the sociohistorical origins of Brazilian Vernacular Portuguese in Recife (2000); from the Associa¸ca˜ o Brasileira de Lingu´ ¨ ıstica to teach a course on language restructuring at their XIV Instituto Lingu´ ¨ ıstico at the Universidade Federal de Santa Catarina in Florianopolis ´ (1999); from the Deutsche Forschungsgemeinschaft to present a paper at the International Symposium on Degrees of Restructuring in Creole Languages at the University of Regensburg (1998); from the Colombian government to present a paper at a colloquium in Cartagena (1996); and from Hunter College of the City University of New York to present a paper with Gerardo Lorenzino at a conference at the Universidad de Oriente, Santiago de Cuba (1992).

Abbreviations

A, AFK AAE AAVE ANT ART ASM AUX BVP CD CE CF CLP CP CS CV CVC D DARE DEM DET DO E EP F HP IO INFL LPV KA NEG

(standard) Afrikaans African American English African American Vernacular English (term used in some quotations) anterior tense article agentive subject marker auxiliary Brazilian Vernacular Portuguese Creole Dutch Creole English Creole French copula-like particle Creole Portuguese Creole Spanish consonant-vowel consonant-vowel-consonant (standard) Dutch Dictionary of American Regional English (see Cassidy and Hall, [eds.]) demonstrative determiner direct object (standard) English European Portuguese (standard) French Helv´ecia Portuguese indirect object inflection L´ıngua dos Pretos Velhos (see section 1.4) Kaaps Afrikaans negator xix

xx

NP NSCS OBJ ORA P PAST PERF POSS PRES PS QW REL S SB SBP SOV SVO SWVE V VLRF VOS VP 1s 3p {}

Languages in Contact

noun phrase Nonstandard Caribbean Spanish object (marker) Orange River Afrikaans (standard) Portuguese past tense perfective aspect possessive present tense Palenquero Spanish (section 1.5) question word relativizer (standard) Spanish; subject Substrate Standard Brazilian Portuguese subject-object-verb word order subject-verb-object word order Southern White Vernacular English verb vernacular lects of R´eunion French verb-object-subject word order verb phrase first person singular, etc. third person plural, etc. boundaries of an embedded clause

Note: the varying orthography used in this volume for VLRF and some other language varieties is in each case that of the author cited.

1

The study of partially restructured vernaculars

Introduction Language is a kind of social behavior, one of the many ways in which individuals interact with those around them. Thus linguistics is a social science, and linguists take pride in thinking of themselves as scientists, with all the objectivity that word denotes. Unfortunately, objectivity is very hard to achieve, especially in the social sciences, and linguistics is no exception. It is hard to imagine any study of language which manages to put away all ideology, but in the case of the languages discussed in this book, the task is unimaginable. African American English – also called AAE, Ebonics, or just Black English – is a good case in point. Until at least the middle of the twentieth century, the overwhelming majority of white Americans saw their country and its culture as the product of their European roots flourishing in a new land. This ideology allowed very little room for the contribution of other cultures, so that even the distinctiveness of the folk ways and speech of African Americans was attributed to their frequent lack of access to education and general ignorance – if not to their very intelligence. Thus well into the 1950s Negro Nonstandard English (as AAE was then called) was usually considered bad English in need of eradication rather than study. In so far as its origins were considered at all, it was assumed to have descended solely from British dialects that had been left untended in America. In the 1960s the civil rights movement sharply changed this ideology: equal citizens could not logically be unequal human beings, and there was a new willingness to reconsider African Americans, as well as the development of their language and culture in the United States. By the 1970s there was widespread agreement – at least among linguists – that the distinctive features of AAE identified it as a post-creole: the descendant of a variety of English that had first been creolized or restructured when it was learned by adult African slaves on plantations (as English had been creolized in Jamaica, for example). Subsequently this speech 1

2

Languages in Contact

underwent decreolization, or the loss of many of its distinctive creole features through contact with standard English. Dillard’s influential book, Black English (1972) popularized this view, convincing many that AAE, like its speakers, was much more African than anyone had realized. This was part of another growing ideology, supported by many blacks, that affirmed a very separate cultural identity for African Americans. But there were problems in explaining AAE as a post-creole. Most importantly, no one could find reliable historical evidence of the widespread, stable creole from which AAE had supposedly decreolized. The known passages purporting to represent the speech of blacks in North America during the eighteenth and nineteenth centuries remain ambiguous. Quite aside from the inherent problem of the authenticity of such fragments, almost all of which were actually written by speakers of standard English, there is an even greater problem in accurately identifying the kind of speech represented. Unless the purported speaker’s background is documented, it is impossible to determine whether it represents the foreigner’s English of Africans, the Caribbean Creole English of slaves imported from the West Indies, a pidginized variety of English from West Africa, or an indigenous creole such as Gullah, the fully restructured variety spoken along the coast of South Carolina and Georgia. Up to this point linguists had generally assumed that decreolization could account for the varying distance between the grammatical structure of different creoles and that of the European language they were based on: AAE and Caribbean creoles based on English were viewed as postcreoles at different stages of decreolization away from some very early fully creolized variety. By the mid-1980s there were growing objections to this all-or-nothing model of creolization and skepticism that it could account for what was becoming known about the earlier structure of AAE (Hancock 1987:264–265; Schneider 1989; Holm 1991:247). Much of the most recent debate focuses on the nature the language of blacks born in North America (outside of the creole-speaking Gullah area): whether it was from its very beginning a fully restructured creole or rather a compromise between the pidgin or creole brought in by slaves from the West Indies and Africa and the regional speech of British settlers (Winford 1997; Rickford 1997, 1999), and whether partial restructuring can account for the known sociohistorical and linguistic facts concerning AAE and some other languages that apparently had a similar genesis, such as nonstandard Brazilian Portuguese and Caribbean Spanish, Afrikaans (the South African language descended from Dutch), and the vernacular French spoken on the island of R´eunion in the Indian Ocean (Holm 1992, 2000). These language varieties, which appear to have grown out of the partial restructuring of older varieties spoken in Europe that came into contact

The study of partially restructured vernaculars

3

with non-European languages, today have some 200 million speakers, placing them among the major languages of the world. They present formidable challenges not only to linguistic theory but also in practical matters like the language-related problems encountered in education by speakers of nonstandardized varieties, which include all of the language varieties discussed here, except for standard Afrikaans. These problems have shown no signs of going away. And each of these languages has been studied through the prism of particular, often local, ideologies, as Heliana Mello has shown for her own language, Brazilian Vernacular Portuguese (Mello 2001). Of course the concept of a partially restructured language as opposed to a post-creole (which was fully restructured but then decreolized through contact with its lexical source language) has its own ideological implications. If the restructuring of the English spoken by blacks in most of North America was only partial, this implies that the transmission of the English language (and, indeed, other aspects of English culture) to African Americans was much more complete than it has been fashionable to assume. The cultural separatism of the 1960s and 1970s may have distorted the issue by insisting on the Africanness of African Americans to the virtual exclusion of their Europeanness. These languages, then, would require new study if only because our sense of identity and ideology shift with time. But there is a more pressing scientific reason for reassessing them. The genesis and development of such partially restructured languages have become one of the most important leading edges of contact linguistics as a whole. The languages discussed here have a number of the structural features of creoles but appear, nonetheless, never to have undergone full creolization. Their reduced inflectional morphology – particularly in the verb phrase and noun phrase – seems to have been transmitted from one generation to another largely like that of unrestructured overseas varieties, rather than having been reacquired by more basilectal varieties during decreolization, which distinguishes them from post-creoles. Some of the most interesting research in this area has been the effort to correlate the synchronic structure of these languages to the sociolinguistic history of their speakers: the demographic balance of native versus non-native speakers of the target language at the beginning of the speech community’s settlement, their relative power, their migrations, and the nature of their contact. There has also been a shift in theoretical perspective that is facilitating progress in this area of inquiry. More of us working in pidgin and creole linguistics are coming to see our field as only one part of a broader area of research: contact linguistics, as defined by Thomason (1997). The scope of this wider field includes language varieties that have resulted not only from pidginization and creolization (to whatever degree) but also from

4

Languages in Contact

such processes as intertwining (Bakker and Muysken 1994), koineization, or indigenization (Siegel 1997). Such studies promise to increase our understanding of the range of possible outcomes of language contact by encompassing varieties that fail to fit neatly into the definitional boxes in which we have often tried to restrict pidgin and creole linguistics. In addition to the five partially restructured varieties mentioned above, which have received considerable scholarly attention, there are a number of less well-studied varieties that seem likely to have undergone a similar process, such as the nonstandard English of American Indians, Australian Aborigines, and others. There are also partially restructured varieties which appear to have evolved solely through community-wide language shift, such as Irish English. Whether these are indeed the same kind of language, which is the position of Winford (2000:216), has yet to be demonstrated. Specialists in Irish English such as Hickey (forthcoming) are not convinced (see section 2.1.1). This chapter examines how scholarship on each of these five varieties – based on five different European languages – has taken its own course, the literature on each being largely in the corresponding standard language. Although language barriers are still surprisingly effective in limiting the horizons of linguists, there has been a certain amount of communication across these barriers so that research on one variety has sometimes cast light on theoretical problems connected with another. After surveying general views on full and partial restructuring from the earliest creolists until the 1980s (section 1.1), this chapter examines scholarship on each variety, beginning with AAE (1.2). To a limited extent (especially in more recent years) AAE studies have provided models for interpreting the historical development of the other varieties, from (a) the model of a purely European dialect reflecting general Western European tendencies such as the loss of inflections; to (b) the model of a post-creole retaining substratal features; to (c) the model of differing degrees of restructuring, varying according to social factors. This review of the theoretical underpinnings of research on AAE will then be compared with that of work on the other four varieties: Afrikaans (1.3); Brazilian Vernacular Portuguese (1.4); Nonstandard Caribbean Spanish (1.5); and the Vernacular Lects of R´eunionnais French (1.6). The final section (1.7) describes recent comparative research in partial restructuring. 1.1

Partial restructuring versus decreolization

The theoretical foundations for the study of fully creolized languages have been developing since the eighteenth century – particularly since the middle of the twentieth century (Holm 1988–89:13–70). However, linguists

The study of partially restructured vernaculars

5

have had more difficulty developing an adequate theoretical model for dealing with partially restructured languages – one that would allow reliable predictions about the interrelationship between the social history of their speakers and the linguistic structure likely to emerge from a particular context. We have long known that fully creolized languages exist – languages whose linguistic structure differs radically from that of the older languages from which they drew most of their lexicon. For example, the generally synthetic structure of the Western European languages used by colonists (Portuguese, Spanish, French, Dutch, and English – which still use a number of inflections to convey grammatical information) was replaced by an analytical structure in the Atlantic creoles derived both from these European languages and from the isolating Niger-Congo languages spoken by Africans brought to the New World. There is fairly general agreement that the isolating structure of the creoles – using free rather than bound morphemes to convey grammatical information – was determined by several factors: (1) the tendency towards isolating structures that was already widespread in the European superstrate languages; (2) the almost categorical use of isolating structures in the African substrate languages; (3) the universal tendency of adults to use isolating structures when learning a second language (e.g. the pidgins that developed into creoles); (4) the internal systematicity that would have spread the use of isolating structures as the creoles developed; and (5) the converging influence of two or more of these tendencies. Of all the structural similarities of the Atlantic creoles, the common trait that indicates most clearly the completeness of their restructuring is the completeness of their analyticity. If we leave aside the non-Atlantic creoles (which have not been compared as systematically), we find that basilectal creoles – those closest to their earliest form – seem to have very few true inflections, and that varieties that do have true inflections seem not to be the same kind of language as basilectal Atlantic creoles (Holm 1989). The existence of fully restructured creoles (whatever they may have been called) has been acknowledged since the early eighteenth century, and references to what can only be interpreted as more and less fully restructured Caribbean varieties date from the latter part of that century: die creolische, oder Negersprache . . . wird aber von den blanken Creolen feiner gesprochen, als von den Negern. [. . . the creole, or language of the blacks . . . is spoken better by the white Creoles than the blacks.] (Oldendorp 1777:263, quoted by Stein 1984:92)

(Of course feiner, translated as ‘better,’ here means more like the European source language.)

6

Languages in Contact

However, it was not until the late nineteenth century that a linguist observed that there were language varieties that combined features of creoles with those of non-creoles. Schuchardt (1889:480) coined the term Halbkreolisch (literally ‘half-creole’) for certain varieties of Indo-Portuguese and Caribbean Creole French that had taken on superstrate features: Ueberall wo eine kreolische Mundart gesprochen wird, liegt Mischung mit der europ¨aischen Muttersprache sehr nahe, mit anderen Worten: es stellt sich leicht ein individuelles oder gelegentliches Halbkreolisch ein. Indem wir die Bedingungen fur ¨ dasselbe n¨aher suchen, bemerken wir einerseits dass Europ¨aer die des Kreolischen nicht wirklich m¨achtig sind, sondern nur dunkle Vorstellungen davon haben, such bemuhen ¨ von den Einheimischen verstanden zu werden – kreolisiertes Europ¨aisch; anderseits dass Europ¨aer die des Kreolischen mehr oder weniger m¨achtig sind, irgend eine Form der Darstellung w¨ahlen, fur ¨ welche das Kreolische nicht ausreicht, oder dass Kreolen die des Europ¨aischen nicht m¨achtig sind, ihren sprachlichen Ausdruck zu verfeinern sich bemuhen ¨ – europ¨aisiertes Kreolisch. (Wherever a creole dialect is spoken, mixture with the European mother tongue lies very close at hand; in other words, an individual or occasional semi-creole easily appears. When we look more closely into the underlying conditions, we see on the one hand that [1] Europeans who do not really know the creole, having only a confused notion of it, may strive to make themselves understood by the natives, producing a creole-influenced variety of the European language. On the other hand, [2] Europeans who are more or less at home in the creole may use constructions not found in it, or [3] Creoles who have not mastered the European language may attempt to refine their creole, producing a European-influenced creole).

The first situation produces a variety similar to what Muhlh¨ ¨ ausler (1982:456–457) calls Tok Masta; the second situation produces a variety like the lects of Negerhollands and Papiamentu spoken by Europeans; the third produces what are now called decreolized varieties. Schuchardt’s idea of Halbkreolisch was interpreted by Tagliavini (1931:834) as a language that was half-way in the process of being creolized, and so he translated the term into Italian as “lingue creolizzanti.” Unfortunately the present-participial ending might suggest that such languages are “creolizing” in the sense of still undergoing restructuring; Reinecke (1937:22) translated the term as “those tending toward the creole, the creolisant dialects.” Schuchardt also noted that African American English seemed to be losing its creole features: The Negro English that is most widely known is spoken in the southern United States . . . those variants which still show a creole-like character are increasingly falling into disuse by being accommodated to the English of the whites by means of an intermediate speech variety. (Schuchardt c. 1893, in Gilbert 1985:42)

The study of partially restructured vernaculars

7

In this view, African American English originated as a full creole that later acquired non-creole features from contact with regional English. Later Bloomfield (1933:474) reasoned that a restructured variety of English had become nativized among Negro slaves in many parts of America. When the jargon has become the only language of the subject group, it is a creolized language. The creolized language has the status of an inferior dialect of the masters’ speech. It is subject to constant leveling-out and improvement in the direction of the latter. The various types of “Negro dialect” which we observe in the United States show us some of the last stages of this leveling. With the improvement of social conditions, this leveling is accelerated; the result is a caste-dialect . . . It is a question whether during this period the dialect that is being de-creolized may not influence the speech of the community – whether the creolized English of the southern slaves, for instance, may not have influenced local types of sub-standard or even of standard English.

This view was not elaborated into a full-blown theory of decreolization until interest in AAE and the English-based Caribbean creoles became widespread in the 1960s. Stewart asserted that the non-standard speech of present-day American Negroes still seems to exhibit structural traces of a creole predecessor. . . . One of the more important changes which have occurred in American Negro dialects during the past century has been the almost complete de-creolization of both their functional and lexical vocabulary. (1968:51–52)

DeCamp (1961, 1971) developed the idea of a continuum of lects for Jamaican, ranging from the most creole-like to the most English-like. Stewart (1965) applied this idea to African American English, introducing the terms acrolect for the variety closest to the standard and basilect for the variety furthest from it, with mesolect for those between. Later the continuum model was further refined by others (e.g. Bickerton 1973, Rickford 1987). By the end of the 1970s there was a general assumption that decreolization explained the varying structural distance between different creoles and their lexical source language: Caribbean creoles based on English, for example, were actually post-creoles at different stages of decreolization away from a very early fully creolized variety that may have resembled the modern Surinamese creoles, which were cut off from contact with English in the seventeenth century. The idea behind the modern meaning of partial restructuring originated in Hesseling (1897), who pointed out that “the Dutch on the Cape was on the way to becoming a sort of creole . . . [but] this process was not completed” (1979 translation, p. 12). Shortly afterwards, Vasconcellos noted that

8

Languages in Contact

les Portugais ont e´ t´e oblig´es d’apprendre quelquefois les langues indig`enes, et les indig`enes d’apprendre la langue du Portugal. Le second fait est le seul qui m’int´eresse pour le moment, parce qu’il en est r´esult´e la formation des dialectes cr´eoles, et d’autres vari´et´es du portugais; entre les uns et les autres, on peut admettre des degr´es (1901:157–158; my emphasis) [The Portuguese were sometimes obliged to learn the indigenous languages and the indigenous people Portuguese. The second fact is my only interest for the time being because it resulted in the formation of creole dialects and other Portuguese varieties. Between the two groups, one could say there is a question of degree.]

The first recognition of a whole category of such languages can be found in Reinecke (1937:61): In several instances the slaves were so situated among a majority or a large minority of whites (and there were other reasons as well for the result), that they, or rather their creole children, learned the common language, not a creole dialect; or the plantation creole dialects that had begun to form never crystallized, never got beyond the makeshift stage. This happened in . . . Brazil, Cuba and the Spanishspeaking Caribbean countries in general, and in the southern United States in general.

Reinecke was also the first to put this meaning together with the term semi-creolized, which he used in reference to Afrikaans (1937:559). He also pointed out that the English-based creoles of the Caribbean did not seem to have been completely restructured: The Surinam dialects, like West African Pidgin English, are unmistakably creole dialects in the sense of being simplified to a purely analytic structure. The other West Indian dialects are not, however, so completely pruned down [. . . and] may be regarded as what Schuchardt called creolizing languages – dialects on the way to complete analytic simplification, but which for various reasons stopped a little short of it. (1937:274–275)

As recently as l962, Stewart considered Suriname to have the only real creoles based on English in the Caribbean area: “Jamaican and other regional varieties of English are best treated as dialects of English” (1962:50–51). In a personal communication, Stewart explained that at the time it seemed more prudent to exclude these varieties from the discussion of creoles since it was unclear whether they were creoles that had acquired non-creole features or vice versa. By 1967, however, he felt confident that additional historical sociolinguistic information had made it clear that the West Indian varieties were in fact post-creoles. However, the fact that this view came to be widely accepted among creolists does not in itself prove that Reinecke had not been right – that these varieties had never been as fully creolized as the Surinamese varieties. An additional possibility that could explain the considerable structural gap

The study of partially restructured vernaculars

9

between the Surinamese and West Indian varieties of creolized English is that Sranan may have been repidginized in the late seventeenth century, leaving it even further from English than it had been prior to 1667. Bloomfield (1933) had indirectly implied that a non-creole language might take on creole features (a process that could lead to partial restructuring) when he asked, “whether the creolized English of the southern slaves, for instance, may not have influenced local types of sub-standard or even of standard English” (1933:474). Later Silva Neto (1950a:12) followed Schuchardt (1889) in referring to re-lusitanized Indo-Portuguese as a semi-crioulo. That same year he extended the use of the term to the Portuguese spoken by non-whites during the early settlement of Brazil: constituiu-se, no primeiro s´eculo da coloniza¸ca˜ o (1532–1632), na boca ˆ de ´ındios, negros e mesti¸cos, um falar crioulo ou semi-crioulo. [. . . there arose during the first century of colonization (1532–1632) a creole or semi-creole language used by Indians, blacks and people of mixed race] (1950b:166)

Although Silva Neto never spelled out the sequence of social and linguistic events that may have led to the partial restructuring of a language variety from the very beginning of its existence, this possibility struck me as worth exploring when I was working on the same problem of the development of Brazilian Vernacular Portuguese (Holm 1984). Thomason, who was then working on a comparative study of a number of different kinds of languages resulting from contact (Thomason and Kaufman 1988), agreed that it would be useful to reserve the term semi-creole for those varieties that appeared never to have been fully creolized. Therefore I contrasted the term with post-creole varieties such as (according to some) American Black English . . . or vernacular Brazilian Portuguese. . . . Others would call these varieties semi-creoles, which also means that they have both creole and non-creole features but does not necessarily imply that they were ever basilectal creoles, since both creoles and non-creoles (e.g. Caymanian English . . .) can become semi-creoles by borrowing features. Thus some believe that Afrikaans . . . particularly the variety spoken by some people of mixed race . . . could safely be called a semi-creole but not a post-creole (Holm 1988–89:9–10)

The term is also used in this sense by Thomason and Kaufman in reference to Afrikaans (1988:148). Around the same time, Mufwene (1987:99) noted that the results of half-creolization and decreolization may look alike, but the processes responsible for the structural likeness of their outcomes are certainly not the same. Whichever is the case for B[lack] E[nglish] still needs to be demonstrated.

10

Languages in Contact

Bickerton (1984:176–178) proposed what he called a pidginization index to explain why the structure of some creoles is quite close to that of their lexical source language (e.g. R´eunionnais) while that of others is quite far from it (e.g. Saramaccan). Although the mathematical formula which he proposed to indicate the degree of restructuring proved “unworkable” (Singler 1990:645), Bickerton did recognize that creoles stand at different distances from their source languages in terms of the degree of restructuring that they have undergone, and that this differentiation could occur at the beginning rather than the end of the process of restructuring (see the introduction to chapter 2). It was during this period that linguists began to question whether decreolization alone could adequately account for the varying distance of the structure of different creoles from that of their lexical source language. Hancock (1987) put it thus: I do not, then, believe that, for example, Black English was once like Gullah, or that Gullah was once like Jamaican, or that Jamaican was once like Sranan, each a more decreolized version of the other along some kind of mystical continuum. . . . My feeling is that most of the principal characteristics that each creole is now associated with were established during the first twenty-five years or so of the settlement of the region in which it came to be spoken: Black English has always looked much the way it looks now . . . (1987:264–265)

The theoretical importance of gradience in creolization was signaled by a conference on “Degrees of Restructuring in Creole Languages” at the University of Regensburg in Germany in 1998, resulting in an entire volume on this topic (Neumann-Holzschuh and Schneider 2000). 1.2

The study of African American English (AAE)

The decreolization theory for the origin of Black English – the “creolist” theory that finally received the imprimatur of Labov (1982) – was a much more satisfactory explanation for that variety’s creole features than earlier hypotheses that traced its origins solely to British dialects. However, my own work on the lexicon of two much more restructured varieties – Nicaragua’s Miskito Coast Creole English (Holm 1978) and Bahamian Creole English (Holm with Shilling 1982) – made it clear to me that archaic and regional British English must have played a primary role in the genesis of all three African American varieties. Research on possible British origins of specific creole grammatical features had been unfashionable in the 1970s, but in the 1980s two such studies – Schneider 1981 (translated in 1989) and Rickford 1986 – had an important impact on the field, reopening the question of the degree to which British syntactic patterns had been preserved in African American varieties.

The study of partially restructured vernaculars

11

I was further led to question some of the basic assumptions of decreolization theory through work with several non-creoles that appeared to have acquired creole features: White Bahamian English (Holm 1980), and Caymanian and Bay Island English (Washabaugh and Warantz in Holm 1983). I concluded that Although long contact with creolized varieties of English has influenced the English spoken by white Caymanians and their kin on the Bay Islands of Honduras, this influence seems to be confined largely to areal contact phenomena such as word-borrowing and phonological shifts. Considering the English system of verbal inflections in the speech of Utila . . . as opposed to the system of preverbal tense and aspect markers that characterizes C[entral] A[merican] E[nglish] creoles . . . the former would seem to be not a creole but rather a regional variety of English influenced by contact with creolized English, much like the folk-speech of the southern United States. (Holm 1983:15)

In 1986 there began a debate as to whether AAE and white varieties of American English were historically converging (through the decreolization of AAE) or diverging (through AAE’s increasing isolation) as argued by Labov and Harris (1986) (see section 2.1.7). The latter interpretation seemed to support the implausible view of Poplack and Sankoff (1987) that early nineteenth-century AAE had been more similar to white varieties than current AAE is. However, what convinced me that decreolization alone could not account for the present structure of AAE was listening to tape recordings of the speech of former slaves (Bailey et al. 1991, see section 2.1.5). Even taking into account that their speech may have shifted considerably between their childhood in the mid-nineteenth century and the time they were recorded in the 1930s and later, it was clear that what I was hearing was a variety of English with some creole features rather than a variety of creole with some English features. The only honest conclusion that I could reach was that The present study supports the view that the language of the ex-slaves, like earlier attestations of the speech of blacks in the American South, indicates in the light of the relevant sociohistorical and demographic data discussed above that the language of blacks born in North America (outside of the Gullah area) was from its very beginning a semi-creole representing a compromise between the creole of slaves imported from the West Indies and the regional speech of British settlers. While American Black English has certainly undergone decreolization over the past 300 years in the sense that it has replaced many of its original creole features with those of English, this is not actually evidence that American Black English itself ever constituted an autonomous creole system. (Holm 1991:247)

A more radical view (which seems inherently unlikely, given what is known about language contact phenomena) is that the very concept of decreolization is misguided, and that it played no role in the development

12

Languages in Contact

of varieties such as Gullah and AAE, which stand at differing distances from English structurally solely due to their having undergone differing degrees of restructuring. Mufwene (1991:382–383) seems to support such a view. Schneider (1990) re-examined the idea of “creoleness” as a graded phenomenon in reference to varieties of English and English-based creoles in the Caribbean area with a view to casting light on the debate over the creole origin of AAE. He concluded that the question as to whether or not a particular variety is a creole can be very difficult to answer: There is a variety of constitutive factors that contribute independently to the notion, and the label applies to some language varieties better than to others, without implying that the latter are necessarily “non-creoles.” We may distinguish prototypical, or full, creoles that combine all or almost all of these features from varieties that are less typical of the category. Even the notion of semi-creoles does not seem to be very helpful in this dilemma, because its applicability, if not defined too loosely, seems limited, and should not be taken to include the nonprototypical – but nevertheless true – creoles . . . In linguistic matters, more and less are frequently more appropriate responses than yes and no. (1990:105–106)

While Schneider considered the term semi-creole unhelpful because of the limited number of languages it could be applied to (despite the numerical importance of the speakers of partially restructured languages, as discussed above), Kaye dismissed the validity of the very notion with an analogy beyond the reach of logic: There can be no such thing, of course, as partial pidginization or partial creolization (this is why the terms post-creole, semi-creole, and creoloid are imprecise), just as there is no such thing as partial pregnancy. (Kaye 1990:301)

More recent work on AAE has focused increasingly on those sociolinguistic factors which have long been considered relevant to the study of full creoles (e.g. demographic figures suggesting the proportion of native versus non-native speakers during the early period of language contact) but which have not been systematically explored for AAE until now. Winford (1997) traces the social histories of Virginia and the Carolinas, citing early demographic figures from Wood (1989), and compares the key structures in Gullah, AAE, and Southern White Vernacular English, concluding that “AAVE was never itself a creole, but it was created by Africans, and bears the distinctive mark of that creation.” Rickford (1997, 1999) has followed a similar methodology and reached a similar conclusion; Mufwene also suggests that AAE “may simply have resulted from a restructuring which was not as extensive as what produced Gullah” (2001). However, Mufwene (2000b) lends his credibility as a creolist to support the position of Poplack (2000:1) that “. . . the many grammatical

The study of partially restructured vernaculars

13

distinctions between contemporary varieties of AAVE and American and British English are relatively recent developments,” i.e. not the result of earlier contact with restructured varieties of English. Such a position allows for less external influence on the development of AAE than the apartheid-era linguists in South Africa allowed on the development of Afrikaans (section 1.3). Hackert and Holm (1997) have shown that the only hard evidence ever offered for the full creolization of AAE resulted from a historical misinterpretation: the creole nature of the folk speech on the southern Bahamian islands should not be interpreted as evidence that AAVE had been fully creolized on the mainland before 1780 (and later decreolized) since the language that was brought there was in all likelihood eighteenth-century Gullah rather than eighteenth-century AAVE.

1.3

The study of Afrikaans

Afrikaans, derived from Dutch, is spoken by some 6 million South Africans; about half are white and the rest are of mixed ancestry. Afrikaans is unique among the language varieties examined here in that it was standardized and made an official state language. Its exhaustive documentation makes it much easier to contrast its structure to that of its lexical source language, which was actually seventeenth-century regional and nautical varieties of Dutch. Also of particular relevance to tracing the development of Afrikaans are its nonstandard varieties spoken by various groups, particularly those of mixed race with little education. The history of the study of Afrikaans and its origins has been summarized by Reinecke et al. (1975:323ff.) and updated and expanded by Roberge (1994), the sources of much of the following. Hahn (1882) claimed that although Afrikaans is “phonetically Teutonic, it is psychologically an essentially Hottentot idiom. For we learn this patois first from our nurses and ayahs. The young Africander on his solitary farm has no other playmates than the children of the Bastard Hottentot servants of his father, and even the grown-up farmer cannot easily escape the deteriorating effect of his servants’ patois.” Viljoen’s 1896 dissertation, focusing mainly on the phonetic system of Afrikaans, claimed it was derived from the dialects of North Holland. Hesseling (1897) provided the first extended discussion of the origins of Afrikaans. Although he recognized the influence of Hottentot (now called Khoi), he emphasized the influence of the Malayo-Portuguese creole of early Indonesian slaves and claimed that “the Dutch on the Cape was on the way to becoming a sort of creole . . . [but] this process was not completed” because of the continuing influence of metropolitan Dutch (1979 translation, p. 12). This

14

Languages in Contact

characterization set off a debate that continued for a century, at times with considerable heat. Hesseling’s Malayo-Portuguese theory was adopted by the Afrikaner Du Toit (1905) and later developed further by Valkhoff (1966, 1972). It was opposed by the “spontaneous development” theory initially proposed by the Dutch linguist Kruisinga (1906), who saw Afrikaans evolving early on out of seventeenth-century Dutch dialects through what was essentially normal language transmission. This model was taken up by Afrikaans-speaking linguists such as Boshoff (1921, 1959) and Smith (1927, 1952), who agreed that Afrikaans had developed according to trends already present in earlier Dutch dialects under minimal influence from other languages. The Dutch linguist Kloeke (1950), usually included in this camp, attributed a strong “founder effect” to the South Holland speech of the first Dutch colonists. The spontaneist model was later revived in a more drastic form by Van der Merwe (1963, 1968), who went so far as to claim that Afrikaans emerged within a half dozen years after the colonists’ arrival (1968:66) due to accelerated drift, and ruled out the possibility that people of color had influenced it in any significant way (1968:29). The approach of Bosman (1923, 1947) is considered eclectic by Reinecke et al. (1975:323), who note that “this view admitted foreign influence, chiefly from Low German colonists and Hottentots, but did not admit a situation favorable to outright creolization (unless of the Dutch spoken by Hottentots).” What den Besten (1987) calls the “South African philological school” came to prevail in that country from the 1960s until majority rule in 1994. Its leading writers were Scholtz (1963, 1980) and Raidt (1974, 1983, 1991), who concerned themselves less with the origins of Afrikaans as such than the history of specific linguistic phenomena. However, their underlying theoretical model was that of ordinary language change within varieties of Dutch accelerated by the influence of non-native speakers in a multilingual setting, whose speech was influenced by their first languages and had interlanguage features, but never underwent outright pidginization and creolization. One of the most complete histories of Afrikaans is Ponelis (1993), which stresses the restructuring resulting from imperfect second language acquisition. During this period, both black and white South African linguists began examining nonstandard varieties of Afrikaans more closely for the light they might cast on the issue of the language’s origins, including “Coloured” Afrikaans and Flytaal (Makhudu 1984), Malay Afrikaans (Kotz´e 1989), and Orange River Afrikaans, including Griqua Afrikaans (van Rensburg 1984, 1989) – the last variety having been studied in a book-length work by Rademeyer (1938).

The study of partially restructured vernaculars

15

What might be called the “Amsterdam school” of Afrikaans scholars has evolved around den Besten (1985, 1986, 1993) and his colleagues, who have focused on the effects of contact with Khoi and other languages. They see the South African school as antiquatedly Eurocentric in its approach: “If a feature can possibly be European, then it must be European,” provoking the opposite caveat regarding Valkhoff ’s approach: “If a feature can be a creolism, it must be a creolism” (Roberge 1994:40). Now that South Africans are reassessing their cultural identity with the advent of majority rule, the composite identity that “creolism” suggests has become increasingly attractive, and the ideological pendulum in linguistics may now be swinging wide of the mark in that direction. At a conference workshop on Afrikaans sociohistorical linguistics at the University of Cape Town (Mesthrie and Roberge, 2001–02) a reference was made to what distinguishes “. . . Afrikaans from other creole languages” (Holm 2001:353). 1.4

The study of Brazilian Vernacular Portuguese (BVP)

The history of the study of BVP from a language-contact perspective has been outlined by Holm (1987) and updated by Mello (1997). The similarity of some of BVP’s structural features to those of Portuguese-based creoles was pointed out over a century ago by Coelho (1880–86 [1967]), who concluded that “it shows a tendency towards creolization” (p. 170), but many Brazilian linguists still resist the view that the development of BVP involved significant restructuring. They have done studies of how the Brazilian lexicon has been influenced by indigenous languages like Tupi (Sampaio 1928; Marroquim 1934) or African languages (Raimundo 1933; Mendon¸ca 1933 [1973]), but in general they have followed the advice of Melo (1946), who cautioned against exaggerating the importance of such external influences when parallels could be found in archaic or regional usages in Portugal. However, Silva Neto (1950b:131) asserted that creole and what he called “semi-creole” (semi-crioulo) varieties of Portuguese had existed in Brazil, defining the latter as closer to the European variety but not speculating as to how they had evolved. R´evah (1963) discounted substratal influence on BVP in favor of a general tendency towards simplification of morphology in Western European languages, a line of thought taken up later by Naro and Lemle (1976). They assumed that BVP was in the process of losing number agreement rules, which were being obscured by certain phonological rules. Among non-Brazilian linguists, Valkhoff (1966) identified BVP features shared by Portuguese-based creoles as evidence of the latter’s influence on it. Jeroslow did a detailed study of a rural dialect (1974) that led

16

Languages in Contact

her to suspect prior creolization (McKinney 1975). Guy (1981) examined the same BVP phenomenon as Naro and Lemle (1976) but reached the opposite conclusion, i.e. that number agreement in BVP was spreading as a final stage in decreolization, comparable to that of AAE in the United States. In 1981 the Brazilian linguist Celso Cunha called for the study of BVP from the perspective of modern creole studies, the goal of Holm (1984, 1987, 1992b), who concluded that partial restructuring was clearly evident in the BVP varieties of Helv´ecia (Silveira Ferreira 1985) and Cear´a (Jeroslow 1974), and began attempting to work out the development of BVP as the product of this process. This approach has been taken up by Baxter (1992, 1997), who evaluates the importance of creole-like features through quantitative methods, and in the recent work of some Brazilian linguists such as Couto (1997) and Careno (1997). The most comprehensive of these is Mello (1997), who concludes that creolization and partial restructuring did not occur throughout colonial Brazil, but mainly in isolated areas which favored these processes . . . later decreolization through contact with B[razilian] P[ortuguese] occurred. In most of settled Brazil, the likeliest scenario was a process of imperfect language shift to Portuguese by the African and Amerindian populations and their descendants. This shift led to the establishment of BVP as the predominant dialect of Portuguese. However, as the shift was taking place, substratum structural features and interlanguage patterns were transferred to the target language, becoming fossilized. (Mello 1997:270)

Lucchesi (2000) basically takes the same position, but calls BVP the product of “irregular language transmission” that was mais leve (‘lighter’) than full creolization. Bonvini (2000) describes a creole-like lect of BVP called L´ıngua dos Pretos Velhos (LPV) traditionally used by Brazilian practitioners of candombl´e religious ceremonies for the light it could shed on earlier language contact. Studies approaching BVP from a language-contact perspective have contributed to the growth of Afro-Iberian linguistics as a distinct field. For obvious historical reasons, scholars working in Afro-Portuguese and Afro-Hispanic studies are natural allies, and there has been a movement to join the two camps since the first conference on Portuguese-based creoles was held in Lisbon in 1991 (d’Andrade and Kihm 1992). Since then the journal Papia: Revista de Crioulos de Base Ib´erica has been published in Brazil in both Portuguese and Spanish, encouraging further research in the coalescing field. An international colloquium on creoles based on Portuguese and Spanish in Berlin (Zimmermann 1999) was an important forum for debating the role of partial restructuring in the emergence of

The study of partially restructured vernaculars

17

both BVP and NSCS, as is the recently founded Associa¸ca˜ o: Crioulos de Base Lexical Portuguesa e Espanhola. 1.5

The study of Nonstandard Caribbean Spanish (NSCS)

NSCS is spoken by a substantial portion of Puerto Ricans, Dominicans, Cubans, and coastal Venezuelans and Colombians, as well as many of the Spanish speakers of New York City and Miami. Research since the 1960s indicates that a number of features in these varieties have parallels in Spanish-based Caribbean creoles. An overview of these studies can be found in de Granda (1975, 1987, 1998) and Green (1997), the sources of part of the following summary. One of the earliest references to external influence on a variety of Caribbean Spanish is that of Sandoval (1627), who describes the language spoken by Africans on the coast of what is today Colombia as “corrupt Spanish . . . influenced by the Portuguese they call the language of S˜ao Tom´e.” This and the emergence of Palenquero Creole Spanish nearby provide sufficient evidence that a Spanish-based pidgin built on Afro-Portuguese did in fact exist in the Caribbean, but it does not confirm the speculation of Bickerton and Escalante (1970:262) that there existed “a Spanish-based creole spoken in many parts of the Caribbean during the sixteenth and seventeenth centuries,” since there is no evidence that pidgins or jargons spoken elsewhere in the the region ever developed into stable creoles. One such pidgin or jargon was the habla bozal spoken by the large influx of Africans brought to Cuba (and elsewhere) in the first half of the nineteenth century to work on sugar plantations (section 2.4.2). Pichardo (1862: vii, iii) described their “mutilated Castilian, without concord, number declension or conjugation,” but noted that “Negroes born in Cuba talk like the local whites.” Van Name (1869–70:125) referred to it as only “the beginning of proper Creole” – an assessment later confirmed by Reinecke (1937:271). Van Name was also among the first to recognize that Cura¸cao’s Papiamentu was a creole language rather than a dialect of Spanish. Although Cuba’s habla bozal was never nativized as a creole, it did leave its traces in the local vernacular; Ortiz (1924) documented its lexicon of African origin. Still, most linguists in the Spanish-speaking Caribbean remained reluctant to admit any significant influence of African languages on local Spanish. Henriquez Urena ˜ (1940:130, 169), for example, described Dominican Spanish as having no more words of African origin than did general Spanish; he argued for an Andalusian origin for

18

Languages in Contact

the alternation of /r/ and /l/, seeing African influence only in the loss of syllable-final -s. Most linguists of this period who recognized the possibility of external influence on local Spanish, such as Wagner (1949), were not from the Caribbean. Cabrera (1954) was an exception; a Cuban anthropologist who interviewed older people of African descent around 1930, her portrayal of their Spanish is considered accurate, revealing a number of African-like features both in their normal speech and in the special language thought to resemble that of their ancestors, used in religious contexts. An equally exceptional linguist was Alvarez Nazario (1961), who examined early texts of bozal Spanish in Puerto Rico, concluding it was a “criollo afroespanol” ˜ linking local Spanish (especially that of black communities such as Loiza Aldea outside San Juan) to African languages via an Afro-Portuguese pidgin. His work served to encourage others studying Caribbean Spanish from the approach of contact linguistics, and he is now recognized as one of the principle founders of Afro-Hispanic linguistics (Ortiz 1999). Another founder, de Granda (1968), identified the speech of Colombia’s Palenqueros as the New World’s other Spanish-based creole, and went on to identify features from African and restructured languages in NSCS, focusing on the theory of an early pan-Caribbean creole that gradually decreolized (1970, 1976, 1978). Meanwhile, Otheguy (1973), working from Cabrera’s Cuban data, identified certain phonological and morphosyntactic traits in the vernacular that had survived from the habla bozal as being specifically creole features supporting the pan-Caribbean creole hypothesis (although he has since retreated from this position). Ziegler (1976, 1977) linked the bozal Spanish of Puerto Rico to that of Cuba, also arguing for decreolization. Megenney has focused on non-Peninsular features in the vernaculars of coastal Colombia (1976), Venezuela (1985), and the Dominican Republic (1990), as well as African-derived vocabulary used in religious rites in Cuba and Brazil (1999). Lipski, coming from within the creolist camp, has offered counter arguments to the pan-Caribbean creole theory (1993, 1994), seeing substratal influence as more likely to have come into Caribbean Spanish through imperfect second language acquisition, and bozal Spanish as never having undergone complete creolization since it was not nativized (2000). Schwegler, another creolist but one working primarily on Palenquero (1993, 1996a), has also studied the effects of restructuring on Caribbean Spanish (1996b). Schwegler and Morton (2002) document the features of the NSCS of bilingual speakers of Palenquero CS (PS), casting crucial new light on the link between restructured varieties like bozal Spanish and modern NSCS. Schwegler helped to organize one of the first international

The study of partially restructured vernaculars

19

conferences on Palenquero and NSCS (Monino ˜ et al. 2002) as well as the first book-length survey of Afro-Hispanic linguistics (Perl and Schwegler 1998). Perl, coming from Afro-Portuguese studies, has worked on the Cuban vernacular from a creolist perspective (Perl 1985, 1988, 1989), as have Ortiz (1998) and Figueroa (1998). Other younger scholars who have dealt with varieties of NSCS as a the product of partial restructuring include Alvarez (1990), Lorenzino (1993, 1998), and Green (1997), the last describing a hitherto unknown basilectal variety of Dominican Spanish. 1.6

The study of Vernacular Lects of R´eunionnais French (VLRF)

The vernacular French of R´eunion, a small island in the Indian Ocean, is spoken or understood by most of the 500,000 inhabitants. Although locally called cr´eole, its structure seems to be descended mainly from that of seventeenth-century French dialects, including maritime varieties; however, it has a number of features also found in creoles. The creole or non-creole identity of R´eunionnais has long been the subject of considerable debate; this identity is the focus of the following brief review of the literature, based largely on Chapuis (forthcoming), rather than the other main point of contention, which is the historical relationship of R´eunionnais to the fully restructured Ile de France creoles of Mauritius and the Seychelles. Adam (1883) grouped R´eunionnais with the French-based creoles: “Creole is the adaptation of French . . . by and for the slaves of Africa . . . in the Antilles, in R´eunion and in Ile de France.” Schuchardt (1885, translated 1979:15–17) analyzed a text of R´eunionnais spoken by whites as “totally French . . . foreign elements merely float on the surface [making it] . . . only an apparent creole.” Reinecke (1937:526) noted that in comparison to the creole of Mauritius, “the dialect of Reunion has not departed quite so widely from the original French.” Valkhoff (1964:724) suspected authors of R´eunionnais texts “of using an artificial and gallicized language which can be called ‘semi-creole’ ” and decided that the language was a form of “Creole-influenced French rather than French-influenced Creole” [“plutot ˆ du fran¸cais cr´eole que du cr´eole fran¸cais”] and that “there are two varieties of this Creole (without counting many intermediary nuances), namely an urban speech form and a popular speech form, and the former is more gallicized.” Vintil˘a-Radulescu˘ (1976:129) was the first to mention the the possible influence of the IndoPortuguese spoken by the wives of the earliest settlers in R´eunion. She also realized that “the mountainous relief of the islands . . . explains the

20

Languages in Contact

dialectal mosaic” (1967:126). Deltel (1969) proposed that there was a continuum among the different lects of R´eunionnais. The most important work on R´eunionnais to date is that of Chaudenson (1974ff.), who was the first to distinguish among what are now understood to be the three principal lects: (1) the Cr´eole des Bas, spoken by the coastal R´eunionnais of African, Malagasy, and Indian origin; (2) the Cr´eole des Hauts, used by the highland whites; and (3) the urban Creole, which is strongly gallicized. He also proposed that Mauritian and Seychellois Creole were derived from Bourbonnais (an earlier form of R´eunionnais), which since decreolized due to the continuing presence of French on R´eunion. He argued that there had been no substratal influence on R´eunionnais, which had simply evolved out of the fran¸cais avanc´ee or colloquial French that had developed beyond the reach of those who would have kept it more in line with the standard. Boll´ee (1977:116) argued for Mauritian and Seychellois having resulted from a “higher degree of reduction . . . than R´eunionnais.” Papen (1978) provided a comparative study of the grammar and social history of all three varieties, concluding that R´eunionnais represented a postcreole continuum. Valdman (1978) was the first to draw parallels between R´eunionnais and the patois of St. Barts, a comparison later furthered by Calvet and Chaudenson (1998). Hull (1979) pointed out the difference in structure between R´eunionnais and the Ile de France creoles, concluding that “Maur[itian] Cr[eole] evolved on Mauritius out of a nucleus of Pidg[in] Fr[ench], with only secondary borrowings from R´eu[nionnais] Cr[eole]. Seych[ellois] Cr[eole] derives from early Maur[itian] Cr[eole], not R´eu[nionnais] Cr[eole].” He noted that “Where black influence was subsequently removed, a somewhat decreolized form of Cr[eole] could remain on the island, as on R´eunion, or on St. Barts. . . . But on the whole Cr[eole] and French remain psychologically distinct . . . [and] No ‘postCreole continuum’ has formed, as in English Cr[eole]-speaking areas” (Hull 1979:211–213). Baker and Corne (1982) also rejected Chaudenson’s Bourbonnais theory, arguing that R´eunionnais settlers were not present in sufficient numbers on Mauritius during the crucial period, while “West Africans formed a majority of the slave population of Mauritius in the period 1730–35” (1982:241), explaining the striking parallels between the French Creoles of the Atlantic and Indian Oceans. In the same work, Corne pointed out that the “verbal system of R[´eunion] C[reole] is fundamentally ‘French’ in its make-up . . . [while its] Creole features . . . are rather marginal” (1982:102). In more recent work, Chaudenson (1992, 1995, 2000) has described creolization as a restructuring process that is not so different from the

The study of partially restructured vernaculars

21

kind of restructuring found in normal language transmission. This would certainly account for the partial restructuring of varieties such as R´eunionnais, from the perspective of the present study. 1.7

The comparison of partially restructured vernaculars

In 1991 I organized a seminar on partial restructuring at the City University of New York (CUNY), followed by another in 1996. A number of talented doctoral students participated, several of whom were themselves native speakers of partially restructured vernaculars or their source languages. These seminars led to a number of publications, ranging from conference papers to journal articles (e.g. Craig 1991 on American Indian English) and dissertations (e.g. Mello 1997 on Brazilian Vernacular Portuguese, and Green 1997 on nonstandard Dominican Spanish), one of which is still in progress (Chapuis forthcoming on R´eunionnais). The goal of these seminars was to identify some of the problems that needed to be solved in developing a workable theoretical model for this linguistic process by tracing the genesis and synchronic morphosyntax of a number of partially restructured languages, comparing the results, and looking for the possible relationship between the social history of the speakers and the linguistic outcome. Some of the initial results were described in Holm (1992a) and are briefly outlined below along with some further developments that grew out of later work on these varieties. The social factors that we considered potentially relevant to the linguistic outcome included the following: 1. the precise origins of superstrate and substrate speakers; 2. the (changing) ratio of superstrate to substrate speakers; if the latter came to outnumber the former, the length of time this took; 3. the degree of intimacy of early social relations between superstrate and substrate speakers (i.e. the likelihood of pidginization as opposed to normal second language acquistion); 4. the likelihood of either group’s contact with a pidgin or creole spoken elsewhere; 5. demographic changes (e.g. immigration, emigration, wars, plagues) and the effect on intergroup relations; 6. social, economic, and political changes and the effect on intergroup relations; 7. the degree of rigidity of any racial caste system; 8. education: accessibility, actual language of instruction; 9. communications: degree of geographical isolation; 10. any changes in the variety’s status (e.g. new domains of use, standardization).

