This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Creole Language Library (CLL) A book series presenting descriptive and theoretical studies designed to add significantly to the data available on pidgin and creole languages.
Editors Miriam Meyerhoff
Umberto Ansaldo
University of Auckland
The University of Hong Kong
Editorial Advisory Board Mervyn C. Alleyne
Pieter Muysken
Kingston, Jamaica
Nijmegen, The Netherlands
Marlyse Baptista
Peter Mtihlhitusler
Ann Arbor, USA
Adelaide, Australia
George L Huttar
Shobha Satyanath
Dallas, USA
Delhi, India
John Holm
John Victor Singler
Coirnbra. Portugal
New York, USA
Silvia Kouwenberg
Norval Smith
Kingston, Jamaica
Amsterdam, The Netherlands
Susanne Michaelis Leip1ig, Germany
Ann Arbor, USA
Sal.ikoko S. Mufwene
Tonjes Veenstra
Chicago, USA
Berlin, Germany
Sarah G. Thomason
Volume37 Variation in the Caribbean. From creole continua to individual agency Edited by Lars Hinrichs and Joseph T. Farquharson
Variation in the Caribbean From creole continua to individual agency
Edited by
Lars Hinrichs University of Texas at Austin
Joseph T. Farquharson University of Technology; Kingston
John Benjamins Publishing Company Amsterdam I Philadelphia
The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences - Permanence of Paper for Printed Library Materials, ANSI z39.48-1984.
Library of Congress Cataloging-in-Publication Data Variation in the Caribbean : From Creole continua to individual agency I edited by Lars
Hinrichs, Joseph T. Farquharson. p. em. (Creole Language Library; ISSN 091.0-901.6 ; v. 37) Includes bibliographical references and index. Creole dialects, English--Caribbean Area--Social aspects. 2. Language and languages-
1.
-Variation. 3· Languages in contact. 4· Sociolinguistics--Caribbean Area. I.
Hinrichs, Lars. II. Farquharson, Joseph T. PM7874·C27V37
John Benjamins Publishing Co.· P.O. Box 36224 • 1020 ME Amsterdam· The Netherlands John Benjamins North America· P.O. Box 27519 ·Philadelphia PA 19118-0519 • usA
Table of contents
Introduction
1
Lars Hinrichs & Joseph T. Farquharson PART 1.
Variation and linguistic systems
Revisiting variation between sa and o in Sranan
13
Donald Winford Inherent variability and coexistent systems: Negation in Bequia
39
James A. Wa.lker e!"' Jack Sidnell Putting individuals back in contact: Accommodation strategies by Barbadians in Ipswich
57
Michelle C. Bra.fia-Straw Relative markers in spoken Standard Jamaican English
79
Ulrike Gut PART 11.
Variation and identity
"Flying at half-mast"? Voices, genres, and orthographies in Barbadian Creole
107
Ja.nina Fenigsen The creole continuum and individual agency: Approaches to stylistic variation in Jamaica
133
Dagmar Deuber Language attitudes and linguistic awareness in Jamaican English
And1·ea Sand
16}
VI
Variation in the Carlbbean
PART III.