22

Languages in Contact

Regarding the linguistic make-up of each variety, the following factors were considered: 1. the sources of lexicon: archaic, regional, or sociolectal usages in superstrate; substrate languages; adstrate languages; pidgins or creoles spoken elsewhere; 2. phonology: contrasts with superstrate; similarities to any varieties in (1) above, i.e. in phonotactic rules or actual phonemes and their allophones; 3. morphosyntax: contrasts with superstrate; similarities to any varieties in (1) above, e.g. the loss or retention of inflections in the NP (e.g. number/gender marking on articles, adjectives; possessive constructions) and VP (bound vs. free tense/aspect morphemes; uses of tense and aspect), as well as any other constructions not found in the lexical source language (e.g. use of prepositions and conjunctions; word order in main clauses; structure of dependent clauses). 4. the typological distance between the superstrate and substrate. The study of each variety concluded with a summary of the scholarship relating to its status as a creole or non-creole, and an assessment of its status as a partially restructured language. The group’s ultimate task was to compare the results of each study to determine whether the similarities and differences among these varieties would justify their inclusion in a group of partially restructured languages, and then to extrapolate the defining sociolinguistic and structural characteristics of that group. Since partial restructuring is a graded phenomenon, any specification of what proportion of features a variety so designated must share with creoles but not the lexifier language has to be intrinsically arbitrary. As a common-sense guideline, it seemed unhelpful to designate any language that has borrowed any creole feature as being partially restructured. For example, standard American English has borrowed some lexical items and even set phrases from various creoles such as go for broke, but this hardly seems to be grounds for classifying it as being partially restructured. On the other hand, the number of features shared with creoles but not with British varieties of English in AAE does seem significant: it has not only lexical items but also frequently occurring phonological and morphosyntactic features. It is difficult to measure objectively the degree of restructuring that a language variety may have undergone. Nonstandard Caribbean Spanish appears to have undergone less restructuring than either Afrikaans (in which one can say “Ons is bly,” literally ‘Us is happy’ in terms of Dutch) or Brazilian Vernacular Portuguese (e.g. “Eu chamei ela,” corresponding to ‘I called she’ in European Portuguese). Of course the literal translations into English do not provide an objective indication of the degree

The study of partially restructured vernaculars

23

of restructuring, since English speakers associate violations of standard pronominal case marking with basilectal Caribbean Creole English, which is not the impact these structures have in modern Brazil and South Africa. (Parkvall 2000 uses a more objective method of measuring degree of restructuring, discussed in the introduction to chapter 2, and another method is discussed in section 6.2). We considered ways of indicating the degree of restructuring, e.g. by designating varieties as having undergone “weak” or “very weak” restructuring if they have few features shared with creoles, and as “strong” or “very strong” if they have many. The idea was to distinguish between varieties like AAE and the nonstandard varieties used by some Southern whites, containing fewer such features (albeit significant ones such as copula deletion). Such distinctions would seem to be helpful in describing the status of vernacular lects of R´eunionnais French as opposed to that of the nonstandard French of Louisiana’s Cajuns or the patois of St. Barts. The criterion of distinguishing partially restructured varieties from post-creoles through the existence of basilects in the latter’s history (Holm 1988–89:9–10) proved problematical, quite aside from the very real difficulty of finding written evidence of extinct basilects. For example, while the ancestors of the Afrikaners apparently never spoke a fully creolized variety and their seventeenth-century Dutch simply underwent creole influence, the African ancestors of the so-called Coloureds apparently did speak a full creole (or at least a variety that underwent a very strong degree of restructuring, to extrapolate from modern Orange River Afrikaans). However, the modern lects of Afrikaans – both standard and nonstandard – have speakers from both groups, and there is a basilect or near-basilect in the history of these lects, although we can say that the nonstandard lects seem to be descended from this (near) basilect more directly than the standard lects. We encountered other problems as well with our own hypotheses and those of others, but recognized that these difficulties could not be adequately dealt with until we found and organized the relevant linguistic and sociolinguistic facts and then tested our theories against them. This volume, which builds on the work of the students participating in those seminars as well as my own research and that of others, is an attempt to do that.

2

Social factors in partial restructuring

Introduction The kind of partial restructuring of languages examined here is clearly related to the more complete kind of restructuring called creolization, and there is a widespread consensus that the defining characteristics of creolization include social as well as linguistic factors (Holm 2000a:68– 71). In fact, Thomason and Kaufman (1988:35) demonstrated that “it is the sociolinguistic history of the speakers, and not the structure of their language, that is the primary determinant of the linguistic outcome of language contact.” It will be argued here that partial restructuring is also a sociolinguistic process, and to understand it we must consider phenomena that are both social and linguistic. One of the crucial social factors in creolization has to do with the proportion of native versus non-native speakers of the lexical source language present at the time the creole language emerged. Although Parkvall (2000) has questioned the centrality of this demographic ratio in determining the degree to which creoles are restructured, his comparative work actually reaffirms the validity of this correlation, which is the topic of the following discussion. Bickerton (1981:4) claimed that the maximum percentage of native speakers of the superstrate language associated with full creolization was 20 percent. He was not the first to ponder this part of the equation; over a century earlier, Van Name surmised that Of the causes which have contributed to the formation of these dialects the chief are: first, the mature age of the slaves . . . secondly, the fact that they constituted the great body of the population, the whites being in a minority seldom as large as one-fourth. (Van Name 1869–70:124, my emphasis)

While there is general agreement that non-native speakers need to constitute a substantial majority of the population for full creolization to occur, there is debate as to whether this ratio is the overriding factor in the process, and what the minimal majority might be. Rickford 24

Social factors in partial restructuring

25

(1999:235) points out that there are known instances where Africans made up less than 80 percent of the population in places that produced a creole: By 1690, 35 years after the settlement of Jamaica, Blacks constituted 75 percent of the population (Williams 1985:31) . . . In Haiti the Black population was only 34.8 percent 25 years after the colony’s founding, and in Martinique, only 51.5 percent. (Singler 1995)

The discrepancy between 75 percent and 80 percent in Jamaica seems too small to worry about; that in Haiti seems much more considerable, but this may have to do with the date simply being too early – raising the issue of the relevance of the amount of time that was necessary for the slave population to reach the required majority. Haiti’s population was over 80 percent non-European by 1713 (Singler 1995:210). The figures from Martinique seem like better counter evidence to Bickerton’s claim because of the existence of a creole text from 1671, when the population was 62 percent non-European (ibid., Carden et al. 1991). To test the correlation of such demographic ratios to the degree of restructuring, Parkvall (2000) designed a restructuring index to quantify the typological distance between some Atlantic creoles and vernaculars, on the one hand, and their lexifiers on the other. The linguistic features he used to measure restructuring included 36 grammatical features often found in Atlantic creoles but not their lexifiers (e.g. no gender distinction in the third person singular pronoun) and 9 phonological ones (e.g. the presence of tonal distinctions). By quantifying the presence or absence of these features, he was able to quantify the degree of restructuring of each language, resulting in scores (see table 1 below) varying from 90 for the completely “radical” creole to −45 for the ideal lexifier. He then compared these to social factors: the first documented slave imports (event 0); the year the non-white population exceeded 20% (20%); the interval in years; the year the non-white population exceeded 50% (50%); the interval between this and event 0; the year the non-white population exceeded 80% – purportedly triggering the bioprogram (Bickerton 1981) – (80%); and the interval between this and event 0. Table 1 was originally published with a number of qualifying footnotes; here we need note only that the demographic data for St. Thomas∗ actually refer to St. Barth´elemy, the source of the former’s French-based patois. Since this table raises a good number of questions, the reader is referred to the original article, which is quite transparent considering the constraints of space: the results of the linguistic comparison are given for each feature in each variety, and the obvious problems of comparability, reliability, etc. are openly discussed.

26

Languages in Contact

Table 1. Degree of restructuring and social factors (Parkvall 2000:205–206) Score

Variety

Event 0

20%

Interval

50%

Interval

80%

Interval

68 63 51 48 48 42 41 41 41 39 38 35 35 23 20 16 −5 −5 −23 −24 n/a

Sranan S˜ao Tom´e St. Kitts Jamaica Haiti Gullah Negerholl. Fr. Guiana Guadeloupe Louisiana Martinique Mauritius Papiamentu Cape Verde Barbados Virginia Dom. Rep. Brazil Bermuda St. Thomas∗ Fr. St. Kitts

1651 1485 1626 1655 1634 1670 1665 1664 1640 1708 1635 1721 1634 1462 1627 1619 1502 1549 1612 1660 1626

1652 1493? 1650? 1657 1635 1672 1668 1665 1652 1711 1636 1722 1652 1515? 1643 1712 ? ? 1618 1671 1626

1 8 24 2 1 2 3 1 12 3 1 1 18 53 16 93 ? ? 6 11 0

1662 1499? 1680 1670 1698 1708 1675 1668 1667 1726 1655 1730 1657 1525? 1657 never ? ? never 1787 1668

11 14 54 15 64 38 10 4 27 18 20 9 23 63 30 n/a n/a n/a n/a 127 42

1672 1546? 1729 1690? 1710 never 1710 1697 1725 never 1733 c. 1750 1685 1540 1705 never 1520s never never never 1710

21 61 103 35 76 n/a 45 33 85 n/a 98 29 51 78 78 n/a c. 20 n/a n/a n/a 84

30

Average

13

33

60

Parkvall concludes that “. . . there is indeed a correlation – whether causal or not – between the degree of restructuring and the demographic ratios in the formative period” (2000:197). Assuming the soundness of his method and execution of this inquiry, however, it is also the case that demographic ratios are not everything, as he himself points out. To deal first with two apparent exceptions: Table 1 indicates that two areas which developed full creoles never had a non-white population exceeding 80% (Gullah, scoring 42; Louisiana, scoring 39). “Gullah” appears to refer to the Gullah country – certain coastal counties of South Carolina and Georgia – rather than to either of these states as a whole. But, if that is the case, we must note that by the nineteenth century blacks made up some 80% of the the population of the tidewater area of South Carolina – and over 95% in its rural districts (Reinecke 1937:491). As for Louisiana, a footnote clarifies that “the non-white population reached an all-high low [sic: ‘all-time high’?] of 77,9% in 1741” (Parkvall 2000:5).

Social factors in partial restructuring

27

In other words, both areas seem to have had a slave population very close to 80%. Having dealt with the apparent exceptions to the 80% rule, we now need to consider the relevance, if any, of the amount of time that it took for the slave population to reach 80% of the total. This may account for the discrepancy noted above, namely the observation of Rickford (1999:235) that in Haiti the slave population was only 34.8% twenty-five years after the colony’s founding, and in Martinique, only 51.5%. According to Table 1, it took 76 years for Haiti’s slave population to reach 80%, and 98 years for Martinique’s. But if we look at the number of years in the final column of Table 1, there seems to be no pattern: the most restructured varieties (Sranan, S˜ao Tom´e, St. Kitts) took 21, 61, and 103 years respectively to reach 80%, while the least restructured creoles (Parkvall draws the bottom line at a score of 20) are Papiamentu, Cape Verde, and Barbados (to follow his nomenclature), taking 51, 78, and 78 years respectively. Thus, if there were a correlation, these results seem counterintuitive: the places developing the most restructured creoles took both the shortest time (Sranan: 21 years) and the longest time (St. Kitts: 103 years) to reach 80%. A concern with the amount of time that was required for the slave population to reach a majority and the possible correlation this might have with the degree of restructuring can be seen in Bickerton’s (1984) discussion of the amount of time between the beginning of immigration of non-native speakers and what Baker (1982:852) had called “Event 1” – the point at which slave and master populations achieved numerical parity. . . . The longer this period, the greater the exposure of early arrivals to the dominant language and hence the richer the second-language version that would be transmitted to the first influx. (Bickerton 1984:177)

This led to Bickerton’s pidginization index (PI): “Y ×

P = PI R

Where Y represents the number of years between colonization and Event 1, P the total substratum-speaking population at Event 1, and R the yearly average of post-Event 1 immigrants. . . . A higher PI indicates a ‘richer’ pidgin, one that retains more features of the dominant language” (1984:178). However, this mathematical formula to indicate the degree of restructuring proved “unworkable” (Singler 1990:645). Later Mufwene (1996) dealt with a similar idea in his article on the founder principle:

28

Languages in Contact

In some colonies, such as South Carolina, Virginia and R´eunion, the Africans remained minorities for the first 30–50 years, whereas in some others such as Suriname, Mauritius (Baker, to appear [= 1996]) and apparently Guyana (Winford, p.c. 1994), the plantation phase came about rather rapidly bringing along an early slave majority. (Mufwene 1996:98)

While the first three colonies gave rise to less restructured vernaculars, the latter three have fully creolized languages. However, the length of time it took for the slaves to become a majority may not be the relevant social factor here. All three of the latter colonies could have imported restructured languages along with their slave force. Many, if not most, of the slaves brought to Suriname between 1651 and 1667 came from Barbados, and many, if not most, brought to Guyana after 1760 came from the British West Indies and West Africa. Even in Mauritius, where the slave population grew from 30 to 1,000 between 1727 and 1730, some 600 came from French entrepots in West Africa and might well have spoken some pidginized French. Thus the relationship between the degree of restructuring and social factors like demographic ratios is not always a straightforward one, a factor Parkvall recognizes: One possible reason why there appears to be a lack of correspondence between the demographic and the linguistic data is that the creoles under investigation did not arise in the areas where they are presently spoken. (2000:193)

This complexity of the interrelationship of social factors has long been recognized in the acknowledgment of the effect of factors such as maroonage or the early withdrawal of the superstrate on a creole’s proximity to the “bioprogram” (Bickerton 1984:177–178) or its proximity to a prototype (McWhorter 2000). As to the existence of a straightforward relationship between the amount of time that it takes for a slave population to become a substantial majority and the degree of restructuring in the ensuing language, this would appear to be a question that has not yet been answered clearly, although Parkvall (2000) has cast some light on the matter. The most we can say with certainty is that it appears to have taken an average of sixty years for the non-native-speaking population to reach its greatest proportion of the total, at least in the history of the language varieties he examined – and possibly the same amount of time for these varieties to reach their present degree of restructuring. While Bickerton (1984) was certainly right to assert that time is a relevant factor, its relevance depends on still other variable, complicating factors. For example, the slave population of R´eunion reached 80 percent within 100 years, but this resulted in partial rather than full restructuring because of the particular

Social factors in partial restructuring

29

sociolinguistic circumstances of the colony’s early history (see section 2.5 below). It will be argued below that, like creolization, partial restructuring is a sociolinguistic process, that a relationship does indeed exist between sociolinguistic factors in the development of these languages and their synchronic typology, and that this relationship is parallel to the one between these factors in fully creolized languages. In order to examine this hypothesis, we will first look at the sociolinguistic setting of the formation of each language variety, focusing on social factors related to its genesis and development – the topic of the following sections of this chapter (the disparate lengths of which reflect in part the disparate amounts of published research to draw on). Chapters 3 to 5 will then compare corresponding areas of their morphosyntax, and the final chapter will attempt to relate the social and linguistic facts in a theoretical model that will account for what we know about these languages. 2.1

AAE: the sociolinguistic setting of its development

2.1.1

AAE: the seventeenth century

The proportion of Africans in most of Britain’s North American colonies was so much lower than the ratio of blacks to whites in the Caribbean that it seems doubtful that full creolization ever took place outside of a few demographically exceptional areas such as coastal South Carolina and Georgia, where Gullah developed. While settlers in both the British Caribbean and North America were largely European-born indentured servants in the early seventeenth century, by 1680 the black population equalled or surpassed the white population in many Caribbean colonies, whereas in North America blacks made up less than 5 percent of the total population (Rawley 1981:329). By the 1730s blacks still made up only 14 percent of the North American population, but 80 percent of the population of the British and French Caribbean (ibid.). The sociolinguistic history of African slaves in the Americas was largely determined by the economic forces shaping the societies to which they were brought, which in turn depended on particular crops. The tobacco plantations in Virginia and Barbados to which the first slaves were brought were relatively small farms worked mostly by indentured servants from Britain. The first Africans, arriving in Virginia in 1619 and in Barbados in 1627, were outnumbered by the British laborers among whom they worked and from whom they most likely learned English largely through normal adult second language acquisition; their contact was much closer at this early period than in the plantation societies that developed later.

30

Languages in Contact

The social conditions necessary for creolization – lack of access to nativespeaker models of the target language – did not occur on Barbados until the cultivation of sugar was introduced in the middle of the seventeenth century. As the supply of tobacco in Europe rose, its price fell and farmers began looking for other products to grow in the West Indies. The production of sugar was more profitable, but this required large plantations and a large workforce. Great quantities of sugar cane had to be produced to offset the initial investment of capital in the mill and other machinery, as well as land and labor. The work was much more grueling than raising tobacco, and indentured servants were unwilling to do it. Their labor was gradually replaced by that of the African slaves: while there were only a few hundred of the latter in Barbados in 1640, by 1645 there were 6,000 – and some 40,000 whites. By 1685 the number of blacks had risen to 46,000 and the number of whites had actually dropped to 20,000. The effect of the sugar revolution on Barbados was to force the smallholders (mostly former indentured servants) off the land as large sugar plantations took over the island. The displaced peasantry chose emigration over starvation, and populated new English colonies in Suriname, Jamaica, the Carolinas, and the Leeward Islands. Thus in the second half of the seventeenth century Barbados played a central role in the dispersal of English in the New World (Holm 1994). The question is to what degree that English reflected contact with the languages spoken by Africans, especially the speech of slaves exported from Barbados and other Caribbean islands such as St. Kitts. Winford (2000:216, 242) considers modern Barbadian an “intermediate” variety, “. . . the result of targeted shift . . . quite comparable with other outcomes of communal shift such as Hiberno English . . .” Yet Hickey (forthcoming) finds that Winford has misinterpreted the facts about Irish English, such as “its preference for invariant does” (Winford 2000:232). Hickey (forthcoming) says “There is no evidence of this in Irish English. Perhaps the claim derives from a misunderstanding . . .” At any rate it seems clear that more basilectal varieties were used in Barbados in the seventeenth century; one such lect was documented in the testimony of a Barbadian slave at the 1692 Salem, Massachusetts, witch trial, e.g. “He tell me he God” (Dillard 1972:79). The cradle of the development of African American English was Virginia, the first and most populous colony in the southern part of British North America. In Virginia the status of Africans was comparable to that of indentured servants during the first part of the seventeenth century; most worked as domestics or on small farms and seem likely to have had sufficient access to local English. Mufwene (2000b:240) reasons that this would have led to their normal acquisition of the British settlers’ English

San Francisco

N E A O C

ke

Albuquerque

de ran oG Ri

Denver

o

Ar k

Kansas City

Houston

Red

0 0

250

Super ior

250

500

500 miles

750 km

Atlanta

Cincinnati

Montgomery

New Orleans

Lake Erie

Cleveland

Detroit

Lake Huron

Chattanooga Memphis

Chicago

ke La

St Louis

Little Rock

Minneapolis

Oklahoma

Dallas

San Antonio

iss

Jacksonville

Charleston

Washington

Philadelphia

Lake Ontario

Boston

New York

O C E A N

A T L A N T I C

Norfolk

St La wr en ce

Map 1 United States: ‘Coloured Population, 1900. Originally published in Darby and Fuller (1978) The New Cambridge Modern History Atlas (Cambridge University Press).

Areas without coloured population or unsettled

Under 10%

o ad lo r Co

Salt Lake City

M

i

10–30%

30–50%

Sn a

m bia

Great Salt Lake

Over 50%

COLOURED POPULATION

Los Angeles

Portland

Lake Michigan

lu Co

Lake Winnipeg

io

Seattle

Mississipp i

P A C I F I C Oh

ur

sas an

32

Languages in Contact

Table 2. Estimated population of (East) Virginia, 1685–1790

Total Red White Black

1685

1700

1715

1730

1745

1760

1790

43,600 7% 87% 6%

63,500 3% 88% 9%

96,300 1% 77% 22%

153,900 0.6% 67% 32%

234,200 0.1% 63% 36%

327,600 0.1% 60% 40%

747,800 0.03% 59% 41%

(Based on Wood 1989:38, cited by Winford 1997.)

as a second language. Africans in Virginia did not become slaves for life until the latter part of the century (2000b:238), when their numbers became more considerable. Although slaves then made up a larger proportion of the local population, Mufwene (2000b:240) believes that these later arrivals tended to adapt to the local language norms rather than imposing their own norms due to the founder principle (Mufwene 1996). However, it was established above that the founder principle can be overridden by the importation of a variety that was restructured elsewhere, and Rickford (1999:241–242) objects to the scenario proposed by Mufwene, since during this period most slaves were imported from England’s Caribbean colonies and were already likely to know English in a pidginized or creolized form. It seems doubtful that they would have abandoned such linguistic habits on arriving in North America since these would clearly have been useful to them and their local interlocutors, not only in facilitating communication but also in establishing sociolinguistic boundaries.

2.1.2

AAE: the eighteenth century

As tobacco farming spread during the eighteenth century, the importation of slaves – increasingly from Africa rather than from the Caribbean – grew as indentured labor became more expensive. The proportion of blacks in Virginia increased from 22 percent in 1715 to 40 percent in 1760. Thus later newcomers were less likely to have as much access to nativespeaker models of the target language as did those who arrived earlier, but whites still made up a substantial majority of the population: In fact, whites made up some 60 percent of the population of all the southern colonies in 1750, although it must be borne in mind that this was a large and demographically heterogeneous area. Rickford (1999:244) points out how these statistics are misleading for Georgia, which in 1750 had been founded only recently: by 1776 blacks

Social factors in partial restructuring

33

Table 3. Estimated proportion of blacks in American colonies in 1750 Colony/region

Total blacks

% of blacks

% in 13 colonies

New England Middle Colonies Southern Colonies Maryland Virginia N. Carolina S. Carolina Georgia

10,982 20,736 204,702

3.1% 7.0% 39.8% 30.8% 43.9% 27.1% 60.9% 19.2%

4.6% 8.8% 86.6%

(Based on Rickford 1999:238.)

made up 48 percent of Georgia’s population and were concentrated on the coast, where Gullah was spreading. In other words, the most likely scenario after 1715 in most parts of the American South is that English was not fully creolized but did undergo some degree of restructuring as it was acquired by speakers of African languages and of more fully restructured varieties of English brought in from the Caribbean, and was increasingly learned from other African Americans speaking an emerging sociolinguistically distinct lect of local English that became the forerunner of modern AAE. While Mufwene is surely right in surmising that “children, regardless of whether they were locally born or imported . . . slowed down the divergence of Africans’ English vernaculars from those of Europeans, because they were more likely to acquire colonial English with minimal restructuring” (2000b:240), the important issue here is the very existence of the divergence, for which the only credible explanation is restructuring. The roots of that divergence are likely to go back to at least the early eighteenth century. The recognition of a distinctively black way of speaking seems implicit in a passage in a journal of 1705, in which one white criticizes another for speaking “Negro” to an Indian (quoted by Dillard 1972:80). However, there is a problem in what the designation of “Negro” might mean. Since colonial times African Americans have spoken many varieties of English, ranging from the foreigner’s English of Africans; the creole or partially restructured English of slaves imported from the West Indies; the pidginized English brought from West Africa; the indigenous creoles of North America (such as Gullah and its offshoot, Afro-Seminole); the partially restructured varieties that had emerged by the eighteenth century; unrestructured regional English (which may have borrowed some interlanguage or substrate features from contact with the previous varieties); and

34

Languages in Contact

the standard American English first spoken by educated freedmen and now spoken by millions of African Americans – usually with AAE phonological features to signal a positive awareness of ethnic identity. This is not conjecture: Kautzsch and Schneider (2000) document idiolects in South Carolina that they identify as creole, semi-creolized, and non-creole. The accuracy of their identification is borne out by the linguistic features in these idiolects. As Schneider (1989) suggests, pockets of greater restructuring developed in localities with demographics more similar to those in South Carolina’s low country that led to the emergence of Gullah, which underwent relatively complete creolization, making AAE the product of a few presumably independent creolization processes in localities with an exceptionally dense black population . . . with a certain degree of leveling, mixture and perhaps loss or even spreading of such forms in the post-emancipation period. (Schneider 1989:278)

2.1.3

The emergence of Gullah

As in other colonies, the sociolinguistic situation of Africans in coastal South Carolina was determined by the crop they were brought there to raise. In this case it was the cultivation of rice that created a setting for the creolization of English – relatively complete if compared to the partial restructuring of AAE, but still far from the fundamental restructuring of English in Suriname. In 1670 Charles II granted the Lords Proprietors of Carolina on the mainland a patent that included the Bahama Islands, creating a single colony that endured as such for the next half century. In the same year Charleston, in what is now South Carolina, began to be settled by some 800 British settlers and 300 slaves. According to Wood (1974:24–25), almost half the whites and more than half the blacks came from Barbados. Other settlers included French Huguenots and religious dissenters from the British Isles and from other colonies in British North America (Baptists, Quakers, and Catholics), as well as settlers from Bermuda, the Bahamas, the Leeward Islands, and Jamaica (ibid; Joyner 1984:13). During the first years of settlement the colonists traded with the Indians for deerskins and also raised livestock, sending meat to Barbados in exchange for slaves and sugar. By the 1690s it became clear that the coastal low country was suitable for raising rice, for which there was a growing market in southern Europe and the Caribbean. Although the British settlers had no experience in cultivating rice, their slaves did since it had long been grown in various parts of West Africa. The intensive labor needed for its cultivation led to a great increase in the importation of slaves: by 1708 the colony’s population of 8,000 was equally divided between whites

Social factors in partial restructuring

35

Table 4. Estimated population of South Carolina, 1685–1775

Total Red White Black

1685

1700

1715

1730

1745

1760

1775

11,900 84% 12% 4%

14,100 53% 27% 20%

19,200 27% 29% 45%

33,400 5% 29% 65%

62,400 2% 33% 65%

97,500 1% 40% 60%

179,400 0.3% 40% 60%

(Based on Wood 1989:38, cited by Winford 1997.)

and blacks, but by 1740 there were 40,000 slaves as opposed to only 20,000 whites (Wood 1974:143–152). The proportion of inhabitants of American Indian, European, and African ancestry underwent the shifts documented in table 4. These figures suggest that during the first generation of colonization, blacks made up a small minority who either learned English as a second language through relatively close contact with native speakers under frontier conditions, or else arrived from Barbados or other parts of the British West Indies with some knowledge of Creole English. In the next generations, however, the black population (coming increasingly straight from Africa with little or no knowledge of English or Creole) became the majority, leaving white native speakers a minority of about one-third to two-fifths of the population. It must be remembered that the above figures are for all of South Carolina not just the coastal low country, where the concentration of blacks was much higher. Segregation was instituted in South Carolina in 1720 after the colony was separated from the Bahamas and North Carolina, and a 1726 survey of two low-country parishes indicated that two-thirds of the slaves lived on plantations, defined as landholdings with 25–100 hands (Wood 1974:160). As noted previously, by the nineteenth century blacks made up some 80 percent of the the population of the tidewater area of South Carolina – and over 95 percent in its rural districts (Reinecke 1937:491). In coastal Georgia (settled after 1733) the proportion was not quite so great, but Gullah came to predominate there as well. The social and linguistic relations in Gullah country differed considerably from those in Virginia and elsewhere in the American South. Quite apart from the continuing existence of creolized English there today, the above historical, social and demographic data support the development of a creole continuum in this area during the eighteenth century – one that developed in a society with a relatively high percentage of native speakers of English, keeping the speech of most creole speakers quite mesolectal in comparison to Caribbean varieties.

36

Languages in Contact

2.1.4

AAE: the early nineteenth century

In 1790 the United States’ 757,000 blacks (only 60,000 of whom were free) made up over 19 percent of the country’s population. Large-scale slavery was restricted mainly to the tobacco-growing area of the Upper South (Virginia, Maryland, and North Carolina) and central Kentucky and Tennessee, as well as the coastal areas of South Carolina and Georgia, where rice and indigo could be grown (Bailey 2001:58). However, in 1793 the invention of the cotton gin made short-staple cotton a viable cash crop, leading to the spread of new cotton plantations westward into the interior South. During the first half of the nineteenth century, up to a million slaves were moved to Alabama, Mississippi, and newly acquired areas as far west as Louisiana and Texas (2001:59). By 1830 parts of the interior South, such as the Mississippi Delta, the Alabama Black Belt, and the lower Brazos Valley in Texas, had ratios of blacks as high as those in Gullah country (2001:60). Mufwene (2000b:247) claims that the cotton plantations were on average smaller than the tobacco plantations in the Upper South and rice plantations in coastal South Carolina and Georgia, but Bailey (2001:60) claims they were often just as large. While many of the slaves brought to these states came from the Upper South and spoke early AAE, many also came from coastal South Carolina and Georgia and spoke Gullah. Still others were brought directly from Africa and, to a lesser extent, the Caribbean until the importation of slaves was made illegal in 1808 – although it continued until the Civil War (1861–65). The latter slaves, speaking restructured or non-native varieties of English learned from the first two groups, seem likely to have contributed to a heterogeneous mixture of AAE varieties, so it is not surprising that later creolists such as Stewart (1968) and Dillard (1972) had the impression that there existed a widespread “Plantation Creole” continuum at this time: evidence of creole features was indeed widespread outside the original Gullah area. Stewart (1968:52) claimed that decreolization began well before the Civil War; however, Mufwene (2000b:247) concludes that this was the period when AAE and Gullah were stabilizing, referring to similarities in how they were represented in fiction then and how they are spoken today. Pointing to the end of the illegal importation of slaves, he surmises that The increasing number of native speakers and the attrition of non-native speakers enabled the vernaculars that had been developing till then among descendants of Africans to stabilize, or at least not to undergo any form of restructuring motivated by language contact. (Mufwene 2000b:247)

Bailey, however, interprets the same social and linguistic evidence as supporting restructuring:

Social factors in partial restructuring

37

Given the fact that a substantial portion of slaves coming into most of these areas [of the interior South] spoke some form of English, we would hardly expect the development of a full-scale creole language, but we might well expect a variety of English showing the clear imprint of the creole and African languages that formed part of the matrix from which it derived. According to Bailey/Thomas (1998a), that is precisely what we find in the early varieties of AAVE in the interior South. (Bailey 2001:60)

2.1.5

AAE: diaspora varieties and the ex-slave narratives

One of the basic problems in reconstructing the history of the development of AAE has been the fact that it was rarely written, and what written fragments that do exist in fiction or travelers’ accounts are of limited reliability, most having been written by non-native speakers of AAE. Moreover, it is not possible to identify exactly which of the many varieties spoken by African Americans is being portrayed (native, nonnative, creole, etc.) unless something is known about the origin and background of the individual speaker (Holm 1988–89:498). One attempt to avoid this dilemma has been to examine the modern speech of the descendants of groups of African Americans who left the United States at earlier points in its history, although the information this provides is usually indirect and sometimes misleading. Such groups include those who went to the Bahamas in the 1780s (Holm 1983); Sierra Leone in the 1790s (Hancock 1971; Huber 2001); Nova Scotia in the nineteenth century (Poplack 2000); the Saman´a peninsula of the Dominican Republic in the 1820s (ibid.); and Liberia, also in the early nineteenth century (Singler 1989). However, there are pitfalls in this method, as I have discovered: for example, Holm (1988–89:501) concluded that Decreolization [of AAE in the United States] would appear to have begun during this same time span [1780–1820] since the descendants of the emigrants to the Bahamas speak creolized English while those in Saman´a do not.

However, a closer examination of the historical records revealed that the creole nature of the folk speech on these southern Bahamian islands should not be interpreted as evidence that AAVE had been fully creolized on the mainland before 1780 (and later decreolized) since the language that was brought there was in all likelihood eighteenth-century Gullah rather than eighteenth-century AAVE. This new interpretation undermines the most compelling evidence we knew of to support the hypothesis that AAVE was ever fully creolized, and it provides further indication that AAVE was – from its beginning – the product of semi-creolization. (Hackert and Holm forthcoming)

Another source of information about earlier forms of AAE is the body of transcribed interviews with elderly former slaves conducted during

38

Languages in Contact

the 1930s and later, known as the Works Project Administration Ex-Slave Narratives (e.g. Schneider 1989). The problem of disentangling the linguistic system of the transcriber from that of the speaker has been partly resolved with the Ex-Slave Recordings, mechanically recorded interviews with African Americans born in the American South between 1844 and 1861 (Bailey, Maynor, and Cukor-Avila 1991). Although there is some controversy over how much speakers’ grammars might change over their lifetimes, these recordings have been of great value in casting light on the structure of AAE since the middle of the nineteenth century. A recent comparative quantitative study of these sources (Kautzsch 2002:257) concludes that properly chosen interviews “can be regarded as very close approximations to speech” – much more so than letters, since literacy is so closely tied to the learning of standard English. 2.1.6

AAE: the late nineteenth century

It is not necessary for AAE ever to have been fully creolized for it to have undergone decreolization. However, if it was never a full creole, “decreolization” may be a misleading term. Mufwene (2000b) prefers “debasilectalization,” and Holm (2000:32) refers to this reapproachment with non-restructured varieties as “secondary leveling.” It has been assumed that African Americans had more freedom of movement and better access to education after emancipation, and that this accelerated the reapproachment of AAE to the non-restructured lects with which they were in contact. Evidence for a straightforward movement in this direction over time, however, is mixed. Poplack et al. (2000:100) admit that the rate of plural -s marking is lower in early AAE than in contemporary AAE (but not that this is due to decreolization; see section 4.4.1), while Kautzsch (2002:260) finds no evidence of decreolization in any of his quantitative studies of earlier AAE. Bailey (2001:64ff.) points out that the area where cotton was grown continued to expand as rapidly after the Civil War as before it, with sharecropping or farm tenancy replacing slavery as the mechanism for meeting the intensive labor demands of cotton. The main difference was that after the war, poor whites were also ensnared in this system and kept there by the lack of alternative sources of credit. Bailey makes his point with two maps of a Georgia plantation: in 1860 laborers are confined to the slave quarters, but in 1881 the shacks of both black and white tenant farmers are scattered over the plantation. He concludes that The spread of tenancy to Whites and the spatial reorganization of the plantation created new contexts for Black–White speech relationships throughout the South, contexts which allowed for interaction among African Americans and Whites that

Social factors in partial restructuring

39

was probably more widespread than before the Civil War and took place among people who were closer to being socioeconomic equals. . . . Hence we might expect their vernaculars to to have much in common and . . . they do share many features that emerged during the late 19th and early 20th centuries. (Bailey 2001:66)

However, the late nineteenth century also saw the rise of white supremacists who enacted Jim Crow laws throughout the South to segregate and disfavor African Americans socially, economically, and educationally. This situation triggered emigration. 2.1.7

AAE: the twentieth century

Almost a million African Americans left the South during the great migration, which began even before the opening of new economic opportunities in the North that came with the First World War and lasted until well after the Second World War. Whatley (1981:93) points out that the increasing migration of blacks to the North as well as the West Coast is reflected in the decreasing proportion left in the South: 85 percent of the national total in 1920, 77 percent in 1940; and 60 percent in 1960. By 1970, over a third of all African Americans lived in just seven major urban areas: New York, Philadelphia, Baltimore, Washington D.C., Detroit, Chicago, and Los Angeles (Bailey 2001:66). These areas became more and more segregated: the African American population of Detroit, for example, increased from 37 percent in the 1960s to 83 percent in 2000 (Wolfram forthcoming). The restricted social environment of blacks in the ghettos of Northern cities fostered the continuation of features brought from the South, but this speech took on a distinctively urban character. Labov (1972b:xiii) described AAE as having a “relatively uniform grammar” from one region to another, as one might expect from a variety that had recently migrated. The question is whether AAE and other American varieties of English have been growing apart. Mufwene speculates that Perhaps the late nineteenth century was also the time when African and European Americans were particularly eager to identify some linguistic peculiarities as ethnic markers and thus made divergent selections of features from the pool of variants that they had shared until then. Such selections would have enhanced ethnic-dialectal differences that may have been minor at the dawn of the twentieth century. (Mufwene 2000b:251)

Poplack (2000:1) interprets this as endorsing her own position that the many grammatical distinctions between contemporary varieties of AAVE and American and British English are relatively recent developments, possibly initiated during the post-Civil War period, as suggested by Mufwene [2000b] . . . in a social context highly propitious to racial segregation and divergence.

40

Languages in Contact

Rickford (1999:xix) notes that “Accelerated AAVE use – especially of salient features like be – and scathing criticism of those who ‘talk white’ are part of a symbolic statement by today’s young people of awareness of and pride in African American identity.” Kautzsch (2002) interprets his quantitative study of earlier AAE negation patterns as evidence that In the first half of the nineteenth century AAE has a largely nonstandard English basis. During the second half of the nineteenth century it gradually – though slowly and partially – moves closer to standard English. Finally, a renewed deviation away from standard English seems to have started in the early twentieth century, which is supported in some cases by present-day data (cf. ain’t for didn’t; negative concord). This return to nonstandard patterns might result from increased segregation of African Americans from the early twentieth century onwards, which at the same time favors the emergence of a black linguistic identity. (2002:88)

Concerning the relationship between Southern varieties of AAE and Southern White Vernacular English (SWVE), Cuckor-Avila (2001:116– 119) found that there were linguistic correlates to the reduced contact between speakers of the two varieties after the Second World War due to the decline of share-cropping, northern migration of blacks, and other factors. Some of the earlier shared features have survived only in AAE (e.g. copula absence after you, lack of third person singular present -s marking of verbs), and others have evolved only in AAE (e.g. be + verb + -ing). However, AAE has retained still other features that had not been shared with SWVE (zero third personal singular copula, remote time been, ain’t for didn’t). It is the last group which makes the position of Poplack (2000) untenable. However, it is still possible that historically (or at least currently) the grammars of AAE and other varieties are diverging rather than converging, as was assumed by those who saw AAE mainly as the result of the decreolization of a full-fledged creole (e.g. Stewart 1968). In 1986 Labov and Harris argued that increasing racial segregation had produced a “BEV [Black English Vernacular] that is more remote from other dialects than . . . reported before” (1986:4). The work of Bailey and Maynor (1987, 1989) (cf. Bailey [2001] above) in Texas appeared to provide independent confirmation of this hypothesis, suggesting that it might be a general pattern in the United States. A symposium, published as Fasold et al. (1987), raised the points “that Labov and his colleagues had failed to provide comparison points in either real or apparent time, and that Bailey and Maynor needed an intermediate age group to minimize the the possibility of age-grading” (Rickford 1999:262). However, Rickford’s own recent work in Palo Alto, California (1999:274) confirms this growing

Social factors in partial restructuring

41

divergence, particularly in the AAE use of invariant be and copula absence (see section 3.1.1). He notes that the first feature is “now very salient as a distinctively black form . . . be is an invariant lexical item, which can be consciously adopted and rapidly spread like slang terms and other lexical items” (1999:276). Finally, he notes that “Black teenagers are less assimilationist than their parents and especially their grandparents, and more assertive about their rights to talk and act their ‘natural way’ ” (1999:274). Whatever the current state of the relationship between AAE and other American varieties of English, this brief historical survey demonstrates that, given what we now know about what is required for full creolization to take place, that is not what happened to AAE in most parts of the American South. On the other hand, it has become clear that the features which modern AAE shares with Caribbean varieties of Creole English but not with Southern White Vernacular English cannot all be explained through the kind of borrowing associated with language contact, i.e. between a less-restructured AAE and more-restructured varieties of English brought in from Africa, from the Caribbean, or from Gullah country. While contact between creolized and non-creolized varieties of English can explain the borrowing of lexical items (including ones with syntactic implications, such as the complementizer say, discussed in section 5.1.2) and even the borrowing of morphosyntactic features, it cannot explain systemic differences like the patterning of zero copulas (section 3.1.4). The only remaining explanation for the origin of such features is that AAE is the result of English having undergone a less-extensive degree of restructuring than that which led to Gullah or the English-based creoles of the Caribbean and Suriname. Such limited restructuring would still have allowed for the features described in chapters 3–5, i.e. both the reduction of inflectional morphology and the retention of certain substrate features such as a correlation between the form of the copula and its following syntactic environment. 2.2

Afrikaans: the sociolinguistic setting of its development

2.2.1

Afrikaans: the seventeenth century

In 1652 the Dutch East India Company established the first European settlement on Africa’s Cape of Good Hope in order to supply their ships with provisions for the long voyage from Holland to the Dutch East Indies. For the first six years the principal languages in contact were Dutch (mainly the dialects of south Holland) and Khoi, spoken by the

42

Languages in Contact

original inhabitants of the Cape with whom the Dutch bartered for cattle. By 1656 it was reported that the Khoi “were beginning to speak Dutch fairly well” (cited by Scholtz 1970:85); they soon began working for the settlers as nursemaids and herdsmen. In 1658 the colony received its first shipload of slaves from West Africa and Angola; at this point “there were 166 whites against 194 slaves and freedmen, together with some Hottentot hangers-on” (Reinecke 1937:597). Thus the slaves found two kinds of Dutch in use: that of the settlers (among themselves); and that of the Khoi (used with the settlers). The latter probably played an important role in shaping the pidgin, which did not have SVO word order as did the Niger-Congo languages of the slaves, but rather SOV, the underlying word order of both Dutch and Khoi (den Besten 1989:220). In 1685 a colonial official reported that the settlers’ children were picking up the broken Dutch of the non-whites (Valkhoff 1972:40). In 1677 190 slaves were brought to the Cape from the Asian colonies that the Dutch had taken from the Portuguese in India, Ceylon, and the East Indies (Reinecke 1937:568); they seem likely to have made up a substantial portion of the colony’s 350 slaves in 1691 (1937:597). As in Batavia, the new Dutch capital of the East Indies, their lingua franca was creole Portuguese, which was spoken throughout Portugal’s former possessions. Hesseling (1897) argued that the use of Malyo-Portuguese at the Cape Colony was both early and widespread, and that this was a much more important factor in the partial creolization of Afrikaans than the influence of Khoi, although this interpretation has been seriously challenged. It has long been asserted that contact with other European languages played a role in the relatively rapid evolution of Dutch into Afrikaans. In the 1660s the Europeans were mostly Dutch (numbering 104), but there were also a considerable number of Germans (40) and individuals of other nationalities (23) (Reinecke 1937:576). However, most of the Germans were men who had come to work for the Dutch East India Company; many knew the Low German dialects closely related to Dutch, and most soon learned Dutch and became assimilated (Scholtz 1970:84). In 1688 some 225 French-speaking Huguenots sought refuge in the colony; they made up a full sixth of the white population at the time, but the Dutch governor ensured their assimilation by requiring their farms to be interspersed among those of the Dutch rather than in a single settlement. The resulting social contact led to intermarriage and their acceptance as “true Afrikaners,” so that their language disappeared within some forty years (ibid.). It seems clear that the crucial period in the formation of Afrikaans was the second half of the seventeenth century, by the end of which “gender distinction had disappeared completely; wij ‘we’ was replaced by ons ‘us’;

Social factors in partial restructuring

Windhoek

43

MOZAMBIQUE

B O T S W A N A

Rehoboth

K a l a h a r i

LIMPOPO

D e s e r t

NAMIBIA Pretoria(A)

Namaland

Griqualand Little Namaqualand

. Ora n ge R

NORTHERN CAPE

(E) . lR a a V FREE STATE

Bloemfontein LESOTHO Or a n ge R .

M MP U

WEST

A

GAUTENG

NORTH Johannesburg

N LA

GA

SWAZILAND

KWAZULUNATAL

Durban (E)

EASTERN CAPE Cape Town (A=E) 0

300 km

0

200 miles

WESTERN CAPE

East London (E) Port Elizabeth (E)

(A) Largely Afrikaans-speaking (E) Largely English-speaking

Map 2 Southern Africa. Originally published in Holm (1989) Pidgins and Creoles Volume 2 (Cambridge University Press).

the present and perfect had coalesced [although Raidt (1983) places this later]; and some verbs had already lost the infinitive suffix -en” (Hesseling 1897, translation of 1979:7). Although further development took place over the next century, during which Cape Dutch became what can now be recognized as Afrikaans (Raidt 1983:28), a major factor in the crucial early period was clearly the restructured Dutch of the Khoi and the slaves. Since the geography of the Cape was unfavorable for the kind of plantation agriculture that arose in the New World, there were no large slave holders except for the Dutch East India Company itself. While the slaves from Mozambique, Madagascar, and Asia tended to live in the Cape Town area, the Khoi often lived out in the country among the Dutch as domestics and farm hands (van Rensburg, personal communication). Their partial assimilation was not only cultural and linguistic but also genetic. Few Dutch women accompanied the first settlers to the Cape, and in the first twenty years of the colony, some 75 percent of the children born to female slaves were fathered by Dutch colonists (Valkhoff 1966:206), as were many children born to Khoi women, so that the Cape

44

Languages in Contact

Coloured population began to grow early on. In the seventeenth century such children received their freedom on reaching adulthood, provided they spoke Dutch and were confirmed members of the church (Raidt 1983:11). During this early period freed slaves of mixed ancestry were not viewed as a separate ethnic group; they had all the civil rights of whites, with whom they intermarried (ibid.), although this situation changed in the eighteenth century. Early “Hottentot Dutch” can be viewed either as a pidgin (den Besten 1989) or as the early stage of a continuum of interlanguages that the Khoi evolved in their acquisition of Dutch (van Rensburg 1989). 2.2.2

Afrikaans: the eighteenth century

During the late seventeenth and early eighteenth centuries many Khoi retreated into the interior to avoid European domination, often taking with them a knowledge of Dutch. The detribalization of those who remained at the Cape was hastened by a severe smallpox epidemic in 1713, which decimated the Khoi. Those who fled took with them what was left of their social system, and those who remained merged into the colony’s population of free blacks and Coloureds. By the end of the eighteenth century most of their descendants were monolingual speakers of Dutch (Combrink 1978:77). From the middle of the century onwards the European settlers began moving east and then north to areas populated by Bantu-speaking peoples, but it is important to bear in mind that within the Cape Colony itself the slaves never greatly exceeded the white population. In 1708 there were 1,723 free burghers, who owned 1,298 slaves; in 1798 there were 21,746 Europeans, with some 25,000 slaves (Reinecke 1937:597). Table 5 summarizes the demographic figures given above. Although these figures are incomplete, they allow some inferences regarding language contact in so far as group origin indicates likelihood of a first or second language knowledge of Dutch, or, in 1798, the forerunner of Afrikaans. (It should be noted that the estimated figure for the Khoikhoi is for the southwestern Cape from the Oliphants to the Breede River; however, the relevant figure is the unknown number of Khoikhoi who were in regular contact with the other groups.) These figures are helpful in assessing the social and linguistic developments from 1652 until the end of the eighteenth century, when Afrikaans had clearly emerged as a system quite distinct from Dutch, which most Afrikaners could no longer speak (Raidt 1983:18). Unlike Negerhollands, the fully creolized Dutch that developed in the Danish West Indies (Holm 1988–89:325–328), Afrikaans developed in a colony that had remained under Dutch control

Social factors in partial restructuring

45

Table 5. Estimated population of the Cape Colony, 1658–1798 1658 White Dutch German French Other Slaves African and freed Asian Total Khoikhoi

1660s

1677

1688

1691

166

1708

1798

1,723

21,746

1,298 ?

25,000

104 40 225 1,125

23 194 190 50,000?∗

?

?

?

350 ?

(Based on Reinecke 1937:568–697; ∗ Elphick and Malherbe 1989:3.)

up to that point. The maintenance of standard Dutch as the language of government, education, and religion undoubtedly played an important role in preventing Cape whites from adopting most of the creole features used by non-whites, as happened in the West Indies. However, there is unambiguous evidence that some such borrowing did in fact take place in the case of Afrikaans, although to a more limited extent (see below). It is also of primary significance that whites continued to make up a full half of the speakers of Afrikaans, compared to the 6 percent white minority of the speakers of Negerhollands. White Afrikaans speakers were able to impose a much more European character on their local culture and language, much like the large minority or majority of whites in the American South. 2.2.3

Afrikaans: the nineteenth century

The French occupation of the Netherlands in 1795 led Britain to seize the Cape Colony, which it gained permanently in 1814. The arrival of 5,000 British settlers in 1820 brought the first large body of Europeans not to be assimilated into Afrikaans culture. As in Ceylon, Guiana, and other Dutch colonies that the British took over at this time, English came to replace Dutch in government and education, although the Afrikaners firmly maintained it in their religion with the 1637 Statenbijbel in standard Dutch. Afrikaners had begun moving north into the interior in search of better pastures from the middle of the eighteenth century, but when the British abolished slavery in 1833, some 12,000 Afrikaners left the Cape for what became Natal in the east, despite Zulu resistance. However, Britain annexed Natal in 1843 and many of the Afrikaner farmers or Boers

46

Languages in Contact

migrated to the interior, establishing the republics of the Orange Free State and the Transvaal in the 1850s, where standard Dutch was made the official language despite the difficulty it presented many Afrikaans speakers. The British, Dutch, and even many of its own speakers looked down upon the Boers’ “corrupt kitchen Dutch” and considered it incapable of serving any serious purpose. However, in the second half of the nineteenth century some Afrikaners, feeling increasingly threatened by the political and cultural dominance of the English, began a movement to promote the use of their language in writing (Deumert 1999). Citing the example of the Negerhollands Bible for the Danish West Indies, it was proposed that the Statenbijbel should be translated into Afrikaans to facilitate its use among the Coloureds and poor whites who had difficulty reading Dutch. Despite long opposition, this was finally accomplished in 1933, and the translation also came into general use among whites. Support for the new literary language remained divided until the Boer War (1899–1902), after which the victorious British forced the Boer republics to join the Union of South Africa. 2.2.4

Afrikaans: the twentieth century

In 1910 “Dutch” was given equal status with English within the Union of South Africa, but in 1925 it was made explicit that “Dutch” included the newly standardized Afrikaans language, i.e. Afrikaans Nederlands or ‘African Dutch.’ The three main regional varieties of the language are (West) Cape Afrikaans, Orange River Afrikaans, and East Cape Afrikaans. East Cape formed the basis of the standard, but also has nonstandard sociolects. West Cape (Kaaps) is spoken in the area that made up the Cape Colony in 1750 and has many features of the nonstandard speech of the Coloureds, who today make up 54 percent of Cape Town’s population (Makhudu 1984:134). Orange River Afrikaans, spoken by by the descendants of the Khoi who retreated to the Orange River area in the eighteenth century, is “widely recognized as a pidgin or creole” variety in South Africa (van Rensburg 1989:136), although it appears to be a post-creole. It is spoken by several groups, including the Griqua, a group of mixed Khoi, slave, and white ancestry that migrated from the Cape Colony north to the Orange River area in the eighteenth century (Rademeyer 1938; van Rensburg 1989). It is also spoken by the Rehoboth Basters, the descendants of Dutch farmers who married Khoi women on the northern fringes of the Cape Colony in the eighteenth century. In the middle of the nineteenth century they were driven north by encroaching white farmers and attacking indigenous peoples; in 1870 they settled in Rehoboth in what was to become the German protectorate of South

Social factors in partial restructuring

47

West Africa in 1884, thus bringing Afrikaans to the area that recently became independent Namibia (van der Merwe 1989, 1990; Luijks 2000). Finally, nonstandard Afrikaans is also spoken by many of the 127,000 Cape Malays, descendants of the slaves brought from Asia. It was their Islamic clergy who first put Afrikaans into writing in the nineteenth century, using Arabic script. Kotz´e (1984, 1989) considers the Cape Malays “a small but linguistically significant core group, since it can be shown that the nonstandard characteristics that are viewed as typical of Cape Afrikaans are quantitatively strongest in the speech of this community” (1984:42). 2.3

BVP: the sociolinguistic setting of its development

2.3.1

BVP: the sixteenth century

Because African slaves were brought to Brazil in such great numbers that they and their descendants came to make up the majority of the population in certain parts of the country – and of the country as a whole at certain points in its history – it would be reasonable to expect a more fully restructured variety of Portuguese to have developed there, much as creoles developed under apparently similar sociolinguistic conditions elsewhere. As Guy (1981) put it: From the social historical standpoint, our question probably would not be “Was Portuguese creolized in Brazil?”, but rather “How could it possibly have avoided creolization?” (1981:309)

However, there is a paucity of evidence to show that a widespread, stable creole ever became firmly established in Brazil. This may be due to the way in which differing sociolinguistic conditions there affected language transmission. Yet certain features of the nonstandard variety indicate the influence of Amerindian, African, and creole languages. Brazil eventually became the world’s greatest importer of slaves, receiving 38 percent of all Africans brought to the New World, as compared to the 4.5 percent who went to British North America (Curtin 1975:41). But the point at which these Africans arrived, the sociolinguistic situation they found, and the portion of the population they made up are factors of crucial importance regarding the degree of restructuring that Portuguese underwent in Brazil. In one important respect Brazil was unlike the first places where Portuguese-based creoles became established: the Cape Verde Islands and S˜ao Tome, which the Portuguese colonized in 1462 and 1485, respectively. These islands off the coast of western Africa had no inhabitants

48

Languages in Contact

when the Portuguese came upon them, so there was no local language to learn when settlers arrived from Portugal and slaves were brought from the African mainland, putting their languages in contact and leading to the development of new creoles within a generation. Brazil, by contrast, was already inhabited. When the Portuguese began to explore its vast coast after 1500, they found it to be inhabited by Amerindians speaking closely related varieties of Tupi. As the Portuguese established settlements from the 1530s onwards, contact among the various Indian subgroups increased and there evolved a common Tupi vocabulary fitting into a shared syntactic framework which was relatively free of complicated morphology. This koine, which the Portuguese also learned for contact with the Indians, came to be called the L´ıngua Geral, or general language for communication throughout the colony. Sampaio (1928:3) claimed that during the first two centuries of colonization, L´ıngua Geral was the principal language of three-quarters of Brazil’s population, albeit with growing bilingualism in Portuguese. Even the Brazilian-born Portuguese settlers, often raised by Tupi-speaking nurses, used the language with ease and seemed to have a strong emotional attachment to it (Reinecke 1937:692). There was a great deal of intermarriage between the Portuguese and Tupis, and in some parts of Brazil people of such mixed ancestry came to predominate; their mother tongue was L´ıngua Geral. One such group was the bandeirantes of S˜ao Paulo, who carried the flag of Portugal ever further inland in their search for slaves and gold, taking their language with them. Thus it is not clear what language(s) African slaves encountered on arriving in Brazil during the colony’s first two hundred years. They may well have had to learn L´ıngua Geral more often than Portuguese, as suggested by Reinecke (1937:549), so that during the linguistically crucial first generations of the colony there was little opportunity for a fully restructured variety of Portuguese to become established among African slaves unless they had brought such a language with them from Africa. If Africans first learned L´ıngua Geral as a second language and their descendants learned it as their mother tongue, their later shift to a Tupi-influenced but unrestructured variety of Portuguese along with the rest of the population during the eighteenth century would have established such Portuguese as a second language to be learned by slaves newly arrived from Africa. Under these circumstances there would have been no need for the establishment of a creole. In any case, the first generations of Africans arriving in Brazil did not encounter the same kind of linguistic situation as those coming to the off-shore islands of Africa did, or indeed as did those arriving on Caribbean islands whose native populations had all but disappeared. Brazil also differed from most of the areas where creole languages developed in that Africans or their descendants made up only a quarter of

Social factors in partial restructuring

49

Dominica Martinique St Lucia Barbados

Caribbean Sea ao raç ire Cu Bona

Aruba

St Vincent Grenada

Tobago

A T

Trinidad

Caracas

VENEZUELA

Georgetown Paramaribo Cayenne Dutch G

L

O

A C

N E

T

A

I

C

N

British

U I French SURA N A

Bogotá

INA

COLOMBIA

M

RIO BRANCO

AMAPÁ

P A R Á MARANHÃO

AMAZONAS

U N I T E D O F

GUAPORÉ

S T A T E S

PIAUÍ

PARAIBA PER

N AM

BUCO ALAGOAS

B R A Z I L

SERGIPE

BAHIA

MATO GROSSO

PERU

RIO GRANDE DO NORTE

CEARÁ

GOIÁS

Salvador

Brasilia

La Paz

BOLIVIA Helvécia

C H I L E

SÃO PAULO

A

Asunción

G PARANÁ

São Paulo

RIO DE JANEIRO

Rio de Janeiro

SANTA CATARINA

ARGENTINA

ESPIRITO SANTO

AY

P A C I F I C

PA R

U

O C E A N

MINAS GERAIS

ATLANTIC O C E A N

RIO GRANDE DO SUL 0

URUGUAY

0

250

500 250

750 km 500 miles

Map 3 South America, 1830–1956. Originally published in Darby and Fuller (1978) The New Cambridge Modern History Atlas (Cambridge University Press).

the population from 1600 to 1650; it was not until the 1770s that they constituted over 50 percent of the population (Marques 1976:359, 435), reaching 65 percent in 1818 (Reinecke 1937:556). Yet whites, the group most likely to speak unrestructured Portuguese, never made up more than a third of the population until the second half of the nineteenth century (see table 6), when slavery ended and the linguistic dice had already been

50

Languages in Contact

Table 6. Estimated population of Brazil, 1538–1890

African-born Creole Africans Integrated Amerindians Mixed European-born Creole whites

1538–1600

1601–1700

1701–1800

1801–1850

1851–1890

20% – 50

30% 20 10

20% 21 8

12% 19 4

2% 13 2

– 30 –

10 25 5

19 22 10

34 14 17

42 17 24

(Based on Mussa 1991:163, cited in Mello 1997:85.)

cast. Thus there were proportionately fewer whites in Brazil than in most of the American South, although they always exceeded 20 percent, the maximum proportion of native speakers associated with full creolization (see introduction to this chapter). Yet national population figures for large countries can obscure the local conditions of particular speech communities. For example, in what became the United States whites soon far outnumbered all other groups, but blacks still predominated in certain areas such as coastal South Carolina and Georgia, where a creolized variety of English developed and survived (section 2.1.3). The same seems likely to have been the case in the sugar-growing areas of Brazil, where labor-intensive production required plantations with a majority of African slaves since Indians proved unsuitable for this work. Lucchesi (2000:42ff.) claims that “the expansion of the Portuguese language in Brazil until the beginning of the 18th century occurred parallel to the expansion of the sugar-plantation society in the Northeast” [my translation]. Marques (1976:362) points out that each sugar mill in Brazil required a minimum of 80 slaves, besides the hundreds needed to work the fields. The number of such mills increased rapidly, from 1 in 1533, to 60 in 1570, 130 in 1585, and 346 in 1629. According to Silva Neto (1963:79), of the 14,000 Africans in Brazil in 1597, almost all were in the sugargrowing settlements of Bahia and Pernambuco. At this time there were 10,000 Africans in Pernambuco, making up fully half of its population, which also included some 8,000 whites and 2,000 Indians. Moreover, these Africans constituted a considerable proportion of the small total population of Brazil at this linguistically crucial early stage (numbering only a few tens of thousands), making it more susceptible to influence from the fully creolized Portuguese brought to Brazil from Africa. Evidence that a more highly restructured variety of Portuguese had been the

Social factors in partial restructuring

51

language of coastal Brazilian sugar plantations rather than L´ıngua Geral can be deduced from the fact that many of the earliest Portuguese sugar planters and their slaves came to Brazil by way of S˜ao Tom´e (Ivens Ferraz 1979:19). The Portuguese had settled this island in the decades following their discovery of it in 1470, intermarrying with the slaves they brought from the mainland. A creolized variety of Portuguese evolved along with the cultivation of sugar on large plantations. The prosperity that this brought to S˜ao Tom´e during the first half of the sixteenth century waned during the second half as slave rebellions and maroon attacks eventually destroyed the island’s economy. The Portuguese began abandoning S˜ao Tom´e in large numbers, many going to Brazil (ibid.). It is unlikely that they would have left behind the greatest financial asset needed to establish sugar plantations in Brazil, namely their creole-speaking slaves from S˜ao Tom´e. Moreover, there are abundant phonological, syntactic, and lexical features linking S˜ao Tom´e Creole Portuguese and BVP (Holm 1992b). 2.3.2

BVP: the seventeenth century

Further evidence of the use of restructured Portuguese on Brazilian sugar plantations comes from New Holland, the empire that the Dutch tried to carve out of northeastern Brazil from the modern states of Sergipe to Maranh˜ao from 1630 to 1654. After the Dutch seized this area from Portugal in their struggle for independence from Spain (united with Portugal under a single monarch from 1580 to 1640), they found allies in many of the marranos (crypto-Jews) who had settled in Brazil to avoid the Inquisition in Spain and Portugal. The more tolerant Dutch allowed them to revert openly to Judaism, and they were joined by other Sephardic Jews whose families had earlier fled the Iberian Peninsula to take refuge in Holland (Goodman 1987). The Dutch and the Sephardic Jews seem likely to have communicated in Spanish or Portuguese (or restructured varieties of these) with one another and with their Brazilian slaves, mistresses, and wives. Few Dutch women came to Brazil, which is why the Netherlanders were unable to establish their language or culture there during the period they held this colony – a full generation (Boxer 1965:227). When the Portuguese regained the area in 1654, the Dutch and most of their Jewish collaborators were forced to leave Brazil. Many resettled in the Carribean regions, particularly in Dutch holdings in the Guianas and on islands such as Cura¸cao (Goodman 1987). Modern Papiamentu, the creole language of the Netherlands Leeward Antilles (Cura¸cao, Aruba, and Bonaire), reveals unmistakable Portuguese influence in its most basic vocabulary (Schuchardt 1882:895; Jeuda

52

Languages in Contact

1991). In fact Hancock (1969b:26) identifies Papiamentu as an offshoot of a Brazilian variety of pidgin or creole Portuguese. However, the Portuguese element in Papiamentu has been obscured by the creole’s three centuries of close contact with the Spanish spoken by the inhabitants of nearby Venezuela. There is also a strong Portuguese element in the lexicon of Saramaccan, an English-based creole spoken in the interior of Suriname (formerly Dutch Guiana). This has taditionally been attributed to the influence of the Portuguese-speaking Jewish plantation owners who fled from Brazil and made up 75 percent of the entire European population of Suriname at the end of the seventeenth century (Rens 1953:22). However, there has been considerable debate as to how Portuguese actually came to influence Papiamentu and Saramaccan. It has been suggested (e.g. Lenz 1928; Voorhoeve 1973) that this influence was via a pidginized variety of Portuguese used in the slave trade between Africa and the New World, which was partly learned by slaves and carried over into the languages which evolved in the Caribbean area. However Goodman (1987) presented considerable evidence supporting the view that the Portuguese element in Papiamentu, Saramaccan, and certain other creoles in the Caribbean area (e.g. the French-based creole of Guyane) was introduced by refugees from Dutch Brazil. If this is indeed the case, then Portuguese, whatever its degree of restructuring, must have been the language of Brazil’s coastal sugar plantations since at least the early 1600s. Under these circumstances, L´ıngua Geral could not have acted as a buffer to prevent the restructuring of Portuguese there. It has recently been debated that “no slaves could have been taken from Brazil to Surinam” (Ladhams 1999:232) and thus “the origin of the Portuguese element in the Surinam Creoles should not be sought in Brazil” (Arends 1999:204). However, Goodman (1987:368) asserted that “According to van Dantzig (1968:77) they took slaves along as well” and pointed out that the 1,200 refugees from Brazil who landed in Guadeloupe and Martinique in 1654 included 300 slaves (1987:391). Smith (1999) reasons that the presence of pidgin Portuguese features found in the Portuguese-derived vocabulary of Saramaccan but not in Sranan supports their origin in Dutch Brazil. To avoid a lengthy digression here, the interested reader is referred to these articles, but it should be pointed out that the transfer of restructured Portuguese from Pernambuco to the Caribbean area does not require the massive transfer of slaves: it is well known that in the seventeenth and eighteenth centuries whites were often proficient in creole languages and appear to have been the agent of their transfer in some cases, such as that of Creole French to Louisiana; cf. Hull (1983).

Social factors in partial restructuring

53

Furthermore, there is linguistic evidence that the Portuguese taken to the Caribbean area arrived via Brazil rather than Africa. Reinecke (1937:467) noted that the Saramaccan word plakkeh (now spelled pulak¨e) “a kind of eel” comes from the north Brazilian form poraquˆe (elsewhere in Brazil piraquˆe) “electric eel,” ultimately from Tupi pura’ke (da Cunha et al. 1982). Since the Saramaccans were not in direct contact with Tupi speakers, this word must have come into their language via Brazilian Portuguese. There is further evidence in the Sranan word kab´ugru (whence Dutch karboeger) “a mixed Amerindian/negro group living in Western Suriname” from the Brazilian Portuguese word caboclo (Smith 2002:151). The latter, meaning “indio; mesti¸co de branco com indio” is derived from Tupi kariuoka “homen branco” (da Cunha et al. 1982). There is also linguistic evidence that the Portuguese brought to the Caribbean area had already undergone some restructuring. Wullschl¨agel (1856:328) notes a proverb in Suriname that he tentatively identifies as “Neger-portugiesisch?”: “Praga beroegoe no mata caballo.” This is apparently the Brazilian proverb, “Praga de burrico n˜ao mata cavalo,” or “The braying of an ass doesn’t kill a horse,” i.e. “The insults of the weak don’t affect the strong” (Diva Penha Lopes, personal communication). The lack of de “of ” in the NP “praga beroegoe” indicates a possessive construction also found in BVP and S˜ao Tom´e CP (4.3.3). It should be noted that beroegoe also shows the effect of vowel harmony (Holm 2000a:151); this and the zero articles in both the Surinamese and modern BVP proverb indicate restructuring (Inverno and Swolkien 2001). Even though the English-based Surinamese creoles could have influenced these last two features, they have a different possessive construction [possessor + possessed], whereas the construction in the proverb is praga beroegoe, or [possessed + possessor]. Thus this proverb is perhaps the best indirect evidence that the language of Brazil’s coastal sugar-growing areas in the sixteenth and seventeenth centuries was Portuguese (rather than L´ıngua Geral) and that this Portuguese had undergone some degree of restructuring. Besides L´ıngua Geral, another factor complicates the reconstruction of language transmission in Brazil as this might relate to restructuring. This factor is the retention of African languages over many generations and among large numbers of people. Such retention was largely absent from other New World societies in which creoles developed, where Africans were often put into linguistically heterogeneous groups to prevent communication in a language that their overseers could not understand in order to foil possible plots to revolt. However, in Brazil linguistic homogeneity seems to have been valued since it enabled older generations of slaves to teach newcomers more easily (Le Page and Tabouret-Keller

54

Languages in Contact

1985:33). Until 1600, most slaves came from West Africa; from then until about 1660, Bantu-speaking Angolan and Congolese slaves predominated, and afterwards they came from both parts of Africa (Marques 1976:361). Influence from both sources is confirmed in the etymological study of African words surviving in Bahian BVP by Castro and Castro (1980:46): of 1,955 words, 967 (49.5 percent) were of Bantu origin, while 988 (50.5 percent) were of West African origin. Mendon¸ca (1933:28) claimed that the West African or Sudanic people predominated in Bahia, while Bantu speakers predominated to the north and south, but this may be an oversimplification. In the northeastern part of Brazil, African languages seem likely to have been used as lingua francas in the quilombos or maroon settlements established by escaped slaves (Reinecke 1937:557). In 1579, a Jesuit priest wrote that the foremost enemies of the colonizers are revolted Negroes from Guin´e in some mountain areas, from where they raid and give much trouble, and the time may come when they will dare to attack and destroy farms as their relatives do on the island of S˜ao Thom´e. (quoted by Kent 1979:174)

That time came when the Portuguese were distracted by the attacking Dutch in 1630 and the great “Negro Republic” of Palmares was established in Pernambuco. Its fortified villages had thousands of inhabitants with well-organized governments; Palmares endured until the end of the seventeenth century (ibid.). Freitas (1978:48, my translation) notes that Os palmarinos falavam uma l´ıngua toda sua, que compreendia formas da l´ıngua portuguesa, das l´ınguas africanas e secundariamente, das l´ınguas ind´ıgenas. Quando queriam se entender com eles, moradores e autoridades recorriam a “l´ınguas” ou seja int´erpretes. [The inhabitants of Palmares spoke a language that was all their own and included elements from Portuguese, African languages and, to a lesser degree, Indian languages. When they wanted to communicate with them, the local Europeans and their authorities took recourse to “linguists” or interpreters.]

Clear evidence that the Portuguese of such maroons was restructured can be found in the text of a quilombo song: Folga nego, branco n˜ao vem c´a Se vi´e, o diabo h´a de lev´a . . . . Pau h´a de lev´a. [Relax, black [friend], Whites won’t come here If they come, the devil take them . . . They’ll get a good beating.] (Freitas 1978:43; translated by Heliana Mello)

Social factors in partial restructuring

2.3.3

55

BVP: the eighteenth and nineteenth centuries

At the beginning of the eighteenth century it was estimated that in Bahia the proportion of blacks to whites [was] twenty to one . . . in the city of S˜ao Salvador alone twenty-five thousand Negroes were catechized and instructed in the Angolan tongue. (Nash 1926:127)

There is further evidence that Portuguese – or a restructured variety of it – was the language of the coastal sugar plantations rather than L´ıngua Geral. The latter clearly predominated in S˜ao Paulo and other areas where Indians and mesti¸cos worked on fazendas raising cattle. The shift from L´ıngua Geral to Portuguese in the interior seems to have been triggered by the gold rush in Minas Gerais during the first half of the eighteenth century. Although the gold had been discovered by bandeirantes who spoke L´ıngua Geral, Portuguese soon become the common language in the mining region (Carvalho 1977:27). This was due not only to the great influx of men from Portugal but also to the great influx of African slaves (to do the actual mining) from the coastal sugar-growing areas where such slaves were plentiful and “the inhabitants were proficient in Portuguese; we believe that they had never actually abandoned the language since their commercial activity kept them in constant contact with Portugal” (Carvalho 1977:27). However, an African-based lingua franca was also used during the gold rush in Minas Gerais, apparently among newly arrived slaves from Africa. From a detailed description of this Lingoa Minna written in 1731 and published two centuries later (Peixoto 1945), Hazel Carter (personal communication) determined that the language was based on Fon, an eastern variety of Ewe. As the mines became exhausted after the 1750s, there was a general movement of population away from Minas Gerais, and this probably played a key role in spreading a newly leveled variety of Portuguese – the forerunner of BVP – throughout the settled parts of Brazil, at the expense of L´ıngua Geral. As Portuguese gradually came to predominate, it is likely that successive generations of bilinguals had decreasing competence in L´ıngua Geral and increasing competence in Portuguese. It would be surprising indeed if there had not been considerable interpenetration of the two languages on all linguistic levels under such conditions. Describing the linguistic situation in the Amazon region in the nineteenth century as this language shift moved deeper inland, Hartt (1872:72) noted that many Portuguese idioms have crept into the Tup´ı; but on the other hand, the Portuguese, as spoken in the Amazonas, besides containing a large admixture of Tup´ı words, is corrupted by many Tup´ı idioms.

56

Languages in Contact

It would seem probable, therefore, that the L´ıngua Geral variety of Tupi left a strong mark on the BVP of rural peasants (Silveira Bueno 1963). The process involved was language shift, which can involve considerable restructuring (Thomason and Kaufman 1988:110–146). Different linguists’ attribution of BVP features to the influence of both Tupi and West African languages presents no real dilemma: such totally unrelated languages may well coincidentally share structural similarities (e.g. syllabic structure rules, progressive nasalization, a lack of many kinds of inflections) which simply converged to reinforce one another in shaping BVP. African languages continued to be spoken by large numbers of people in Brazil during the eighteenth and nineteenth centuries. Yoruba became established as a lingua franca in the state of Bahia and as late as 1900 an observer noted that The Nagoˆ [i.e. Yoruba] language is in fact much spoken at Bahia by almost all the old Africans of different nationalities and by a large number of [black] creoles and mulattoes. When in this state it is said that a person speaks the L´ıngua da Costa, invariably the Nagoˆ is meant. (Nina Rodrigues, quoted by Reinecke 1937:553)

Although it is no longer spoken natively, Yoruba is still used as a liturgical language in Bahia (Holm and Oyedeji 1984). I recorded it there in 1983 during a religious ceremony, and its identity was later confirmed by Oyedeji, a native speaker from Nigeria who listened to the tapes. Remnants of Bantu languages have also been identified (Fry, Vogt, and Gnerre 1981). Such long-term survival of African languages as lingua francas among different African ethnolinguistic groups and their descendants seems likely to have worked against the retention of an early Portuguese creole for such communication since it was not needed for this purpose. It is important to stress that European Portuguese continued to reassert itself in Brazil. Silva Neto (1963:68ff.) notes that when the Portuguese court fled to Rio de Janeiro during the Napoleonic wars, “tudo se foi re-europeizando” [“everything was re-Europeanized”], from houses to clothes. In 1808 Rio had 50,000 inhabitants and the number of whites was much lower than blacks; by 1835 the population was 110,000 with many more whites, given the arrival of 24,000 Portuguese and a large number of other Europeans (ibid.). To this day, Rio’s inhabitants are known for their palatalization of plural -s in the European Portuguese manner, a prestige pronunciation that occurs only sporadically outside Brazil’s former captial. Political independence from Portugal in 1822 saw the creation of an empire that lasted until the abolition of slavery in 1888. However, the establishment of independent Brazilian cultural institutions did not mean that the grammar of European Portuguese lost any

Social factors in partial restructuring

57

status in education. Silva Neto (1963:90) quotes a priest in the northeastern province of Maranh˜ao writing in 1819: At present the language of this country is Portuguese; educated people speak it very well; however, among the rustics there is a certain dialect, which I believe is the result of the mixture of the languages of the diverse nations that have inhabited Maranh˜ao.

A sociolinguistic factor relevant to the survival of a restructured vernacular was race relations. Although observers from societies with more rigid racial caste systems sometimes believe Brazil to be a country without racism, black Brazilians are usually quick to disabuse one of this notion. Although Brazilians of all races have been equal before the law since abolition, without any form of officially sanctioned segregation, blacks still tend to be poor and powerless, while the rich and powerful still tend to be white or light skinned. An important historical factor determining the present racial structure of Brazilian society was the official sixteenthcentury colonial policy of encouraging Portuguese men to marry native women in their colonies in order to establish a local part-Portuguese community with cultural and political loyalty to Portugal. In Portuguese India the founding governor, Afonso de Albuquerque, carried out this policy by granting such couples a state-subsidized dowry (Marques 1976:249). As a country with a small population – just over a million during this period (Boxer 1969:49) – it was the only way Portugal could maintain its far-flung trading empire, extending from Brazil to what is today Indonesia. Portuguese women simply remained at home in Portugal during this early period. Marques (1976) notes that on the whole the [early] social patterns of Brazil copied that of the Atlantic islands, particularly Cape Verde and S˜ao Tom´e. . . . cross-breeding derived from the needs of nature, not from race equality. The whites were always considered superior to the others and held most offices of leadership, although tolerance and respect toward both mulattoes and mamelucos [Portuguese-Tupi mesti¸cos] reached a higher level in Brazil than probably anywhere else (1976:255, 360)

2.3.4

BVP: the twentieth century

The greater frequency with which Portuguese men fathered children by their slave women probably led to the far higher rate of manumission in Brazil than in the Caribbean. Racial mixing worked against the maintenance of the rigid caste system that helped to preserve creolized language varieties elsewhere. Later, as Afro-Brazilians had fewer barriers to face in rising socially, they also had more incentive to learn Standard Brazilian

58

Languages in Contact

Portuguese as a mark of their standing; such circumstances seem likely to have accelerated the loss of non-Portuguese features in their speech. As education started to become more widely available, Sousa de Silveira noted in 1921 that “even the Negroes speak better today than they used to” (reprinted in Pimentel Pinto 1981:27). The language-related problems facing modern Brazil are still daunting: out of a total population of 165 million, 17 percent, or some 28 million Brazilians, are illiterate (Famighetti 1997:746) and likely to speak only nonstandard BVP. Azevedo (1989:869) notes that Vernacular speakers must learn to read and write in a dialect they neither speak nor fully understand, a circumstance that may have a bearing on the high dropout rate in elementary schools: as recently as 1983 only 70 out of every thousand students reportedly finished the 8th grade.

Today BVP is the language usually spoken by lower class Brazilians with little education. It differs considerably from Standard Brazilian Portuguese (SBP), the literary language usually spoken by educated middleand upper-class Brazilians, particularly in formal circumstances. It should be noted that BVP is not a variety spoken only by black Brazilians. The most divergent varieties of BVP and SBP are the extremities of a continuum of lects that relate more to social class and education than to race. Although blacks are certainly overrepresented in the lower class and underrepresented in the upper class, the structure of Brazilian society is such that all sociolects have speakers of all races, just as other aspects of Brazilian culture are shared by all ethnic groups. Social variables for language usage are further complicated by the fact that most middle- and upper-class Brazilians are in fact bidialectal: Educated speakers are often ambivalent about the vernacular, caught as they are between a prescriptive norm based largely on written models . . . and a linguistic reality that departs considerably from that ideal model. In a society where literacy and grammaticality of language use are held as the hallmark of education, such ambivalence is inevitable when one is aware that the occurrence of stigmatized features in one’s own speech is too spontaneous and frequent to be dismissed as resulting from occasional slips of the tongue. It is apparent that rules considered typical of the vernacular are present in the native linguistic repertory of educated speakers, who acquire the standard largely through normative coaching, which includes not only formal school instruction but also pressure from family and peers (Azevedo l989:862)

Thus educated Brazilians are comparable to educated African Americans, who use the standard in writing and speaking in formal situations but often use the nonstandard in other social situations to signal intimacy or solidarity (1989:868). It is often not clear which variety is a speaker’s

Social factors in partial restructuring

59

first or dominant language. There are rural usages that are unlikely to be found in the speech of educated Brazilians, although many such usages have become part of urban sociolects through immigration from the countryside. 2.3.5

Helv´ecia Portuguese

The most extreme case of nonstandard rural usage that has been documented is the restructured Portuguese spoken in the village of Helv´ecia in the state of Bahia (HP), recorded by Silveira Ferreira in 1961 but published some years later (1985). HP is described by Baxter (1987, 1992, 1997), and Lucchesi (2000). The paucity of the information available about the social history and linguistic structure of HP makes it difficult to extrapolate about the general sociolinguistic conditions that shaped this variety. Discovered during the course of a dialect survey of the state of Bahia, HP was described in a short article (Silveira Ferreira 1985) that received little attention until recently. Helv´ecia (from the Latin name for Switzerland) is an isolated, onestreet village in the southern part of the state of Bahia, which Silveira Ferreira reached by jeep, traveling for seven hours over a rough road from Nanuque. She stayed there only one night, but the young people she first met (who spoke the usual variety of regional BVP) diziam que naquela cidade havia muita gente que falava diferente, “engra¸cado,” principalmente os mais velhos, e acrescentavam ainda que muitas vezes era dif´ıcil, para eles mismos, filhos da terra, entenderem. [“They said that in that town there were a lot of people who spoke differently, ‘funny,’ mainly the oldest people, and they added that many times it was difficult even for them – natives of the town – to understand them.”] (Silveira Ferreira 1985:23)

Helv´ecia was founded in the eighteenth century by Swiss (speaking both German and French) and Germans, who bought slaves in the state of Bahia in order to raise coffee. It was first referred to in 1818 under the name Colonia ˆ Leopoldina, and described forty years later (Tolsner ¨ 1858, cited in Neeser 1951) as a prosperous community of 40 plantations with 200 whites (mostly German and Swiss, with some French and Brazilians) and 2,000 blacks. Lucchesi (2000) notes that Helv´ecia had a relatively high proportion of African-born slaves (some 40 percent) at this time. Silveira Ferreira concluded that With the decline of the coffee industry, the whites gradually abandoned the area, leaving as their legacy the name of the town, mixed descendants, and some family names. The blacks, however, remained isolated, continuing to speak a creole that must have had general currency among them since vestiges of it persisted until 1961. (1985:22)

60

Languages in Contact

Lucchesi (2000:84) describes HP as the result of semi-creolization. From what is known about the community’s history, its main distinguishing feature is that it was founded by non-native speakers of Portuguese. German seems likely to have served as a lingua franca among the first generations of Europeans, who probably learned much of their Portuguese from their slaves in order to communicate with them. If this was the case, the Europeans’ Portuguese could not have served – as seems likely elsewhere in Brazil – as a model for decreolization. Since the slaves were originally purchased in Bahia in the eighteenth century, they must have already known some Portuguese – apparently in a restructured or second language form, to judge from the variety that survived there until the twentieth century. It is possible that this variety was also influenced by the second language Portuguese of the first generation of Europeans, although their children must have become proficient in the Portuguese of the slaves. The isolation of the community was certainly not complete, since some Brazilians joined it, slaves from Africa were brought in (at least until abolition in 1888), and trade with the outside brought it prosperity. Contact with the less-restructured BVP of the region seems likely to have increased over the years, bringing the speech of younger people more in line with that of other communities in the area. Although the sociolinguistic history of Helv´ecia Portuguese is not typical of BVP varieties in a number of respects, and its existence cannot be viewed as unambiguous evidence that fully creolized Portuguese was ever widely spoken in Brazil, it does demonstrate that Portuguese was restructured there to varying degrees. 2.4

NSCS: the sociolinguistic setting of its development

Today the Caribbean basin constitutes a generally definable dialect area of American Spanish, although there is some disagreement on exactly how far inland this dialect is spoken on the mainland. It includes the varieties found in the Greater Antilles (Cuba, the Dominican Republic, and Puerto Rico), pockets in Trinidad and Guyana, the Caribbean coasts of Venezuela and Colombia, and some coastal areas of Central America and Mexico, as well as offshoots in the Miami and New York City areas of the United States. 2.4.1

NSCS: the sixteenth to eighteenth centuries

The Spanish colonized many parts of this area for over a century before the establishment of British, French, and Dutch colonies in the seventeenth century. The indigenous Arawakan languages of the Greater

Social factors in partial restructuring

61

U . S . A . AT L A N T I C OCEAN Miami Gulf of Mexico

Mexico Puerto Rico

Cuba Mexico Belize Coastal area of Guerrero and Oaxaca

Guatemala El Salvador

Small groups of Afro-Hispanics without numerical importance.These groups probably never spoke an Afro-Hispanic variety

Jamaica Honduras & Nicaragua

Haiti Dominican Republic Caribbean Sea Aruba, Curaçao, Bonaire Trinidad

Cartagena Panama

Caracas

Costa Rica

Barlovento Parts of the

Bogotá state of Zulia The lowlands of Colombia (Chocó, Cauca Valley, Nariño)

Colombia Popayán

Ecuadorian coast (especially the province of Esmeraldas)

State of Miranda

Venezuela

Patía (Colombia)

Chota Valley (Ecuador)

Quito

Ecuador

Guayaquil P A C I F I C

O

C

E

A

N

Brazil Peruvian coast (especially between Trujillo and Chiclayo)

Per u Lima

Peruvian coast (especially between Chincha and Ica)

B ol i v i a

Map 4 ‘Localizacion ´ aproximada de a´ reas afro-hispanas en Am´erica’. Originally published in Perl/Schwegler (eds) (1998) Am´erica Negra: panor´amica actual de los estudios ling¨u´ısticos sobre variedades hispanas, portuguesas y criollas (Iberoamericana). Reproduced by kind permission of Iberoamericana.

Antilles died out during the course of the sixteenth century as their speakers were virtually exterminated by forced labor and disease or absorbed into the Hispanicized population. During the same period the large-scale importation of African slaves began, but in most parts of the Caribbean basin the populations of the Spanish colonies did not become overwhelmingly black by the middle of the eighteenth century as did those of the British, French, and Dutch, who in the main concentrated more on laborintensive sugar plantations. Until the nineteenth century the Spanish islands generally had small farms raising various crops (e.g. tobacco, coffee,

62

Languages in Contact

cattle) on which a small number of slaves worked alongside poor whites and a steadily increasing number of free mulattos. It has been claimed that “Whites in Iberian America felt less hostility toward and a closer affinity with people of color, especially if they were light-skinned, than did the French or Anglo-Saxons” (Cohen 1980:108). Conditions on the Spanish islands were thus more conducive to the Africans’ linguistic and cultural assimilation than they were on the overwhelmingly black sugar islands with the rigid racial caste system of plantation slavery (Mintz 1971). Spanish slave laws were also more liberal; under the Cuban system called coartaci´on, slaves could purchase their own freedom by paying instalments. Through this and more prevalent manumission, over 40 percent of Cuba’s non-white population was free by 1774 (Laurence 1974:492) – although these figures were to change in the next century. At this time Cuba’s population was 56 percent white, 18 percent free colored, and 26 percent slave (1974:495). During this period slaves accounted for only 10 percent of Puerto Rico’s population (Reinecke 1937:266), although at the the beginning of colonization in the sixteenth century whites were outnumbered four to one by Indians and African slaves (Reinecke et al. 1975:144). De Granda (1970b:76) interprets the linguistic data in Alvarez Nazario (1961) as evidence of a Puerto Rican creole that existed until the nineteenth century. In the Dominican Republic, where over 80 percent of the population has some African ancestry (Reinecke et al. 1975:146), similar social conditions seem to have led to fairly thorough linguistic assimilation. However, on the Caribbean coast of Colombia, which also has a sizeable population of African descent, there is a direct reference to an early pidgin in Sandoval (1627), a Jesuit who lived there from 1605 to 1652 (de Granda 1970a). He refers to a language used between Spaniards and African slaves of various ethnolinguistic groups, which was “corrupt Spanish, as the Negroes commonly speak it . . . influenced by the Portuguese they call the language of S˜ao Tom´e” (Sandoval 1627: p. 94 of 1956 edn). This and the later emergence of Palenquero Creole Spanish in northern Colombia provide sufficient evidence that a Spanish-based pidgin based on Afro-Portuguese did in fact exist in the Caribbean, but it does not confirm the speculation of Bickerton and Escalante (1970:262) that there existed “a Spanish-based creole spoken in many parts of the Caribbean during the sixteenth and seventeenth centuries” since we do not know (apart from Cura¸cao, where there is an unresolvable question regarding the continuity of Spanish after the Dutch takeover) the fate of pidgins and jargons that may have been spoken elsewhere in the region.

Social factors in partial restructuring

2.4.2

63

NSCS: Habla bozal in the nineteenth century

However, there is fairly clear evidence that a pidgin developed on Cuba during the nineteenth century, although it is less clear that it ever became a true creole. Large-scale sugar plantations were rapidly developed in Cuba towards the end of the eighteenth century when the Spanish attempted to gain the markets for sugar lost by the French in Saint Domingue after independent Haiti plunged into chaos. This caused the importation of slaves into Cuba to increase steadily: the proportion of the population that they made up grew from 26 percent in 1774, to 36 percent in 1817, and 43 percent in 1841. Cuba’s bozales or African-born slaves developed a language which Pichardo (1862:vii) described thus: Este lenguaje es comun ´ e id´entico en los Negros, sean de la Nacion ´ que fuesen, y se conservan eternamente, a m´enos que hayan venido mui ninos: ˜ es un Castellano desfigurado, chapurrado, sin concordancia, numero, ´ declinacion, ´ ni conjugacion ´ . . . en fin, una jerga m´as confusa mi´entras m´as reciente la inmigracion. ´ [This speech is uniform among the Negroes, no matter from which nation they come, and is retained throughout life, unless they came very young. It is distorted and mutilated Castilian, without concord, number, declension, or conjugation . . . in sum, a jargon the more confused the more recent the arrival.]

Based on transcriptions of the speech of aged Cubans of African descent recorded earlier in the twentieth century and published later (Cabrera 1969), Otheguy (1973) concluded that the variety was a creole (1973:335). Although he was undoubtedly right in considering it to be “related to other Caribbean Creoles . . . through a common West African origin” (1973:332), this may have been via the typological similarity of the languages which influenced both rather than via a creole family tree, since there is evidence that bozal Spanish was not a creole. Its use was confined to slaves born in Africa, so it was not a mother tongue; Pichardo (1862:iii) states quite specifically that “Los Negros Criollos hablan como los blancos del pa´ıs de su nacimiento o vecinidad” [“the Negroes born in Cuba talk like the local whites”]. However, certain features of bozal Spanish survived among rural black Cubans (Reinecke 1937:271), suggesting that there was partial nativization resulting in partial restructuring (Figueroa 1999; Ortiz 1998). 2.4.3

NSCS: the twentieth century

More recently, Green (1997:190ff.) analyzed the partially restructured variety of Dominican Spanish referred to as Poror´o in Villa Mella and also

64

Languages in Contact

Table 7. Estimated proportion of whites in the Greater Antilles in the late eighteenth century Colony

Year

Percent

Source

Cuba Puerto Rico Santo Domingo

1774 1776 1794

56% 45% 34%

(Laurence 1974:495) (Alvarez Nazario 1974:76) (Lorenzino 1991:13)

spoken in Cambita, two small rural communities in the San Cristobal ´ district near Santo Domingo. A considerable proportion of their population is descended from African slaves who worked on local sugar plantations during the colonial period. Today the survival of the Afro-Hispanic roots of their culture is evident not only in the local speech variety but also in their cofrad´ıas, or brotherhoods, showing syncretism with African religious fraternities (1997:6). Green based her analysis of the grammar of this variety of NSCS on her recordings of thirty-three speakers living in these communities, ten of whom spoke basilectal varieties (1997:4). It should be said that John Lipski (personal communication, October 9, 2002) has suggested that the speech of two of these thirty-three speakers may be abnormal because they may have “a congenital and probably genetically-based problem.” This seems a somewhat impressionistic diagnosis, but I have not quoted the speech of these informants without noting that they, rather than others, used a particular structure. Finally, important new light on the relationship between fully and partially restructured Spanish in the Caribbean comes from a recent study (Schwegler and Morton 2002) documenting morphosyntactic features hitherto unknown to linguists in the regional vernacular Spanish of bilingual speakers of Palenquero Creole Spanish in Colombia. In assessing the historical likelihood of partial restructuring in the region as a whole, it has to be borne in mind that sociolinguistic conditions on each Spanish island differed to some degree. However, each colony was part of the same political, economic, and cultural system, and the similarity of the distinctive features of the Spanish spoken in each country today reflects the strength of their ties to one another during the Spanish colonial period. Although colonial census figures for the Greater Antilles are disparate, taken in different years with demographic groupings that are not always comparable, we can get an overall impression of the sociolinguistic setting for the development of NSCS if we compare the proportion of whites on each island at the end of the eighteenth century. These figures will be discussed in the concluding section, 2.6.

Social factors in partial restructuring

2.5

VLRF: the sociolinguistic setting of its development

2.5.1

VLRF: the seventeenth and eighteenth centuries

65

R´eunion, a small volcanic island lying some 400 miles east of Madagascar in the Indian Ocean, was uninhabited when the Portuguese first came upon it in 1511. It was settled in 1663 by some Frenchmen and Malagasies from Madagascar. The island, called Bourbon until 1848, became a permanant colony as an outpost of the French East India Company in 1665. In 1671 the population consisted of 36 whites, 37 Malagasies, and 3 mixed children born on the island (Papen 1978:9–10). In 1674 the colonists wrote to the company and “asked for wives, most of them having had to marry Negresses, their slaves” (quoted by Cohen 1980:49). Some young Indo-Portuguese women were brought from Daman and Surate (Chapuis forthcoming) and from Pondicherry, the last a French enclave in southeast India earlier held by Portugal (Stein 1984:90), and by 1686 the island’s population of 269 included 144 children born to French fathers and mothers who were either Indo-Portuguese, French, or Malagasy (Baker and Corne 1986:167). Chaudenson (1974:455) notes that of the 36 Frenchmen with families in 1686, 12 had Indo-Portuguese wives, 14 had Malagasy wives, and 10 had French wives. They and their children made up 75 percent of the island’s population, the rest being Malagasy and Indian slaves. The vernacular French of R´eunion appears to have taken something approaching its present form during this early period. In court testimony taken some time between 1715 (Chaudenson 1974:1147) and 1723 (Baker and Corne 1982:6), a sentence appears whose features are those of modern VLRF (Hull 1979:212): “Moin la parti maron parce qu’Alexis l’homme de jardin l’´etait qui fait a` moin trop l’amour,” “I ran away because Alexis, the gardener, made amorous advances” (my emphasis). In modern VLRF la marks the perfect (or “pass´e perfectif ” according to Watbled 2002) and lete ki marks the durative past (or “pass´e imperfectif” [ibid.]; see section 3.5.1). Because of the latter construction, Baker and Corne (1982:111) referred to the vernacular of the early colonists as lete ki French. According to their interpretation, this variety was influenced by the French spoken as a second language by the non-French wives and slaves, and it became the first language of the children (particularly of mixed marriages) who were born on the island. During this early period of settlement, free people made up the majority of R´eunion’s population, and they usually had access to at least one parent who was a native speaker of French. Although some restructuring took place under these conditions, it resulted more from the borrowing of isolated non-French features

66

Languages in Contact IN D I A N

St Denis 0 0

10 km

Ste Marie Bellepierre La Montagne

N1

5 miles

Le Port

Ste Suzanne

OC E A N

La Possession

W

Dos-d'Ane Cirque de Mallate

St Paul

IN D I A N

St Andre

in

2277

dw

a

rd

St Benoit

Salazie

id

2992 3000

2896 Trois Bassins e e w

Ste Rose

N1

a

St Leu

Plaine des Makes

LEGEND (a) Areas above 400 meters

Les Avirons

e

L

O C E A N

S

Grand Ilet

Cilaos

r

Plaine des Palmistes

d

1606

La Plaine des Caffres

Entre Deux

Etang Sale

2631

i

Tampon d

Predominantly (East) Indian

Area

e

Pierrefonds N3

St Louis

Predominantly White

Vo l c a n i c

S

Predominantly Mixed and Black Scattered groups of (East) Indians predominant Main roads

Plaine des Gregues

Petite Ile

St Pierre

N2

St Philippe St Joseph

St Denis

LEGEND (b)

0

1

Basilectal varieties of VLRF

0

3

5 miles

5

6

7

Acrolectal varieties of VLRF

10 km

2

4

Le Brulé

LePort

8 9

Dos-d'Ane 11

Grand Ilet 14

IN D I A N

18

Grand Place

12

13

17

10

20

O C E A N

Lanouvelle Trois Bassins 23

22

21

Lachaloupe

Cilaos

26

25

Tevelave 28

AF RIC A

27

29

33

30

32 34

asc ar

38

36 37

St Pierre

d ag Ma

Bois Blanc

Entre Deux

35

Réunion

24

Plaine des Cafres

Riv. St Louis

31 IND I A N OC E A N

Saint Benoit

16

19

38 40

St Philippe

Map 5 (a) ‘Carte de L’ˆıle de la R´eunion’. R´eunion Ethnic groups. Originally appeared in the Chaudenson (1998) Cr´eolisation du fran¸cais et francisation du cr´eole: Le cas de Saint-Barth´elemy et de la R´eunion (previously unpublished paper presented at the Conference on Degrees of

Social factors in partial restructuring

67

rather than the complete reanalysis of the base language’s structure that characterizes full-fledged pidginization and creolization. The cultivation of coffee on R´eunion started in 1715, beginning a plantation system that had far-reaching effects on the island’s demography. Over the next half-century the island’s population grew from 2,000 in 1717 (of whom 45 percent were whites according to Chaudenson 1974:458–459) to 27,700 in 1767, of whom only 5,300 were free as opposed to 22,400 slaves (Papen 1978:13), brought in mainly from Madagascar and East Africa, but also from India and even West Africa. Baker and Corne (1982:111, 1986:166) hypothesize that as the proportion of the slave population rose from 50 percent to 80 percent of the total, the new slaves began a period of pidginization and creolization, leading to the emergence of the predecessor of the modern basilectal Cr´eole des Bas (see below) and a continuum of varieties ranging from the regional varieties then spoken in France to the partially restructured vernacular lects of the island, which had already stabilized before this period began but which may have taken on additional creole features. However, the sociolinguistic situation on R´eunion during the eighteenth century differed in one important respect from that of Mauritius during the same period, or that of Haiti in the late seventeenth century. In both of the latter colonies, which developed fully creolized varieties of French, there was a rapid influx of a large, non-French-speaking slave population with little access to the French used by native speakers. On R´eunion, on the other hand, the first half century of settlement had led to a variety of French that had undergone only limited restructuring, which was accessible to the quarter or more of the population that consisted of slaves. This group had sufficient contact not only with the early colonists but also with the later slaves to act as an effective agent of language transmission, so that the restructuring of the local French spoken on R´eunion (i.e. lete ki French in Baker and Corne’s terminology) was not as complete as the pidginization and creolization of French on Mauritius and Haiti. <

Map 5 (cont.) Restructuring in Creole Languages, University of Regensburg, June 24– 27 1998). Reproduced by kind permission of Robert Chaudenson. (b) ‘R´eseau ALR´e’. R´eunion Lects. Originally appeared in the Chaudenson (1998) Cr´eolisation du fran¸cais et francisation du cr´eole: Le cas de Saint-Barth´elemy et de la R´eunion (previously unpublished paper presented at the Conference on Degrees of Restructuring in Creole Languages, University of Regensburg, June 24–27 1998). Reproduced by kind permission of Robert Chaudenson.