Variation and the community
The varilingual repertoire ofTobagonian speakers
191
Valerie Youssef On the emergence of new language varieties: The case of the Eastern Maroon Creole in French Guiana Bettina Migge & Isabelle Uglise 'Creole' and youth language in a British inner-city community
231
Susa.n Dray c. . Mark Sebba Le Page's theoretical and applied legacy in sociolinguistics and creole studies
251
John R. Rickfoni Name index Subject index
273 275
Introduction Lars Hinrichs & Joseph T. Farquharson
Creole studies and Caribbean linguistics have always been closely linked with each other. Much influential theory on pidgin and creole languages, as well as a significant portion of descriptive work. emerged from work on this part of the world. For example, a clear majority of the papers contained in the seminal collection edited by Hymes (1971), Pidginization and Creolization of Languages, draw on empirical facts and data from the Caribbean. In the same volume, Hymes and contributors showed that they considered the study of variation as a constitutive subfield of creole linguistics by allotting to it one of the book's six sections, "Variation and use: a range of English-linked cases" (297-419). John Rickford even asserts: "If one were to ask what sub field of linguistics the study of pidgins and creoles was most closely associated with, the most likely answer would be sociolinguistics" (1988: 51). The 1971 Hymes volume, and the 1968 conference from which it emerged, fundamentally shaped creole linguistics and the way it would come to be practiced in the decades to come. At the initial stages, the field drew considerable energy from a sense that it had the potential to fill in some of the blind spots that the study oflanguage in the Chomskian paradigm was seen to be developing - and the study of variation was located at a particularly important nexus in that endeavor (Kouwenberg & Singler 2008). In Hymes's formulation, Unguists wishing to treat their w01k as referring to an ideally fluent speaker-listener in a homogeneous speech community (Chomsky 1965) often can ignore variability as peripheral. In these cases [i.e. in creole speech communities] variability is central. (Hymes 1971: 300)
The motivating and unifying power of these calls to theoretical-syntactic arms in creole language studies has waned. 1 Instead, creolists are now engaging sociolinguistic questions as worthwhile objects of study in their own right, not as a weapon in the struggle between formal and empirical linguistics. Compared to the idealism of Hymes and other creole sociolinguists of the first hour, Peter Patrick (2008) recently spoke of the relationship between variationist creole linguistics and transformational syntax in rather more matter-of-fact tones. 1.
2
Lars Hinrichs & Joseph T. Farquharson
Caribbean sociolinguistics looks back on a history of considerable achievement, which includes substantial impact on sociolinguistics in general (see papers in Rickford, ed., 1988). Since creoles rarely exist without other language varieties being present in the same sociolinguistic space, from which they usually are very different structurally, creole-speaking communities often make the relations between grammar, variation, and social context visible in particularly stark relief As a result, theoretical models have emerged from the study of creoles that have been widely adopted by sociolinguists for non -creole settings. By causing us to re-think how we approach the social and cognitive organization of different grammatical systems in one community, the field has yielded, for example, a formulation of implicationally scaled vernacular-standard continua (DeCamp 1971; Rickford 1987; Escure 1997), and the Acts ofidentity model (LePage and Tabouret-Keller 1985), both of which have been widely influential. Patrick (2008) also points out that research on creoles (especially on the creole origins of African -American Vernacular English) has played a considerable role in establishing and developing the VARBRUL framework, a multivariate statistical method that dominated quantitative sociolinguistics in the 1980s and 1990s, as a disciplinary standard.2 Though it is creole linguistics that has traditionally been most naturally interested in Caribbean languages, the somewhat younger study of English as a World Language is now equally engaged in the study of Caribbean variation. Through its focus, then, this volume intentionally unites scholars from several sub-branches of linguistics. The varieties it covers include both creoles and local standard varieties, and while a majority of the varieties discussed are English-based, creoles involving other lexifiers are also addressed. Thereby, and by including a look at diasporic Caribbean varieties (Braiia-Straw; Dray and Sebba), the volume reflects the breadth of topics and approaches in current work on Caribbean variation. The subtitle, "From creole continua to individual agency," acknowledges the range of theoretical and methodological approaches that these works take to the linguistic variation which exists in Caribbean speech communities. For a long time, discussions of (post-)creole continua in a variationist framework and continuum theory have dominated Caribbean sociolinguistics. Meanwhile, general variation studies underwent a profound reorientation, or rather: addition of interests, turning to language and identity as well as pursuing microapproaches to variation. This shift in interests has been referred to as the "interactional turn" in sociolinguistics. While linguistic anthropology engaged concepts
2. For introductions to the variable rule (VARBRUL) framework and the GoldVarb software Implementation see fur example Cedergren and Sankoff (1974), Bayley (2002), and Tagliamonte (2006). Bayley's article includes a helpful dosing section on future directions in research using variable rules (Bayley 2002: 134-136). For a critique of the statistical method see Johnson (2009).