68

Languages in Contact

2.5.2

VLRF: the nineteenth and twentieth centuries

The French Republic’s emancipation of slaves in 1794 led to an open rebellion among the colonists on R´eunion, but this was soon diverted by the attacks of the British. Napoleon re-established slavery in 1803 and renamed the island Bonaparte. By 1810 the British had captured R´eunion, Mauritius, and the Seychelles; in 1814 only R´eunion was returned to the French, who promptly renamed it Bourbon. Freed from the administrative and economic domination of Mauritius (Papen 1978:16), R´eunion began a rapid expansion of the cultivation of sugar cane, which soon replaced coffee as the island’s main crop. When slavery was permanently abolished by the French Republic in 1848 (which renamed the island R´eunion), its population of 110,000 included some 60,000 slaves. Most of the freed blacks refused to continue working for their former masters and established themselves on the island’s high central plateaus, also inhabited by the descendants of Maroons who had been fleeing there to escape slavery on the lowland plantations since the beginning of the eighteenth century. The economic situation of whites holding small farms had been deteriorating since the late eighteenth century. Land was becoming increasingly scarce because of the continuing immigration of French peasants and a high birth rate. Holdings became even smaller since land was inherited not by the eldest son but rather by all children, among whom it was divided equally. Many whites became impoverished and took refuge in the higher elevations of the volcanic mountain slopes, where they lived like the Maroons who had fled there before them by hunting, fishing, and poaching. The absorption of small holdings into large sugar plantations aggravated this situation, and by 1836 two-thirds of the island’s white population was described as “indigent” (Papen 1978:13). The colonial authorities viewed with increasing alarm the deforestation carried out by the Petits Cr´eoles, the poor whites who were joined by freed slaves in the highlands. Chapuis (forthcoming, citing Grondin 1998) notes that later the whites came to be referred to as Petits Blancs to set them apart from the non-white populations of this area. In 1869 it had been observed that “They have tried, through the centuries – after an initial racial mixing that is beyond any doubt – to remain as white as possible, despite the fact that some Petits Blancs . . . are practically black” (quoted in Bourquin 1994:114, trans. by Chapuis). The modern local connotations of the word Cr´eole in reference to people (which is certainly related to the use of the same term in reference to their language, regardless of what this term might mean elsewhere) is made clear in the following passage:

Social factors in partial restructuring

69

ici . . . Cr´eole est devenu synonyme de R´eunionnais, avec nuances: la majorit´e des Cr´eoles ne sont pas exactement blancs mais tous les natifs de la R´eunion ne sont pas admis a` la dignit´e de Cr´eole; ainsi l’enfant du “zoreil” de passage ou celui du Chinois . . . [Here . . . Creole has become the synonym of R´eunionese, with nuances: the majority of Creoles are not exactly white, but not all the natives of R´eunion can be called Creoles; e.g. the child of a passing metropolitan Frenchman, or that of a Chinese . . .] (Vaxelaire 1992:147, my translation)

This complex social identity surely bears a relationship to the group’s speech, with its mixture of features associated both with Europeans and non-Europeans. Baker and Corne (1982:111) see the Cr´eole des Hauts or highland vernacular as descended from metropolitan and the earlier local lete ki French, but influenced by the Cr´eole des Bas or lowland vernacular that developed during the eighteenth century among slaves and others, and was then learned as a second language by the indentured servants from southern India who were brought in after the abolition of slavery in 1848 to supply the need for labor on the sugar plantations. Today the poor white population (30 percent of the total of 692,000 in 1997) typically inhabits the mountainous interior, while the coastal lowlands are occupied by people whose ancestry is Indian (25 percent) or African and mixed (45 percent). The latter groups tend to speak more basilectal lowlands VLRF, while the whites tend to speak a less basilectal archaic highlands variety (Baker and Corne 1982:12, 110), although there is a considerable overlap of features. Chaudenson (2000) confirms this opposition, noting that speaking the lowlands basilect is called koz´e an nwar, “talking black” (2000:365), although he notes that non-whites living among whites talk like whites and vice versa. He refers to the highlands variety as fran¸cais cr´eolis´e, or ‘Creole-influenced French’, and the lowlands variety as “un cr´eole moins basilectal que dans d’autres aires,” ‘a creole that is less basilectal than those found elsewhere’ (2000:373). The lowlands basilect is marked by the unrounding of the French rounded vowels and the depalatalization of French palatal fricatives, as well as certain morphosyntactic constructions (see section 3.5.1). These markers are well known and there is no problem in mutual comprehension (2000:368). However, Chapuis (forthcoming) asserts that . . . a clear distinction between the lect of . . . a Petit Blanc and a Kaf or dark-skinned speaker (a mixture of African and Malagasy origin) cannot be established linguistically. . . . What has, however, been considered a watershed between acrolectal and basilectal speech varieties is the use of the long form of the verb for the past progressive by the Petits Blancs and the use of the t´e i form by the rest of the

70

Languages in Contact

population. Not that the same speaker cannot use both in consecutive sentences or even the same sentence (Cellier 1985a).

This opposition of lects was valid until the 1960s; there is now a complex interrelationship of these varieties with more acrolectal urban speech – a continuum with two distinct basilectal extremities. Although the most basilectal lects of R´eunion French are clearly more restructured than the other four languages examined in this volume, one of the linguists who knows the structure of the VLRF best concludes that “. . . dans l’´evolution du fran¸cais au cr´eole [r´eunionnais] la cr´eolisation a e´ t´e incompl`ete . . .” [In the evolution of French to R´eunion Creole, creolization has been incomplete] (Cellier 1985a:135). Yet Watbled (forthcoming) is also right to assert of VLRF (“cr´eole”): “En d´epit des origines fran¸caises, les structures cr´eoles ob´eissent a` des principes grammaticaux diff´erents” [‘In spite of their French origins, R´eunionnais structures follow different grammatical principles.’] 2.6

Common sociolinguistic factors

We now return to the discussion of the role of sociolinguistic factors in determining the structure of these new language varieties resulting from contact. If we focus on the proportion of whites in the five settings outlined above, an important factor in the partial restructuring of languages emerges. Since race was usually an indicator of whether an individual spoke the European lingua franca as a first or second language during the earliest period, it can be inferred that in all of these situations the maximum percentage of native speakers associated with full creolization – i.e. 20 percent according to Bickerton (1981:4) – was considerably exceeded. If we compare the above figures to the proportion of Europeans in the late eighteenth century in some speech communities where full creoles developed (Holm 1988–89), we find a considerable discrepancy (table 8). In the case of R´eunion, the figures on the ethnicity of the population in the late eighteenth century would seem to be at variance with the above pattern: in 1767 over 80 percent of the population were slaves and only 20 percent were free. This apparent discrepancy stems from the unusual circumstances of the early development of the VLRF: as pointed out in section 2.5.1, the first half century of settlement had led to an only slightly restructured variety of French that was quite accessible to the slave population. In 1717 R´eunion’s population was comparable to those in table 8 where partially restructured languages emerged, being 45 percent white and 55 percent non-white. The non-whites had had sufficient contact with the French settlers during the earlier period to

Social factors in partial restructuring

71

Table 8. Estimated proportion of whites in various societies in the late eighteenth century Colony

Developing language

Percent

Virgin Islands Jamaica (rural) Cura¸cao Virginia Brazil Cape Colony
Cuba Puerto Rico Santo Domingo

Negerhollands Creole Dutch Jamaican Creole English Papiamentu Creole Spanish/Portuguese African American English Brazilian Vernacular Portuguese Afrikaans

c. 6% c. 8% c. 7% c. 59% c. 32% c. 47% c. 56% c. 45% c. 34%

Nonstandard Caribbean Spanish

learn their language with only minor restructuring and then act as an effective agent in transmitting this local French to the new slaves who arrived in the great influx after 1717, with whom they also had sufficient contact to allow for something approaching normal language learning. Thus, the partially restructured French that had developed up to 1717 remained the island’s language, although it surely acquired more non-French features over the next half century, particularly on the lowland plantations. To summarize, partially rather than fully restructured languages developed in those societies with a higher proportion of native speakers of the European lingua franca. This important point will be discussed more fully in the concluding chapter. Before proceeding with a structural comparison of the five languages examined here, it should be pointed out that no claim is being made that these newer language varieties are any more closely related to one another genetically than the older languages out of which they grew, but it is being claimed that they underwent a similar kind of partial restructuring triggered by parallel social and linguistic factors. To keep the comparison of their structure manageable, we will deal first with the morphosyntax of the verb phrase (chapter 3), then with the morphosyntax of the noun phrase (4), and finally with the syntax of clauses (5), presenting the facts about each language variety in turn in each area of grammar so they can be more readily compared. In our conclusion (6), we will propose a theory to account for the the sociohistorical and linguistic facts that we have surveyed.

3

The verb phrase

Introduction The verb phrase has been of central importance in contact linguistics. While it is true that no particular set of syntactic features can identify a language as having undergone restructuring without reference to its sociolinguistic history, it is also true that the structure of the verb phrase has been of primary importance in distinguishing creole varieties (e.g. Jamaican) from non-creole varieties (e.g. Caymanian English) of the same lexical base. In the Caribbean, the non-creoles have their European system of tense marking (e.g. auxiliary verbs and verbal inflections) more or less intact, whereas the creoles have a radically different way of dealing with tense and aspect. With few exceptions, basilectal Atlantic creole verbs have no inflections; instead, they are preceded by particles indicating tense (the time of an action’s occurrence) or aspect (referring to its duration, recurrence, completion, etc.). These often have the outer form of auxiliary verbs from the lexical source language (which occupy a similar position and usually serve a similar function), but semantically and syntactically they are much more like the preverbal tense and aspect markers in many of the creoles’ African substrate languages. The data below show that verbs in the partially restructured languages have few if any inflections – far fewer than their European source languages. In some cases their auxiliary verbs can take on the semantic or even syntactic uses of preverbal markers in fully creolized languages of the same lexical base, giving rise to constructions quite unlike those in their source languages. Further non-European constructions found in these varieties include exotica like discontinuous double negation and non-verbal predicates without copulas.

72

The verb phrase

3.1

The AAE verb phrase

3.1.1

AAE verbal morphology

73

In AAE the simple present tense is usually indicated by the verb stem without any -s inflection in the third person singular, e.g. (1)

AAE: Where Miss Annie . . . live

now. (Schneider 1989:65)

Working-class white Americans usually confine such deletion to a single lexical item (“he don’t”) and only in the negative (although “he do” is found in the dialects of southwestern England according to DARE). In contrast, Wolfram (1969:36) found that lower working-class African American males in Detroit deleted the third person present ending 74 percent of the time. The -s inflection can also occur with other persons in AAE: (2)

AAE: I members de first shoes I ever had. (ibid.; Labov et al. 1968).

The omission of the third person singular -s is highly stigmatized and considered a marker of social class. In AAE a verb with past reference does not need to be marked for the past tense; speakers may alternate between inflected and uninflected forms: (3)

AAE: They taught me mighty good, they teach me good. (Holm 1991:235)

Schneider (1989:81) found that of some 8,000 verbs in a past-tense context in the ex-slave narratives (recordings made between 1935 and 1974 of AAE speakers born before 1861; see section 2.1.5), 75 percent were morphologically marked, indicating that the category of past tense is part of the grammar of AAE, although actual marking of verbs is optional. 3.1.2

AAE auxiliaries/preverbal markers

Mufwene (1983) shows that the semantics of time reference in the AAE verbal system bears out a kinship to the English-based creoles of the Caribbean. However, the use of been as a creole-like preverbal marker of anterior tense or remote past is relatively rare in AAE, even in the ex-slave recordings, e.g. (4)

AAE: I got on a cowboy shirt now that I brought from Texas. Been have it all my days. (Holm 1991:235)

74

Languages in Contact

Decreolizing varieties often replace anterior been with did, had, or was; these are frequently less deviant from standard usage and thus less stigmatized. AAE been plus verb can be made to conform more closely to standard English morphology as an auxiliary in a progressive construction, despite the semantic mismatch: (5)

AAE: I hear jus’ as good now as I ever been hearing. (ibid.)

Some older AAE speakers occasionly use unstressed did to mark the past: (6)

AAE: Let me see how that did come up. (ibid.)

Another remnant of preverbal been may be the AAE use of had in constructions that do not conform syntactically or semantically to non-AAE usage: (7)

AAE: Today I had went to work. (Cukor-Avila 2001:105)

In Gullah and the Caribbean creoles the completive aspect marker done is followed by the uninflected form of the verb, but in the ex-slave narratives it is followed by the past participle: (8)

AAE: Bout eight o’clock he done been all around. (Cukor-Avila 2001:238)

In decreolizing Caribbean varieties, progressive aspect is indicated not by the basilectal preverbal marker de but by the English verbal suffix -ing (without be as an auxiliary verb); this is the construction found throughout the ex-slave recordings, e.g. (9)

AAE: They all going home now. (Cukor-Avila 2001:236)

Although -ing is clearly an inflectional morpheme in English, its status in decreolizing varieties is less unambiguous. In a number of English-based creoles unstressed does marks habitual aspect, a dialectal usage also found in England and Ireland. A parallel construction is found in the ex-slave recordings: (10)

AAE: An’ I does enjoy certain of his show. (Cukor-Avila 2001:237)

In the Bahamas, this habitual preverbal marker does has the reduced forms is and ’s: (11)

Bahamian CE: They is be in the ocean. (Holm with Shilling 1982:111)

The verb phrase

75

Rickford (1980) suggests that the complete loss of these reduced forms left be itself with habitual force in some varieties: (12) (13)

Bahamian CE: Sometimes you be lucky. (Holm 1988–89:160) Bahamian CE: They just be playing. (ibid.)

Invariant be has also taken on the force of a habitual marker in modern AAE: (14)

AAE: Those boys be messing with me. (Cukor-Avila 2001:105–107)

This usage is not usually found in the speech of whites, and has become much more widespread in AAE as a marker of ethnic identity since the middle of the twentieth century (ibid.). The habitual meaning of this construction can be emphasized with steady: (15)

AAE: Them brothers be rappin steady. (Baugh 1983:86)

Another apparent AAE innovation is the combination be done for the future perfect (Cukor-Avila 2001:104–107): (16)

AAE: We be done washed all the cars by the time JoJo gets back with the cigarettes. (Baugh 1983:78)

3.1.3

AAE negation

Some nonstandard features of AAE negation, such as most uses of ain’t, are also found in other nonstandard varieties of British and American English, and their origin is not connected to contact with creolized varieties of English. However, AAE ain’t can also be used to negate verbs understood to refer to past action, such as the following: (17)

AAE: He ain’t do it. ‘He didn’t do it.’ (Rickford 1999:8)

Kautzsch (2002:45) notes that AAE “preverbal ain’t appears to have lost its potential to occur in the present tense and has been restricted to past tense contexts.” This use of ain’t is not found in the speech of whites in the American South who use ain’t in other contexts (Cukor-Avila 2001:105–107). Taken together, the AAE uses of ain’t seem related to the use of ain’t in decreolizing varieties to replace the preverbal negator no found in more basilectal varieties: (18)

Bahamian CE: Stone at sea bottom no know sun hot. (Holm with Shilling 1982:143)

76

Languages in Contact

(19)

Bahamian CE: Bookie ain’ know who do it yet. (Holm with Shilling 1982:3)

This seems likely to have converged with the use of ain’t in other nonstandard varieties corresponding to standard haven’t in the present perfect tense (with partially parallel past reference). Similarly, AAE multiple negation or negative concord (negating not only the auxiliary verb but also all the indefinite pronouns in the sentence) is also found in white speech (Cukor-Avila 2001:105–107). However, AAE and creole English can extend negation to noun phrases as definite as proper nouns: (20)

AAE: We don’ want no six-month investigation! (AAE speaker, Euronews, 11/7/02)

(21)

Bahamian CE: They can’t sell that in no Haiti (Holm with Shilling 1982:143)

For emphasis, AAE can invert the negative auxiliary and the indefinite subject: (22)

AAE: Don’t nobody like him. ‘Nobody likes him.’ (Sells, Rickford, and Wasow 1996)

Negative concord can also be transferred across clauses: (23)

AAE: It ain’t no cat can’t get in no coop. ‘There isn’t any cat that can get into any coop.’ (Labov 1972a:130)

Howe and Walker (2000:110) point out that such clause-external concord (like negative inversion) is also found in non-AAE varieties of nonstandard English but is apparently not documented in any English-based creole. Sentence (23) is particularly difficult for monodialectal speakers of standard English to parse because it also contains a zero subject relative pronoun (section 5.1.2) and AAE it’s ‘there is’, also found in SWVE. This construction seems likely to be linked to Bahamian CE it have idem. via the use of it’s for both ‘it is’ and ‘it has’ (Holm 2000a:200). There are parallels in Bantu languages as well as creoles based on French, Spanish, and Dutch; among partially restructured varieties, BVP uses tem ‘[it] has’ (versus EP ha) ´ and NSCS uses tiene ‘[it] has’ (versus S hay) (ibid.). 3.1.4

AAE non-verbal predicates

Non-verbal AAE predicates have received particular attention in the literature. Labov (1969) did a quantified study of the absence of forms of

The verb phrase

77

be in certain phonological and syntactic environments, which he related to social variables. Holm (1984) related the AVE patterns to those in Atlantic creoles and the African languages that influenced them to trace the role of restructuring in AAE’s genesis and development. Expressed forms of the copula are normally required in standard English and all British dialects. Poplack (2000:20) concedes that “zero copula is perhaps the only variant studied in this volume which cannot be identified as a legacy of English.” Walker (2000:67) implies he has counter evidence regarding this point: “Regardless of the lack of historical examples, zero copula does exist in other nonstandard varieties of English, in locales such as Alabama (Feagin 1979), Mississippi (Wolfram 1974), and Yorkshire (Tagliamonte, p.c.).” Aside from dealing with the likelihood that zero copulas came into the non-AAE varieties of Alabama and Mississippi through contact with AAE, Walker needs to provide more precise information than reference to a personal communication, since many varieties of informal English can omit copulas when they are inverted auxiliaries (e.g. “You going?”). There is massive variation of the AAE zero form of the copula with the expressed forms; sometimes this variation occurs almost within the same sentence: (24)

AAE: They all dead. All of them’s dead. at . . . where they is. Where they The Yankee be to the landing, they drunk. (Holm 1991:239)

The AAE pattern has parallels in Gullah, Jamaican, and ultimately Yoruba (ibid.), providing evidence that AAE resulted from the partial restructuring of English under the influence of similar creole and African languages. 3.2

The Afrikaans verb phrase

3.2.1

Afrikaans verbal morphology

Unlike Dutch, Afrikaans verbs have no inflections to indicate person or number, as can be seen in this comparison of present tense forms: (25)

A: om te help ‘to help’ ek help ons help jy help julle help hy help hulle help

cf. D: helpen idem ik help wij helpen ‘I/we help’ jij helpt jullie helpen ‘you help’ hij helpt zij helpen ‘he helps/they help’

78

Languages in Contact

The relative simplicity of Afrikaans morphology is especially noticeable in its regularization of suppletive forms, such as in the verb meaning ‘to be’: (26)

A: om te wees ‘to be’ ek is ons is jy is julle is hy is hulle is

cf. D: zijn idem ik ben wij zijn ‘I am/we are’ jij bent jullie zijn ‘you are’ hij is zij zijn ‘He is/they are’

There has been a tendency towards morphological simplification in such West Germanic languages as English and Dutch for at least the past half millennium, and there are regional varieties of both languages in which verbal morphology is even simpler than in the standard. However, the loss of inflections is also characteristic of language contact, and there can be little doubt that this tendency in Dutch was accelerated when the language came into contact with other languages at the Cape. Evidence for this can be found in the fact that while the base form of the Afrikaans verb is usually derived from the ik form of the Dutch verb (e.g. help above), this is not always the case. For example, some Afrikaans verbs are based on the Dutch infinitive, e.g. A gaan ‘go’ (cf. D gaan ‘to go’ vs. D ik ga ‘I go’). Such a mixture of loss and retention of Dutch inflections in Afrikaans is strongly suggestive of contact with languages lacking such inflection (e.g. Khoi or Creole Portuguese); as in full creoles, the most frequent superstrate forms were taken to be the base forms regardless of the presence or absence of what had been inflections in the source language. The Dutch simple past tense, with its inflections and many irregular forms, is – with very few exceptions – not found in Afrikaans, which uses for the past a construction derived from the Dutch present perfect, i.e. the auxiliary het ‘have’ plus a past participle beginning with the prefix ge- (as in Dutch) but followed by a stem with almost none of the Dutch irregularities, e.g. “Ek het geskryf ” ‘I wrote; I have written.’ Lockwood (1965:210) notes that in Afrikaans The past part[iciple] is formed from the present stem, so that ablaut no longer plays any role. Thus the most typical feature of the Germanic verbal system, the division into strong and weak classes, has vanished from Afrikaans. This has occurred in no other Germanic language – Pidgin English excepted.

Afrikaans can indicate past action not by a morphological change in the verb but by a temporal adverb (especially toe ‘then, at that time’) used with the present tense form of the verb (except for the few verbs with irregular past tense forms, e.g. was):

The verb phrase

(27)

79

A: Ek het op die systraat gestaan. Toe kom die motorkar vinnig om die hoek en ry vas teen die lamppaal. Daar was ‘n harde geraas . . . ‘I was standing on the pavement. Then the motor-car came rapidly round the corner and rode slap into the lamppost. There was a loud noise.’ (Burgers 1963:118–119)

Such a use of the present-tense form of the verb for the past is unknown in Dutch, but pidgins characteristically indicate tense with adverbials rather than inflections. 3.2.2

Afrikaans auxiliaries/preverbal markers

Finally, a more highly restructured nonstandard variety, Orange River Afrikaans (ORA, which has other non-European features also found in the eighteenth-century Afrikaans of non-Europeans), sporadically uses ga or ge as a preverbal marker of past tense, comparable in form and meaning to both the substrate Nama marker gye or go and the Dutch past participle prefix ge- (Roberge 1994:73–74). Another creole-like preverbal marker in ORA is gedaan, indicating completive aspect: (28)

ORA: Jij mijn Cameraat gedaan vast maken you my comrade PERF-ANT fast make ‘You (have) tied up my comrade.’ (den Besten 1989:225, quoting Franken 1952:50)

The ORA preverbal marker lˆe indicates durative aspect (Rademeyer 1938:78–79); it also occurs in the non-standard Kaaps variety of Afrikaans (KA), e.g. (29)

KA: Hy lˆe wag daar. ‘He always waits there.’ (Makhudu 1984:88–80)

3.2.3

Afrikaans negation

Afrikaans has a postverbal negator (as does Dutch); if any other element follows the verb, there is a second negator at the end of the sentence (not found in Dutch): (30) (31)

A: Sy she A: Sy she

eet eat eet eat

nie. ‘She isn’t eating.’ NEG (Combrink 1978:79) nie pap nie. ‘She does not eat porridge.’ NEG porridge NEG (ibid.)

80

Languages in Contact

Combrink (1978:84) notes that the second element of the double negation found in Afrikaans seems most likely to have been influenced by the postverbal (and therefore sentence-final) negator in Khoi languages like Nama. A study of written documents from the period 1830–44 (Nienaber 1965) revealed that among Coloured speakers of Afrikaans whose ancestors had spoken Khoi, 40 percent of all cases of negation had the second element, but among white speakers only 13 percent did. Later, when Afrikaans was being standardized, its structural differences from Dutch were emphasized to increase its Abstand, or distancing from the other standard, and double negation became fully accepted in white Afrikaans.

3.3

The BVP verb phrase

3.3.1

BVP verbal morphology

There is a drastic reduction of verbal inflection in BVP, which cannnot be due solely to a phonological tendency towards the loss of final -s. The intimate second person forms used in Portugal (e.g. the singular tu partes and the obsolescent plural v´os partis for ‘you leave’) have been replaced by polite forms taking third person endings (singular vocˆe parte and plural vocˆes partem) in most parts of Brazil, except in prayer. The feature distinguishing singular from plural in the third person is nasalization, which often does not occur. The most coherent explanation for this variation is partial restructuring. Without the intimate second person forms, the six distinct verbal inflections for person in the present tense in European Portuguese (EP) are reduced to four in Standard Brazilian Portuguese (SBP) – confining the discussion to one conjugation and noting that there are parallel contrasts in the other two: (32)

SBP: eu vocˆe/ele nos ´ vocˆes/eles

parto parte partimos partem

‘I leave’ ‘you leave/he leaves’ ‘we leave’ ‘you/we leave’

However, in BVP denasalization of the third person plural ending can yield a three-way contrast, with distinctive endings only in the first person singular and plural. This might be attributed to a phonological rule for denasalization. In the data I collected there were a number of sentences like the following:

The verb phrase

(33)

81

BVP: Os alunos . . . que n˜ao conhece . . . ‘The students . . . who don’t know . . .’ [cf. SBP conhecem /konyes˜e/]

However, the loss of contrasting inflections is unlikely to be due to a purely phonological rule because in some lects of BVP the first person plural ending is also nondistinctive (Marroquim 1934:115–116), an unambiguous indication of restructuring: (34)

BVP: eu vocˆe/ele nos ´ vocˆes/eles

parto parte parte parte

‘I leave’ ‘you leave/he leaves’ ‘we leave’ ‘you/they leave’

A further confirmation of the morphological rather than phonological nature of the rules that are needed to account for the loss of endings for person on BVP verbs is found in the rural variety of Cear´a, where forms following even the first person singular pronoun can take what in SBP is the third person singular ending, e.g. eu da ‘I give’ (cf. SBP eu dou) or eu sabe ‘I know’ (cf. SBP eu sei) according to Jeroslow (1974:142, 171). Because of the reduced verbal paradigm, BVP makes greater use of subject pronouns than EP, which is a Pro-drop language that uses them for emphasis only. In BVP subject pronouns are required for all persons except the first singular, since it usually maintains its own distinctive verbal inflection (Mello 1997). It should be noted that most Atlantic creole verbs, which take no inflections, appear to be derived from the imperative form of the verb in the lexical donor language, rather than the infinitive (or possibly the third person singular form, although this is could not be the case in the Englishbased creoles). This can be seen more clearly in the case of many irregular verbs; for example, Guin´e-Bissau CP bay and Papiamentu bai, both meaning ‘go’, are from P vai ‘go!’ (or possibly [ele] vai ‘[he] goes’). It is significant that this is also the form found in BVP, e.g. “Nos ´ vai l´a’ ‘We go there’ (cf. SBP “Nos ´ vamos l´a”) (Rodrigues 1974:208). This two-way inflectional contrast can also be found in the preterit: (35) eu vocˆe/ele nos ´ vocˆes/eles

SBP parti partiu partimos partiram

BVP parti partiu partiu partiu

‘I left’ ‘you/he left’ ‘we left’ ‘you/they left’

In the imperfect tense, a single form (partia) can be used for all persons, replacing the three-way contrast in SBP. In BVP other inflected tense

82

Languages in Contact

forms are rarely used, and the subjunctive mood tends to be replaced by the indicative (Azevedo 1989:866–867). The tense system of Helv´ecia Portuguese appears to have been quite divergent from that of even rural varieties of BVP. For example, one informant in Silveira Ferreira (1985) used e´ (a SBP present-tense form for ‘is’) with what could only be past-time reference. 3.3.2

BVP auxiliaries/preverbal markers

The Portuguese-based creoles use preverbal markers to signal categories of tense and aspect; there are some partially parallel markers in rural varieties of BVP such as foi ‘PAST’ (literally ‘was’) and vivia ‘HABITUAL PAST’ (literally ‘lived’): (36) (37)

BVP: Eli foi dis . . . ‘He PAST said. . . .’ (McKinney 1982:6) BVP: Eli vivia trabayava ‘He HABITUAL-PAST worked’ (McKinney 1982:7)

3.3.3

BVP negation

Negation of the BVP verb can be handled three ways (Schwegler 1996b), as in the following sentences, each of which means ‘He doesn’t know’: (38)

l. Before the VP: BVP: Ele nao ˜ sabe. 2. Before and after the VP: Ele nao ˜ sabe nao. ˜ 3. After the VP: Ele sabe nao. ˜

In pattern 2, utterance-final nao ˜ alternates with nada ‘nothing’: (39)

BVP: El’ nao ˜ faloˆ issu nada. ‘He didn’t say that.’

Older Portuguese had only pattern l, while modern EP has 1 and 2, the latter for emphasis only. Schwegler concludes that in BVP there is a change in progress towards pattern 3. One possible external factor may have been contact with creoles having patterns 2 and 3. S˜ao Tomense Creole Portuguese has pattern 2: (40)

S˜ao Tomense CP: I’n˜e na ka ‘tlaba na’i fa they not ASPECT work here not ‘They do not work here.’ (Ferraz 1976:36)

while the closely related creole of nearby Pr´ıncipe has pattern 3 with only utterance-final fa. Discontinuous double negation is also found in some African languages such as Ewe, which surrounds the verb with the disjunctive negators me . . . o, the first element of which can sometimes be omitted (Boretzky 1983:102).

The verb phrase

3.3.4

83

BVP non-verbal predicates

Some rural varieties of BVP have non-verbal predicates consisting of noun phrases, adjective phrases, and locative phrases, all without copulas: (41)

BVP: Eu Ela i eli

mininu. ‘I [was] a child’ loka pur eli. ‘She [is] crazy about him’ ali ‘and he [was] there’ (McKinney 1975:15)

These constructions are parallel to “zero copula” structures in AAE, (section 3.1.4), the Atlantic creoles, and their substrate African languages (Holm 1988–89:175–178).

3.4

The NSCS verb phrase

3.4.1

NSCS verbal morphology

While scholars of NSCS disinclined to accept any external influence on its structure have long resorted to postulating the internal development of phonological rules to account for the loss of inflectional morphology in the VP, the similar but more pervasive restructuring of some varieties of rural BVP (see section 3.3 above) suggests that phonology alone does not provide an adequate explanation for the reduced verbal inflections in either BVP or NSCS, forcing the issue of morphological restructuring resulting from contact: (42) I buy you buy s/he buys we buy you buy they buy

Spanish

NSCS

Portuguese BVP

compro compras compra compramos compran compran

compro compra compra compramo compra compra

compro compras compra compramos compram compram

compro compra compra compra⇐ compra compra

Traditionally the loss of inflectional distinctions in NSCS was explained as purely phonological: syllable-final /s/ weakens to /h/ or disappears altogether, as in the Spanish of Andalusia (which, of course, was in contact with Arabic for many centuries); and syllable-final /n/ weakens to vowel nasalization or disappears altogether. However, there appear to be no regional varieties of European Portuguese that have undergone parallel phonological changes in any comparable way (pace Naro and Scherre 2000), and the BVP loss of an entire syllable of inflection (P compramos becoming BVP compra) can only be accounted for by the same kind of

84

Languages in Contact

morphological restructuring which, on a more extensive scale, produced the Portuguese-based creoles. Because of the sociolinguistic parallels in the history of NSCS and BVP, it is reasonable to deduce by analogy that it was contact-influenced morphological simplification rather than only internally motivated phonological rules that led to the reduced inflectional distinctions of not only BVP but also NSCS. In fact, the same phonological rules may well have been initiated by externally motivated morphosyntactic rules, as discussed in section 3.3.1 above. 3.4.2

NSCS auxiliaries/preverbal markers

NSCS is less restructured than BVP and there are no unambiguous cases of syntactically induced loss of inflections in most modern varieties of NSCS as there are in BVP. However, some nineteenth-century texts from Puerto Rico and Cuba reveal exactly this kind of morphological change. Note that in the first clause of (43) below, the verb has the third person singular ending quiere despite its first person subject, yo ‘I.’ In the second clause, the form ta (from the Spanish auxiliary esta, ´ as in esta´ cantando ‘s/he is singing’) seems to function as a preverbal marker of durative aspect, as in creoles based on Spanish and Portuguese around the world. (43)

NSCS: y mientre ma te quiere yo . . . tu´ no ta quer´e a m´ı ‘and the more I love you . . . you don’t love me’ cf. S: y mientras mas te quiero . . . tu´ no me quieres ´ (Alvarez Nazario [1974]:190)

This variety of Afro-Caribbean Spanish was like most Spanish and Portuguese-based creoles in that it used the third person singular form of the present tense as the general form of the verb. Green (1997:136ff.) found some sporadic contemporary examples of this in the unusually basilectal variety of Dominican Spanish she studied (see section 2.4.3): (44)

NSCS: . . . nosotroh iban . . . ‘we went’ (Green 1997:138) (cf. S: nosotros ´ıbamos)

Spanish is a Pro-drop language in which verbal inflections indicate person, usually making subject pronouns unnecessary except for emphasis. However, because NSCS has a reduced verbal paradigm like BVP, it requires greater use of subject pronouns than other varieties of Spanish to make clear which person is the subject. 3.4.3

NSCS negation

Like AAE, NSCS has extended the rules of multiple negation or negative concord from those of the standard (negating the verb and all the

The verb phrase

85

indefinite pronouns in the sentence) to negating nouns as definite as a proper noun: (45)

NSCS: Jos´e no fue a ningun ´ Boston. ‘Jos´e didn’t go to Boston.’ (Puerto Rican informant [February 1988], cf. S: Jos´e no fue a Boston.

The verb in NSCS can take double negation (Schwegler 1996b), as in (46): (46)

NSCS: Pero yo no me acueldo na deso no. ‘But I don’t remember anything about that.’ (Ortiz 1996:200)

This construction occurs in the vernacular Spanish spoken not only in the Greater Antilles but also in coastal Colombia and Venezuela – but not in European dialects of Spanish. There is a parallel construction in Palenquero, the creolized Spanish of northern Colombia (Schwegler 1991; Dieck 2000): (47)

Palenquero CS: Nu abla ma nu. NEG speak more NEG ‘Don’t say any more.’ (Bickerton and Escalante 1970:259)

3.4.4

NSCS non-verbal predicates

Non-verbal predicates are rare in NSCS, but do occur sporadically. Green (1997:183ff.) found zero copulas in the unusually basilectal variety of Dominican Spanish that she studied: (48)

NSCS: Yo Bido Dobe. (Green 1997:185) 1s [am] Bido Doble ‘I am a Bido Doble.’ [family names]

Zero copulas are also documented for the NSCS of Venezuela (Alvarez 1990:125).

3.5

The VLRF verb phrase

3.5.1

VLRF verbal morphology

The verbal systems of the various vernacular lects of R´eunion French form a continuum, with uninflected single-form verbs with preverbal

86

Languages in Contact

auxiliaries (or possibly tense-mood-aspect markers more similar to those in the verbal systems of the fully creolized languages) usually occurring more frequently at the basilectal end of the continuum. Apparently inflected verbs and auxiliaries analogous to forms in the verbal system of European varieties of French (possibly reanalyzed as single morphemes) are more frequent at the acrolectal end of the VLRF continuum – although a single individual may use both of the alternative forms. Thus Chaudenson (2000:362) contrasts basilectal VLRF moin te ki dans ‘I was dancing’ with acrolectal mi dans´e idem. (cf. F je dansais [d˜ase]). Some of the apparently conjugated VLRF verb forms can be seen in the following table from Baker and Corne 1982:27 (using their orthographic system, but omitting some allomorphs): (49) Selected VLRF verb forms Present tense Past tense Future (negative) Past participle Infinitive

eat

serve

put

say come

know

be

mˆaz mˆaze mˆazra mˆaze mˆaze

serv serve servira servi servir

me(t) mete metra met/mi met(r)

di dize dira di dir

kone konese konetra koni konet(r)

le lete s(o)ra (e)te et(r)

viˆe vne viˆe(n)ra vni vnir

Despite the phonemic representation, the inflected French verb forms from which these VLRF forms derive are usually unambiguous (e.g. le from il est ‘he is’). However, the status of the above VLRF “inflections” as distinct morphemes is dubious, given sentences like VLRF i f´emalra pa (cf. F C ¸ a ne fera pas mal) ‘That won’t hurt’ (Cellier 1985a:86). In fact, VLRF rapa seems to function as an unbound postverbal marker of (negative) future tense (Chapuis forthcoming). Corne asserts that The morphosyntax of R[´eunion] C[reole] derives, as will be readily apparent, from (mainly seventeenth-century varieties of) French . . . the semantics of the R[´eunion] C[reole] verbal system are also basically “French” (Baker and Corne 1982:13, 29)

However, Ramassamy’s 1985 dissertation was not yet available when Corne made this assertion. Her study, based on the speech of monolingual speakers of the most basilectal variety of VLRF, the Cr´eole des Bas, demonstrates the profound (if subtle) influence on the VLRF of Malagasy, its earliest and most important substrate language. Chapuis (forthcoming) notes how this study reveals that the VLRF (like Malagasy, but unlike French) distinguish processive predicates (i.e. with a

The verb phrase

87

verb that is semantically a process, taking a subject that is the agent of the action), by preceding such verbs with the agentive subject marker (ASM) i: (50)

VLRF: Zan i vyen manz´e aswar. Jean ASM come eat tonight ‘Jean is coming to dinner tonight.’ (Ramassamy 1985:146)

Such processive predicates are opposed to non-processive predicates (semantically not a process but a quality, situation, etc., taking a nonagentive subject), which are preceded by the copula-like particle (CLP) le (cf. F il est ‘he is’) with various forms marked for past (te) or future tense (sra). (51)

VLRF: Mon rob l´e rouz. Is dress CLP red ‘My dress is red.’ (Ramassamy 1985:154)

The agentive subject marker i also has various forms conveying different tenses and aspects, but these are distinct from the corresponding forms of le. Although the distinction between the above two predicates is easy to understand, corresponding to the distinction between a (non-stative) verb and a copula in French and other Western Indo-European languages, other such distinctions are more subtle: (52)

(53)

VLRF: La kaz i e´ kl´er´e. DET house ASM light up ‘The house was lighting up.’ (Ramassamy 1985:68) VLRF: La kaz l´e e´ kl´er´e. DET house CLP light up ‘The house was lit up.’ (Ramassamy 1985:69)

Chapuis (forthcoming) points out that VLRF verbs are inherently stative, and the agentive subject marker i destativizes them. For other interpretations of VLRF i, see B´eraha (2002), Caid (2002), Cellier (1985a), and Watbled (forthcoming). 3.5.2

VLRF auxiliaries/preverbal markers

There is no doubt that the VLRF are characterized by a considerable amount of restructuring. Forms of auxiliaries (e.g. avuar ‘have’ and etr ‘be’ above) combine with the verb to form tenses such as the following (Baker and Corne 1982:26):

88

Languages in Contact

(54)

Present Past

m i mˆaz muˆe lete ki mˆaze m i mˆaze Future: definite m i sava mˆaze indefinite m ava mˆaze negative m i mˆazra pa

Future in past muˆe te i sa mˆaze Conditional muˆe nore mˆaze
Perfect muˆe la mˆaze muˆe la fin mˆaze
Pluperfect muˆe lave mˆaze muˆe te fin mˆaze

‘I eat’ ‘I ate’ ”” ‘I’m going to eat’ ‘I’ll eat’ ‘I won’t eat’

Future Perfect

‘I’ll have eaten’

mi sra fin mˆaze

‘I was going to eat’ ‘I have eaten’ ‘I had eaten’

It should be noted that the English glosses in (54) were provided not by Baker and Corne (1982:26) but rather by the present author with the help of J.-Ph. Watbled, who analyzes the VLRF verbal system somewhat differently (Watbled 2002). This system includes periphrastic constructions expressing progressive aspect in various tenses (muˆe l’apre mˆaze ‘I am eating’), a recent past (mi i sort mˆaze ‘I have just eaten’), and an imminent future (muˆe le pur mˆaze ‘I am about to eat’). But it is not just the pronominal form, the agentive subject marker i, and the lack of agreement for person and number that make the above structures seem un-French. There are two alternative perfect markers, one which is from French (la < il a; lave < il avait) and the other (fin < F finir ‘to finish’) which is French in form only, that are used semantically and syntactically as in fully creolized varieties based on French. Moreover, some of the semantic oppositions (e.g. definite versus indefinite future) appear not to be French at all. (However, Watbled notes in a personal communication that he believes that the VLRF do not, in fact, make this last distinction, which is, to be sure, found in Mauritian Creole French.)

3.5.3

VLRF negation

To negate a verb, standard French places discontinuous negators (e.g. ne . . . pas) around the first verbal element in the verb phrase, which can be either the main verb (55) or an auxiliary (56): (55)

F: Je ne sais pas. ‘I don’t know.’ Is NEG know NEG

The verb phrase

(56)

89

F: Ils n’ ont pas dormi. ‘They haven’t slept.’ 3p NEG have NEG slept

VLRF maintain the pattern of colloquial metropolitan French, in which the first negator can be omitted: (57) (58)

VLRF: Mi kon´e plu. ‘I don’t know any more.’ I know NEG (Cellier 1985a:135) VLRF: Zot la pa vu. ‘They haven’t seen.’ They PERF NEG seen (ibid.)

Fully creolized varieties like Mauritian CF have developed an alternative system of verbal negation, in which the single negator pa always precedes the entire verb phrase. Comparing the three systems, it is striking how close the syntax of VLRF remains to that of standard French: (59)

Standard French

VLRF

Mauritian CF

Je ne mange pas. Mi mˆaz pa. Mo pa manz´e. Je n’ai pas mang´e. Muˆe la pa mˆaze. Mo pa ti manz´e. ‘I don’t eat.’ ‘I haven’t eaten.’ (Watbled, personal communication) Cellier (1985a:136) notes that this similarity of the surface structure of VLRF to that of French suggests the “. . . simplification du fran¸cais plus que . . . la cr´eation d’un nouveau syst`eme linguistique” (‘simplification of French rather than the creation of a new linguistic system’). 3.5.4

VLRF non-verbal predicates

If forms of etr ‘be’ are considered auxiliary verbs in the structures above (e.g. muˆe le pur mˆaze ‘I am about to eat’), then VLRF have verbal predicates with a copula followed by an adjective: (60)

VLRF: muˆe le malad. ‘I am sick.’ muˆe lete malad. ‘I was sick.’ (Baker and Corne 1982:26)

This VLRF structure contrasts with the zero copula in the non-verbal predicates of fully creolized varieties: (61)

Mauritian CF: Mo morisien. ‘I am Mauritian’ (Cellier 1985a:136)

90

Languages in Contact

However, if the above forms derived from F eˆ tre simply indicate stativity (or non-processiveness) and tense, then malad in sentence (60) above could be considered a verb meaning ‘to be sick,’ specified by a preverbal marker. This interpretation is supported by the fact that Malagasy, the principal substrate language, has no copula, although it does have a topicalizer that can occur between the subject and a nominal predicate (Keenan 1978:301). Zero copulas can also occur in VLRF non-verbal predicates that consist of an NP: (62)

VLRF: Sa eˆ bug fite. ‘That is one cunning fellow.’ (Baker and Corne 1982:30)

Cellier (1985a:85) notes that VLRF sa (cf. F c¸ a ‘that’) varies with VLFR se (cf. F c’est ‘that is’) as a “pr´esentatif,” or highlighter, before NPs, as above. However, there are non-verbal predicates not only after sa but also after NPs: (63)

VLRF: L´e chip`ek, kom inn papang. (Cellier 1985a:73) “Grasshoppers” are the same as a “buzzard” [referring to an unpleasant woman].

3.6

A comparison of the Verb Phrase

While no claim is being made that the newer language varieties examined above are any more closely related genetically or typologically than the older languages out of which they developed, it is being claimed that they underwent a similar kind of partial restructuring triggered by partly similar social and linguistic factors. For this reason, the structural differences between the verb phrases of the newer and older languages are compared with one another below in an effort to identify any common patterns of restructuring. 3.6.1

A comparison of verbal morphology

The five varieties examined above all have highly reduced inflectional morphology to indicate personal endings on verbs in comparison to the European languages from which they evolved: AAE, Afrikaans, and VLRF need no such endings at all (although AAE can have -s for all persons) and rural varieties of BVP can have only one distinctive ending for the first person singular, whereas NSCS has distinctive endings in both the singular and plural of the first person. Indication that this morphological simplication was due to language contact as well as a general tendency in the European language towards the loss of inflection can be

The verb phrase

91

seen in the fact that sometimes inflected forms in the source language were selected as base forms in the newer variety (e.g. AAE is, Afrikaans gaan, BVP vai, NSCS ta or VLRF le). 3.6.2

A comparison of auxiliaries/preverbal markers

Although past-tense marking is categorical in BVP, NSCS, and VLRF, it is optional in both AAE and Afrikaans. This optionality would appear to be due to the influence of earlier varieties that were more highly restructured in which tense was indicated not by inflections but by adverbials or preverbal markers. Remnants of the latter can be detected in modern conservative varieties (AAE been, ORA ge, rural BVP foi, basilectal NSCS ta, and VLRF lete) along with preverbal aspect markers that appear to be the remnants of an earlier, more creole-like, verbal system. 3.6.3

A comparison of negation

Superstrate influence can be seen in multiple negation in AAE (as well as BVP and NSCS), but other aspects of negation suggest the influence of language contact. The coincidental similarity of different substrates seems to have led to discontinous double negation in Afrikaans and in BVP and NSCS, as in some of the creole languages which influenced them. 3.6.4

A comparison of non-verbal predicates

Although expressed copulas have apparently always been categorical in Afrikaans, such copulas vary with non-verbal predicates in AAE, some varieties of BVP and NSCS, as well as VLRF. Non-verbal predicates seem likely to be due to the influence of partially overlapping sets of substrate languages, some of which are not shared with Afrikaans.

4

The noun phrase

Introduction None of the inflectional morphology of the noun phrase in the European source languages is preserved – at least functioning as such – in fully creolized languages. This inflectional morphology can be relatively complex. In the English NP, nouns are marked for plural number; while they take no inflectional marking for grammatical gender, they do take a possessive inflection. Dutch nouns, in addition to taking plural inflections, co-occur with articles marked for gender as well as number, and these and other modifiers agree with the head; however, Dutch possessive inflections have largely given way to periphrastic constructions. In Spanish and Portuguese modifiers such as determiners and adjectives must be marked by inflections to agree with the head of the NP in gender (usually -o for masculine and -a for feminine) and number (zero for singular, -s for plural). While number agreement is audible in spoken Spanish and Portuguese, it is often not audible in spoken French, in which the pluralizing inflection (usually -s) is often silent. Still the covert system of plural marking of French, in which nouns co-occur with articles and other modifiers marked for plurality, is nearly as salient as the overt system of plural marking in Spanish and Portuguese. Furthermore, personal pronouns in all five European languages must agree with the noun to which they refer in number and gender – the latter not only in the third person singular, but also in all persons of the plural in Spanish. These pronouns are also usually marked for case to indicate their function as subjects, objects, and possessives. All of these inflections are considerably reduced in the partially restructured varieties, except for number marking in standard Afrikaans and gender marking in BVP and NSCS. Moreover, new constructions with free morphemes have arisen to indicate possession and grammatical categories unknown in the European source languages, such as associative plurals.

92

The noun phrase

4.1

93

The AAE noun phrase

AAE nouns are variably marked for number and possession, although in some varieties the possessive marker is categorically absent. AAE personal pronouns usually mark case. 4.1.1

Number in the AAE noun phrase

As in many West African languages, Caribbean English Creole nouns are not inflected for number, although when relevant plurality can be indicated by juxtaposing a noun with a morpheme that is homophonous with the pronoun meaning ‘they’: (64) (65)

Yoruba: awo ` . n o.kunrin ` [literally ‘they men’] i.e. ‘the men’ (Rowlands 1969:195–197) Jamaican CE: dem bwai / de bwai-dem ‘the boys’

This plural marking usually implies definiteness and is confined to animate nouns. Of course the form is related to the British and American dialectal demonstrative “them boys,” but creoles frequently derived their definite articles not from those of their lexical source languages (which tend not to receive emphasis) but rather from the latter’s demonstratives (Holm 1988–89:191). Some parallel constructions can be found in the language of the ex-slaves: (66)

AAE: them wagon (Holm 1991:240)

However, the English -s inflection also occurs frequently: (67)

AAE: two looms (ibid.)

In quantitative studies of contemporary AAE, the -s pluralizer is nearly always present (Wolfram 1969:143). Rickford (1999:7) notes that its absence is “much less frequent” than the absence of the homophonous verbal or possessive inflection. However, in earlier varieties of AAE such as the ex-slave narratives, there is so much variation that it is not always clear that the -s morpheme is anything more than a stylistic variant: (68)

AAE: had hounds . . . them hound . . . . . six mens . . . six mans . . . six men (Holm 1991:240)

Poplack et al. (2000:100) claim that this grammatical approachment of AAE to standard English is not evidence of decreolization since the varieties of English out of which early AAE grew (which they claim were not influenced by African or creole languages) also had zero marking for some plurals (e.g. “two bushel ”). However, they are unable to offer any

94

Languages in Contact

evidence that the rate of zero plural marking in these varieties was in any way comparable to that in early AAE. Thus the lower rate of plural -s absence in contemporary AAE is indeed evidence of a leveling of features between AAE and other American varieties, particularly in light of the finding of Rickford (1999:273) that his recent studies of this feature in Palo Alto show “no appreciable change” in the rates found by Wolfram (1969) or Labov et al. (1968) that might support a growing divergence between AAE and other varieties. It has been suggested that such nonstandard forms as mans and childrens resulted from decreolization, i.e. the acquisition of the English pluralizing morpheme and its use with what were considered monomorphemic lexical items. Schneider (1989:161) rejects this hypothesis, citing such British dialect forms as foots and feets as likelier sources; however, the burden of proof would seem to be upon him to demonstrate that there were British dialect models for all or even most such AAE forms, and that their use was widespread among southern whites. In addition to the simple plural, AAE and other varieties of English have an associative plural after names of persons: (69)

AAE: Felicia an’ them done gone. (Mufwene 1998:73)

The construction, often pronounced /nεm/, here means ‘Felicia’s friends or family or associates.’ Mufwene (ibid.) notes that AAE shares this construction “with English creoles, rather than with other varieties of English,” although he adds in a footnote that it is also used by whites in the American South. It does indeed seem to be related to a syntactically and semantically parallel construction in a number of African and Atlantic creole languages of various lexical bases: (70)

Yoruba: awo ` . n T´a´ıwo` [literally ‘they Taiwo’], i.e. ‘Taiwo and his family, schoolmates or friends’ (Rowlands 1969:196)

(71)

Miskito Coast CE: di sukya dem [literally ‘the medicineman they’], i.e. ‘the medicine man and his lot’ (Holm 1988–89:193)

Actually the AAE construction is also found in informal use in many parts of the United States (DARE) and England (R. Hudson, personal communication), although it may well have originated in Africa: (72)

Nonstandard E: Mary an’ them came over yesterday.

The tendency of emphasized and conjoined pronouns to take the object case in informal English (whatever their function in the sentence) probably facilitated the borrowing of the AAE associative plural structure into other varieties of English.

The noun phrase

4.1.2

95

Gender in AAE

For all practical purposes, standard English lacks grammatical gender; even British bureaucrats are now nervous about calling a ship she, let alone a country (Michael Pye, personal communication). For this reason, gender agreement between elements in the verb phrase is as irrelevant to AAE as it is to standard English. Dillard (1972:56) mentions AAEspeaking children occasionally producing sentences like “He a nice little girl,” but this is not a feature of adult AAE, although it can be found in Gullah and other varieties of creole English. 4.1.3

Possession in the AAE noun phrase

The English-based creoles indicate possession by juxtaposition rather than inflection: (73)

Miskito Coast CE: di uman biebi ‘the woman’s baby’ (Holm 1978:286)

This is also found in AAE: (74)

AAE: the white folk kitchen (Holm 1991:241)

However, Schneider (1989:162) found the possessive morpheme present in over 90 percent of the 377 cases in which it was possible in the exslave narratives. He notes that while there is variable use of the possessive inflection in all northern urban varieties of AAE, there are southern varieties in which the morpheme’s absence is categorical, suggesting that the suffix has been gaining ground as a part of decreolization (1989:164). Cukor-Avila (2001:106–107) lists the absence of both plural and possessive -s as features found only in AAE, not found in the vernacular speech of whites in the American South. 4.1.4

Pronouns in AAE

In the original pronominal system of most of the Caribbean creoles, it would appear that no distinction was made for gender or case, and the same form also served as a possessive determiner. Possible remnants of such a system can be found in the language of the ex-slaves, although these usages are less frequent in current AAE: (75) (76)

AAE: Well the master had promise’ to, to give we all forty dollars a month in pay. (Holm 1991:241–242). AAE: We had we own lawyers. (ibid.)

96

Languages in Contact

The use of standard English object pronouns in subject position is also encountered in the ex-slave narratives: (77)

AAE: When us all leaves dis old world. (Schneider 1989:66)

Schneider concludes that “there is some evidence that this grammatical variable was more subject to some degree of creolization than others” (1989:177). AAE’s second person plural pronoun, you all /yɔ/, is likely to be an innovation. The pronoun of the same person and number is derived from various African languages in Caribbean Creole English (Holm 1988:203–204), probably motivated by the need for a form of you that is unambiguously plural (cf. similar dialectal forms, e.g. yous, you guys, you chaps, etc.). In fact Eastern Caribbean CE has all you; the AAE form, generalized throughout Southern American English, may be a calque on Twi mo´ nyina, literally ‘you all.’ In AAE (but seldom in the speech of Southern whites) there is a corresponding possessive: (78)

AAE: It’s y’all ball. (Rickford 1999:7)

Like other varieties of English including CE, AAE has pleonastic subject pronouns: (79)

AAE: That teacher, she yell at the kids. (Fasold and Wolfram 1970:81)

These pronouns seem likely to facilitate the parsing of AAE sentences containing relative clauses without subject relative pronouns (see section 5.1.2). Their existence in CE may also have facilitated the adoption of the postnominal dem pluralizer (cf. Cassidy and Le Page 1980:147). Finally, Wolfram (forthcoming) notes that “The regularization of mine to mines in ‘The book is mines’ is quite robust in most varieties of AAVE, though it appears more typical of preadolescent speakers than older speakers.” 4.2

The Afrikaans noun phrase

In comparison to the seventeeth-century Dutch dialects from which it developed, Afrikaans reveals the effects of language contact through the loss of inflectional morphology and the replacement of bound by unbound morphemes in the noun phrase. Although the distinction between common gender (e.g. D de man ‘the man’) and neuter (D het huis ‘the house’) was lost in Afrikaans (die man, die huis), the distinction between singular and plural survived, but a number of morphosyntactic rules were lost or reinterpreted.

The noun phrase

4.2.1

97

Number in the Afrikaans noun phrase

Afrikaans generally marks nouns with a plural inflection, usually either -s after certain sounds or -e elsewhere. However, like many earlier Dutch dialects, Afrikaans uses -s to mark the plural on many nouns whose etyma form the plural with -en (often pronounced as a schwa) in standard Dutch, which also uses -s to mark other plurals. Like Dutch and other Germanic languages, Afrikaans also has irregular plurals with a vowel change: (80)

A: stad ‘city’ versus stede ‘cities’ (Donaldson 1993:69–75)

The final consonant clusters of earlier Dutch nouns often underwent simplification: (81)

D: kast > A: kas ‘cupboard’ (ibid.)

However, the Afrikaans plural ending -e usually preserves the consonant lost in the singular: (82)

D: kasten > A: kaste ‘cupboards’ (ibid.)

However, this is not always the case: (83)

D: kost ‘food’ > A: kos ‘food’, kosse ‘foods’ (ibid.)

This is comparable to AAE des’ ‘desk,’ which has also undergone consonant cluster simplification (Rickford 1999:4), taking the plural form for words ending in a sibilant: desses. In Afrikaans this phenomenon has led to non-historical hypercorrect forms in nonstandard varieties: (84)

D: bus, bussen > A: bus, buste ‘bus, busses’ (Donaldson 1993:69–75) (cf. D: buste ‘bust, bossom’)

The Afrikaans plural markers -s and -e can sometimes both be used on the same noun: (85)

A: beddens ‘beds’ (cf. D: bedden, plural of bed) (ibid.)

The plural endings can also alternate in Afrikaans with different meanings, e.g. man ‘man’ becomes manne in the sense of ‘units (e.g. in war, sports)’ but mans in the sense of gender, e.g. mans en dames (cf. D mannen for both) (Burgers 1963:54). Moreover, a free morpheme unknown in standard Dutch is used in Afrikaans to form an associative plural (section 4.1.1): hulle ‘they, them’ (from dialectal Dutch hun-lui, literally ‘them-people,’ according to Lockwood 1965:210). In informal standard Afrikaans hulle can be

98

Languages in Contact

used only after [+human] nouns like pa ‘dad’ and ma ‘mom’, e.g. pahulle ‘Dad and the others’ (Makhudu 1984:64). However, in Kaaps, the nonstandard Afrikaans of Coloureds, hulle can also be used with nouns that are [−human] but [+animate], e.g. die bome- hulle ‘the trees’ (ibid.). The word goed ‘goods’ can also be used as an associative plural marker in nonstandard and Orange River Afrikaans, e.g. Ma goed ‘Mom and the others.’ Webb (1989) notes that in Khoi the reduced form of the masculine plural pronoun gu ‘they’ can be used after a noun to give it plural meaning, e.g. khoe ‘man’ becomes khoegu ‘the men.’ 4.2.2

Gender in the Afrikaans noun phrase

As mentioned above, the Dutch distinction between common gender (e.g. D de man ‘the man’) and neuter (D het huis ‘the house’) was lost in Afrikaans (die man, die huis). Because of this, the rules governing the marking of these distinctions in Dutch became opaque in Afrikaans. In Dutch, for example, attributive adjectives usually take an -e ending: (86)

D: de jonge man ‘the young man’

There is an exception to this rule before neuter singular nouns with certain determiners: (87)

D: een goed

boek ‘a good book.’

This morphosyntactic rule became a phonological rule in Afrikaans: polysyllabic adjectives take an -e (aangename weer ‘pleasant weather’) but monosyllabic adjectives (except those ending in d, f, g, s, and certain others) do not (A ‘n sterk man ‘a strong man’). The problem is that the phonological rule in Afrikaans has scores if not hundreds of exceptions; Donaldson admits, “It is impossible to formulate water-tight rules for the inflection of adjectives” (1993:170; see also Lass 1990). 4.2.3

Possession in the Afrikaans noun phrase

Possession is no longer normally indicated with inflectional morphemes on common nouns in modern Dutch (cf. Vondels werken ‘Vondel’s works’), bur rather with the preposition van ‘of ’ (de werken van Vondel). The latter is a frequent structure in standard Afrikaans as well, which also has another method of indicating possession with the particle se after nouns: (88)

A: pa se hoed ‘father’s hat.’

This is apparently an extension of the function of colloquial Dutch z’n (cf. D zijn ‘his’); z’n can be used only after masculine nouns (e.g. vader z’n hoed idem), although there is no such gender restriction on the Afrikaans

The noun phrase

99

particle. Kotz´e (personal communication) notes that there is a variant of the Afrikaans structure with se which is not restricted to animates: (89)

A: Die stoel wat by die tafel staan se poot. the chair which by the table stands POSS leg ‘the leg of the chair standing by the table’

Interestingly, the Afrikaans possessive marker se, like the possessive ’s in (colloquial) English, marks the end of an NP and is not a nominal inflection, as in my aunt who lives in New Jersey’s son, or (possibly more salonf¨ahig) the King of England’s crown. 4.2.4

Pronouns in Afrikaans

Like the European Portuguese clitic object pronouns that became free morphemes in BVP (section 4.3.4), the unemphatic personal pronouns of spoken and written Dutch were not preserved in Afrikaans, which has only the forms corresponding to the Dutch emphatic pronouns – with the sole exception of the set phrase “jy weet” ‘you know,’ in which the vowel of the pronoun tends to be reduced to a schwa, like that of the corresponding Dutch unemphatic pronoun – and, coincidentally, that of the English filler phrase, “y’ know” (Donaldson 1993:126). Otherwise singular personal pronouns in Afrikaans are quite similar to those of Dutch, marking not only case distinctions but also gender distinctions in the third person (rare in fully restructured Atlantic creoles). However, the plural personal pronouns make fewer of the Dutch case distinctions for subjects, objects, and possessives: (90)

Subject

Dutch:

1. wij 2. jullie 3. zij 1. ons 2. julle 3. hulle

Afrikaans:

Object ‘we’ ‘you’ ‘they’ ‘we’ ‘you’ ‘they’

ons jullie hen/hun ons julle hulle

Possessive ‘us’ ‘you’ ‘them’ ‘us’ ‘you’ ‘them’

ons/onze jullie hun ons julle hulle

‘our’ ‘your’ ‘their’ ‘our’ ‘your’ ‘their’

The Afrikaans plural possessive determiners above can form possessive pronouns when followed by s’n: (91)

A: ons s’n ‘ours’

Afrikaans can also form possessives through the use of se after the relative pronoun wat (section 5.2.2) and the relative and interrogative pronoun wie ‘who’:

100

Languages in Contact

(92)

A: Wie se boek is dit? ‘Whose book is this?’ (Burgers 1963:101)

When no noun is being modified, this construction also serves as a possessive pronoun: (93)

A: Wie s’n is dit? ‘Whose is this?’ (ibid.)

Compare standard Dutch “Wiens boek is dit?” or “Van wie is dit boek?” with colloquial Dutch “Wie z’n boek is dit?” (Donaldson 1984:72). The Afrikaans possessive relative pronoun construction (“Die man wie se boek ek geleen het . . .” ‘The man whose book I borrowed . . .’) is also from spoken Dutch (“De man die z’n boek ik geleend heb . . .”) rather than written Dutch (“De man wiens boek ik geleend heb . . .”). However, there is also convincing evidence that coloquial Dutch was not the only source of the se possessive construction. The converging influence of parallel particles in Khoi and Creole Portuguese is more clearly seen in the use of se and s’n after personal pronouns – constructions unknown in Dutch. In standard Afrikaans possessive pronouns are formed with s’n after the pronouns u ‘you,’ ons ‘we,’ julle ‘you (plural),’ and hulle ‘they’: (94)

A: Die boek is ons s’n. ‘The book is ours.’

Moreover, nonstandard varieties of Afrikaans use this particle to form possessive determiners in an equally un-Dutch construction: Makhudu (1984) notes the frequent occurence in Coloured Afrikaans of julle se ‘your (plural)’ (but not jy se ‘your [singular]’), while Griqua, or Orange River Afrikaans, has parallel forms: (95)

ORA: hy se huis ‘his house’ (den Besten 1978:31)

In the Khoi language Nama, the genitive particle di also occurs after pronouns as well as nouns: (96)

Nama: //ˆeib di omi ´ [literally ‘he ’s house’], i.e. ‘his house’ (ibid.)

(The symbol // represents a voicless lateral click.) The Malayo-Portuguese genitive marker sa is also used this way: (97)

Papia Kristang CP: eli-sa mai ‘his mother’ (Hancock l969:41, 1975:229)

Furthermore, a parallel marker is used the same way in Bazaar Malay (Hancock 1975:229). Although the use of se after pronouns to form possessive determiners in nonstandard Afrikaans could be viewed as an extension of the rule for its use after nouns in standard Afrikaans (which

The noun phrase

101

could then be traced to colloquial Dutch), the case for its survival from the ancestral languages of the Griqua and Cape Malay in their restructured varieties of Afrikaans seems more convincing. Of course these sources are by no means mutually exclusive, and the most likely scenario is that their influence converged in the development of the modern Afrikaans constructions with se (den Besten 1978:38). In Afrikaans the reflexive pronouns are identical to the object pronouns: (98)

A: Die professor trek hom in sy studeerkamer terug ‘The professor retreats into his study.’ (Brachin 1985:141)

However, -self can be added for emphasis (Donaldson 1993:290). In modern Dutch, the third person singular and plural reflexive pronoun is zich (Donaldson 1984:172), a borrowing from German that did not become prevalent in Dutch until the eighteenth century (Brachin 1985:13), well after the beginning of Afrikaans. Thus Brachin (1985:141) comments on “the archaic nature of Afrikaans.” 4.3

The BVP noun phrase

The BVP noun phrase preserves the categorical gender agreement found in SBP, but tends to mark plural number only on the first item in an NP. Gender and number agreement are categorically absent in the Portuguese-based creoles and variably absent in Helv´ecia Portuguese, although the relevant data from the latter is very limited. BVP object pronouns differ from those of SBP in their case marking and word order, in which they resemble more the pronouns of the Atlantic creoles. 4.3.1

Number in the BVP noun phrase

SBP requires that all determiners, nouns, and adjectives in a noun phrase be marked for plural number: (99)

SBP: os livros velhos ‘the old books’

However, BVP often indicates plurality by adding -s to only the first element (usually a determiner), leaving the plural -s inflection optional on following nouns and adjectives: (100)

BVP: um dos mais velho orix´as ‘one of the most ancient deities’ (Holm 1987:417) SBP: um dos mais velhos orix´as

However, it is not always the first element that is marked:

102

Languages in Contact

(101)

BVP: o meus irm˜ao ‘[the] my siblings’ (ibid.) SBP: os meus irm˜aos

(102)

BVP: todo os mais velho ‘all the most ancient [ones]’ (ibid.) SBP: todos os mais velhos.

Regarding the tendency of BVP noun phrases to mark only the initial element for number – Guy (1989) found that over 95 percent had such marking – this pattern may, as he suggested, represent a survival of the system of marking plurality at the beginning of noun phrases in many Niger-Congo languages. An early variety of restructured Portuguese in Brazil may have had an optional system of marking plurality comparable to that of S˜ao Tom´e CP, in which n˜e, the word for ‘they,’ is used before nouns, e.g. n˜e mwala ‘the women.’ This is comparable to the parallel use for the word for ‘they’ in various African and creole languages discussed in section 4.1.1. Through decreolization, this plural marker could have been replaced by a plural form of the definite article, os (which also functions as the object pronoun ‘them’). This is suggested by the attestation of earlier BVP forms such as osˆele ‘they,’ apparently a combination of a pluralizing os plus ele ‘he,’ instead of SBP eles (Mendon¸ca 1933:67). Support for this interpretation can be found in the parallel use of Cape Verdean CP uʃ ˜ ʃ (cf. the P plural indefinite article uns ‘some’) as a pluralizer: (103)

Cape Verdean CP: uʃ ˜ ʃ rapaz´˜ı ‘(some) boys’ (Almada 1961:92)

It is clear that the variable rule for -s is both phonological and morphosyntactic. As a phonological rule it operates on (synchronically) single morphemes: (104)

BVP: somo ‘[we] are’ SBP: somos

Guy (1989) found clearly monomorphemic instances such as BVP onibu ‘bus’ (SBP onibus) (personal communication). He goes on to point out that at the same time a variable syntactic rule of NP plural marking is required to account for phrases such as as vez, os espanhol, as na¸cao, ˜ because if they resulted from simple S-deletion, they should be as veze, os espanhoi, as na¸coe ˜

due to certain irregularities in the formation of some SBP plurals. Viewed historically, the variable marking for number in modern BVP makes sense only if the variety evolved from an uninflected variety which began borrowing inflections from SBP at a stage when the latter’s system of number agreement within noun phrases and between subjects and verbs was still

The noun phrase

103

opaque to speakers of BVP. The inflections were probably first applied randomly (as in decreolizing English the boy go / the boys goes / the boy goes / the boys go) in free variation. The syntactic rules of the more frequent inflections (-s and nasalization) alternating with their absence in turn led to BVP phonological rules for the same alternation that could then be applied to single morphemes. Naro and Scherre (2000) have claimed that the BVP phonological rules for the variability of -s and the nasalization of vowels were inherited from European Portuguese, but such variation is very unusual in Portugal. It is not clear from the available data whether number agreement in HP is absent or resembles that of BVP. Regarding NP number agreement in the colloquial speech of educated Brazilians, Azevedo (1989:867) notes that it is frequently that of BVP, e.g. “Prova uns paozinho” ˜ ‘Try some rolls’ (SBP paezinhos). ˜ 4.3.2

Gender in the BVP noun phrase

Gender marking appears to be categorical in both SBP and BVP. However, it is not a part of the grammar of the Atlantic creoles or most of their Niger-Congo substrate languages (Holm 1988–89:195). There is evidence that in HP there is no gender agreement between nouns (or pronouns) and adjectives: (105)

HP: ’Ela E ’mu˜itu sa’idu ‘she is very meddlesome.’ Silveira Ferreira (1985:30–31) SBP: ela e´ muito saida

Silveira Ferreira (ibid.) notes that there is an absence of gender agreement between nouns and articles in HP as well, and that nouns do not necessarily take the gender marking of SBP: (106)

HP: ‘ua ˜ a’bota ‘an abortion’ (ibid.) SBP: um aborto ˆ

The variability of gender agreement in HP is the topic of a dissertation by Lucchesi (2000), who notes that such variation is disappearing among younger speakers. He points out that its origin can only be explained by what he calls the “irregular transmission” of rules of morphology, since there are no known instances of the phonological variation of /o/ and /a/, marking masculine and feminine forms, respectively, in the history of the Portuguese language. Bonvini (2000:402) notes that in the ritual “L´ıngua dos Pretos Velhos” gender agreement within the NP is variable:

104

Languages in Contact

(107) (108)

LVP: mia povu ‘my people’ (cf. SBP: meu povo) (ibid.) LVP: tera foi kiadu ‘the earth was created’ (cf. SBP: a terra foi criada) (ibid.)

Finally, Amaral (1928 [1976]:73) notes that in the language of ex-slaves, the pronoun ele (SBP ‘he’) could also be used in reference to females, like the third person singular pronoun in basilectal Atlantic creoles and in most of the Niger-Congo languages that formed their substrate. 4.3.3

Possession in the BVP noun phrase

Possession is normally indicated by the preposition de in BVP, as in standard Portuguese: (109)

SBP: A casa de Maria ‘Maria’s house’

However, in some rural varieties of BVP the preposition can be omitted: (110)

BVP: kaza

Maria [literally ‘house Maria’] (Jeroslow 1975)

There is a parallel possessive construction in S˜ao Tom´e Creole Portuguese, in which the preposition di can also be omitted: (111)

S˜ao Tom´e CP: donu di losa or donu losa ‘[the] owner [of the] farm’ (Ivens Ferraz 1979:69)

4.3.4

Pronouns in BVP

One of the most striking features of BVP is the use of personal pronoun forms that can be used only for emphatic subjects in EP as direct objects in BVP, replacing the clitics of the standard: (112)

BVP: Ela chamou eu. ‘She called me’ (Azevedo l989:863) SBP: Ela chamou-me.

This usage is also frequently found in the casual speech of educated Brazilians; Azevedo recorded a linguist saying the following: (113)

BVP: . . . impediu eles de passar ‘prevented them from passing’ (Azevedo 1989:864) SBP: . . . impediu-os de passar

In many Atlantic creoles there is no distinctive case marking for subject and object pronouns. Moreover, these creoles always preserve their basic SVO word order with object pronouns, unlike Romance languages, in

The noun phrase

105

which direct and indirect object pronouns usually occur before the verb. BVP preserves this word order as well: (114)

BVP: Esses porco a´ı, nos ´ ganhemo eles. ‘Those pigs, we got them as a gift’ (ibid.) SBP: Esses porcos, nos ´ os ganhamos

Turning from personal to reflexive pronouns, the latter are an integral part of SBP grammar: (115)

SBP: Jo˜ao cortou-se com faca. (Mello 1997:153) ‘John cut himself with a knife.’

However, these do not commonly occur in BVP; instead, the following constructions occur: (116)

BVP: a. Jo˜ao cortou com faca. b. Jo˜ao cortou ele com faca. c. Jo˜ao cortou ele mesmo com faca. (ibid.)

The first BVP structure without any pronoun is found throughout the Atlantic creoles and many of their substrate languages, in which any transitive verb can have not only an active meaning but also a passive one if its subject is a plausible object: (117)

Papiamentu CS: E

yama Maria. ‘She is called Mary.’ (Holm 1988:83) cf. S: Ella se llama Mar´ıa.

(118) (119)

S˜ao Tom´e CP: E ple’de.. ‘He got lost.’ (Ivens Ferraz 1979:72) cf. EP: Ele se perdeu. Bambara: To` dun Mali la. ‘Millet porridge is eaten in Mali.’ millet-porridge eat Mali in (Holm 1988:84)

4.4

The NSCS noun phrase

Plural marking in the noun phrase is much reduced in NSCS when compared to the standard, but in most modern varieties this appears to be through phonological rather than morphological rules – although earlier contact could well have played a role in the development of these phonological rules.

106

Languages in Contact

4.4.1

Number in the NSCS noun phrase

The tendency of NSCS to delete word-final -s, which marks the plural forms of nouns, as well as their modifiers such as articles and adjectives in the standard, appears to be the result of a phonotactic tendency in NSCS to simplify a CVC syllabic structure to CV. At least this has been the prevailing interpretation in Spanish dialect studies. This tendency does not always do away with plural marking, which can be indicated by a weakening of /s/ to /h/, especially in feminine forms: (120)

NSCS: lah nina ˜ ‘the girls’ (Lorenzino in Holm, Lorenzino and Mello 2000:203) S: las ninas ˜

However, the total loss of /s/ throughout the NP is more frequent when the noun is masculine because the NSCS plural masculine article lo [cf. S los] still contrasts with the singular masculine form el: (121)

NSCS: lo hombre ‘the men’ [cf. S: los hombres] (ibid.)

A distinct plural form is also retained in masculine nouns ending in a consonant, such as dictador ‘dictator,’ which in standard Spanish require an e before the plural -s: (122)

NSCS: lo dictadore esto ‘those dictators’ S: los dictadores estos (ibid.)

In a more highly restructured variety of NSCS spoken in the Choco´ region of Colombia, the occurrence of the plural form lo dictador without the -e of the allomorph of the Spanish plural morpheme (Ru´ız Garc´ıa 2001) is as clear evidence of morphological restructuring (as opposed to mere phonological rules) as BVP plurals forms like as vez instead of as veze (cf. SBP as vezes ‘the times’) (Guy, personal communication; see section 4.3.1). Studies by Poplack (1978), Terrel (1979), and Nunez ˜ Cedeno ˜ (1980) analyze such plural marking in the noun phrase in the regional Spanish of Puerto Rico, Cuba, and the Dominican Republic, respectively, which in each case is more pervasive the lower the socioeconomic class of the speaker. More recently, Green (1997:190ff.) analyzed the use of -se as a plural marker in an unusually basilectal variety of Dominican Spanish (see section 2.4.3). The weakening and loss of final /s/ in plural nouns ending in a stressed syllable appears to have motivated an analogical reinterpretation of -se as a plural marker in these vernacular Dominican varieties: (123)

NSCS: lo

caf´ese ‘caf´es’ (cf. S: los caf´es)

The noun phrase

107

by analogy with NSCS “lo mese ” (cf. S “los meses”) (Lorenzino in Holm, Lorenzino, and Mello 1999:46). Although the extent of the use of pluralizing -se is unclear in other varieties of Antillean Spanish in Cuba and Puerto Rico, Green (1997:190– 196) indicates that in the Dominican varieties she studied, its use has spread to nouns not stressed on the last syllable: (124)

NSCS: much´achose ‘boys’; l´atase ‘tin cans’ (ibid.)

4.4.2

Gender in the NSCS noun phrase

Nonstandard Spanish throughout the Caribbean retains gender agreement in the noun phrase, which is not found in creoles based on Spanish and other Romance languages. These creoles usually derive nouns and adjectives from the masculine form in the superstrate: (125)

Palenquero CS: cabeza malo ‘bad head’ (cf. S: cabeza mala) (Lorenzino 1993:118).

(126)

Papiamentu CS: baka gordo ‘fat cow’ (cf. S: vaca gorda) (ibid.)

However, Schwegler and Morton (2003) note the sporadic lack of gender agreement in the Spanish spoken by bilingual Palenqueros: (127)

PS: cosa important´ısimo ‘important things’ (ibid.) cf. S: cosas important´ısimas

4.4.3

Possession in the NSCS noun phrase

Possession in NSCS is formed with the preposition de as in standard Spanish, and no other structure (such as that in BVP discussed in section 4.3.3) occurs. 4.4.4

Pronouns in NSCS

NSCS personal pronouns show no reduction of case or gender marking. However, in the NSCS of bilingual speakers of Palenquero (section 1.5), Schwegler and Morton note that reflexive pronouns can be omitted (128d), as they are in Palenquero and other Atlantic creoles based on Romance languages (cf. section 4.3.3), or take the subject form of personal pronoun plus mimo (cf. S mismo as in the emphatic construction yo mismo ‘I myself.’ Of the structures below, they note that (128a) and (128b) are common, while (128c) and (128d) are sporadic:

108

Languages in Contact

(128)

PS: a. Yo me lavo. ‘I wash myself ’ (ibid.) b. Yo me lavo yo mimo. c. Yo lavo yo mimo. d. Yo lavo. (2003: 118)

In the Dominican Republic, the structure in (126d) was used by one of the informants whose speech was rejected as abnormal by Lipski (Green 1997:207). However, Lipski (1998) notes the absence of reflexive pronouns in earlier bozal Spanish. 4.5

The VLRF noun phrase

4.5.1

Number in the VLRF noun phrase

In VLRF nouns and their modifiers have “no rules of gender or number agreement” (Cellier 1985a:36). For example, the following could be interpreted as either singular or plural: (129) VLRF: marmay t´e i sort l´ekol (ibid.) child PAST ASM leave school ‘The child was/the children were leaving the school.’ Above VLRF marmay comes from a collective noun in French which can only be singular (la marmaille ‘troupe de petits enfants’; cf. marmot ‘petit gar¸con’ [familiar]), much like E cousinage. The F word marmaille cannot refer to an individual, but VLRF marmay can: (130)

VLRF: . . . lo marmay evidaman li kon´e pa tousal`a . . . (C. Warnecke, personal communication) ‘. . . the child obviously doesn’t know all that . . .’

However, in VLRF if an indication of plurality is essential, there is a free morpheme bann (cf. F bande ‘band, gang’) which can be placed before the noun as a pluralizer, as in the fully creolized varieties of French in the Indian Ocean: (131)

VLRF:

bann marmay t´e i sort l´ekol. (ibid.) ‘The children were leaving the school.’

This unbound pluralizer can also form an associative plural, as in Bann Pay`et ‘the Payet family’ (Honor´e 2002:37). 4.5.2

Gender in the VLRF noun phrase

Although articles in VLRF appear to match those of standard French regarding gender (e.g. la kaz ‘the house’; cf. the feminine article in F

The noun phrase

109

la case ‘the hut’), there is no gender agreement between VLRF nouns and their other modifiers such as possessive determiner and adjectives, as there is in French: (132)

VLRF:

mon kaz l´e gran (Cellier 1985a:19) my house is big

(Cf. the feminine forms in the equivalent F “ma maison est grande [gr˜ad]” versus the masculine forms mon ‘my’ and grand [gr˜a] ‘big’.) Cellier (1985a:19) notes that “this calls into question the existence of lexical gender despite the apparent analogy with French [regarding articles]; thus la is only a frozen noun indicator and kaz does not behave like a feminine noun because there is no agreement between the adjective and the noun.” While the VLRF certainly do not have the same kind of gender agreement as French, the VLRF may indeed have true articles: Chapuis (forthcoming) notes that Malagasy has definite articles that are distinct from demonstrative adjectives (Keenan 1978:297) and suggests that this helped to ensure the survival of definite articles in VLRF. Chaudenson (1974:366) notes the use of some VLRF adjectives based on the ˜ l´egliz ‘a large church’ (cf. F une grande F feminine form (e.g. e grad e´glise), but these may be acrolectal forms influenced by the standard language; they appear to be sporadic and most usual in set phrases such as ˜ ravin; cf. F Grande Ravine). toponyms (e.g. grad From this perspective, the VLRF articles that seem to indicate number and gender as in French may simply preserve the form of the French article, as do the initial syllables of certain words in fully creolized varieties of French: (133)

Seychellois CF: cf. F:

Fer

letour

lakaz.

‘Take a tour of the house.’ Faites le tour de la maison. (D’Offay and Lionnet 1982:238)

The existence of analogous forms in VLRF such as lakaz ‘house’ (Chaudenson 1974:349), however rare, suggests remnants of earlier stages of more intense restructuring (e.g. the mid-eighteenth century). Nonetheless, the synchronic co-occurrence of VLRF in ‘a’ and lo ‘the’ (cf. F un ‘a [masculine]’ and le ‘the [masculine singular]’ on the one hand, coupled with the synchronic co-occurrence of VLFR inn ‘a’ and la ‘the’ (cf. F une ‘a’ [feminine] and la ‘the’ [feminine singular] on the other hand, make the question of the existence of grammatical gender and number more complicated to determine in VLRF than in fully restructured creoles like Seychellois. Still, Ramassamy (1985:249) observes that although the VLRF possessive pronouns (l´e-myenn ‘mine,’ l´e-tyenn ‘yours’ etc.) are derived from the feminine plural French forms (les miennes, les tiennes), the

110

Languages in Contact

VLRF forms do not carry the gender and number references of their etyma, making clear that the very concepts of grammatical gender and number are at least much more marginal in VLRF than they are in French. 4.5.3

Possession in the VLRF noun phrase

Possession, usually indicated by the prepositional de in French, is indicated by a noun complement in VLRF, i.e. without a preposition: (134)

VLRF: la kaz son momon (Ramassamy 1985:227) the house [of] her mother (cf. F la maison de sa maman)

This construction is also found in fully creolized varieties: (135)

Mauritian CF: lakaz lapay house straw house’

mo tohtoh (Baker 1972:83) my uncle, i.e. ‘my uncle’s straw

This construction is extended to more abstract connotations of de, such as association, juxtaposition, etc.: (136) (137)

VLRF: lankazman ti-fiy (Baker 1972:373) ‘the girl’s engagement’ (cf. F l’engagement de la fille) VLRF: bo la mer (Baker 1972:392) ‘by the sea’ (cf. F au bord de la mer)

In fact, Cellier (1985a:106) states that the French preposition de does not exist in VLRF except in fossilized forms like “piedmang” (
Pronouns in VLRF

VLRF have the following personal pronouns (Ramassamy 1985:231– 239): (138)

VLRF:

Singular

Plural

1st person mwen ‘I’ nou ‘we’ 2nd person ou/tou´e ‘you’ zot ‘you’ 3rd person li ‘s/he, it’ zot ‘they’ bann-la ∼ banna ‘they’ Most are derived from the French disjunctive or emphatic pronouns, while zot appears to come from F vous autres ‘you [others]’ or les or eux autres ‘the others,’ and can refer to either ‘you [plural]’ or ‘they.’

The noun phrase

111

An alternative term for ‘they’ is bann-la (with the variant banna) from F bande ‘band, gang’ (cf. the pluralizer bann, section 4.5.1). There is a parallel here to those Atlantic creoles and their substrate languages in which the pronoun meaning ‘they’ is juxtaposed to a noun to emphasize plurality (see section 4.1.1). The VLRF pronoun bann-la was used by slaveholders to refer to their slaves and can have a pejorative connotation (Ramassamy 1985:237). Note that unlike their French counterparts (e.g. il, elle; ils, elles), none of the above VLRF personal pronouns is marked for gender. However, Cellier notes that VLRF has borrowed e` l (cf. F elle ‘she’): “`el is a gallicism that usually (but not always) has a [+human] referent; this loan can be basilectal; it is marked for natural gender and does not imply the introduction of grammatical gender” (Cellier 1985b:335, my translation). When personal pronouns are used as subjects before the agentive subject marker i, there is an obligatory contraction of the two morphemes as follows (Cellier 1985b:234): (139)

VLRF:

Singular Plural 1st person mi 2nd person wi/ti 3rd person li

ni zot i zot i

Before the future marker va there is a parallel contraction to ma, wa, ta, li va; na, zot va, but this is optional and the full forms (e.g. mwen va) are found as well (ibid.). The full pronouns can function not only as subjects but also as objects. They simply follow prepositions (`ek li ‘with him/her’), but as direct (140) and indirect objects (141), all except bann-la are preceded by a-: (140) (141)

VLRF: boug la lap rogad a-nou. ‘That man is looking at us.’ (Ramassamy 1985:235) VLRF: mwen la donn a-li en liv. ‘I gave him/her a book.’ (ibid.)

Note, however, that a- can also precede non-object pronouns: (142)

VLRF: manz a-ou. ‘[you singular] eat!’ (Ramassamy 1985:330)

(143)

VLRF: manz a-zot. ‘[you plural] eat!’

(ibid.)

This a- seems to reinforce the emphatic nature of the pronoun, and thus can be optional in preverbal position; it is obligatory in the post-verbal position (Watbled, personal communication).

112

Languages in Contact

(144)

VLRF: a-ou e` k a-li va geny´e. ‘You and he will win.’ (Ramassamy 1985:385)

(145)

VLRF: ou e` k

li va geny´e. idem.

(ibid.)

While this a- is compatible (at least in form) with the French preposition a` ‘to,’ it should be noted that the stressed pronouns in a number of varieties of restructured Portuguese are also marked by a- (< P a mim ‘to me’) (Holm 1988:203), and these forms may have entered VLRF via Indo-Portuguese, reinforced by the initial a- in free-standing object and possessive (but not subject) personal pronouns in Malagasy (Chaudenson 1974:954). Finally, it should be noted that the VLRF pronouns often do not follow the case marking and word order of the non-emphatic French pronouns. This makes them morphosyntactically more similar to pronouns in fully creolized varieties of French (although none of these takes an initial a-), which is compatible with the origin of the VLRF pronouns in the French disjunctive pronouns. VLRF does not have the reflexive verbs found in French (Cellier 1985a:68, 73). When the subject and object refer to the same person, VLRF can express the latter simply by using the normal object pronoun (146) or le ko:r (147); cf. F le corps ‘the body,’ used this way in archaic and regional French, with parallels in many languages of the world including creoles (Holm 1988:205), and Malagasy tena ‘body, self ’ (Chapuis forthcoming). (146) (147)

VLRF: li bal˜as ali. (cf. F il se balance) ‘He balances himself.’ (Chaudenson 1974:345) VLRF: zot ` i p´e pan zot ` ko:r. (cf. F ils peuvent se pendre) ‘They can [go] hang themselves.’ (ibid.)

Reflexive verbs in French can also correspond to intransitive verbs in VLRF: (148)

VLRF: mi ap`el Paul. (cf. F je m’appelle Paul) ‘My name is Paul.’ (Cellier 1985a:175)

Note the similarity of VLRF ap`el in (148) above to the corresponding verb in a fully creolized variety of French: (149)

Haitian CF: Li r´el´e Mari. ‘She is called Mary.’ (Holm 1988:83)

4.6

A comparison of the noun phrase

The differences between the noun phrases of the partially restructured languages examined here and their source languages are compared with one another below to identify any common patterns of restructuring.

The noun phrase

4.6.1

113

A comparison of number in the noun phrase

In comparison with their source languages, number marking in the noun phrase is considerably reduced in all the varieties under discussion except standard Afrikaans. Plural marking mainly involves nouns in AAE and Afrikaans, but in BVP and NSCS there is variable number agreement between head nouns and their modifiers. Phonological rules could account for most of this variation in NSCS, but in BVP the variation clearly originated in morphosyntactic restructuring. In VLRF nouns take no plural inflections whatsoever; they co-occur with article-like morphemes that can indicate number, but there is no number agreement between nouns and adjectives or other modifiers. In AAE, Afrikaans, and VLRF there is a free morpheme of the same form as the pronoun meaning ‘they’ which can serve as a plural or an associative plural marker. A parallel construction is found in a number of Atlantic creoles and their African substrate languages. The vestigial BVP form osˆele ‘they’ suggests that a parallel form may also have existed in BVP. 4.6.2

A comparison of gender

Like English, AAE has no rules of gender agreement in the noun phrase. Afrikaans nouns have not retained the covert gender marking of their Dutch etyma, and the Dutch morphological rule for gender agreement between adjectives and nouns has been reinterpreted as an (often unworkable) phonological rule in Afrikaans. However, gender agreement in the noun phrase is categorical in BVP (except for the Helv´ecia variety and the L´ıngua dos Pretos Velhos) and NSCS. It is interesting that the BVP and NSCS mark number variably but gender categorically, while in Afrikaans number marking is almost categorically present (except in nonstandard varieties) but gender marking is categorically absent (except with personal pronouns). Lucchesi (2000:329ff.) tries to account for this in BVP by pointing out that the phonological variability of /o/ and /a/ in the marking of gender is less plausible than the phonological variation of /s/ and the nasalization of vowels in marking number in subject–verb agreement, but it is equally implausible that either of these rules was inherited from EP (sections 4.3.1 and 4.3.2). The explanation may lie in the evolutionary tendencies of the older language varieties out of which the newer ones developed. While number marking is still categorical in standard Dutch, Portuguese, and Spanish, the marking of gender distinctions between masculine and feminine was already breaking down in seventeenth-century Dutch (Brachin 1985:67). This historical fact seems likely to have contributed to the complete loss

114

Languages in Contact

of gender marking on modern Afrikaans nouns, although contact with African and Asian languages lacking this distinction is surely relevant as well (note that modern Dutch nouns retain a distinction between common gender and neuter). Similarly, the optional marking of number in many of the languages with which Spanish and Portuguese were in contact in the New World (and with which Dutch was in contact in the Cape Colony) seems to have worked against the categorical marking of number in BVP and NSCS – and in nonstandard varieties of Afrikaans. VLRF lacks number and gender agreement between nouns and adjectives altogether. This seems likely to be related to substrate influence, and it can be seen as evidence that VLRF underwent considerably greater restructuring than BVP or NSCS. However, in VLRF, the synchronic co-occurrence of morphemes derived from the singular versus plural and masculine versus feminine forms of French articles and nouns makes the question of the existence of grammatical gender and number more complicated to determine than in the other restructured varieties discussed here. 4.6.3

A comparison of possession

In the seventeenth century the inflectional marking of possession with -s was being lost in Dutch but not in English; however, this construction did not survive in Afrikaans or some varieties of AAE, probably because there was no parallel inflection in the other languages in contact. Juxtaposition [possessor + possessed] appears to have prevailed in early AAE, being both a universal and a substrate strategy for indicating possession and associated relationships. However, Afrikaans adopted the possessive marker of colloquial Dutch [possessor HIS possessed], which converged with constructions in the languages with which it was in contact. BVP and NSCS took over the standard Portuguese and Spanish construction [possessed OF possessor], but the sporadic absence of the preposition in some rural varieties of BVP suggests that in earlier stages the preposition was variably absent, as it is in some modern varieties of creole Portuguese. VLRF use a parallel construction without the French preposition de. 4.6.4

A comparison of pronouns

In all the restructured varieties except NSCS the personal pronouns of the source languages underwent some loss of case distinctions, apparently because of the influence of substrate languages or the restructuring associated with adult second language acquisition. Reduction of case marking was also influenced by the fact that Afrikaans, BVP, and VLRF

The noun phrase

115

all selected the emphatic or disjunctive personal pronouns of their source languages (a characteristic strategy in language contact); these are less often marked for case and more often follow SVO word order. Due to decreolization the case discrepancies between AAE and standard English pronouns have largely disappeared, but the partly similar discrepancies between Afrikaans and Dutch have become fixed with the standardization of Afrikaans. BVP appears to have reformulated a morphosyntactic rule favoring emphatic (i.e. subject) pronouns as free morphemes following verbs over the object-case-marked clitics of standard European Portuguese. Reflexive pronouns are omitted in VLRF and in most varieties of BVP (and an unusually basilectal variety of NSCS, Palenquero Spanish) as they are in the Atlantic creoles, corresponding to constructions in substrate languages. The difference in reflexive pronouns between Dutch and Afrikaans appears to have nothing to do with language contact, resulting instead from an innovation in Dutch after Afrikaans had already come into existence.

5

The structure of clauses

Introduction With the exception of Afrikaans, the partially restructured languages examined here have the subject-verb-object word order for declarative sentences that is found in both their Western European superstrate and Niger-Congo substrate. In fact, this is also the basic word order in all of the Atlantic creoles. Brazilian Vernacular Portuguese favors the SVO order even for object pronouns, where European Portuguese can have SOV order. Afrikaans is unlike the other varieties in that it has basic SOV word order, but it is like them in so far as it follows the word order of its superstrate, Dutch, which coincides with the SOV order of some of Afrikaans’ substrate languages. While the European lexical source languages can require the inversion of the subject and the verb (or auxiliary) to transform a statement into a question, this is not a part of the syntax of Niger-Congo languages or full-fledged creoles. Instances of creole-like non-inversion of subject and verb (or auxiliary) that would be unacceptable in the source language can occur freely in all of the partially restructured varieties under discussion except Afrikaans, which adheres strictly to Dutch word order.

5.1

The structure of AAE clauses

5.1.1

AAE word order

African American English has the usual English subject-auxiliary inversion (or lack of it) in questions that can be answered “yes” or “no,” e.g. “Can I go?” (Burling 1973:68). However, unlike standard English, AAE has optional inversion with question words in the main clause, i.e. both of the following occur: (150) 116

AAE: Where can I go? Where I can go? (ibid.)

The structure of clauses

117

In embedded questions, which have no subject-auxiliary inversion in standard English, inversion is again optional in AAE: (151)

AAE: I wonder {where can I go}. I wonder {where I can go}. (ibid.)

In embedded yes/no questions, AAE may have no connecting if or whether but does have inversion: (152)

I wonder {can I go}. (ibid.)