Introduction
like individual identity/agency and language ideology as factors in variation considerablyearlier than sociolinguistics, the latter adopted them from the early 1980s onward. A culmination of this development has been the adoption of ethnographic approaches and concepts by sociolinguists who were trained in the macro-variationist, Labovian tradition. 3 A major theoretical stepping stone in this development came right out of creole linguistics: Le Page and Tabouret-Keller's Acts of Identity (1985). For a variety of reasons, however, this model had a lesser effect on the field of creole linguistics than on the rest of sociolinguistics, where it facilitated and supported studies of variation in a non-stratified, identity-based framework (Mair 2003a). This volume, as well as the 2006 workshop at the Sociolinguistics Symposium in Limerick from which it sprang, is intended as a place for Caribbean data to meet with the broadest possible range of sociolinguistic approaches to variation, which includes a chance to introduce very recent sociolinguistic thinking. That spirit is reflected in the selection of papers. The first group of papers looks at variation in linguistic features, with three papers devoted to (morpho-)syntax (Winford, Walker and Sidnell, and Gut) and one to phonological variation in diasporic dialect contact (Brafta-Straw). Don Winford addresses an issue of purported variation in Sranan: sa and o have both been analyzed as future markers, but the question of their interchangeability has so far been unresolved. Winford uses historical writings on Sranan as well as native speaker judgments to work out the semantic differences between both markers in different contexts. Walker and Sidnell address negation in one Caribbean setting from a classic variationist standpoint: binary variation analysis in the VARBRUL framework. Their contribution shows the power of such analysis in the study of an issue frequently faced by Caribbeanists: island-internal variation. The grammatical environment of their response variables lends support to a view of Bequia as a setting that features co-existing systems. Michelle Brafta-Straw's paper expands this volume's geographical scope to the Caribbean diaspora. In an analysis of the speech of Barbadians in Ipswitch (UK), she presents in detail the effects of a new majority accent on an immigrant communitfs vowel system. She demonstrates that the outcome of immigrants' dialect contact, rather than taking place on a simple continuum between accommodation and maintenance, is complicated both by phonological mechanics and the social dynamics of the diaspora. There, Caribbean varieties compete for social status in ways that differ from those in native contexts: the great symbolic load of}amaican 3· Key publications in this type of ethnographic variation studies include Eckert (2000), Fought (2003), Zhang (2005; 2008), and Mendoza-Denton (2008).
3
4
Lars Hinrichs & Joseph T. Farquharson
Creole makes it encroach upon the usage domains of other English-based Caribbean creoles. Ulrik.e Gut shifts the focus to an acrolectal variety: Standard Jamaican English. In her choice of data, she addresses a longstanding problem in the sociolinguistics of creole settings, which Alison Irvine describes as a kind of circularity in the definition of the polar varieties on the (Jamaican) continuum that, typically, takes this form: a) Since the acrolect is not unlike other standard Englishes, phonologically as in [khat] or morpho-syntactically as in did not take; then b) the basilect reflects structures that are maximally divergent from the acrolect - [kjat] or neva tek; as such, c) acrolectal speech is maximally nonbasilectal and can therefore be e:xpected to produce the forms [khat] or did not take. (Irvine 2004:45 and 2008: 10)
To circumvent this circularity in matching data and definitions, Gut draws on the Jamaican component of the International Corpus of English (ICE-JA, see e.g. Mair 2009). This collection of data targets acrolectal speech as defined primarily by social contexts of use, i.e. as the speech of educated Jamaicans in dominantly formal settings, not by linguistic criteria. 4 This solution is very similar to the one chosen by Irvine, who collected her data for the study of the spoken acrolect in Jamaica - in a government agency - a domain in which there was maximal pressure on speakers to use standard speech. Gut's study shows that despite the ongoing endogenization of the acrolectal norm, which is taking place in many post-colonial settings following political independence, infusion of the acrolect by the local creole cannot be observed, at least for the aspect of the syntax of the Jamaican acrolect examined in this contribution. The second set of papers addresses the individual as a cause and site oflinguistic variation. Fenigsen's paper explores the written representations of speech in a multilectal Caribbean setting, Barbados. She considers, in a qualitative-interpretative approach to written language, authorial stance as a factor in framing discourse and thus in defining the social meanings of nonstandard (English) spellings, which, in Caribbean contexts, are never just used to signal that (orthographically non-standardized) Creole is being used (Miihleisen 2002; Hinrichs and WhiteSustafta forthcoming). Deuber studies Creole influences on the English of educated Jamaicans in a set of transcribed conversations, drawn from ICE-JA. Her two-pronged study design 4· Note, however, that this definition of acrolectal speech through the combination of domain of occurrence and education of speaker does not, in practice, amount to an ex:dusion of all creole speech (Mair 1992). Creole does occur in educated speech and is therefore included in the ICE corpus. The stylistics of this particular outcome oflanguage contact are addressed in Deuber's contribution to this volume.