In this respect AAE is unlike English-based creoles, which have no such inversion at all and therefore happen to match standard English word order in embedded questions: (153)

Jamaican CE: Dem aks mi {if a want i}. ‘They asked me {if I wanted it}.’ (Hancock 1979b:14)

A case might be made for AAE being more similar to Irish English, in which direct questions are also embedded: (154)

Irish English: I don’t know {is that right or not}. (Barry 1982:108)

While Irish English might well have served as a model for AAE at an earlier period (Rickford 1986), the AAE pattern of subject-auxiliary inversion could also be the result of partial restructuring or decreolization. Bahamian English, which seems to be either more restructured or at an earlier stage of decreolization than AAE (or both), has no subjectauxiliary inversion in the basilect but frequent inversion in the upper mesolect, even in embedded questions. Thus one finds (155) varying with (156): (155) (156)

Bahamian CE: I can go? (Holm 2000a:236) Bahamian CE: Can I go? (ibid.)

The same occurs in embedded questions: (157)

Bahamian CE: I don’t know {where I can go}. ∼ I don’t know {where can I go}. (ibid.)

5.1.2

Dependent clauses in AAE

The structure of many AAE relative clauses is parallel to their equivalents in standard English, but this is not always the case. In AAE the zero form of the relative pronoun can be used not only for the object of the verb as

118

Languages in Contact

in standard English (e.g. “The man { also for the subject: (158)

AAE: He got a gun {

he is hiring} is my uncle”) but

sound like a bee}. (Dillard 1972:68)

This structure is also found in Creole English (CE), e.g. (159)

Jamaican CE: De man { (Cassidy 1961:57)

owe me money} gone a Cuba.

According to Cassidy, Jamaican CE, “like the Niger-Congo languages . . . gets along with paratactic constructions” (ibid.). However, Yoruba, which is a Niger-Conger language, does have a relativizer (t´ı ‘who, which’) but in “spoken Yoruba t´ı is often omitted,” e.g. (160)

Yoruba: as.o. { mo r`a l´anˇa} n’`ıy´ı (Rowlands 1969:90) cloth I bought yesterday this-is ‘This is the cloth I bought yesterday.’

However, “It cannot be omitted where its omission would produce ambiguity,” e.g. (161)

Yoruba: m`alu´ u` {t´ı ko` n´ı `ıru} ` cow which no has tail ‘a cow which has no tail’ (ibid.)

In fact, the Yoruba relativizer t´ı sometimes seems more like a European subordinator than a relative pronoun, in that it can introduce a clause which requires its own subject pronoun even though the relativizer itself would have this function in a European language: (162)

Yoruba: e` mi {t´ı mo fun ´ e. n´ı gbogbo owo´ yˇı} 1s REL 1s give 2s OBJ all money DEM ‘I who [I] gave you all this money.’ (Rowlands 1969:88)

Bickerton (1981:63) speculates that creoles may have been “born without surface relativizers” and gives examples of zero subject relative pronouns in Guyanese CE, Seychellois CF, and Annobon ´ CP (1981:62–63). AAE’s zero subject relative pronouns have been used to support its creole history, but Tottie and Harvie (2000) provide convincing evidence that the varieties of English that British settlers brought to the New World also contained this construction. Not only was it the predominant form in Middle English, as in Chaucer’s “I saugh a beest was lyk an hound” (Tottie and Harvie 2000:202), but it is also found throughout British regional varieties (Orton et al. 1978, map S5). This construction apparently converged with the zero subject relativizers in African and creole languages to favor the selection of this form in AAE. Although there is a lack of published research in this area, Tottie and Harvie point out that

The structure of clauses

119

a study of one individual’s AAE (McKay 1969) reveals that “zero is the most frequently used relativizer, with 54% of all cases, more than half of which are subjects; that comes second with 38%, and what accounts for 9%; except for quotations from the Bible, who and which do not occur” (Tottie and Harvie 2000:200). While zero subject relativizers account for 41 percent of the tokens in the ex-slave recordings, they account for only 2 percent and 5 percent in the modern spoken English of Americans and Britons, respectively (2000:224). However, it is ironic that Tottie and Harvie (2000:223) find the Englishness of AAE confirmed by the fact that “the Gullah relativizers wuh and weh are totally lacking in our data” since these can be traced to England’s Northcountry dialect (Holm with Shilling 1982:218). As a matter of fact, Kautzsch (2002:213) notes that in his early AAE data there are indeed a few sporadic occurrences of “relative clauses introduced by non-spatial where, which makes it hard to categorically deny any creole influence on AAE relative constructions.” Dillard notes that some speakers of AAE seem to hypercorrect clauses without relative pronouns, supplying not only an object relative pronoun but also the clause’s original non-relative object pronoun: (163)

AAE: Dem little bitty hat {what dey wearin’ dem now}. (Dillard 1972:68)

In AAE ambiguity in sentences without a subject relative pronoun can often be cleared by a pleonastic subject pronoun (section 4.1.4) marking the verb of the main clause: (164)

AAE: The boy { won} he did a three. ‘The boy who won did a three.’ (Smith 1973:94)

Regarding subordinate clauses, most in AAE are identical in structure to those in standard English, with the notable exception of the use of say to introduce a quotation: (165)

AAE: They told me {say they couldn’t get it}. (Rickford 1977:212)

This construction is also found in a number of English-based creoles: (166)

(167)

Krio CE: A yεri {se Olu de fes di buk kam}. ‘I heard that Olu is bringing the book along.’ (Yillah forthcoming.) Gullah CE: dε lɔ {sε wi tu ol}. ‘They admit that we’re too old.’ (Turner 1949:211)

Turner (1949:201) pointed out the formal and syntactic similarity of Gullah sε and Twi sε ‘that, saying’ and English say. Cassidy (1961:63)

120

Languages in Contact

noted that the pronunciation of Jamaican se is /sε/ when it means ‘that’ rather than /sey/, leading him to support the connection with Akan se. Boretzky (1983:177) finds the lexical borrowing of se into the creoles an inadequate explanation in light of the fact that the Surinamese creoles have completely different forms, i.e. Sranan tak(i) and Saramaccan taa, ´ leading him to believe that the substrate influence on this construction lay in the grammar rather than the lexicon. (168)

Sranan CE: M sab {tak a tru}. ‘I know that it’s true’ (Voorhoeve 1962:26)

The existence of parallel structures in creoles not based on English (Holm 2000a:208–209) supports Boretzky’s conclusion. 5.2

The structure of Afrikaans clauses

5.2.1

Afrikaans word order

As noted above, Afrikaans differs from the other varieties in this study, following the SOV word order of its superstrate Dutch and some of its substrate languages such as Khoi and Indo-Portuguese (den Besten 1986:187). Like Dutch, Afrikaans is considered an SOV language because this is the word order found in its dependent clauses, whether these are subordinate (169) or relative (170): (169)

(170)

A: Jy weet {dat ek dit more ˆ doen}. 2s know that 1s 3s tomorrow do ‘You know I’m doing it tomorrow.’ (Donaldson 1993:365) A: Dit is die mense {wat langsaan bly}. 3s be ART people REL next door live ‘These are the people who live next door.’ (ibid.)

SVO can be found in independent clauses with one verb: the verb occupies the second position, which (as in German) usually follows the subject (171), although time adverbials can also occupy the first position, leaving the verb second and the subject third (172): (171)

(172)

A: Hy is siek vandag. 3s be sick today ‘He is sick today.’ (Donaldson 1993:362) A: Vandag is hy siek. today be3s sick ‘Today he’s sick.’ (ibid.)

However, the fact that the main verb occurs at the end of the main clause whenever there is an auxiliary verb (which then occupies the second

The structure of clauses

121

position) is taken as evidence that Afrikaans, like Dutch, has an underlying SOV order: (173)

A: Ek sal dit more ˆ doen. 1s will 3s tomorrow do ‘I will do it tomorrow.’ (Donaldson 1993:363)

This word order sets Afrikaans apart from the Atlantic creoles, which all have strict SVO word order – including Negerhollands CD. However, it confirms Afrikaans’ status as one of the partially restructured languages, which follow the main word order of their superstrates except for some creole-like variation in questions. Note that in yes/no questions, Afrikaans follows the (inflected) verbsubject word order of Dutch (and English in sentences with auxiliaries and the equivalent of be): (174)

(175)

V S A: Voel jy naar? feel 2s nauseous ‘Do you feel nauseous?’ (Donaldson 1993:370) A: Sal jy dit asseblief vir my doen? will 2s that please for me do ‘Will you please do that for me?’ (ibid.)

However, if there is a question word, Afrikaans follows (inflected) verbsecond word order (again like Dutch or English auxiliaries). (176)

(177)

QW V S A: Waar bly jy? where live 2s ‘Where do you live?’ (Donaldson 1993:323) A: Waar is die wildtuin? where be the game park ‘Where is the game park?’ (Donaldson 1993:327)

Note, however, that in indirect questions Afrikaans has two possible word orders. Following the word order of embedded questions in Dutch, the verb can be final (the order normally found in written Afrikaans): (178)

QW S V A: Hy sal seker weet {waar die wildtuin is}. ‘He’ll know for sure where the game park is.’ (ibid.)

Or the embedded question can follow the inverted V-S order of the direct question, a syntactic innovation also found in AAE (see section 5.1.1):

122

Languages in Contact

(179)

QW V S A: Hy sal seker weet {waar is die wildtuin}. (ibid.)

5.2.2

Dependent clauses in Afrikaans

As noted above, Afrikaans follows Dutch word order in both main clauses (verb or auxiliary second) and dependent clauses (verb and auxiliaries last). Relative clauses in Afrikaans differ from those of Dutch, which has various relative pronouns depending on gender and number, in that Afrikaans makes almost exclusive use of a single relative pronoun wat (which can never be omitted) for all antecedents: (180)

A: Ek wil die boek hˆe {wat jy in jou hand het}. ‘I want the book which you have in your hand.’ (Donaldson 1993:146)

This wat combines with the possessive marker se (section 4.2.3) to form the relative ‘whose’ or ‘of which’: die tafel wat se poot af is ‘the table, the leg of which is off’ (den Besten 1996:12) – although [+human] antecedents can take the relative wie se. The only other exception to the exclusive use of wat is when the relative is the object of a preposition, in which case the Dutch forms are used when there is pied-piping (181a); if the preposition is stranded (181b), then wat must be used (181b): (181a) A: die probleem waarvan jy praat (den Besten 1996:12) ‘the problem whereof you speak’ (181b) A: die problem wat jy van praat (ibid.) The origin of the form of the Afrikaans relative is Dutch wat, which can occur as a relative pronoun after an indefinite (e.g. alles wat je zegt ‘everything you say’). This form (and the parallel Negerhollands CD relativizer wa) may have been selected over competing forms for grammaticalization as the relativizer due to the influence of nonstandard or archaic Dutch usage, or simply because it is more salient. Den Besten analyzes it as a “gespecialiseerd voegwoord” [specialized conjunction] (1996:13) that is a “marker for noninterrogative WH-movement in finite clauses” (1981:141). The indeclinability of wat can sometimes lead to ambiguity: (182)

A: Die Engelse soldate {wat di´e Boere verslaan het} ‘The English soldiers whom those Boers defeated . . .’ OR ‘The English soldiers who defeated those Boers . . .’ (Donaldson 1984:64)

The structure of clauses

123

However, the latter meaning can be made unambiguous by using the personal direct object marker vir: (183)

A: Die Engelse soldate {wat vir di´e Boere verslaan het} . . . (ibid.)

This use of vir is an Afrikaans innovation unknown in Dutch. Brachin (1985:140) attributes its origin to Malay, but it could also come from the object-marker ku in Malayo-Portuguese or per in Indo-Portuguese: (184)

Malayo-Portuguese: Eli-sa mai mand´a ku e´ li bai skola. ´ ‘His mother sent OBJ him to school.’ (Holm 1988–89:293).

In Afrikaans subordinate clauses, the subordinator dat is omitted as often as the English subordinator that, unlike the equivalent Dutch form, which is never omitted: (185)

A: Ek weet {dat jy dit gedoen het}. ‘I know that you did it.’

(186)

A: Ek weet {jy het dit gedoen.} ‘I know you did it.’ (Donaldson 1993:146)

(Note that without dat, the subordinate clause takes main clause word order.) Regarding the possible omission of the Afrikaans subordinator dat, Brachin (1985:141) again attributes this to Malay. 5.3

The structure of BVP clauses

5.3.1

BVP word order

While Portuguese is basically an SVO language, it is like other Romance languages in that object pronouns can precede the verb, particularly in European Portuguese and standard Brazilian Portuguese: (187)

EP, SBP: N˜ao os vi. ‘I didn’t see them.’ (Cunha 1982:279) NEG 3p saw-1s

In these varieties of Portuguese, object pronouns can also occur after the verb as unemphasized clitics: (188)

EP, SBP: Ela chamou-me. ‘She called me.’ (Azevedo 1989: 863) 3s called 1s

However, in BVP these clitics are usually replaced with emphatic subject pronouns, requiring SVO word order: (189)

BVP: Ela chamou eu. ‘She called me.’ (ibid.) 3s called 1s

124

Languages in Contact

Moreover, “subject-verb inversion in question constructions is not commonly found in BVP” (Mello 1997:167): (190)

BVP: Onde vocˆe mora? where 2s live

‘Where do you live?’

This structure is paralleled in the Atlantic creoles and many of their African substrate languages (Holm 1988–89:212–214), but not in the Portuguese of Europe. There subject-verb inversion is required (191)

EP: Onde mora vocˆe? where live you

unless the question word is emphasized: (192)

EP: Onde e´ que vocˆe mora? where is-it that you live

Lemle (1976:77) notes that the following question structures can be observed in urban Rio de Janeiro for ‘Where did you fall?’: (193)

MOST FORMAL:

a Onde vocˆe caiu? b Onde caiu vocˆe? CAREFUL COLLOQUIAL: c Vocˆe caiu onde? d Onde e´ que vocˆe caiu? e Onde foi que vocˆe caiu? f Onde e´ que foi que vocˆe caiu? MOST INFORMAL: g Onde que foi que vocˆe caiu? h Onde que vocˆe caiu?

The last structure, (193h) without e´ , is unknown in European Portuguese but etymologically derivable from the structure illustrated in (190) above. Here BVP que is parallel to a creole highlighter used after question words: k’ in Capeverdean CP and ki in Guin´e-Bissau CP (Baptista, Mello, and Suzuki forthcoming). However, the most surprising result of Lemle’s survey is that one of the most formal constructions (193a Onde vocˆe caiu?) – presumably considered standard Brazilian Portuguese – is not possible in European Portuguese. 5.3.2

Dependent clauses in BVP

In BVP almost all relative clauses are introduced by the relative pronoun que, unlike the standard, which has a number of different relativizers such as quem ‘whom’ (as the object of a preposition), cujo ‘whose,’ etc. In some rural or isolated varieties of BVP, such as that of Cear´a, the relativizer que (which takes the form ki) can be omitted, whether

The structure of clauses

125

it is functioning as the subject of the relative clause (195) or as the object (196). (194)

(195)

(196)

BVP: moʃtr´e u p˜au {ki eu kumia} ‘I showed the bread that I was eating.’ (Jeroslow 1974:193) BVP: u fradi morava nu sobradu { era ‘The priest lived on-the second-floor, [which] was muitu ´ a´ utu} very high.’ (1974:194) BVP: i tudus kuienu us pedasu { kir´ıa kum´e} ‘And all choosing the pieces [that] they-wanted to-eat.’ (1974:195)

While the zero form of the relativizer is apparently unusual in other varieties of BVP, it is also found in Helv´ecia Portuguese (Baxter and Lucchesi 1997:78). To make clear the syntactic function of BVP que within the relative clause, a resumptive pronoun can be used, e.g. BVP can preserve the direct object in its original position even though it is also represented by que – a structure unknown in standard Portuguese (but cf. a parallel structure in AAE sentence [163] above): (197)

BVP: Esse rapaz, {que eu conheci ele} . . . ‘This guy that I met [him]’ (Amaral 1928 [1976] 78)

The resumptive pronoun can also be the object of a preposition, which remains in its original position: (198)

BVP: O aluno {que eu conhe¸co o pai dele.} ‘the student that I know the father of him’ ‘The student whose father I know . . .’

However, the latter construction (198) is also found in colloquial European Portuguese. To work out its origin, it would be relevant to know if it is attested in Europe before 1500; if not, it may have been imported from Brazil. Standard Portuguese uses the relative determiner cujo to indicate possession: (199)

SBP: O aluno {cujo pai eu conhe¸co} . . . ‘The student whose father I know . . .’

Like colloquial European Portuguese, BVP can delete the resumptive pronoun in sentences like (198) above, a construction which is preferred

126

Languages in Contact

because it avoids a stigmatized structure. While this deletion can sometimes result in a standard construction, as in the case of resumptive object pronouns (197), it can also result in nonstandard structures, as when the prepositional phrase in (198) is deleted: (200)

BVP: O aluno {que eu conhe¸co o pai } the student that I know the father

(201)

BVP: auk˜aso´ a kaza {ki tij˜a duxmidu} ‘He-reached the house [in] which he had slept.’ (Jeroslow 1974:193)

Such deletion of a stigmatized construction resulting in a structure that is still nonstandard is characteristic of decreolization. Although relative clauses with deleted resumptive pronouns are considered less socially marked, they are frequently ambiguous, as seen in the possible interpretations of the following: (202)

BVP: a menina que falei . . . SBP: a menina {com quem falei} . . . ‘The girl I talked to’ a menina {de quem falei} . . . ‘The girl I talked about’ a menina {por quem falei} . . . ‘The girl I spoke for’

As Tarallo (1986) notes, relative clauses beginning with a preposition plus quem are only found in written Portuguese in Brazil: (203)

BVP: Esse fulano a´ı, {com quem eu nunca tive aula} ‘This guy with whom I never had a class’

In speech que is the relativizer, and the prepositional phrase either occurs in its original form and position (204) or not at all (205): (204)

BVP: Esse fulano a´ı, {que eu nunca tive aula com ele}

(205)

BVP: Esse fulano a´ı, que eu nunca tive aula.

Baxter (1987) notes that relative clauses structurally parallel to the BVP clauses above can be found in the Portuguese-based creoles of Africa, such as S˜ao Tom´e CP: (206)

S˜ao Tom´e CP: omi {ku zo˜ sa ka fla ne} man who John PROG talk about him ‘the man that John is talking about’

To this could be added parallel constructions in other fully creolized varieties of Portuguese and Spanish: (207)

Angolar CP: OmE {ki m ba kw E} . . . the man that I went with [him] (Lorenzino forthcoming)

The structure of clauses

(208)

(209)

(210)

127

Cape Verde CP: kel ome {ke n fala k’ el} . . . That man that I spoke with [him] (Baptista, Mello, and Suzuki forthcoming) Guin´e-Bissau CP: N mora na kasa {ku bu mora-ba n el} I live in house that you live ANT in [it] ‘I live in the house that you used to live in.’ (ibid.) Papiamentu CS: E homber {ku m’a papia kun’ e} The man that I PAST speak with [him] a papia malu. PAST speak bad ‘The man that I spoke with spoke badly.’ (Michel forthcoming)

Given the significant number of speakers of such creolized varieties of Portuguese in Brazil during the first centuries of settlement from Europe and Africa while the predecessor of modern BVP was taking form, their influence in making this structure a part of the vernacular seems very likely. It should be noted that the above data provide clear counter evidence to the position of Lucchesi (2000:34): “. . . n˜ao consideramos prov´avel a transferˆencia de estructuras das l´ınguas do substrato para o portuguˆes brasileiro” [‘we consider it improbable that any structures were transfered from substrate languages to Brazilian Portuguese’]. Regarding subordinate clauses, in some rural or isolated varieties of BVP, such as that of Cear´a, the complementizer que (which takes the form ki) can be omitted: (211)

BVP:

(212)

BVP:

eli sabi {ki n˜au se} ‘He knows that I do not know.’ (Jeroslow 1974:199) eu se { eu ko˜jesu a m˜aga} (ibid.) ‘I know [that] I am familiar with the range.’

While the zero form of the subordinator ‘that’ is apparently unusual in other varieties of BVP, it is also found in Helv´ecia Portuguese (Baxter and Lucchesi 1997:78) as well as Cape Verdean and Guin´e-Bissau Creole Portuguese (Baptista, Mello, and Suzuki forthcoming). 5.4

The structure of NSCS clauses

5.4.1

NSCS word order

NSCS has nothing like the BVP use of subject pronouns as direct objects after verbs, although some basilectal varieties achieve SVO word order by

128

Languages in Contact

replacing direct object pronouns before the verb with quasi-disjunctive object pronouns marked by a after the verb. NSCS is unlike most other dialects of Spanish (but like the Atlantic creoles and their substrate languages) in that it usually requires expressed subject pronouns (see section 4.3.4) and permits these pronoun subjects to take SVO word order after non-subject question words, which in standard Spanish require subject–verb inversion: (213)

NSCS: Qu´e tu´ dices? What you say, i.e. ‘What do you say?’ cf. S: Qu´e dices (tu)? ´ (Holm 1989:308)

In the Atlantic creoles (but not in NSCS) SVO can be found in questions even when the subject is a noun: (214)

Papiamentu CS: Unda e buki ta? ‘Where is the book?’ where DET book is (Maurer 1988:44)

5.4.2

Dependent clauses in NSCS

The form que predominates over other relative pronouns in NSCS, including zero: even nineteenth-century Afro-Puerto Rican texts indicate its fairly consistent presence: (215)

NSCS: ¿Como ´ ba quer´e senorita ˜ que son tan bonita, uno hombre as´ı tan feo? ‘How could you, who are so beautiful, love a man as ugly as I am?’ (Alvarez Nazario 1961 [1974]: 385)

Zero relative pronouns, which do not occur in standard Spanish, did occur in the unusually basilectal variety NSCS studied by Green (1997) in the speech of the informant that Lipski thought to be mentally deficient (section 2.4.3). (216)

NSCS: No, ah´ı hay una ta mara. ‘No, there is one that is bad there.’ (Green 1997:166)

This structure is parallel to that in AAE sentence (164) above, with a zero relative pronoun acting as the subject of a relative clause. The NSCS subordinator que can also take a zero form, which never occurs in standard Spanish; in the bozal Spanish of Cuba, the complementizer que was often omitted: (217)

NSCS: Dice cf. S:

jaguey ¨ t´a chiquito. ‘He says that the liana is small.’ Dice que el jaguey ¨ est´a chiquito (Cabrera 1969, cited in de Granda 1978:486)

The structure of clauses

129

However, the presence of the complementizer is no less frequent. Similarly, there is a variable absence of the complementizer que introducing subordinate clauses in the basilectal variety of Dominican Spanish studied by Green (1997): (218)

NSCS: . . . dice se pusien acechalo ‘. . . they say that they started to spy on him’ cf. S: . . . dicen que se pusieron a acecharlo. (Green 1997:166).

Another speaker of the same lect used para as a complementizer to mark an infinite, like English to, which has no equivalent in standard Spanish: (219)

NSCS: Yo no temo para habla, ´ convers´a con nalie. (ibid. 168) I NEG fear to speak converse with no one. ‘I do not fear speaking, conversing with anyone.’

Kihm (1994) notes that (unlike Spanish and Portuguese) Guin´e-Bissau CP uses pa (cf. P para ‘for’) to mark infinitive-like nominal uses of verbs: (220)

Guin´e-Bissau CP: Pa lei i yera un tarbaju dif´ısil. ‘To read was a hard job.’ (Cited in Holm 1988:170)

In many Atlantic creoles based on Spanish and Portuguese, pa can mark a tensed clause with a subject and verb, e.g. (221)

Papiamentu CS: Mi ke pa bo bai. (Goilo 1972) 1s want COMP you go ‘I want you to go.’

This construction is also found in the NSCS of bilingual Palenquero speakers: (222)

PS: La abuelita m´ıa no quer´ıa pa m´ı ir al monte. ‘My grandmother didn’t want me to go to the fields.’ cf. S: La abuelita no quer´ıa que yo fuera al monte.’ (Schwegler and Morton 2002:150)

5.5

The structure of VLRF clauses

5.5.1

VLRF word order

Unlike French, in which subject-verb inversion can occur in questions, in VLRF it does not occur:

130

Languages in Contact

(223)

VLRF:

QW S V ousa vi r`es?

cf. F:

QW V S Ou` habitez-vous? (Cellier 1985a:20)

‘Where do you live?’

However, it should be noted that strategies for avoiding subject-verb inversion were well established in French before the eighteenth century, when VLRF emerged (A. Kihm, personal communication): (224)

F:

QW S V Ou` est-ce que vous habitez?

In yes/no questions, the interrogative nature of the VLRF structure is signaled solely by rising intonation (Ramassamy 1985:323): (225)

S VP VLRF: Zan la vni zordi? ‘Did John come today?’ (ibid.) John PAST come today

While this is also possible in standard French, the latter has other syntactic strategies for marking such questions. It should be noted that Malagasy, which played a significant role as a substrate for the VLRF, has VOS sentence structure (Chapuis forthcoming). However, a frequent Malagasy existential construction followed by a relative clause lent itself to reinterpretation as SV(O) word order in VLRF, which – like Malagasy – has optional relative pronouns (see section 5.5.2): (226)

Malagasy: misy zaza izay mitomany PRES exist child (who) PRES cry ‘There is a child who cries.’ ‘Some child is crying.’ (ibid.)

(227)

VLRF:

Nana zanfan i pl`er. Exist child ASM cry ‘There is a child who is crying.’ ‘Some child is crying.’ (ibid.)

In fact, VLRF existential nana (cf. F il y en a ‘there are some’) became reinterpreted as ‘to have’ in sentences like the following: (228)

VLRF:

Larivy`er sinni nana bokou dlo zordi river St.-Denis exist much water today ‘There is a lot of water in the St.-Denis River today.’ ‘The St.-Denis River has a lot of water today.’ (ibid.)

The structure of clauses

131

As Chapuis (forthcoming) notes, “Indeed, here, from a Popular French point of view, rivy`er sinni appears to be the subject: ‘The Saint-Denis River,’ while, from the basilectal point of view, it constitutes an adverbial phrase: ‘in the Saint-Denis River,’ since no preposition is used with place names in the VLRF.” Another feature of VLRF word order appears also to have developed due to substrate influence. In French, the direct object must precede the indirect (indicated by the preposition a): ` (229)

F:

S V DO IO Jean donne l’argent a` Paul ‘John gives the money to Paul’; ‘John gives Paul the money’ (Chapuis forthcoming)

In the VLRF, however, no preposition is used and either object can come first when neither is a pronoun: (230)

(231)

IO DO VLRF: Zan la di momon tout mon s´egr´e John PAST tell mother all 1s secrets ‘John told Mother all my secrets.’ (Ramassamy 1985:271) DO IO VLRF: Zan la di tout mon s´egr´e momon.

(ibid.)

Malagasy has the same lack of constraints on the order of direct and indirect objects (Chapuis forthcoming citing Keenan 1975:251, 267– 268). 5.5.2

Dependent clauses in VLRF

Ramassamy (1985:229) notes that in VLRF relative clauses immediately follow the noun they modify without a relativizer: (232)

VLRF: mang vert {mwen la manz´e} la t´e maf. mango green 1s PAST eat DET was over-ripe ‘The green mango that I ate was over-ripe.’ (ibid.)

While relative pronouns are obligatory in French, the VLRF pattern follows that of Malagasy, in which the relative pronoun izay can also be omitted (Chapuis forthcoming citing Keenan 1978:296). Ramassamy further notes (1985:230) that the end of the NP can be indicated by la (as above), which demarcates the syntactic unit and facilitates the parsing of the sentence. Similarly, Malagasy “frames” noun phrases with

132

Languages in Contact

an initial and final demonstrative (Chapuis forthcoming citing Keenan 1978:299). Cellier (1985a:150) notes that ke (cf. F qui) can function as an overt relativizer, which Watbled (personal communication) describes as acrolectal: (233) (234)

VLRF: l´et´e k`et soz {ke l´et´e dur}. ‘It was something that was hard.’(ibid.) VLRF: na dmoun i gard´e le feu asoir (ibid.) EXIST people ASM watch ART fire this-evening ‘There are people who watched the fire this evening.’ (ibid.) cf. F: il y a des gens qui gardaient le feu ce soir.

Ramassamy (1985:378) does note that relativizers like ousa ‘where’ can occur but are not obligatory. The fact that such relativizers occur optionally as objects of prepositions (which are themselves optional) again suggests influence from French: (235)

VLRF: mwen la pran le ti-santy´e {(par-ousa) nou la pas´e lot zour}. ‘I took the little path by which we went the other day.’ (ibid.)

Again unlike French, VLRF can omit ke (cf. F que) functioning as a subordinator. Ramassamy (1985:364) treats the juxtaposition of the subordinate clause without a subordinator as the normal VLRF structure after certain verbs, such as those of saying or perception: (236)

VLRF: li krwa { ou ri [sic: kri?] a-li}. ‘He thinks [that] you are calling him.’ (ibid.)

However, Cellier (1985a:150) notes that this kind of sentence can take the subordinator ke (or its variant k), again possibly under the influence of standard French: (237)

VLRF: mi kroi {k i apel´e “tandif.”} ‘I thought that it was-called “tandif”.’ (ibid.)

From the perspective of French, the subordinator is also missing in clauses following what seems to be a preposition, pou (cf. F pour ‘for’): (238)

VLRF: Mwen la di a-zot sa {pou li kon`et}. (Ramassamy 1985:349) ‘I told you that so he would know.’ cf. F: Je vous ai dit c¸ a pour qu’il le sache.

The structure of clauses

133

Ramassamy (1985:350) calls this a “pseudo-infinitif” since there are no other grounds to justify positing the existence of infinitives in VLRF. This construction, found throughout the Atlantic creoles – see sentence (221) above – lends itself to translation with an infinitive in a number of European languages (‘. . . for him to know’), but it is actually a tensed clause and pou is best treated as a subordinator meaning ‘so that’ (Holm 1988:168–70). Watbled (personal communication) considers sentence (238) as containing an infinitive clause; he points out that a corresponding finite construction, more like the French equivalent of (238) above, also exists: (239)

VLRF: Mwen la di a-zot sa {pou k(e) li kon´e.}

5.6

A comparison of clause structure

To reiterate, no claim is being made here that the newer language varieties discussed above are any more closely related genetically or typologically than the older languages out of which they developed. However, it is being claimed that they underwent a similar kind of partial restructuring triggered by partly similar social and linguistic factors. For this reason, the differences in the structure of clauses of the newer and older languages are compared with one another below to identify any common patterns of restructuring that might characterize this kind of partial restructuring. 5.6.1

A comparison of word order

The five language varieties examined here have the word order for declarative sentences that predominates in both their superstrate and substrate languages. For all but Afrikaans this is SVO, which is also the basic word order of all of the Atlantic creoles; for Afrikaans, that word order is SOV. Although the inversion of the subject and the verb (or auxiliary) to transform a statement into a question can or must occur in all the superstrate languages, this is not a part of the syntax of full-fledged creoles or their substrate languages. Non-inversion can or must occur in all the partially restructured varieties except Afrikaans, which adheres strictly to Dutch word order. 5.6.2

A comparison of dependent clauses

AAE relative clauses differ from those of standard English mainly in that subject relative pronouns can be omitted; this structure is found in some English-based creoles as well as earlier and nonstandard varieties of

134

Languages in Contact

English, which appear to have had a converging influence on AAE. On the other hand, AAE’s use of the subordinator say to introduce a quotation is clearly of creole and African origin. Dependent clauses in Afrikaans closely follow the structure found in Dutch, except that Afrikaans has a single flectionless relative pronoun wat (except for complex relatives) and has a zero subordinator as an allomorph of dat. BVP relative clauses differ from those of standard Portuguese in that BVP has only one form of the relative pronoun, que, corresponding to the single simplex form of the Afrikaans relativizer, wat. The structure of BVP relative clauses follows patterns found in Portuguese-based creoles and nonstandard European Portuguese. The avoidance of stigmatized resumptive pronouns suggests the camouflaging strategies characteristic of decreolization. NSCS and BVP differ from standard Spanish and Portuguese in that the complementizer que can be omitted both as a relativizer and as a subordinator, but this occurs only in unusually basilectal or archaic varieties. Zero seems to be the default form of both kinds of VLRF complementizers, although ke or k also occurs, apparently under the influence of French. In summary, clauses in the partially restructured varieties examined here have many of the structural features of their European source languages, although they also reflect morphosyntactic simplification and have some features found in fully creolized languages and their substrates.

6

Conclusions

Introduction This concluding chapter will attempt to relate the social information in chapter 2 to the linguistic information in chapters 3–5 in order to provide a basis for a theory that can account for the facts known about the five languages examined here: what led to their partial restructuring and how this process affected their structure.

6.1

Social factors in partial restructuring

Chapter 2 pointed to a single, overriding social factor in the development of African American English, Afrikaans, Brazilian Vernacular Portuguese, Nonstandard Caribbean Spanish, and the Vernacular Lects of R´eunion French as varieties distinct from both unrestructured overseas varieties of their source languages (e.g. the English of Ontario, the extinct Dutch of New York and New Jersey, the Portuguese of Madeira, the Spanish of Chile, or the French of Quebec) and completely restructured creole languages (e.g. Guyanese Creole English, the extinct Creole Dutch of the Virgin Islands, Guin´e-Bissau Creole Portuguese, Palenquero Creole Spanish, or Mauritian Creole French). That social factor is the demographic balance, during the first century of a new language’s development, of native speakers versus non-native speakers of the European source language. Parkvall (2000) is certainly correct in his conclusion that this demographic ratio is not the only social factor that determines the degree of language restructuring: there are other relevant factors, such as an incoming population already having some fluency in a common restructured language brought in from elsewhere, such as the English-Creole-speaking slaves imported into the southern American colonies during the seventeenth century (introduction to chapter 2 and section 2.1.1). Moreover, time is certainly a relevant factor: the importance of a society’s final demographic balance can be overridden by the earlier emergence of a common 135

136

Languages in Contact

partially restructured language spoken by a population that is fully accessible to a later incoming population, as was the case in Barbados and R´eunion (sections 2.1.1 and 2.5.1–2.6, respectively). However, we can conclude from both Parkvall (2000) and chapter 2 of the present study that the ratio between native and non-native speakers of the source language during the first century of a new language’s development is indeed the most important social factor in determining the structure of that language. Where native speakers made up a strong majority in the new society, unrestructured overseas varieties developed. Where non-native speakers made up a strong majority, fully restructured creole languages developed. Neither process was particularly pretty: it is well known that many plantations that depended on slave labor were in fact death camps that simply consumed the men and women brought there to work. However, the ethnic changes necessary for unrestructured overseas language varieties to flourish (except for those on previously uninhabited islands like Madeira, of course) were not much more attractive: massive immigration of European colonists, who controlled the wealth, government, and cultural institutions of the colonies; coupled with the extermination, absorption, or retreat of speakers of indigenous languages. The partial restructuring of languages occurred in new societies where neither group – neither native nor non-native speakers, which in the beginning meant neither Europeans nor non-Europeans – were numerous enough completely to overwhelm the other group culturally. The fact that partially rather than fully restructured languages developed in societies with a higher proportion of native speakers of the European lingua franca is logical: there were simply more native speakers to provide non-native speakers with samples of the language from which the latter could derive the rules needed to speak it. Despite social stratification, learners still had better access to the target language than they did in those plantation societies where fully creolized languages developed. This led to two defining characteristics of the resulting languages: first, the non-native version of the European language was never as completely restructured as a fully creolized language; second, as the partially restructured language acquired native speakers – often among the descendants of Europeans as well as non-Europeans – it developed into an identifying community language that could draw on features not only from the nonnative lingua franca but also from native-speaker varieties of the European language. The five partially restructured languages examined here continued to be in contact not only with native-speaker varieties of the European language (either acquired by local whites via normal transmission or brought in by new arrivals from the colonizing country) but also with

Conclusions

137

the fully restructured pidgins and creoles spoken by newly arrived slaves: by Caribbean and African slaves brought to the American South; by Asians and Africans brought to the Cape Colony; by African slaves arriving via Cape Verde and S˜ao Tom´e to Brazil and the Spanish Caribbean; and by Africans, Malagasies, and Indians brought to R´eunion. Furthermore, these partially restructured languages were in varying degrees of contact with the pidgins and subsequent creoles that developed in nearby areas where sociolinguistic conditions were favorable to fuller restructuring, producing the forerunners of Gullah in the American South, Orange River Afrikaans in the Cape Colony, Helv´ecia Portuguese in Brazil, Habla Bozal and Poror´o in the Spanish Caribbean, and the Cr´eole des Bas on R´eunion. Thus, even after the local pidgin or jargon of the earliest contact period ceased to be used, non-European features could still be borrowed into the version of the local language used by the monolingual descendants of the non-European groups. And, because humans signal their social identity and solidarity with others through their choice of linguistic variables (Le Page and Tabouret-Keller 1985), and because the social identities of the descendants of both the European and the non-European groups evolved over time, as did their relationships to one another, the variety spoken by each group tended to grow less dissimilar to that of the other through the two-way borrowing of features on all linguistic levels.

6.2

Linguistic factors in partial restructuring

Unlike the social factors discussed above, the linguistic facts surveyed in the three preceding chapters are so many and various that they lend themselves less readily to generalizations. Table 9 is presented in order to allow a better overview of the morphosyntactic changes that characterize partial restructuring. Space constraints have made necessary a good deal of encoding in this table; the following explanation is intended to help crack the code. First, the features are described by number without abbreviations below, followed by examples in AAE wherever possible (since readers know the grammar of modern standard English, which is often close to AAE’s source language), followed by the number of the sentence in the preceding chapters, where further discussion of the morphosyntactic feature in question can be found. 1. The absence of an inflection indicating the third person singular form of the present tense: AAE: Where Miss Annie . . . live now. (1)

138

Languages in Contact

Table 9. Key morphosyntactic features in partial restructuring AAE Sb AFR Sb BVP Sb NSCS Sb VLRF Sb E D P S F Verb phrase Verbal morphology∗ 1. Zero 3s PRES infl. 2. Zero 1p PRES infl. 3. Zero PAST infl. Aux./ preverbal marker 4. Semantic influence Negation 5. Negative concord 6. Discontinuous double 7. Non-verbal predicates Noun phrase Number 8. Zero plural infl.∗ 9. Unbound pluralizer 10. Associative plural Gender 11. No agreement in NP Possession 12. [possessor 0 possessed] 13. [possessed 0 possessor] Pronouns 14. Reduced case marking 15. Zero reflexive pronoun Clauses Word order 16. QW S-V/Aux (direct)∗∗ Dependent clauses 17. Zero subject REL 18. Zero subordinator “that” Total number of +’s

+ + +

+ + +

+ + +

+ + +

0 + 0

+ + +

+

0 0 0

+ + +

+

+ + 0

+ +

00 00 00

000 000 000

+

00

000

+

+

+ 0 +

+

0 + 0

+

+ + +

+ +

+ + +

+ +

+ 0 +

+

0 0 +++ 00 +0+ 00 000

+ + +

+ + +

0 0 +

+ + +

+ + 0

+ + +

+ 0 0

+ + +

+ + 0

+ + +

00 00 00

000 000 000

+

+

+

+

0

+

0

+

+

+

+0

000

+ 0

+ +

0 0

+ +

0 +

+ +

0 0

+ +

0 +

+ +

00 00

000 000

+ 0

+ +

+ 0

+ +

+ +

+ +

0 +

+ +

+ +

+ +

00 00

000 000

+

+

0

+

+

+

+

+

+

+

00

000

+ + 15

+ +

0 + 9

+ +

+ + 13

+ +

+ + 9

+ +

+ + 14

+ +

00 +0

000 000

Sb = relevant substrate language(s) + = attested presence of feature infl. = inflection 0 = attested absence of feature see also List of Abbreviations, p. xix = not applicable or unknown ∗ refers to spoken, non-suppletive forms only ∗∗ with S immediately following QW (non-echo question)

2. The absence of an inflection indicating the first person plural of the present tense: BVP: n´os parte ‘we leave’ (cf. SBP n´os partimos ibid.) (34) 3. The absence of an inflection indicating past tense: AAE: They taught me mighty good, they teach me good. (3) It has to be stipulated that the above three features (as well as feature 8, the absence of a plural inflection on nouns etc.) refer only to spoken, non-suppletive forms in order to exclude irregular standard English

Conclusions

139

past forms like put or silent standard French inflections like (il ) vient ‘he comes.’ Of course, it is arbitrary to assign the inflection on E we go or F il parle ‘he speaks’ either + or 0. 4. Semantic influence of a (creole) preverbal marker on a (source language) auxiliary verb: AAE: I got on a cowboy shirt now. . . . Been have it all my days. (4) 5. The presence of a negator before a verb requires the negative form of indefinite determiners, pronouns, etc. AAE: We don’ want no six-month investigation! (20) Assigning the presence of negative concord a plus in this chart seems to imply that it results from restructuring, which is not at all clear. It is a standard feature in the Romance languages and in earlier, nonstandard, varieties of English and Dutch (Brachin 1985:22). It is found in Atlantic creoles of all lexical bases (Holm 2000:195–196), but its presence in substrate languages is unknown. Apropos of double negatives, part of the problem in decoding Table 9 has to do with negative designations for certain features, e.g. 1 (Zero 3s PRES infl.) or 11 (No agreement in NP). This encoding was necessary to avoid the opacity of having the same symbol (e.g. +) randomly encoding both restructuring and the lack of it. As the table stands, + encodes restructuring except in a few cases where the European superstrates (E D P S F) differ among themselves structurally (e.g. feature 11: all except English have gender agreement within the NP). In sum, no negative (or zero feature, such as the absence of an inflection) equals a negative (0, e.g. the absence of the Zero 3s PRES infl. in standard English), while its presence equals a positive (+, e.g. the presence of the Zero 3s PRES infl. in AAE). 6. The presence of a negator both before and after the verb: ˜ sabe nao. ˜ literally ‘He doesn’t know no.’ (38) BVP: Ele nao 7. The absence of the equivalent of a form of ‘be’: AAE: They all dead. (24) 8. The absence of a plural inflection on nouns (or other elements of the noun phrase): AAE: them hound (68) 9. A separate word (often meaning ‘they’) to indicate plurality: AAE: them hound (68) 10. The use of a pluralizing word to indicate a person’s usual associates: AAE: Felicia an’ them done gone. (69) 11. The absence of inflections indicating gender agreement between a noun and its modifiers: VLRF: mon kaz l´e gran ‘my house is big’ versus F ma maison est grande (132)

140

Languages in Contact

12. Nouns have no inflection identifying them as the possessor of the following noun: AAE: the white folk kitchen (74) 13. Nouns do not follow a preposition indicating that they possess the preceding noun: BVP: kaza Maria ‘Maria’s house’ (cf. SBP a casa de Maria) (110) 14. The forms of personal pronouns do not necessarily indicate their grammatical function in the sentence as in the source language: AAE: We had we own lawyers. (76) 15. Reflexive pronouns required in the source language can be omitted: BVP: Jo˜ao cortou com faca. ‘John cut [himself] with a knife.’ cf. SBP: Jo˜ao courtou-se com a faca. (115) 16. In direct questions with a question word, the subject can precede the verb or auxiliary: AAE: Where I can go? (150) It should be explained that the subject must immediately follow the question word to exclude constructions like Where [is it that] I can go? Furthermore, it is a direct question and not the repetition of an indirect question for confirmation: Where I can go? 17. A relative pronoun functioning as the subject of the clause can be omitted: AAE: He got a gun { sound like a bee}. (158) 18. The equivalent of ‘that’ introducing a subordinate clause can be omitted: NSCS: Dice jag¨uey t´a chiquito. ‘He says that the liana is small.’ (217) This survey of the salient morphosyntactic features that distinguish these five partially restructured varieties from the standard variety of their source language (which does not imply that the latter was necessarily the most relevant source of the former) focuses on those features found in a number of the restructured varieties rather than in just one, such as the unbound possessive marker se in Afrikaans. This is because the point of table 9 is to provide an overview of the general structural tendencies of languages that have undergone partial restructuring. Many of these tendencies can be characterized as structural reduction: reduced morphological marking for person or tense on verbs; for number on nouns and other elements in the NP; or for case on personal pronouns. Sometimes this reduction means the loss of syntactic complexities (such as subject-verb inversion in questions) or the loss of function words (the reflexive pronoun, the preposition equivalent to ‘of ’ indicating

Conclusions

141

possession, the subordinator equivalent to ‘that’). However, the loss of these particular features rather than others does not seem to be random: the losses that took place tend to make the partially restructured varieties more like their substrate languages. Of course isolating Niger-Congo languages formed an important part of the substrate of most of the partially restructured languages examined here, so it is not surprising that these tend to be more isolating, too. However, it is also a universal in second language acquisition that adults tend, when possible, to isolate grammatical information in unbound morphemes rather than inflections. Yet there is no need to choose one of these linguistic processes over the other in accounting for the structure of the new varieties: each one obviously reinforced the other. Furthermore, the fully restructured creoles of the same lexical base with which the partially restructured varieties may have been in contact bear the mark of their substrates even more strongly, but again there is no need to choose among these three forces pulling in the same direction. In the final analysis, there is no question that the partially restructured varieties bear the stamp of their substrate, since they have innovative structures that represent not a reduction of the structure of their superstrate languages but rather an addition to it from their substrate: features 4 (the semantic influence of preverbal markers on auxiliaries), 6 (discontinuous double negation), 7 (non-verbal predicates), 9 (unbound pluralizers), and 10 (associative plurals). And there are many more such features that occur in only one or two partially restructured varieties: the AAE complementizer say (5.1.2), the VLRF agentive subject marker i (3.5.1), the resumptive pronouns in BVP relative clauses (5.3.2), etc. Regarding the last feature, it is probably not entirely coincidental that this feature in BVP sentence (197) is also found in AAE sentence (163): it is the same strategy for repairing ambiguity in a variety that has undergone restructuring with all the potential for impairment of communication that this may have brought about. Other such parallels include the option for subject-verb (or auxiliary) inversion found in embedded questions in both AAE (151) and Afrikaans (179). The cunning that languages appear to have to repair and renew themselves makes August Schleicher’s nineteenth-century conceit – that languages are living organisms in a Darwinian world – seem less quaint; of course, such cunning is that of a language’s speakers, who are indeed such organisms. Finally, table 9 confirms subjective impressions about the degree of restructuring which each of these varieties has undergone (cf. section 1.7). Arranging them according to their total number of positive features (indicating restructuring) yields the following hierarchy:

142

Languages in Contact

15 African American English 14 Vernacular Lects of R´eunion French 13 Brazilian Vernacular Portuguese 9 Afrikaans 9 Nonstandard Caribbean Spanish Of course, the accuracy of such a quantitative indication of their degree of restructuring depends on the representativeness of the features chosen for table 9. Still, this hierarchy is confirmed by a comparison of the two varieties with the most closely related source languages, BVP and NSCS. Features indicating restructuring that are not found in NSCS are present in BVP: e.g. features 2, 9, 13, and 14. 6.3

Linguistic processes in partial restructuring

I proposed an earlier version of the theory below (Holm 2000b) to account for what is known about partial restructuring, based on the results of the research of my students, myself, and others. This model includes the findings of Mello (1997), who observes that a number of linguistic processes must have combined to trigger the partial restructuring of Brazilian Vernacular Portuguese; and Green (1997), who in her study of Nonstandard Caribbean Spanish, saw that, in addition to contact with more fully restructured varieties, what sets this kind of partial restructuring apart from other kinds of language shift (cf. Thomason and Kaufman 1988) is that it occurs among a shifting population speaking a number of different first languages. Both of these factors place the kind of partial restructuring discussed here closer to full creolization than language shift by linguistically homogeneous groups, as Thomason acknowledges: in ordinary shift situations there is no negotiation among shifting speakers, because they all share the same L1; and if they learn the T[arget] L[anguage] imperfectly, they are likely to make similar or identical “errors.” These “errors” may include (among other things) marked features of their original L1. (Thomason 2001:255)

However, the kind of partial restructuring discussed here is more like full creolization in that it presupposes shift by a linguistically heterogeneous population. This implies negotiation in the acceptance of substrate features, so that highly marked features lacking transparency are less likely to be accepted into the emerging vernacular. However, marked features that are transparent – e.g. AAE say introducing quotations, as in sentence (165) – are acceptable. More importantly, the existence of a number of different first languages among the shifting population is crucial in that it precipitates “. . . a forced shift because of an urgent need for a contact medium in a new multilingual contact situation” since bilingualism in

Conclusions

143

two of the substrate languages cannot provide a solution to the crisis in communication (Thomason and Kaufman 1988:307). The theoretical model for partial restructuring presented here presupposes a population with different first languages shifting to a typologically distinct target language (itself an amalgam of varieties in contact, including fully restructured ones) under social circumstances that partially restrict access to the target language as normally used among a minority or weak majority population of native speakers. It predicts that some or all of the following linguistic processes will shape the resulting restructuring: 1. language drift, following internal tendencies within the target language, particularly phonotactic, morphological, or syntactic simplification; 2. primary leveling, preserving lexical or structural features that are archaic, regional, or rare in the target language, sometimes extending them to new contexts; 3. imperfect language shift by the entire population, perpetuating structural features from ancestral languages and interlanguages in the speech of monolingual descendants; 4. language borrowing, incorporating structural features from fully pidginized or creolized varieties of the target language spoken by newcomers or found locally but confined to areas where sociolinguistic conditions were favorable to full restructuring; 5. secondary leveling, or the possible loss of features not found in the target language (from any of the above processes) if there is continued contact with the target language and it is perceived to have more prestige. These processes result in a new variety with a substantial amount of the target language’s structure intact, but also with a significant number of substrate or interlanguage structural features, i.e. a partially restructured vernacular. The order in which these processes are listed above follows the order in which they occur. The first two refer to the Western European lexical source languages taken overseas: all had been undergoing drift (1) in Europe; when their different regional and social varieties came into closer contact in an overseas colony, they underwent the first stage of dialect leveling (2). The overridingly important process in partial restructuring is language shift (3) of the dominated non-European population to the language of the dominating Europeans. After this there is language borrowing; it is convenient to distinguish the new variety’s borrowing of what are largely substrate features (4) from its borrowing of features of the lexical source language (5), parallel to the decreolization of fully creolized varieties.

144

Languages in Contact

6.4

Comparative studies in restructuring

The aim of this book has been to demonstrate the usefulness of the concept of the partial restructuring of languages. That this concept has eluded linguists up to now is not surprising: partially restructured languages had simply fallen between the cracks of theory, being neither unrestructured overseas varieties nor fully restructured creoles. Each was compared only to varieties of its lexical source language, so it was not possible for any pattern to emerge. It is the comparison of such varieties not only with their source languages but also with one another – focusing on their sociolinguistic histories as well as their synchronic structure – that makes it clear that despite their dissimilar vocabularies they are, in a very important sense, the same kind of language, resulting from the same sociolinguistic processes. Recent research comparing different partially restructured languages (e.g. Holm, Lorenzino, and Mello 1999) clearly demonstrates that insights gained from the study of one can cast light on others. While scholars of NSCS disinclined to accept any external influence on its structure have long resorted to postulating internally motivated phonological rules to account for the loss of inflectional morphology on verbs and within the noun phrase, the similar but more pervasive restructuring of BVP indicates that phonological rules alone are an inadequate explanation, forcing the issue of morphological restructuring resulting from contact (see section 3.4.1). Because of the sociolinguistic parallels in the history of NSCS and BVP, it can be argued by analogy that what led to the reduced inflectional morphology of not only BVP but also NSCS was contact-induced morphological simplification rather than just phonological rules, which themselves seem more likely to have been motivated by such contact than random language-internal forces. Moreover, the difference in the degree of restructuring evident in NSCS and BVP, objectively confirmed in section 6.2, points to a simple but important observation: that restructuring can indeed take place to differing degrees. This issue is now settled. Another insight from comparative study can be found in Afrikaans, whose sociolinguistic history has been shown to have defining parallels with the other four language varieties examined in this study. There is a salient contrast between the lack of conjugational inflections in Afrikaans and their presence in its source language, Dutch (section 3.2.1). At first glance, this too might seem attributable to phonological rules such as the simplification of final consonant clusters. In fact, such a phonological rule has long been evoked to account for the loss of certain inflections

Conclusions

145

in another speech variety with important parallels in its sociolinguistic history, AAE. The phonological constraints on the deletion of the morpheme {-s} (marking the plural or possessive of nouns or the third person singular present tense of verbs or even the contracted form of is) or of {-d} (marking the past and past participle form of verbs) do not in themselves provide an adequate explanation for this deletion, as Labov (1969) implied. Indeed, the flectionless simplicity of AAE habitual be is the mirror image of Afrikaans is, and the only credible explanation is that both varieties underwent morphological restructuring. Since there is no convincing evidence that AAE itself was ever fully creolized (and considerable demographic evidence to the contrary), we can safely conclude that AAE underwent partial restructuring as it is defined above. Such comparative work demonstrates the usefulness of this theoretical model in accounting for the known facts of the historical development and synchronic structure of the five partially restructured languages examined in this study: identifying common sociolinguistic patterns in their histories should lead us to examine their linguistic structures aided by insights gained from the study of other varieties that fall into the same category. The concept of a partially restructured language is a useful generalization about a group of languages that facilitates predictions about all members of this category. The discovery of such falsifiable generalizations is what science is supposed to be about. As noted in the introduction to this book, there has been a recent movement to place pidgin and creole linguistics within the broader scope of contact linguistics (Thomason 1997), including language varieties resulting not only from the kind of restructuring associated with pidginization and creolization (to whatever degree) but also those resulting from such processes as intertwining (Bakker and Muysken 1994), koineization, and indigenization (Siegel 1997). Such studies promise to increase our understanding of the range of possible outcomes of language contact, and this understanding will surely shed new light on many kinds of restructuring. For example, Siegel (1997) examined immigrant koines (e.g. overseas Hindi), indigenized varieties (such as Singapore English), and even renativized Hebrew, concluding that the adoption of features in the leveling that took place in all of these was affected by certain factors: frequency, regularity, salience, and transparency. This suggests a solution for the long search for principles that guide the selection of substrate features into pidgins, creoles, and partially restructured varieties during their genesis, and their adoption of other features during their development: the likelihood of a lexical or structural feature being selected is greater if it is more frequent, more regular, more salient, or more transparent. Siegel’s

146

Languages in Contact

observation of this tendency in contact between different varieties of the same language is supported by Boretzky’s (1986) observation of the same tendency in contact between different languages. Contact linguistics should continue to apply the insights gained from the study of one kind of restructuring to the study of other kinds of restructuring. Certainly it is time to re-examine what we understand not only about Romani but also Romanian, Old French, and Middle English, as well as all the other languages whose histories are known to have involved substantial contact.

References

Adam, L. 1883. Les idiomes n´egro-aryen et mal´eo-aryen: essai d’hybridologie linguistique. Paris: Maisonneuve. Almada, M. D. de Oliveira. 1961. Cabo Verde: contribui¸ca˜ o para o estudo do dialecto falado no seu arquip´elago. Estudos de ciˆencias pol´ıticas e sociais, No. 55. Lisbon: Junta de Investiga¸coes ˜ do Ultramar. Alvarez, Alexandra. 1990. Venezuelan Spanish: the search for a creole continuum. Ph.D. dissertation, Georgetown University. Alvarez Nazario, Manuel. 1961. El elemento afronegroide en el espa˜nol de Puerto Rico. San Juan de Puerto Rico: Instituto de Cultura Puertorriquena ˜ (2nd edn., 1974). Amaral, Amadeu. 1928. O Dialeto Caipira. (3rd edn. 1976, S˜ao Paulo: HUCITEC, Secretaria da Cultura, Ciˆencia e Tecnologia). Arends, Jacques, ed. 1995. The Early Stages of Creolization. Amsterdam Philadelphia: John Benjamins. 1999. The origin of the Portuguese element in the Surinam Creoles. In M. Huber and M. Parkvall, eds., pp. 195–208. Arends, J., P. Muysken, and N. Smith, eds. 1994. Pidgins and Creoles: an introduction. Amsterdam/ Philadelphia: John Benjamins. Azevedo, Milton M. 1989. Vernacular features in educated speech in Brazilian Portuguese. Hispania 72, pp. 862–872. Bailey, Guy. 2001. The relationship between African American and White vernaculars in the American South: a sociocultural history and some phonological evidence. In S. Lanehart, ed., pp. 53–92. Bailey, Guy, and Natalie Maynor. 1987. Decreolization? Language in Society 16, pp. 449–473. 1989. The divergence controversy. American Speech 64, pp. 12–39. Bailey, Guy, Natalie Maynor, and Patricia Cukor-Avila, eds., 1991. The Emergence of Black English: texts and commentary. Amsterdam: John Benjamins. Baker, Philip. 1972. Kreol: a description of Mauritian Creole. London: Hurst. 1982. The Contribution of Non-Francophone Immigrants to the Lexicon of Mauritian Creole. Ph.D. diss., University of London. 1996. Pidginization, creolization, and fran¸cais approximatif. Review of Des ˆıles, des hommes, des langues, by Robert Chaudenson. Journal of Pidgin and Creole Languages 11:1, pp. 95–120. Baker, Philip, and Chris Corne. 1982. Isle de France Creole: affinities and origins. Ann Arbor: Karoma. 147

148

References

Bakker, Peter, and Pieter Muysken. 1994. Mixed languages and language intertwining. In Arends et al., eds., pp. 41–84. Baptista, Marlyse, Heliana R. de Mello, and Miki Suzuki. Forthcoming. Capeverdean and Guin´e-Bissau Creole Portuguese. In J. Holm and P. Patrick, eds. Barry, M. V. 1982. The English language in Ireland. In R. Bailey and M. Gorlach, ¨ eds., English as a World Language. Ann Arbor: University of Michigan Press, pp. 84–133. Baugh, John. 1983. Black Street Speech: its history, structures, and survival. Austin: University of Texas Press. Baxter, Alan N. 1987. Creole-like traces in rural Brazilian Portuguese dialects. Department of Spanish, La Trobe University, Bundoora, Australia. 47 pp. MS. 1992. A contribui¸ca˜ o das comunidades afro-brasileiras isoladas para o debate sobre a criouliza¸ca˜ o pr´evia: um exemplo do estado da Bahia. In E. d’Andrade and A. Kihm, eds., pp. 7–35. 1997. Creole-like features in the verb system of an Afro-Brazilian variety of Portuguese. In A. Spears and D. Winford, eds., pp. 265–288. Baxter, Alan N., and Dante Lucchesi. 1997. A relevˆancia dos precessos de pidginiza¸ca˜ o e criouliza¸ca˜ o na forma¸ca˜ o da l´ıngua portuguesa no Brasil. Estudos ling¨u´ısticos e liter´arios 19, pp. 65–84. 1999. Un paso m´as hacia la definicion ´ del pasado criollo del dialecto afrobrasileno ˜ de Helv´ecia (Bahia). In K. Zimmermann, ed., pp. 119–141. B´eraha, Jeanette. 2002. La particle verbal

i en r´eunionnais. Paper presented ´ at the Colloque International des Etudes Cr´eoles, R´eunion. Bhattacharjya, Dwijen. 2000. The Genesis and Development of Nagamese: its social history and linguistic structure. Ph.D. dissertation, City University of New York. Bickerton, Derek. 1973. On the nature of a creole continuum. Language 49, pp. 640–649. 1981. Roots of Language. Ann Arbor: Karoma. 1984. The language bioprogram hypothesis. Behavioral and Brain Sciences 7, pp. 173–221. Bickerton, Derek, and Aquiles Escalante. 1970. Palenquero: a Spanish-based creole of northern Colombia. Lingua 24, pp. 254–267. Bloomfield, Leonard. 1933. Language. London: Allen and Unwin. Boll´ee, Annegret. 1977. Zur Entstehung der franz¨osischen Kreolen-dialekte im Indischen Ozean: Kreolisierung ohne Pidginisierung. Geneva: Droz. Bonvini, Emilio. 2000. La langue des

pretos velhos (vieux noirs) au Br´esil: un cr´eole a` base portugaise d’origine africaine? Bulletin de la Soci´et´e de Linguistique de Paris XCV, pp. 389–416. Boretzky, Norbert. 1983. Kreolsprachen, Substrate und Sprachwandel. Wiesbaden: Harrassowitz. 1986. Verbkategorien im Fantse und im Jamaican Creole. Ms. Boshoff, S. P. E. 1921. Volk en taal in Suid-Afrika. Pretoria: De Bussy. 1959. Die wonder van die ontstaan van Afrikaans. In M. S. du Buisson, ed., Die wonder van Afrikaans. Johannesburg: Voortrekkerpers, pp. 25–35.

References

149

Bosman, Daniel B. 1923. Oor die ontstaan van Afrikaans. Amsterdam: Swets and Zeitlinger. 1947. Die ontstaan van Afrikaans. In C. M. van den Heever and P. de Villiers Pienaar, eds., Kultuurge skiedenis van die Afrikaner. Cape Town, Bloemfontein, and Port Elizabeth: Nasionale Pers 2, pp. 1–22. ´ Bourquin, A. 1994. Etude d’une cat´egorie sociale: les Petits Blancs de la R´eunion. Th`ese de doctorat. Universit´e J. Moulin, Lyon. Boxer, C. R. 1963. Race Relations in the Portuguese Colonial Empire, 1415–1825. Oxford: Clarendon Press. 1965. The Dutch Seaborne Empire. New York: Knopf. 1969. The Portuguese Seaborne Empire, 1415–1825. New York: Knopf. Brachin, Pierre. 1985. The Dutch Language: a survey. Leiden: Martinus Nijhoff. Burgers, M. P. O. 1963. Teach Yourself Afrikaans. London: The English Universities Press. Burling, R. 1973. English in Black and White. New York: Holt, Rinehart, and Winston. Byrne, F., and J. Holm, eds. 1993. Atlantic Meets Pacific. Amsterdam/Philadelphia: John Benjamins. Cabrera, Lydia. 1969 [1954]. El Monte. Miami: Ediciones C.R. Caid, Leila. 2002. Le morph`eme verbal

i en cr´eole r´eunionnais. Paper pre´ sented at the Colloque International des Etudes Cr´eoles, R´eunion. Calvet, Louis-Jean and Robert Chaudenson. 1998. Saint-Barth´elemy. Une e´nigme linguistique. Paris: Didier Erudition. Carden, G., M. Goodman, R. Posner, and W. A. Stewart. 1991. A 1671 French creole text from Martinique. Paper presented to the Society for Pidgin and Creole Linguistics, Chicago. Careno, Mary F. do. 1997. Vale do Ribeira: a voz e a vez das comunidades negras. S˜ao Paulo: Editora Arte e Ciˆencia. Carlin, Eithne B., and Jacques Arends, eds. 2002. Atlas of the Languages of Suriname. Koninklijk Instituut voor Taal-, Land- en Volkenkunde, Caribbean Series 22. Leiden: KITLV Press. Carvalho, Felix de. 1977. Falares crioulos de Brasil: um tema em debate. Rio de Janeiro: Dpto. de Letras, Pont´ıfica Universidade Catolica, ´ disserta¸ca˜ o de mestrado. Cassidy, Frederic G. 1961. Jamaica Talk: three hundred years of the English language in Jamaica. London: Macmillan. Cassidy, Frederic G., and Joan H. Hall, eds. 1985 . . . Dictionary of American Regional English. Cambridge, Mass: Belknap Press. Cassidy, Frederic G., and R. B. Le Page. 1980. Dictionary of Jamaican English. Cambridge: Cambridge University Press. Castro, Y. de Pessoa de, and G. A. de Souza Castro. 1980. Culturas africanas nas Am´ericas: um esbo¸co de pesquisa conjunta na localiza¸ca˜ o dos empr´estimos. Afro-Asia 13, pp. 27–50. Cellier, Pierre. 1985a. Comparaison syntaxique du cr´eole r´eunionnais et du fran¸cais (R´eflexions pr´e-p´edagogiques). Universit´e de la R´eunion. 1985b. Description syntaxique du cr´eole r´eunionnais: essai de standardisation.Th`ese de doctorat d’Etat. Aix: Universit´e de Provence.

150

References

Chapuis, Daniel. forthcoming. Aspects of Restructuring in Vernacular Lects of R´eunion French. Ph.D. dissertation, City University of New York. Chaudenson, R. 1974. Le lexique du parler cr´eole de la R´eunion. Paris: Champion. 2 vols. 1992. Des ˆıles, des hommes, des langues. L’Harmattan: Paris. 1995. Les Cr´eoles. Que sais-je? Paris: Presses Universitaires de France. 2000. Cr´eolisation du fran¸cais et francisation du cr´eole: Les cas de SaintBarth´elemy et de la R´eunion. In I. Neumann-Holzschuh and E. W. Schneider, eds., pp. 361–381. Coelho, F. Adolpho. 1880–1886. Os dialectos romˆanicos ou neolatinos na Africa, Asia, e Am´erica. Boletim da Sociedade de Geographia de Lisboa. Reprinted 1967 in Estudos Ling¨u´ısticos Crioulos, J. Morais Barbosa. Lisbon: Acadˆemia Internacional de Cultura Portuguesa. Cohen, W. B. 1980. The French Encounter with Africans: white response to blacks, 1530–1880. Bloomington London: Indiana University Press. Combrink, J. 1978. Afrikaans: its origin and development. In L. W. Latham and K. P. Prinsloo, eds., Language and Communication Studies in South Africa. Oxford: Oxford University Press, pp. 69–95. Couto, Hildo H. do. 1996. Introdu¸ca˜ o ao estudo das l´ınguas crioulas. Bras´ılia: Editora da Universidade de Bras´ılia. 1997. Falar capelinhense: um dialeto conservador do interior de Minas Gerais. Col´oquio internacional: substandard e mudan¸ca no portuguˆes do Brasil. Berlin: Ibero-Amerikanisches Institut. Craig, Beth. 1991. American Indian English. English World-Wide 12:1, pp. 25– 62. Crystal, David. 1986. A Dictionary of Linguistics and Phonetics. Oxford: Blackwell. Cukor-Avila, Patricia. 2001. Co-existing grammars: The relationship between the evolution of African American and White Vernacular English in the South. In Sonja L. Lanehart, ed. 2001, pp. 93–127. Cunha, Celso Ferreira da. 1981. L´ıngua, Na¸ca˜ o, Aliena¸ca˜ o. Rio de Janeiro: Editora Nova Fronteira (Cole¸ca˜ o Logos). 1982. Gram´atica da L´ıngua Portuguesa. Rio de Janeiro: FENAME (Funda¸ca˜ o Nacional de Material Escolar, Minist´erio da Educa¸ca˜ o e Cultura). Curtin, Philip D. 1975. The Atlantic slave trade. In A. M. Pescatello, ed., The African in Latin America. New York: Knopf, pp. 35–48. da Cunha, A. G. et al., 1982. Dicion´ario etimol´ogico Nova Fronteira da l´ıngua portuguesa. Rio de Janeiro: Editora Nova Fronteira. d’Andrade, Ernesto, and Alain Kihm, eds. 1992. Actas do col´oquio sobre crioulos de base lexical portuguesa. Lisbon: Colibr´ı. DeCamp, David. 1961. Social and geographical factors in Jamaican dialects. In R. B. Le Page, ed., Creole language studies II. Proceedings of the Conference on Creole Language Studies (University of the West Indies, Mona, 1959). London: Macmillan. 1971. Towards a generative analysis of a post-creole speech continuum. In Hymes, ed., 1971, pp. 349–370. DeGraff, Michel, ed., 1999. Language Creation and Language Change. Cambridge, Mass: MIT Press.

References

151

2001. Against creole genesis as “abnormal transmission.” Paper presented to the Linguistic Society of America, Washington, DC. de Granda, Germ´an. 1968. La tipolog´ıa “criolla” de dos hablas del a´ rea lingu´ ¨ ıstica hisp´anica. Thesaurus 23, pp. 193–205. ´ 1970a. Un temprano testimonio sobre las hablas “criollas” en Africa y Am´erica. Thesaurus 25, pp. 1–11. 1970b. On the study of the creole dialects in Spanish-speaking areas. Orbis 19, pp. 72–81. 1975. Planteamientos y necesidades actuales de los estudios lingu´ ¨ ısticos afrohispano-americanos. Anuario de Letras 12, pp. 53–82. 1976. Algunos rasgos morfosint´acticos de posible origen criollo en el habla de a´ reas hispanoamericanas de poblacion ´ negra. Anuario de Letras 14, pp. 5–22. 1978. Estudios ling¨u´ısticos hisp´anicos, afrohisp´anicos y criollos. Madrid: Editorial Gredos. 1987. Situacion ´ actual de los estudios lingu´ ¨ ısticos afrohispanoamericanos. Thesaurus 31, pp. 481–501. 1998. Estudios sobre el espanol ˜ caribeno ˜ (1985–1995): resumen y evaluacion ´ de resultados. In M. Perl and A. Schwegler, eds., pp. 62–69. de Kleine, Christa. 1997. The verb phrase in Afrikaans: Evidence of creolization? In D. Winford and A. K. Spears, eds., pp. 289–307. Deltel, J. R. 1969. Le Cr´eole de la R´eunion. Annales de Nice 7, pp. 29–34. den Besten, Hans. 1978. Cases of possible syntactic interference in the development of Afrikaans. Amsterdam Creole Studies 2, pp. 5–56. 1981. Marking WH-Movement in Afrikaans. In P. Muysken, ed., pp. 141– 179. 1985. Die doppelte Negation im Afrikaans und ihre Herkunft. In N. Boretzky et al. eds., Akten des 1. Essener Kolloquiums u¨ ber Kreolsprachen und Sprachkontakte vom 26.1.1985 an der Universit¨at Essen. Bochum: Studienverlag Dr. N. Brockmeyer, pp. 9–42. 1986. Double negation and the genesis of Afrikaans. In P. Muysken and N. Smith, eds., Substrata versus Universals in Creole Genesis. Amsterdam: John Benjamins, pp. 185–230. 1987. Review article of Raidt 1983. Journal of Pidgin and Creole Languages 2, pp. 67–92. 1989. From Khoekhoe foreigner talk via Hottentot Dutch to Afrikaans: the creation of a novel grammar. In Putz ¨ and Dirven, eds., pp. 207–250. 1993. The Khoikhoi basis for the Afrikaans associative construction. Paper presented to the Society for Pidgin and Creole Linguistics, Amsterdam. 1996. Afrikaans relatief “wat” en de West-Germanse relativisatiesystemen. In H. den Besten, J. Goossens, F. Ponelis, and P. van Reenen, eds., Afrikaans en vari¨eteiten van het Nederlands. Amsterdam: P. J. Meertens-Instituut. Taal en Tongval, Thema nummer 9, pp. 12–34. Denis, B´eatrice. 1998. St. Thomas French: a semi-creole? 62-p. ms., Department of Linguistics, City University of New York. Deumert, Andrea. 1999. Variation and Standardisation: the case of Afrikaans (1880–1922). Ph.D. dissertation, University of Cape Town.

152

References

Dieck, Marianne. 2000. La negaci´on en Palenquero: an´alisis sincr´onico, estudio comparativo y consecuencias te´oricas. Frankfurt: Vervuert/Madrid: Iberoamericana. Dillard, J. L. 1972. Black English: its history and usage in the United States. New York: Random House. D’Offay, Danielle, and Guy Lionnet. 1982. Diksyonner kreol-franse/Dictionnaire cr´eole seychellois-fran¸cais. Hamburg: Buske. Donaldson, Bruce C. 1984. Dutch Reference Grammar. Leiden: Martinus Nijhoff. 1993. A Grammar of Afrikaans. Berlin, New York: Mouton de Gruyter. du Toit, P. J. 1905. Afrikaansche studies. (dissertation, Ghent). Ghent: A. Siffer. Elphick, R., and V. C. Malherbe. 1989. The Khoisan to 1828. In R. Elphick and H. Giliomee, eds., The Shaping of South African Society, 1652–1840. Middletown, Conn.: Wesleyan University Press, pp. 3–65. Famighetti, Robert, ed., 1997. The World Almanac and Book of Facts 1998. Mahwah, NJ: World Almanac Books. Fasold, Ralph W., and Walt Wolfram. 1970. Some linguistic features of Negro dialect. In R. Fasold and R. Shuy, eds., pp. 87–104. Fasold, Ralph W., and Roger W. Shuy, eds., 1970. Teaching Standard English in the Inner City. Washington, D.C.: Center for Applied Linguistics. Fasold Ralph, W., William Labov, Fay Bayel Vaughn-Cooke, Guy Bailey, Walt Wolfram, Arthur K Spears, and John Rickford. 1987. Are black and white vernaculars diverging? American Speech 62:1, pp. 3–80. Feagin, C. 1979. Variation and Change in Alabama English: a sociolinguistic study of the White community. Washington: Georgetown University Press. Ferguson, C. A., and S. B. Heath, eds., 1981. Language in the USA. Cambridge: Cambridge University Press. Ferraz, Luiz. 1976. The origin and development of four creoles in the Gulf of Guinea. African Studies 35:1, pp. 33–38. Figueroa A., V. Jesus. ´ 1998. Rasgos semi-criollos en el espanol ˜ no est´andar de la region ´ suroriental cubana. In K. Zimmermann, ed., pp. 411–440. Folb, Edith A. 1980. Running Down Some Lines: the language and culture of Black teenagers. Cambridge, Mass.: Harvard University Press (see glossary, pp. 227–260). Franken, J. L. M. 1952. Taalhistoriese bydraes. Amsterdam, Capetown: Balkema. Freitas, D´ecio. 1978. Palmares: A Guerra dos Escravos. Rio de Janeiro: Graal. Fry, P., C. Vogt, and M. Gnerre. 1981. Mafamba e caxapura: na encruzilhada da identidade. Revista de Ciˆencias Sociais 24:3, pp. 373–389. Gilbert, Glenn G. 1985. Hugo Schuchardt and the Atlantic creoles: a newly discovered manuscript “On the Negro English of West Africa.” American Speech 60:1, pp. 31–63. Gilbert, Glenn. ed. 1987. Pidgin and Creole Languages: essays in memory of John E. Reinecke. Honolulu: University Press of Hawaii. ed. 2002. Pidgin and Creole Linguistics in the Twenty-first Century. New York: Peter Lang. Goilo, E. R. 1972. Papiamentu Textbook. Aruba: de Wit. Goodman, Morris. 1987. The Portuguese element in the American creoles. In G. Gilbert, ed., pp. 361–405.

References

153

Grondin, R.-C. 1998. La construction historique des cat´egories sociales. L’exemple des Petits Blancs des Hauts. M´emoire de DEA. Universit´e de la R´eunion. Green, Katherine. 1997. Non-standard Dominican Spanish: evidence of partial restructuring. Ph.D. dissertation, City University of New York. Ann Arbor: University Microfilms International. Guy, Gregory R. 1981. Linguistic Variation in Brazilian Portuguese. Ph.D. dissertation, University of Pennsylvania. 1989. On the nature and origins of Popular Brazilian Portuguese. In Estudios sobre espa˜nol de Am´erica y ling¨u´ıstica afroamericana. Bogot´a: Instituto Caro y Cuervo, publication no. 83, pp. 227–245. Hackert, Stephanie, and John Holm. Forthcoming. Southern Bahamian: transported AAVE or transported Gullah? In John Lipski ed., African-American English and its Congeners. Amsterdam Philadelphia: John Benjamins. Hahn, Theophilus. 1882. On the Science of Language and its Study, with Special Reference to South Africa. Cape Town: Herrmann Michaelis. Hancock, Ian F. 1969. The Malacca Creoles and their language. Afrasia 3, pp. 38–45. 1971. A Study of the Sources and Development of the Lexicon of Sierra Leone Creole. Ph.D. thesis, School of Oriental and African Studies, University of London. 1975. Malacca Creole Portuguese: Asian, African, or European? Anthropological Linguistics 17(5): 211–236. 1979. The relationship of Black Vernacular English to the Atlantic creoles. Working paper of the African and Afro-American Studies and Research Center, University of Texas at Austin. 1987. A preliminary classification of the Anglophone Atlantic creoles, with syntactic data from 33 representative dialects. In G. Gilbert, ed., pp. 264– 334. Hartt, Charles F. 1872. Notes on the Lingoa Geral or modern Tupi of the Amazonas. Transactions of the American Philological Association 3, pp. 58–76. Healy, Maureen. 1993. The parallel continuum model for Suriname: a preliminary study. In F. Byrne and J. Holm, eds., pp. 279–289. Henriquez Urena, ˜ Pedro. 1987 [1940]. El espa˜nol en Santo Domingo. Santo Domingo: Editora Taller. Hesseling, D. C. 1897. Het Hollandsch in Zuid Afrika. De Gids 61, pp. 138–62; reprinted as Dutch in South Africa: In T. L. Markey and Paul T. Roberge, eds. and trans., On the Origin and Formation of Creoles: a miscellany of articles, Ann Arbor: Karoma, 1979. Hickey, Raymond. 1997. Arguments for creolisation in Irish English. In R. Hickey and S. Puppel, eds., Language History and Linguistic Modeling: a festschrift for Jacek Fisiak on his 60th Birthday. Berlin: Mouton de Gruyter, pp. 969–1038. Hickey, Raymond, ed. forthcoming. Legacies of Colonial English. Cambridge: Cambridge University Press. Highfield, Arnold R. 1979. The French Dialect of St. Thomas, U.S. Virgin Islands. Ann Arbor: Karoma. Holm, John. l978. The Creole English of Nicaragua’s Miskito Coast: its sociolinguistic history and a comparative study of its lexicon and syntax. Ph.D. thesis, University College, London. Ann Arbor: University Microfilms International.

154

References

1980. African features in White Bahamian English. English World-Wide 1:1, pp. 45–65. ed. 1983. Central American English. Heidelberg: Groos/Amsterdam: Benjamins. 1983. On the relationship of Gullah and Bahamian. American Speech 58:4, pp. 303–318. 1984. Variability of the copula in Black English and its creole kin. American Speech 59:4, pp. 291–309. 1987. Creole influence on Popular Brazilian Portuguese. In G. Gilbert, ed. pp. 406–430. 1988–89. Pidgins and Creoles. 2 vols. Cambridge: Cambridge University Press. 1989. Do creoles have inflectional morphology? Paper presented at the Workshop on Creole Morphology, University of Amsterdam. 1991. The Atlantic creoles and the language of the ex-slave recordings. In G. Bailey et al., eds., pp. 231–247. 1992a. A theoretical model for semi-creolization. Paper presented to the Society for Caribbean Linguistics, Barbados. 1992b. Popular Brazilian Portuguese: a semi-creole. In E. d’Andrade and A. Kihm, eds., pp. 37–66. 1994. The history of the English Language in the West Indies. In The Cambridge History of the English Language, vol. V, R. Burchfield, ed., English in Britain and Overseas: Origins and Development, pp. 328–381. 2000a. An Introduction to Pidgins and Creoles. Cambridge: Cambridge University Press. 2000b. Semi-creolization: problems in the development of theory. In I. Neumann-Holzschuh and E. W. Schneider, eds., pp. 19–40. 2001. The semi-creole identity of Afrikaans lects: parallel cases of partial restructuring. Journal of Germanic Linguistics 13:4, pp. 353–379. 2002. The study of semi-creoles in the 21st century. In G. Gilbert, ed. Pidgin and Creole Linguistics in the Twenty-first Century. Peter Lang: New York, pp. 173–197. Holm, J., G. A. Lorenzino, and H. R. de Mello. 1999. Diferentes grados de reestructuracion ´ en dos lenguas vern´aculas: el espanol ˜ caribeno ˜ y el portugu´es brasileno. ˜ In Lu´ıs Ortiz, ed., El Caribe hisp´anico: perspectivas ling¨u´ısticas actuales. Frankfurt: Vervuert, pp. 43–60. Holm, John, and A. Goke Oyedeji. 1984. The Yoruba Language in the New World. Oso 3:1, pp. 83–90. Holm, John, and Peter Patrick, eds. forthcoming. Comparative Creole Syntax. London: Battlebridge Press. Holm, John, with Alison Shilling. 1982. Dictionary of Bahamian English. Cold Spring, NY: Lexik House. Honor´e, Daniel. 2002. Dictionnaire d’Expressions Cr´eoles–Semi-lo-mo. R´eunion: Editions Udir. Howe, Darin M., and James A. Walker. 2000. Negation and the creole-origins hypothesis: evidence from Early African American English. In S. Poplack, ed., pp. 109–140. Huber, Magnus. 2001. The Nova Scotia–Sierra Leone connection: new evidence on an early variety of African American Vernacular English in the diaspora.

References

155

Paper presented at the meeting of the Society for Pidgin and Creole Linguistics, Coimbra. Huber, Magnus, and Mikael Parkvall, eds. 1999. Spreading the Word: the issue of diffusion among the Atlantic Creoles. London: University of Westminster Press. Hull, Alexander. 1979. On the origin and chronology of the French-based creoles. In I. Hancock et al., eds., Readings in Creole Studies. Ghent: Story-Scientia, pp. 201–216. 1983. Cr´eole louisianais et cr´eole ha¨ıtien: ressemblances et diff´erences. Paper presented at the 4e Colloque International du Comit´e International des Etudes Cr´eoles, Lafayette. Hymes, Dell, ed. 1971. Pidginization and Creolization of Languages. Cambridge: Cambridge University Press. Inverno, Liliana and Dominika Swolkien. 2001. O artigo definido zero em dois contextos espec´ıficos no Portuguˆes do Brasil e no Crioulo de Cabo Verde. Paper presented at the meeting of Associa¸ca˜ o: Crioulos de Base Lexical Portuguesa e Espanhola, Coimbra. Ivens Ferraz, L. 1979. The Creole of S˜ao Tom´e. Johannesburg: Witwatersrand University Press. Jeroslow, Helen [McKinney]. 1974. Rural cearense Portuguese: a study of one variety of non-standard Brazilian Portuguese. Dissertation: Cornell University, Ithaca, NY. Jeuda, David M. 1991. The Portuguese element in the Papiamentu lexicon. Paper presented to the Society for Pidgin and Creole Linguistics, Chicago. Joyner, C. 1984. Down by the Riverside: a South Carolina slave community. Urbana, Chicago: University of Illinois Press. Kautzsch, Alexander. 2002. The Historical Evolution of Earlier African American English: an empirical comparison of early sources. Berlin, New York: Mouton de Gruyter. Kautzsch, Alexander, and Edgar W. Schneider. 2000. Differential creolization: some evidence from Earlier African American Vernacular English in South Carolina. In I. Neumann-Holzschuh and E. W. Schneider, eds., pp. 247– 274. Kaye, Alan S. 1990. Observations on pidginistics and creolistics. Semiotica 78:3/4, pp. 285–348. Keenan, E. L. 1975. Remarkable subjects. In C. N. Li, ed., Subject and Topic. New York: Academic Press, pp. 249–301. 1978. The syntax of subject-final languages. In W. P. Lehman, ed., Syntactic Typology: studies in the phenomenology of language. Austin: University of Texas Press, pp. 266–327. Kent, R. K. 1979. Palmares: an African state in Brazil. In R. Price, ed., Maroon Societies: rebel slave communities in the Americas. Baltimore: Johns Hopkins University Press, pp. 170–190. Kloeke, G. G. 1950. Herkomst en groei van het Afrikaans. Leiden: Universitaire Pers. Kotz´e, E. F. 1984. Afrikaans in the Malay Quarter: a diachronic perspective. Tydskrif vir Geesteswetenskappe. 24:1, pp. 41–73.

156

References

1989. How creoloid can you be? Aspects of Malay Afrikaans. In M. Putz ¨ and R. Dirven, eds., pp. 251–264. Kruisinga, E. 1906. De oorsprong van het Afrikaans. Taal en Letteren 16, pp. 417–439. Labov, William. 1969. Contraction, deletion and inherent variability of the English copula. Language 45:4, pp. 715–751. 1972a. Negative attraction and negative concord. In Language in the Inner City: Studies in the Black English Vernacular. Philadelphia: University of Pennsylvania Press, pp. 130–196. 1972b. Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press. 1982. Objectivity and commitment in linguistic science: the case of the Black English trial in Ann Arbor. Language in Society 11, pp. 165–201. Labov, William, Paul Cohen, and John Lewis. 1968. A Study of the Non-standard English of Negro and Puerto Rican Speakers in New York City. Report on Cooperative Research Project 3288. New York: Columbia University. Labov, William, and Wendell A. Harris. 1986. De facto segregation of Black and White vernaculars. In David Sankoff, ed., Diversity and Diachrony. Amsterdam, Philadelphia: John Benjamins. Ladhams, John. 1999. The Pernambuco connection? An examination of the nature and origin of the Portuguese elements in the Surinam Creoles. In M. Huber and M. Parkvall, eds., pp. 209–240. Lanehart, Sonja L. 2001. Sociocultural and Historical Contexts of African American English. Amsterdam, Philadelphia: John Benjamins. Laurence, Kemlin M. 1974. Is Caribbean Spanish a case of decreolization? Orbis 23:2, pp. 484–499. Lass, R. 1990. How to do things with junk: exaptation in language evolution. Journal of Linguistics 26, pp. 79–102. Lemle, Miriam. 1976. Heterogenidade dialetal: um apelo a pesquisas. In E. M. P. Lobato, ed., Ling¨u´ıstica e ensino do vern´aculo. Rio de Janeiro: Tempo Brasileiro. Lenz, Rodolofo. 1928. El papiamento, la lengua criolla de Curazao (la gram´atica m´as sencilla). Santiago de Chile: Balcells. Le Page, R. B. and A. Tabouret-Keller. 1985. Acts of Identity: creole-based approaches to language and ethnicity. Cambridge: Cambridge University Press. Lipski, John M. 1993. On the non-creole basis for Afro-Caribbean Spanish. Albuquerque: Research Paper Series no. 24. Albuquerque: The University of New Mexico. 1994. A new perspective on Afro-Dominican Spanish: the Haitian contribution. Albuquerque: Latin American Institute of the University of New Mexico. 1998. Perspectivas sobre el espanol ˜ bozal. In M. Perl and A. Schwegler, eds., pp. 293–328. 2000. Bozal Spanish: Restructuring or creolization? In I. Neumann-Holzschuh and E. W. Schneider, eds., pp. 437–468. Lockwood, W. B. 1965. An Informal History of the German Language, with chapters on Dutch and Afrikaans, Frisian and Yiddish. Cambridge: Heffer.

References

157

Lorenzino, Gerardo. 1991. On the semi-creole status of Dominican Spanish. City University of New York, Dept. of Linguistics, 41-p. MS. 1993. Algunos rasgos semicriollos en el espanol ˜ popular dominicano. Anuario de Ling¨u´ıstica Hisp´anica 9, pp. 111–126. 1998. El espanol ˜ caribeno: ˜ antecedentes sociohistoricos ´ y lingu´ ¨ ısticos. In M. Perl and A. Schwegler, eds., pp. 26–39. Forthcoming. Angolar. In J. Holm and P. Patrick, eds. Lucchesi, Dante. 2000. A varia¸ca˜ o na concordˆancia de gˆenero em uma comunidade de fala afro-brasileira: Novos elementos sobre a forma¸ca˜ o do portuguˆes popular do Brasil. Ph.D. diss., Universidade Federal do Rio de Janeiro. Luijks, Carla. 2000. The Realisation of Syntactic Principles in Non-standard Afrikaans: the Correspondence of Jan Jonker Afrikaner (1820–1889). Ph.D. thesis, University of Cape Town. Makhudu, Dennis. 1984. Is Afrikaans a creole language? M.A. thesis, Department of Linguistics, Southern Illinois University at Carbondale. Markey, Thomas. 1982. Afrikaans: creole or non-creole? Zeitschrift f¨ur Dialektologie und Linguistik 49:2, pp. 169–207. Marques, A. H. de Oliveira. 1976. History of Portugal. New York: Columbia University Press. Marroquim, M´ario. 1934. A l´ıngua do Nordeste (Alagˆoas e Pernambuco). S˜ao Paulo: Companhia Editora Nacional. Maurer, Philippe. 1988. Les modifications temporelles et modales du verbe dans le papiamento de Cura¸cao (Antilles N´eerlandaises). Hamburg: Helmut Buske Verlag. 1995. L’angolar: un cr´eole afro-portugais parl´e a` S˜ao Tom´e. Hamburg: Helmut Buske Verlag. McKay, J. R. 1969. A Partial Analysis of a Variety of Nonstandard Negro English. Ph.D. dissertation, University of California, Berkeley. McKinney, Helen [Jeroslow]. 1975. Creole characteristics in rural Brazilian Portuguese. Paper presented at the International Congress on Pidgins and Creoles, University of Hawaii. 1982. Serial verbs in a non-standard Brazilian dialect: possible creolization. Paper presented at the Conference of the American Association of Teachers of Spanish and Portuguese, New Orleans. McWhorter, John. 1998. Identifying the creole prototype: vindicating a typological class. Language 74:4, pp. 788–818. 2000. Defining “creole” as a synchronic term. In I. Neumann-Holzschuh and E. W. Schneider, eds., pp. 85–123. ed., 2000. Language Change and Language Contact in Pidgins and Creoles. Amsterdam, Philadelphia: John Benjamins. 2002. The rest of the story: restoring pidginization to creole genesis theory. Journal of Pidgin and Creole Languages 17:1, pp. 1–48. Megenney, William W. 1976. El elemento sub-sah´arico en el l´exico costeno ˜ de Colombia. Revista Espa˜nola de Ling¨u´ıstica 6, pp. 405–451. 1985. Africa en Venezuela: su herencia lingu´ ¨ ıstica y su cultura literaria. Revista Montalb´an 15, pp. 1–54. 1990. Africa en Santo Domingo: su herencia ling¨u´ıstica. Santo Domingo: Editorial Tiempo.

158

References

1999. Cuba y Brasil: etnohist´oria del empleo religioso del lenguaje afroamericano. Miami FL: Ediciones Universal. Mello, Heliana Ribeiro de. 1997. Brazilian Vernacular Portuguese: evidence of partial restructuring. Ph.D. dissertation, City University of New York. Ann Arbor: University Microfilms International. 2001. Situando o Portuguˆes Brasileiro atrav´es da perspectiva da an´alise cr´ıtica do discurso. Paper presented at the meeting of Associa¸ca˜ o: Crioulos de Base Lexical Portuguesa e Espanhola, Coimbra. Melo, Gladstone Chaves de. 1946. A L´ıngua do Brasil. Rio de Janeiro: Livraria AGIR Editora. Repr. 1975. Rio de Janeiro: Editora da Funda¸ca˜ o Getulio ´ Vargas. Mendon¸ca, Renato. 1933. A influˆencia africana no portuguˆes do Brasil. Repr. 1973. Rio de Janeiro: Civiliza¸ca˜ o Brasileira/MEC. Mesthrie, Rajend, and Paul T. Roberge, eds., 2001–02. Focus on Afrikaans Sociohistorical Linguistics. Journal of Germanic Linguistics 13:4–14:1. Michel, Abigail. Forthcoming. Papiamentu. In J. Holm and P. Patrick, eds. Mintz, S. W. 1971. The socio-historical background to pidginization and creolization. In D. Hymes, ed., pp. 481–498. Monino, ˜ Yves and Armin Schwegler, eds., 2002. Palenque, Cartagena y AfroCaribe: hist´orica y langua. Tubingen: ¨ Max Niemeyer. Montgomery, M., and G. Bailey, eds. 1986. Language Variety in the South: perspectives in black and white. University, AL: University of Alabama Press. Mufwene, Salikoko. 1983. Some observations on the verb in Black English Vernacular. African and Afro-American Studies and Research Center Papers, series 2. University of Texas at Austin. 1987. Review of Montgomery and Bailey, eds., 1986. Journal of Pidgin and Creole Languages 2:1, pp. 93–110. 1991. Review of Holm 1988–89. Language 67, pp. 380–387. 1994. On decreolization: the case of Gullah. In M. Morgan, ed., Language and the Social Construction of Identity in Creole Situations. Center for African Studies, UCLA, Los Angeles. 1996. The founder principle in creole genesis. Diachronica 13, pp. 83–134. 1997. Jargons, pidgins, creoles, and koines: what are they? In A. Spears and D. Winford, eds., pp. 35–70. 1998. The structure of the noun phrase in AAVE. In S. Mufwene et al. eds., pp. 69–81. 2000a. Population contacts and the evolution of English. The European English Messenger 9, pp. 9–15. 2000b. Some sociohistorical inferences about the development of African American English. In S. Poplack, ed., pp. 233–263. 2000c. Creolization is a social, not a structural, process. In I. NeumannHolzschuh and E. W. Schneider, eds., pp. 65–84. 2001. African-American English. In John Algeo, ed., Cambridge History of the English Language. Vol. 6: English in North America. Mufwene, Salikoko S., John R. Rickford, Guy Bailey, and John Baugh, eds., 1998. African-American English: structure, history and use. London, New York: Routledge.

References

159

Muhlh¨ ¨ ausler, Peter. 1982. Tok Pisin in Papua New Guinea. In R. Bailey and M. Gorlach, ¨ eds., English as a World Language. Ann Arbor: University of Michigan Press, pp. 439–466. Mussa, A. B. N. 1991. O papel das l´ınguas africanas na historia ´ do portuguˆes do Brasil. Unpublished master’s thesis, Universidade Federal do Rio de Janeiro. Muysken, Pieter, ed. 1981. Generative Studies on Creole Languages. Dordrecht: Foris. Naro, Anthony J., and Miriam Lemle. 1976. Syntactic diffusion. In S. B. Steever, C. A. Walker, and S. S. Mufwene, eds., Papers from the Parasession on Diachronic Syntax, Chicago: Chicago Linguistic Society, pp. 221–240. Naro, Anthony J., and M. Marta P. Scherre. 2000. Variable Concord in Portuguese: the situation in Brazil and Portugal. In J. McWhorter, ed., pp. 235–255. Nash, Roy. 1926. The Conquest of Brazil. New York: Harcourt, Brace. Neeser, Hermann. 1951. A Colˆonia Leopoldina. Bahia: Centro de Estudos Baianos, Publica¸ca˜ o no. 5. Neumann-Holzschuh, Ingrid, and Edgar W. Schneider, eds. 2000. Degrees of Restructuring in Creole Languages. Amsterdam, Philadelphia: John Benjamins. Nienaber, P. J. 1965. Iets naders oor die ontkenning in Afrikaans. In P. J. Nienaber, ed., Taalkundige opstelle. Cape Town: Balkema, pp. 22–38. Nunez ˜ Cedeno, ˜ Rafael. 1980. La fonolog´ıa moderna y el espa˜nol de Santo Domingo. Santo Domingo: Taller. Oldendorp, C. G. A. 1777. Geschichte der Mission der evangelischen Br¨uder auf den caraibischen Inseln S. Thomas, S. Croix und S. Jan. Barby: C. F. Laur. Ortiz, Fernando. 1924. Glosario de afronegrismos. Havana: El Siglo XX. Ortiz Lopez, ´ Luis A. 1996. El contacto ling¨u´ıstico afro-hisp´anico en la g´enesis y evoluci´on del espa˜nol de Cuba. Ph.D. dissertation, University of New Mexico, Albuquerque. 1998. Huellas etno-socioling¨u´ısticas bozales y afrocubanas. Frankfurt-am-Main: Vervuert/Madrid: Iberoamericana. ed. 1999. El Caribe hisp´anico: perspectivas ling¨u´ısticas actuales. Homenaje a ´ Manuel Alvarez Nazario. Frankfurt-am-Main: Vervuert/Madrid: Iberoamericana. Orton, H., S. Sanderson, and J. Widdowson. 1978. The Linguistic Atlas of England. London: Croom Helm. Otheguy, Ricardo. 1973. Spanish Caribbean: a creole perspective. In C. Bailey and R. Shuy, eds., New Ways of Analyzing Variation in English. Washington: Georgetown University Press, pp. 323–339. Papen, Robert A. 1978. The French-based Creoles of the Indian Ocean: an analysis and comparison. Ph.D. dissertation, University of California at San Diego. Parkvall, Mikael. 2000. Reassessing the role of demographics in language restructuring. In I. Neumann-Holzschuh and E. W. Schneider, eds., pp. 185– 213. 2002. Cutting off the branch. In G. Gilbert, ed., pp. 355–367. Peixoto, Antonio ˆ da Costa. 1945. Obra nova de L´ıngua Geral de Mina. Lisbon: Agˆencia Geral das Colonias. ˆ

160

References

Perl, Matthias. 1985. El fenomeno ´ de descriollizacion ´ del “habla bozal” y el lenguaje coloquial de la variante cubana del espanol. ˜ Anuario de Ling¨u´ıstica Hisp´anica 1, pp. 191–202. 1988. Rasgos poscriollos l´exicos en el lenguaje coloquial cubano. Thesaurus 43, 47–64. ˜ Anuario de 1989. El “habla bozal”: ¿una lengua criolla de base espanola? Ling¨u´ıstica Hisp´anica 5, pp. 205–220. Perl, Matthias, and Armin Schwegler, eds. 1998. Am´erica negra: panor´amica actual de los estudios ling¨u´ısticos sobre variedades h´ıspanas, portuguesas y criollas. Frankfurt: Vervuert/ Madrid: Iberoamericana. Pichardo, Esteban. 1862. Diccionario provincial casi razonado de voces cubanas. Havana: Imprenta la Antilla. Pimentel Pinto, Edith, ed. 1981. O portuguˆes do Brasil: textos cr´ıticos e te´oricos. Rio de Janeiro: Livros T´ecnicos e Cient´ıficos. Ponelis, F. A. 1993. The Development of Afrikaans. (Duisburger Arbeiten zur Sprach- und Kulturwissenschaft, 18). Frankfurt-am-Main: Peter Lang. Poplack, Shana. 1978. On Deletion and Disambiguation in Puerto Rican Spanish: a study of verbal /n#/. New York: Centro de Estudios Puertorriquenos. ˜ 2000. Introduction. In S. Poplack, ed., pp. 1–32. Poplack, Shana, ed. 2000. The English History of African American English. Oxford: Blackwell. Poplack, Shana, and David Sankoff. 1987. The Philadelphia story in the Spanish Caribbean. American Speech 62:4, pp. 291–314. Poplack, Shana, Sali Tagliamonte, and Ejike Eze. 2000. Reconstructing the source of early African American English plural marking: a comparative study of English and Creole. In S. Poplack, ed., pp. 73–105. Putz, ¨ Martin, and Ren´e Dirven, eds. 1989. Wheels within Wheels: Papers of the Duisburg Symposium on Pidgin and Creole Languages. Frankfurt-am-Main: Peter Lang. Rademeyer, J. H. 1938. Kleurling-Afrikaans: Die taal van die Griekwas en RehobothBasters. Amsterdam: Swets & Zeitlinger. Raidt, Edith. 1974. Nederlandse en Kaapse spreektaal in die 17de en 18de eeu. In F. Odendal, ed., Taalkunde–’n lewe. Studies opgedra aan W. Kempen. Cape Town: Tafelberg, pp. 90–104. 1983. Einf¨uhrung in Geschichte und Struktur des Afrikaans. Darmstadt: Wissenschaftliche Buchgesellschaft. 1991. Afrikaans en sy europese verlede. Cape Town: Nasou. Raimundo, Jacques. 1933. O elemento afro-negro na l´ıngua portuguesa. Rio de Janeiro: Renascen¸ca Editora. Ramassamy, Ginette. 1985. Syntaxe du cr´eole r´eunionnais: analyse de corpus d’unilingues cr´eolophones. Th´ese pour le Doctorat de 3`eme cycle, Universit´e Ren´e Descartes, Paris V. Rawley, J. A. 1981. The Transatlantic Slave Trade: a history. New York, London: Norton. Reinecke, John. 1937. Marginal Languages: a sociological survey of the creole languages and trade jargons. Ph.D. dissertation, Yale University. Ann Arbor: University Microfilms International.

References

161

Reinecke, J. E., S. M. Tsuzaki, D. DeCamp, I. F. Hancock, and R. E. Wood, eds., 1975. A Bibliography of Pidgin and Creole Languages. Honolulu: University Press of Hawaii. Rens, L. L. E. 1953. The Historical and Social Background of Surinam’s Negro English. Academisch proefschrift. Amsterdam: North Holland. R´evah, Israel S. 1963. La question des substrats et superstrats dans le domaine linguistique br´esilien: les parler populaires br´esiliens doivent-ils eˆ tre consider´es comme des parlers “cr´eoles” ou “semi-cr´eoles”? Romania 84, pp. 433–450. Rickford, John R. 1977. The question of prior creolization in Black English. In A. Valdman, ed., Pidgin and Creole Linguistics. Bloomington: Indiana University Press, pp. 190–221. 1986. Social contact and linguistic diffusion: Hiberno English and New World Black English. Language 62:2, pp. 245–289. 1987. Dimensions of a Creole Continuum. Stanford: Stanford University Press. 1997. Prior creolization of African-American Vernacular English? Sociohistorical and textual evidence from the 17th and 18th centuries. Journal of Sociolinguistics 1:3, pp. 315–336. 1999. African American Vernacular English: Features, Evolution, Educational Implications. Oxford: Blackwell. Roberge, Paul T. 1994. The formation of Afrikaans. Stellenbosch Papers in Linguistic PLUS, 23. Department of General Linguistics, University of Stellenbosch. Roberts, Ian. 1999. Verb movement and markedness. In M. DeGraff, ed., pp. 287–327. Rodrigues, Ada Natal. 1974. O Dialeto Caipira na Regi˜ao de Piracicaba. S˜ao Paulo: ´ Editora Atica. Rowlands, E. C. 1969. Yoruba. London: The English Universities Press. Ru´ız Garc´ıa, Martha E. 2001. El espa˜nol popular de Choc´o, Colombia: evidencia de una reestructuraci´on parcial. Ph.D. dissertation. Ann Arbor: University Microfilm International. Sampaio, Theodoro. 1928. O Tup´ı na geografia nacional. Ba´ıa: Sec¸ca˜ o da Escola de Aprendizes. Sandoval, Padre A. de. 1627. De instauranda Aethiopum salute. Seville: F. de Lira. Repr. 1956, Bogot´a: Empresa Nacional de Publicaciones. Schneider, Edgar W. l981. Morphologische und syntaktische Variablen im amerikanischen Early Black English. Frankfurt am Main: Peter Lang. 1989. American Earlier Black English: morphological and syntactic variables. Tuscaloosa: University of Alabama Press. 1990. The cline of creoleness in English-oriented creoles and semi-creoles in the Caribbean. English World-Wide 11:1, pp. 79–113. Scholtz, J. du Plessis. 1963. Taalhistoriese opstelle. Pretoria: J. L. van Schaik. 1970. Internal History of Afrikaans. Standard Encyclopedia of Southern Africa. Cape Town: Nasou, pp. 80–113. 1980. Wording en ontwikkeling van Afrikaans. Cape Town: Tafelberg. Schuchardt, Hugo. 1882. Kreolische Studien I: Ueber das Negerportugiesische von S. Thom´e (Westafrika). Sitzungsberichte der kaiserlichen Akademie der Wissenschaften zu Wien 101:2, pp. 889–917.

162

References

1885. Review of Focard’s Du patois cr´eole de l’ˆıle de R´eunion. Literaturblatt f¨ur germanische und romanische Philologie 6, pp. 513–515. Translated into English as Further remarks on R´eunionnais 1979. In T. L. Markey, ed. and trans., The Ethnography of Variation: selected writings on pidgins and creoles. Ann Arbor: Karoma, pp. 15–17. 1889. Beitr¨age zur Kenntnis des kreolischen Romanisch V. Allgemeineres uber ¨ das Indoportugiesische (Asioportugiesische). Zeitschrift f¨ur romanische Philologie 13, pp. 476–516. Schwegler, Armin. 1991. Negation in Palenquero: synchrony. Journal of Pidgin and Creole Languages 6, pp. 165–214. 1993. Subject pronouns and person/number in Palenquero. In F. Byrne and J. Holm, eds., pp. 145–161. 1996a. Chi ma nkongo: lengua y rito ancestrales en al Palenque de San Basilio (Colombia). Frankfurt: Vervuert. 1996b. La doble negacion ´ dominicana y la g´enesis del espanol ˜ caribeno. ˜ Hispanic Linguistics 8:2, pp. 247–315. Schwegler, Armin, and Thomas Morton. 2003. Vernacular Spanish in a microcosm: Kateyano in El Palenque de San Basilio (Colombia). Revista Internacional de Ling¨u´ıstica Iberoamericana 1:1, pp. 97–159. Sells, Peter, John R. Rickford, and Thomas A. Wasow. 1996. Negative inversion in African American Vernacular English. Natural Language and Linguistic Theory 14, pp. 591–627. Siegel, Jeff. 1997. Mixing, leveling, and pidgin/creole development. In A. Spears and D. Winford, eds., pp. 111–149. Silva Neto, Serafim da. 1950a. Falares Crioulos. Bras´ılia 5, pp. 1–28. 1950b. [2nd edn. 1963]. Introdu¸ca˜ o ao estudo da l´ıngua portuguesa no Brasil. Rio de Janeiro: Departamento de Imprensa Nacional. Silveira Bueno, F. 1963. Les langues indig`enes du Br´esil et leur influence sur le portugais. Orbis 12, pp. 226–240. Silveira Ferreira, Carlota da. 1985. Remanescentes de um falar crioulo brasiliero. Revista Lusitana 5 (Lisbon), pp. 21–34. Singler, John V. 1986. Remarks in response to Derek Bickerton’s “Creoles and universal grammar: the unmarked case?” Journal of Pidgin and Creole Languages 1:1, pp. 141–145. 1989. Plural marking in Liberian Settler English. American Speech 64:1, pp. 40–64. 1990. On the use of sociohistorical criteria in the comparison of creoles. Linguistics 28:4, pp. 645–660. 1995. The demographics of creole genesis in the Caribbean: a comparison of Martinique and Haiti. In J. Arends, ed., pp. 203–232. Slomanson, Peter. 1993. On a problem in the grammar of Orange River Afrikaans. Paper presented to the Society for Pidgin and Creole Linguistics, Amsterdam. Smith, Johannes J. 1927. The evolution and recognition of the Afrikaans language. Official Year Book of the Union 8 (1925), pp. 14–23. 1952. Theories about the origin of Afrikaans. Johannesburg: Witwatersrand University Press.

References

163

Smith, Norval. 1999. Pernambuco to Surinam 1654–1655? The Jewish slave controversy. In M. Huber and M. Parkvall, eds., pp. 251–298. 2002. The history of the Surinamese creoles II. In E. B. Carlin and J. Arends, eds., pp. 131–151. Smith, Norval, and Tonjes Veenstra, eds., 2001. Creolization and Contact. Amsterdam/ Philadelphia: John Benjamins. Smith, Riley B. 1973. Interrelatedness of certain deviant grammatical structures in Negro Nonstandard Dialects. In R. H. Bentley and S. D. Crawford, eds., Black Language Reader. Glenview, Ill. and Brighton, England: Scott, Foresman and Company, pp. 90–96. Spears, Arthur K., and Donald Winford, eds. 1997. Pidgins and Creoles: structure and status. Amsterdam Philadelphia: John Benjamins. Stein, P. 1984. Kreolisch und Franz¨osisch. Tubingen: ¨ Niemeyer. Stewart, William A. 1962. Creole languages in the Caribbean. In F. A. Rice, ed., Study of the Role of Second Languages in Asia, Africa and Latin America. Washington, D.C.: Center for Applied Linguistics, pp. 34–53. 1965. Urban Negro speech: sociolinguistic factors affecting English teaching. In R.W. Shuy, ed., Social Dialects and Language Learning. Champaign, Ill.: National Council of Teachers of English, pp. 10–18. 1967. Sociolinguistic factors in the history of American Negro dialects. Florida FL Reporter 5. Repr. in Wolfram and Clarke, eds., 1971, pp. 74–89. 1968. Continuity and change in American Negro dialects. Florida FL Reporter 6. Repr. in Wolfram and Clarke, eds., 1971, pp. 51–73. Tagliavini, Carlo. 1931. Creole, Lingue. Enciclopedia Italiana di scienze, lettere ed arti. Milan: Istituto Giovanni Treccani. Vol. 11, pp. 833–835. Tarallo, Fernando. 1986. On the alleged creole origin of Brazilian Portuguese: untargeted syntactic change. Paper presented at the Creole Workshop, LSA Institute, City University of New York. In Ian Roberts and Mary A. Kato, eds., 1993. Portuguˆes brasileiro: uma viagem diacrˆonica. Campinas: Editora da UNICAMP, pp. 35–68. Terrel, T. 1979. Final /s/ in Cuban Spanish. Hispania 62, pp. 599–612. Thomason, Sarah G. 1997. A typology of contact languages. In A. Spears and D. Winford, eds., pp. 71–88. 2001. Contact-induced language change and Pidgin/Creole genesis. In N. Smith and T. Veenstra, eds., pp. 249–262. 2002. Column: Creoles and genetic relationship. Journal of Pidgin and Creole Languages 17:1, pp. 101–109. Thomason, S. G., and T. Kaufman. 1988. Language Contact, Creolization and Genetic Linguistics. Berkeley: University of California Press. Tolsner, ¨ Carl A. 1858. Die Colonia Leopoldina in Brasilien. Gottingen: ¨ Dieterichsche Universit¨ats-Buchdruckerei. Tottie, Gunnel, and Dawn Harvie. 2000. It’s all relative: relativization strategies in Early African American English. In S. Poplack, ed., pp. 198– 230. Turner, L. D. 1949. Africanisms in the Gullah dialect. Repr. 1974. Ann Arbor: University of Michigan Press. Valdman, Albert. 1978. Le cr´eole: structure, statut et origine. Paris: Klincksieck.

164

References

Valkhoff, Marius. 1964. Notes socio-linguistiques sur le parler cr´eole de la R´eunion. Revue Philologique 17, pp. 723–735. 1966. Studies in Portuguese and Creole with special reference to South Africa. Johannesburg: Witwatersrand University Press. 1972. New light on Afrikaans and Malayo-Portuguese. Louvain: Peeters. van Dantzig, A. 1968. Het Nederlandse aandeel in de slavenhandel. Bussem: Van Dishoek. van der Merwe, H. J. J. M. 1963. Die ontstaan van Afrikaans. In H. J. J. M. van der Merwe, ed., Inleiding tot die Taalkunde. Pretoria: J. L. van Schaik, pp. 67–88. 1968. Die ontstaan van Afrikaans. In H. J. J. M. van der Merwe, ed., Afrikaans– sy Aard en Ontwikkeling. Pretoria: J. L. van Schaik, pp. 15–66. van der Merwe, Elizabeth G. 1989. Die ontwikkeling, stand en toekoms van Afrikaans in S.W.A./Namibi¨e. Ph.D. diss., Potchefstroomse Universiteit. 1990. Oorlamafrikaans. Paper presented to the Linguistic Society of Southern Africa, University of Stellenbosch. Van Name, Addison. 1869–70. Contributions to creole grammar. Transactions of the American Philological Association 1, pp. 123–167. van Rensburg, M. C. J. 1984. Oranjerivier-Afrikaans. In M. C. J. van Rensburg, ed., Die Afrikaans van die Griekwas van die tagtigerjare. Bloemfontein: University of the Orange Free State, 1:2, pp. 513–518. 1989. Orange River Afrikaans: a stage in the pidgin/creole cycle. In M. Putz ¨ and R. Dirven, eds., pp. 135–151. Vasconcellos, Jos´e Leite de. 1901. Esquisse d’une dialectologie portugaise. Paris: Aillaud. Vaxelaire, Daniel. 1992. Vingt-et-un jours d’histoire: Ile de La R´eunion. Saint-Denis, R´eunion: Azal´ees Editions. Viljoen, W. J. 1896. Beitr¨age zur Geschichte der cap-holl¨andischen Sprache. Strassburg: K. Trubner. ¨ Vintil˘a-R˘adulescu, I. 1976. Le cr´eole fran¸cais. The Hague: Mouton. Voorhoeve, Jan. 1962. Sranan Syntax. Amsterdam: North Holland. 1973. Historical and linguistic evidence in favor of a relexification theory in the formation of creoles. Language in Society 2, pp. 133–145. Wagner, Max L. 1949. Lingua e dialetti dell’America spagnola. Florence: Le Lingue Estere. Walker, James A. 2000. Rephrasing the copula: contraction and zero in Early African American English. In S. Poplack, ed., pp. 35–72. Warantz, Elissa. 1983. The Bay Islands English of Honduras. In Holm, ed., pp. 71–94. Washabaugh, William. 1983. Creoles of the off-shore islands: Providencia, San Andr´es, and the Caymans. In Holm, ed., pp. 157–179. Watbled, Jean-Philippe. 2002. Le syst`eme verbal du cr´eole r´eunionnais. Paper ´ presented at the Colloque International des Etudes Cr´eoles, R´eunion. Forthcoming. Grammaire cr´eole et grammaire fran¸cais. In Gudrun Ledegen, ed., Anciens et nouveaux plurilinguismes: Les actes de la 6e table ronde du Moufia. Paris: L’Harmattan.

References

165

Webb, V. N. 1989. The presumed pidgin origins of Orange River Afrikaans. Paper presented at the 18th conference on New Ways of Analyzing Variation in English, Duke University. Wentworth, Harold. 1944. American Dialect Dictionary. New York: Crowell. Whatley, Elizabeth. 1981. Language among Black Americans. In C. A. Ferguson and S. B. Heath, eds., pp. 92–110. Williams, Jeffery P. 1985. Preliminaries to the study of the dialects of White West Indian English. Nieuwe West-Indische Gids 59, pp. 27–44. Winford, Donald. 1997. On the origins of African American Vernacular English – A creolist perspective, Part I: The sociolinguistic background. Diachronica 14:2, pp. 305–344. 1998. On the origins of African American Vernacular English – A creolist perspective, Part II: Linguistic Features. Diachronica 15:1, pp. 99–154. 2000. “Intermediate” creoles and degrees of change in creole formation: the case of Bajan. In I. Neumann-Holzschuh and E. W. Schneider, eds., pp. 215–246. Wolfram, Walt. 1969. A Sociolinguistic Description of Detroit Negro Speech. Washington, D.C.: Center for Applied Linguistics. 1974. The relationship of white southern speech to vernacular black English. Language 30, pp. 498–527. Forthcoming. Urban AAVE grammar. In Edgar W. Schneider, ed., English in the Americas and the Caribbean. Varieties of English Handbook, Vol. 1. Berlin etc.: Mouton de Gruyter. Wolfram, W., and N. H. Clarke, eds. 1971. Black–White Speech relationships. Washington D.C.: Center for Applied Linguistics. Wood, Peter H. 1974. Black Majority: Negroes in colonial South Carolina from 1670 through the Stono Rebellion. New York: Knopf. 1989. The changing population of the colonial South: an overview by race and religion, 1685–1790. In P. H. Wood et al., eds., Powhatan’s Mantle: Indians in the colonial Southeast. Lincoln: University of Nebraska Press, pp. 35–103. Wullschl¨agel, H. R. 1856. Deutsch-Negerenglisches W¨orterbuch. Repr. 1965. Amsterdam: S. Emmering. Yillah, Sorie. Forthcoming. Krio. In J. Holm and P. Patrick, eds. Ziegler, Douglas V. 1976. A Preliminary Study of Afro-Cuban Creole. M.A. thesis, University of Rochester. 1977. Some Notes on Puerto Rican Creole Spanish. MS. Zimmermann, Klaus, ed. 1999. Lenguas criollas de base lexical espa˜nola y portuguesa. Frankfurt-am-Main: Vervuert/Madrid: Iberoamericana.

Index

a- (VLRF pronoun prefix), 111–112 abbreviations, xix–xx, 138 abolition of slavery in Brazil, 56 in British Empire, 45 in R´eunion, 68 in USA, 36, 38 Abstand, 80 acrolect, 7, 69, 86 Adam, L., 19 African American English, 8, 28 attitudes towards, 1 auxiliary verbs, 73–75 “creolist” theory, 10, 12 decreolization of, 6 dependent clauses, 120 diaspora varieties, 37 ex-slave narratives, 37 historical development, 29–41, 71, 135, 137 negation, 75–76 non-verbal predicates, 76–77 number in NP, 93–94 possession, 95 pronouns study of, 4, 10–13 varieties, 33–34 verbal morphology, 73 word order, 116–117 African languages in Brazil, 53–56 Bantu versus West African, 54 Ewe-Fon, 55 Yoruba, 56 Africans (see also West Africans) in American South, 36, 47 in Brazil, 47, 48–49, 50, 55 in Cape Colony, 45 in R´eunion, 67, 137 in South Carolina, 35, 50 in Spanish Caribbean, 61–62 in Virginia, 32

166

Afrikaans, 2, 7, 8, 9, 13, 23 auxiliaries, 79 “Coloured” Afrikaans, see Kaaps dependent clauses, 122 East Cape versus West Cape, 46 Flytaal, 14 gender in NP, 98 Griqua Afrikaans, 14, 46, 100, 101 historical development, 41–47, 71, 137 Kaaps, 14, 23, 46, 79, 98 Malay Afrikaans, 14, 47, 101 negation, 79–80 number in NP, 97–98 Orange River Afrikaans, 14, 23, 46–47, 79, 98, 100, 137 possession, 98–99 pronouns, 99–101 standardization of, 46 study of, 13–15 verbal morphology, 77–79 word order, 120–122 Afro-Iberian linguistics, 16, 18 Afro-Portuguese, 62 Afro-Seminole, 33 age-grading, 40, 95, 96 agentive subject marker, 86–87, 88, 111, 141 agreement gender, 101, 103–104, 108–110, 139 number, 103, 106–107, 108 subject-verb, 73, 88 agriculture, see plantation slavery ain’t (AAE ‘didn’t’), 40, 75–76 Alvarez Nazario, M., 18 ambiguity, 126, 141 American Indian English, 3–4, 21 American Indians in Brazil, 50 in South Carolina, 35 in Virginia, 32 American South, 33, 39, 41, 137 analyticity, 5

Index Andalusian Spanish, 17, 83 Angola, 42, 54 Angolar CP, 126 animate, 93, 98, 99 anterior tense, 73 Arawakan languages, 60 archaic features, 10, 101 articles, see determiners aspect, 72, 86 assimilation, 41, 42, 43, 61, 62 associative plural, 92, 94, 97, 98, 108, 141 Atlantic creoles, xiv, 5, 81, 103, 107, 111, 116, 121, 124, 139 attrition, xiv Australian Aboriginal English, 4 auxiliary verbs, 72, 120 comparison, 91 influence of preverbal markers on, 72, 73–75, 79, 82, 84, 85, 87–88 Bahamian (C)E, xii, 10, 37, 117 link to Gullah, 13, 34, 35, 37 of whites, 11 Bahia, 50, 55, 56, 59 Bajan see Barbados CE Baker, P., 20 Bambara, 105 bandeirantes, 48, 55 Bantu languages, 44 Barbados, 29–32, 34, 35 CE, 26, 27, 136 basilect, 3, 5, 7, 9, 23, 30, 64, 69, 72, 86 Baxter, A., 16 Bay Island English, 11 be (AAE habitual), 40, 41, 75 be done (AAE future perfect), 75 been (AAE remote past), 40, 73–74 Bermuda, 26, 34 Bickerton, D., 9, 24, 27 bidialectalism, 58 bioprogram, 25, 28 Black English, 1 Boer War, 46 Boll´ee, A., 20 Boomfield, L., 7, 9 borrowing, 137, 143 Bourbonnais, 20, 65 bozal Spanish, see Habla Bozal Brazilian Vernacular Portuguese, 2, 8, 18 auxiliaries, 82 dependent clauses, 124 gender in NP, 103–104 historical development, 26, 47–60, 71, 137

167 negation, 82 non-verbal predicates, 83 number in NP, 101 possession, 104 pronouns, 104–105 study of, 15–17 verbal morphology, 80–82 British Caribbean, 28, 29, 32, 35 British colonists, 45 British dialects, 1, 10, 77, 119, 139 British North America, 33, 47 Cabrera, L., 18 Cajun French, 23 camouflaging, 134 Cape Colony, 41–46, 137 population, 45 Cape Coloured, 43, 44, 46 Cape Dutch, see Afrikaans Cape Malays, 42, 43, 47 Cape of Good Hope, 41–46 Cape Town, 43 Cape Verde CP, 26, 27, 47, 57, 102, 127, 137 Caribbean Spanish, see Non-standard Caribbean Spanish Carolinas, 30, 34 case marking, 92, 95, 99, 101 caste dialects, 7, 21 Caymanian English, 9, 11, 72 Central American (C)E, 11 Central American Spanish, 60 Ceylon, see Sri Lanka Chapuis, D., 19, 21 Chaudenson, R., xv, 20 Choco, ´ 106 civil rights movement, 1 Civil War (in USA), 36, 39 clause structure in AAE, 116–120 in Afrikaans, 120–123 in BVP, 123–127 in NSCS, 127–129 in VLRF, 129–133 coartaci´on, 62 code switching, xiv Coelho, A., 15 cognition, xv Colombian Spanish, 17, 18, 60, 62, 85, 106 Colonia ˆ Leopoldina, 59 Coloured, see Cape Coloured comparative method, xiv comparative studies in restructuring, 144–146

168

Languages in Contact

comparison of auxiliaries/preverbal markers, 91 of dependent clauses, 133–134 of gender, 113–114 of historical development, 70–71 of negation, 91 of number marking in NP, 112–113 of partially restructured vernaculars, 21–23 of possession, 114 of pronouns, 114 of verbal morphology, 90–91 of word order, 133 completive aspect, 74, 79 complexification, 141 consonant cluster simplification, 97 contact linguistics, 3, 145, 146 continuum, 7, 20, 85 convergence/divergence debate, 11, 40 convergence of influence, 5, 118 copula, 41, 87 absence of (see also non-verbal predicates), 23, 40, 41, 74, 90, 139, 141 comparison, 91 in AAE, 76–77 in BVP, 83 in NSCS, 85 in VLRF, 89–90 copula-like particle, 87 Corne, C., 20 Couto, H., 16 Craig, B., 21 creole, xiii influence, 9, 11, 21 prototype, xv, 28 versus non-creole features, 22–23 Cr´eole des Bas, see VLRF Cr´eole des Hauts, see VLRF Creole English in Caribbean, 2, 7, 8, 41 in North America, 30, 32, 33, 34–35, 41 in Suriname, see Surinamese creoles Creole French in Caribbean, xvi, 6 in Louisiana, 26, 52 in Indian Ocean, see Mauritian, Seychelles CF Creole Portuguese in Africa, 47 in Asia, 42 in Brazil, 47, 50, 55 in South Africa, 13–14, 42 Cr´eoles (i.e. people) in R´eunion, 68

Creole Spanish (see also Palenquero, Papiamentu) in Cuba? 63 pan-Caribbean? 18 creolistics, xi creolisant, 6, 8 creolization, xi, xiv, 3 full versus partial, 4–5, 12, 24, 134, 135, 141, 145 crops, see plantation slavery Cuban Spanish, 8, 17, 18, 60, 62, 63, 71, 84, 106 Cura¸cao, 51, 62, 71 de (CE progressive marker), 74 “debasilectalization”, 38 DeCamp, D., 7 decreolization, 2–7, 9, 10, 11, 16, 18, 20, 38, 60, 74, 75, 93, 102, 103, 126, 143 versus partial restructuring, 4, 11 definite articles, 93 definite future, 88 definiteness, 93 DeGraff, M., xiv de Granda, G., 17, 18 degrees of restructuring, see gradience dem, see pluralizer demographic ratio, see native-speaker ratio demonstratives, 93 denasalization, 80 den Besten, H., 14 depalatalization, 69 dependent clauses comparison, 133–134 in AAE, 120 in Afrikaans, 122 in BVP, 124 in NSCS, 128 in VLRF, 129–133 derivational affixes, xv destativizer, 87 determiners articles, 109 absent, 53 gender agreement, 108–110 number agreement, 101 detribalization, 44 diaspora varieties of AAE, 37 Dillard, J., 2, 36 direct object marker, 123 word order, 123, 131 does (CE habitual marker), 74

Index Dominican Spanish, 17, 18, 19, 26, 60, 62, 63, 71, 84, 85, 106, 108 done (CE, AAE completive marker), 74 drift, xv, 5, 15, 78, 90, 113, 143 durative aspect, 79, 84 Dutch, 7, 41–46, 116, 120 archaic, 122 colloquial, 98 dialects, 97, 122, 139 Dutch Brazil, 51, 52 Dutch East India Company, 41, 42, 43 East Africans in R´eunion, 67 East Indies see Indonesia Ebonics, 1 education, 3, 21, 58 emancipation see abolition of slavery embedded questions, 117 emphasis, 81, 82, 94, 124 ethnic linguistic marking, 32, 34, 39, 40 Eurocentrism, 15 European Portuguese, 56, 80, 83, 123, 125 European Spanish, see Peninsular Spanish Europeans (see also whites) in Brazil, 49–50 in Cape Colony, 45 in South Carolina, 35 in Spanish Caribbean, 64 in Virginia, 32, 71 Ewe-Fon, 55, 82 existential constructions, 76, 130 ex-slave narratives, 11, 37, 73, 93, 119 foreigner’s English, 2, 33 fossilized forms, 78, 86, 91, 110 founder effect, 14, 27, 32 fran¸cais avanc´ee, 20 fran¸cais cr´eolis´e, 69 freed slaves, 34, 44, 62 French, 42, 59 French East India Company, 65 frequency, 145 full creolization, see creolization Geechee, xii gender, 25, 42, 92, 95, 107, 108–110, 139 common, 96 comparison, 113–114 lexical, 109 natural versus grammatical, 111 genetic relation, xiv–xv, 71 Georgia, 32, 33, 35

169 German, 14, 59 dialects, 42 gradience, xv, 4, 5, 8–10, 12, 22, 25, 26, 141, 144 Green, K., 17, 19, 21, 142 Guadeloupe, xvi, 26 Guin´e-Bissau CP, 81, 127, 129 Gullah, xii, 2, 26, 28, 29, 33, 36, 37, 41, 77, 95, 119 historical development, 26, 35 Guy, G., 16 Guyanais CF, 26 Guyanese CE, 28 Guyanese Spanish, 60 habitual aspect, 75 Habla Bozal, 17, 18, 108 historical development, 63, 137 had in AAE, 74 Haiti, 25, 26, 27, 67 Halbkreolisch, 6 Hancock, I., 10 Helv´ecia Portuguese, 16, 59–60, 82, 101, 103, 125, 127, 137 Hesseling, D., 7 Hiberno English, see Irish English highlighter, 90, 124 Hispanicization, 61 historical fragments of language, 2, 37 historical linguistics, xiv Hottentot, see Khoi Huguenots, 34, 42 hypercorrection, 97, 119 i (VLRF verb marker), 69, 86–87, 111, 141 ideology, 1–3 Ile de France CF, 19 imperative, 81 identity and language, 136, 137 indentured servants from British Isles, 29, 30 from India, 69, 137 India, 42, 69 indigenization, 4, 145 indigenous languages, replacement of, 136 indirect object, 131 Indo-European languages, xiv Indonesia, 42 Indo-Portuguese, 6, 9, 19, 65, 112, 120, 123 infinitive, 78, 81 marker, 129 pseudo-infinitif, 133

170

Languages in Contact

inflections, xv fossilized, 78, 86, 91 loss of, xv, 3, 4, 5, 22, 41, 43, 72, 78, 90, 140, 144 verbal, 11, 72, 86 influence of full creoles, 137 -ing (AAE progressive marker), 74 innovation and partial restructuring, 121, 141 interior South, 36, 37 interlanguage, 14, 16, 33, 44 intermarriage, 42, 44, 48, 57 “intermediate” varieties, 30 intertwining, xiii, 4, 145 intransitives from reflexives, 112 invariant be, see be inversion of S/V (or Aux), 116–117, 121, 124, 128, 129–130, 133, 140, 141 Irish English, xiii, 4, 30, 117 isolating languages, xv, 5–10, 141 it’s (AAE existential), 76 Jamaican CE, 25, 26, 30, 34, 71, 72, 77, 118, 120 Jeroslow, H. (see also McKinney, H.), 15, 16 Jews in Brazil, 51 in Cura¸cao, 51–52 in Suriname, 52–53 juxtaposition, see [possessor + possessed] Kaaps, see Afrikaans Kaf, 69 Khoi, 13, 14, 15, 41–46, 78, 80, 100, 120 koineization, 4, 48, 145 Kotz´e, E., 14 Krio, 119 la (VLRF perfect marker), 65 Labov, W., 10, 11 language borrowing, see borrowing language change, xiv, xvi language contact, xvi, 78, 90, 91, 144, 145 language drift, see drift language shift, see shift language transmission, see transmission of language Leeward Islands, 30, 34 lete ki (VLRF durative past marker), 65, 67, 69 leveling, 34, 143, 145 primary, 143 secondary, 38, 143

lexicon, 22 Liberia, 37 L´ıngua dos Pretos Velhos, 16, 103 L´ıngua Geral, 48, 51, 52, 55–56 linguistic processes in partial restructuring, 142–143 linguistic traits of partial restructuring, 22, 142 Lipski, J., 18 Lorenzino, G., 19 Louisiana CF, 26, 52 Madagascar, 43, 65, 67 main clauses, 120 Makhudu, D., 14 Malagasy, 65, 86, 90, 109, 112, 130–131, 137 Malay Bazaar, 100 Malayo-Portuguese, 13–14, 42, 78, 100, 123 manumission, 57 mamelucos, 57 Maranh˜ao, 51, 57 marked features, 142 Markey, T., xiv maroons, 28, 51, 54–68 Martinique, xvi, 25, 26, 27 Maryland, 33 Mauritian CF, 19, 20, 26, 28, 67, 88 McKinney, H. (see also Jeroslow, H.), 15 McWhorter, J., xiii, xv Megenney, W., 18 Mello, H., 3, 15, 16, 21, 142 mesolect, 7, 117 Mexican Spanish, 60 Miami Spanish, 17, 60 migration of speakers, 3, 28, 30, 39, 40, 135–136 Minas Gerais, 55 mines (AAE possessive pronoun), 96 Miskito Coast CE, 10, 94 morphological versus phonological rules, 15, 80, 81, 83, 98, 102, 103, 105, 106 Mozambique, 43 Mufwene, S., xii, xiv, xv, 9, 12 Muhlh¨ ¨ ausler, P., 6 Nama, 79, 80, 100 Namibia, 47 nasalization (see also denasalization), 56, 80 Natal, 45

Index native-speaker ratio, 3, 12, 21, 24–29, 70, 135–136 in American South, 33, 145 in Brazil, 50 in Cape Colony, 45 in R´eunion, 67, 70–71 in South Carolina, 35 in Spanish Caribbean, 64 in Virginia, 32 negation comparison, 91 discontinuous double, 72, 80, 82, 85, 88, 91, 139, 141 in AAE, 75–76 in Afrikaans, 79–80 in BVP, 82 in NSCS, 84–85 in VLRF, 88–89 negative concord, 76, 84, 91, 139 clause-external, 76 negative inversion, 76 postverbal, 79, 80, 82 Negerhollands, 6, 26, 44, 45, 46, 71, 121, 122 Negro Non-Standard English, 1 New York City Spanish, 17, 60 Niger-Congo languages, xiv, 5, 42, 102, 103, 116, 141 no (CE verbal negator), 75 Nonstandard Caribbean Spanish, 2, 8 auxiliaries, 84 dependent clauses, 128 gender in NP, 107 historical development, 60–64, 71, 137 negation, 84–85 non-verbal predicates, 85 number in NP, 106–107 possession, 107 pronouns, 107–108 study of, 17–19 verbal morphology, 83–84 word order, 127–128 non-stative (see also stative), 87 non-verbal predicates (see also copula absence), 23, 40, 41, 72, 74, 90, 141 comparison, 91 in AAE, 76–77 in BVP, 83 in NSCS, 85 in VLRF, 89–90 North Carolina, 33, 35 noun phrase in AAE, 93–96 in Afrikaans, 96–101

171 in BVP, 101–105 in NSCS, 105–108 in VLRF, 108–112 Nova Scotia, 37 number marking in NP, 92 comparison, 112–113 in AAE, 93–94 in Afrikaans, 97–98 in BVP, 103 in NSCS, 106–107 in VLRF, 108 objects: order of direct/indirect, 131 Orange Free State, 46 Orange River Afrikaans, see Afrikaans orthography, xx palatalization (see also depalatalization), 56 Palenquero CS, 17, 18, 19, 62, 85, 107 Palenquero Spanish, 18, 64, 107, 108, 129 Palmares, 54 Papen, R., 20 Papia, 16 Papia Kristang CP, 100 Papiamentu, 6, 17, 26, 27, 71, 81, 105, 107, 127, 129 Portuguese element in, 51–52 parameter settings, xv Parkvall, M., xiii, 23, 24–27 partial restructuring, xiii, 2 and substrate influence, 141 theoretical model, 5, 7–10, 142–143 versus decreolization, 4, 23 past tense in AAE, 73, 138 in Afrikaans, 78–79 Peninsular Spanish, 85 periphrasis, 88 Pernambuco, 50, 52, 54 petits blancs (see also poor whites), 68, 69 phonological versus morphological rules, 15, 80, 81, 83, 98, 102, 103, 105, 106 internal motivation of phonological rules, 144 phonotactic rules, 106 Pidgin Dutch, 42, 43, 44 Pidgin English, 2, 34–35 Pidgin French, 28 Pidgin Portuguese, 52 Pidgin Spanish, 63 pidginization, 3 pidginization index, 10, 27

172

Languages in Contact

pied-piping, 122 plantation agriculture, 43 coffee, 59, 61, 67 cotton, 36, 38 rice, 34, 36 sugar, 30, 50, 61, 68 tobacco, 29, 30, 32, 36, 61 “Plantation Creole”, 36 pleonastic subject pronoun, 96 pluralizer = ‘they’, 93, 108, 111, 139, 141 in associative plural, 94, 97, 139 plural marking in NP, 92 irregular, 97 Ponelis, F., 14 poor whites in Cape Colony, 46 in R´eunion, 68–69 Poplack, S., 12 Poror´o, 63 Portuguese, 47 (see also Brazilian Vernacular Portuguese, European Portuguese, Helv´ecia Portuguese, Standard Brazilian Portuguese) influence on Guyanais CF, 52 influence on Papiamentu, 51–52 influence on Saramaccan, 52–53 [possessed (OF) possessor], 53, 98, 104, 110, 140 possession comparison, 114 in AAE, 95 in Afrikaans, 98–99 in BVP, 104 in NSCS, 107 in VLRF, 110 possessive determiner, 95, 99–101 marker on nouns, 92, 98 pronoun, 99–101 [possessor + possessed], 53, 140 in CE, AAE, 95 [possessor HIS possessed], 98–99, 100–101 in colloquial Dutch, 98 post-creole, 1–9, 20, 23, 46 post-verbal negation, see negation post-verbal marker, 86 power relationships, 3 prepositional phrase deletion, 126 prepositions absence before place names, 131 absence before indirect object, 131 preposition stranding, 122 present for past, 73, 78–79, 82, 91

preverbal markers, 11 influence on auxiliary verbs, 72, 73–75, 79, 82, 84, 85, 87–88, 139, 141 primary leveling, 143 prior creolization, 16 processive, 86–87 Pro-drop, 81, 84, 128 progressive aspect, 74 pronouns case marking, 92, 95, 99, 101, 104, 112, 140 clitic, 99, 104, 115, 123 comparison, 114 conjoined, 94 disjunctive, 110, 112, 115, 128 emphatic, 94, 104, 110, 111, 115 gender marking, 95, 99, 104, 111 in AAE, 95–96 in Afrikaans, 99–101 in BVP, 104–105 in NSCS, 107–108 in VLRF, 110–112 interrogative, 100 object, 101, 112 pleonastic subject, 96, 119 possessive, 99–101, 122 reflexive, 101, 105, 107, 112, 115, 140 relative, 76, 96, 122, 125 resumptive, 125, 134, 141 subject, 81, 84, 128 word order, 104, 112, 123 prototype, creole, xv, 28 Puerto Rican Spanish, 17, 18, 60, 62, 71, 84, 85, 106 quantitative studies, 93 questions direct, 121 embedded, 121–122 word order, 116, 121, 128, 130 question words, 124 quilombos, see maroons race relations in Brazil, 57 in South Africa, 43, 44 in Spanish Caribbean, 61–62 in USA, 32, 39, 41 racial segregation, 35, 39, 40, 57 Raidt, E., 14 Rademeyer, J., 14 “radical” creole, 25 rapidity of creole genesis, xvi regional speech forms, 10, 11 regularity, 145

Index Rehoboth Basters, 46 Reinecke, J., xi, 8, 13, 19 relative clauses, 117–119, 120, 122, 124–127, 131, 133 determiners, 125 pronouns, 122, 124, 131, 132 complex/simplex, 134 resumptive, 125, 134 subject, absent, 76, 96, 117–119, 125, 128, 130, 133, 134, 140 religious rites, language of, 16, 18, 56 remote past, 73 renativization, 145 repidginization, 9 restructuring index, 25, 26 R´eunionnais, see Vernacular Lects of R´eunion French Rickford, J., 10, 12, 24, 27 Rio de Janeiro, 56 Roberge, P., 13, 15 Romance languages, 104, 139 -s (E verbal inflection: third person singular), 40 in AAE, 73, 137 (E, D noun inflection: plural), 97, 139 in AAE, 93–94 in British dialects, 93, 94 (NP inflection: plural) in French, 108–112 in Portuguese, 56, 101 in Spanish, 106–107 St. Barts or St. Barth´elemy, 20, 23 St. Kitts, xvi, 26, 27, 30 St. Thomas, 25, 26 salience, 145 Saman´a, 37 S˜ao Paulo, 48 S˜ao Tom´e CP, 17, 26, 27, 47, 57 influence on BVP, 51, 82, 102, 104, 105, 126 influence on Palenquero CS, 62 Saramaccan, 7, 120 Portuguese element in, 52–53 say (CE, AAE quotative), 41, 120, 134, 141, 142 Schneider, E., 10, 12 Scholtz, J., 14 Schuchardt, H., 6, 19 Schwegler, A., 18 se (Afrikaans possessive), 98–99, 122, 140 -se (NSCS plural marker), 106–107 second language acquisition, 5, 14, 18, 35

173 secondary leveling, 38, 143 segregation, see racial segregation semi-creolization, xi–xiii, 8, 9, 12, 15, 19, 37, 60 Seychelles CF, 19, 20, 109 share-cropping, 38, 40 shift, 4, 16, 56 of heterogeneous language community, 142, 143 of homogeneous language community, 30 Siegel, J., 4 Sierra Leone, 37 Silva Neto, S., 9, 15 slaves, 36 from Africa, 32, 33, 36, 43, 45, 67 from Caribbean, 32, 33, 36 from Asia, 42, 43, 45, 67 status, 30 social factors in partial restructuring, 21, 24–71, 135–137 sociolinguistics, xiv, 106 sociolinguistic history, 3, 24 of AAE, 29–41 of Afrikaans, 41–47 of BVP, 47–60 of Gullah, 34–35 of Helv´ecia Portuguese, 59–60 of NSCS, 60–64 of VLRF, 65–70 South, see American South South Africa, 13, 46 South African philological school, 14, 15 South Carolina, 33, 35 Southern White Vernacular English, 23, 40, 41, 75, 76, 95, 96 SOV word order, 42, 116, 120, 133 speech relationships black-white, 39, 40, 41 Spanish, 52 Spanish Caribbean, 84–85, 137 demographics, 64 language contact in, 60–64 “spontaneous development” theory, 14 Sranan CE, 26, 27, 53, 120 Sri Lanka, 42, 45 Standard Brazilian Portuguese, 57–59, 80–81 standardization, 13, 21, 46 stative (see also non-stative), 87 Stewart, W., 7, 8, 36 stigmatized forms, 73, 74 strong restructuring, 23 structural distance, 7–10, 12, 25 structural reduction, 140

174

Languages in Contact

subject/verb (or Aux) inversion, see inversion subjunctive mood, 82 subordinate clauses, 120, 123, 127, 129 subordinator, 118 subordinator, absent, 123, 127, 128, 132, 134, 140, 141 substrate, xiv, 4, 5, 139 influence of, 15, 16, 18, 33, 41, 72, 86, 134, 141, 143 superstrate, xiv, xv, 5, 6 Surinamese creoles (see also Saramaccan, Sranan), 7, 8, 28, 30, 34, 41, 52, 120 SVO word order, 42, 104, 116–119, 120, 121, 123, 130, 133 syllable structure, 56 syntax-internal model, xv, xvi synthetic languages, xv, 5 systematicity within language, 5 Tagliavini, C., 6 target language, access to, 136 te (VLRF verb marker), 69, 87 tense, 72, 86 theoretical model for partial restructuring, 142–143 Thomason, S. G., xiv, 3–4, 9, 24, 142 time factor in restructuring, 24, 27–29, 135 Tok Masta, 6 tone, xv, 25 topicalizer, 90 transmission of languages, xv, 3, 21, 67, 71 normal, xiv, 71 irregular, 103 transparency, 142, 145 Transvaal, 46 Trinidadian Spanish, 60 Tupi, 15, 48, 53, 55–56 typology, xiii, xiv–xv unbound pluralizer, see pluralizer unrestructured overseas varieties, 3, 33, 34, 135, 136 universals, xv, 5 unrounding of vowels, 69 Upper South, 36 urban ghettos, 39 Valkhoff, M., 14, 15, 19 Van Name, A., xvi, 17, 24 van Rensburg, M., 14 Vasconcellos, J., 7

Venezuelan Spanish, 17, 18, 60, 85 verbal morphology comparison of, 90–91 in AAE, 73 in Afrikaans, 77–79 in BVP, 80–82 in NSCS, 83–84 in VLRF, 20, 85–87 verb phrase in AAE, 73–77 in Afrikaans, 77–80 in BVP, 80–83 in NSCS, 83–85 in VLRF, 20, 85–90 verb-last word order, 122 verb-second word order, 120, 121, 122 Vernacular Lects of R´eunion French, 2, 10 auxiliaries, 87–88 Cr´eole des Bas, 20, 67, 69, 86, 137 Cr´eole des Hauts, 20, 69 dependent clauses, 131–133 historical development, 65–70, 136, 137 negation, 88–89 non-verbal predicates, 89–90 number marking in NP, 108–112 possession, 110 study of, 19–21 time factor in restructuring, 28 urban Creole, 20 verbal morphology, 85–87 word order, 129–131 Virginia, 29, 30, 32, 33, 35, 36 Virgin Islands, 71 VSO word order, 130 weak restructuring, 23 West Africans in Cape Colony, 42 in Barbados, 29 in Guyana, 28 in Mauritius, 20, 28 in R´eunion, 67 in South Carolina, 35 in Virginia, 29, 32 whites as creole speakers, 52 in Brazil, 49–50 in Cape Colony, 45 in South Carolina, 35 in Spanish Caribbean, 64 in Virginia, 32, 71 white supremacists, 39 WH-movement, 122

Index Winford, D., 4, 12 withdrawal of superstrate, 28 word order, 101, 104, 116–117, 119 comparison, 133 in AAE, 116–117 in Afrikaans, 120–122 in BVP, 123–124 in NSCS, 127–128 in VLRF, 129–131

175 you all (AAE, SWVE pronoun), 96 Yoruba, 77 associative plural, 94 in Brazil, 56 pluralizer, 93 relative clauses, 118 zero copula, see copula absence zoreil, 69 Zulu, 45