Introduction addresses an old problem in the analysis of stylistic variation in a (post-)creole continuum: while creolists are aware of the scalable nature ofvariation between acrolect and basilect, native speakers in a creole continuum usually think of their linguistic resources as discrete, i.e. composed of one prestige code and another non-prestige code; in the case of Jamaica: "English" and "non-English~ Deuber shows that both approaches can be productively combined in the analysis of stylistic variation, when the creole continuum approach is used as the descriptive backdrop of feature-based variation, and a binary-resources view is taken as the necessary basis for understanding the indexical values of code choices in a creole continuum. Sand's contribution investigates language attitudes toward local forms in the emerging standard Jamaican English. In addition to the discussion over newlyendonormative standards in the Caribbean, she argues, the dimension of educated native speakers' notions of appropriateness must be figured into overt definitions of such new norms. Doing so further expands the methodological inventory that can help in the interruption of the circularity sketched by Irvine (cited above). The final section of the volume deals with descriptive and theoretical ways of understanding variation in the community. Youssef's study discusses the role of individual variation in our understanding of how creole continua are generated and constituted. Her combined focus on individual style and language attitudes/ ideologies in Tobago forms the basis for a critique of DeCamp's original formulation of the creole continuum (1971), which saw decreolization as inevitable once the creole exists alongside its lexifier. A second paper that interrogates the creole continuum model is Migge and :Uglise's, which discusses the social organization of varieties in a practice-based approach. They too question the unidirectionality implicated in DeCamp's model, in which decreolization is the necessary outcome of sociolinguistic evolution in creole societies. Rather, they argue, sociolinguistic practices lend support to multidimensional views of dynamic shifts in the relationship between varieties in a multi-varietal setting like French Guiana. Dray and Sebba once again turn the view toward the Caribbean diaspora in Britain. Their study of Creole use among youth of Caribbean heritage in Manchester takes an interest in what happens to a creole in the diaspora when the influx of Ll-speaker immigrants subsides, and thus serves as an update to Sebba's work on London Jamaican. While Sebba (1993) found diasporic Caribbeans in England using Creole in conversational codeswitching that frequently drew on the potential of Creole to index ethnic identity, the younger generation of British youth (of Caribbean heritage) uses Creole without any notions of the "authentic speaker" or ethnic prerequisites regulating access to the code. In Manchester, the transformation of the function of Creole from that of "language" in the homeland to that of a "symbol" in the diaspora (Mair 2003b) is completed, though it has not
5
6
Lars Hinrichs & Joseph T. Farquharson
ended in a flat indexical relationship with ethnic identity. In the community studied by Dray and Sebba, speakers "are legitimate users of these vernacular styles as a result of their lived experience in the inner city and their interest in and exposure to cultural forms such as the global hip-hop culture." Rickford's paper closes this volume with a broad appraisal of Robert B. Le Page's contribution to creole studies. His paper asserts the importance of variationist work within creole linguistics. Rickford reminds us that an understanding of the identity politics of variation in creole settings, as promoted by the Acts of Identity model, is the purview of Caribbean sociolinguistics just as much as active participation in the discourse on language in education and language rights. He points to a continuing need for more work on Caribbean variation, both of the theoretical and the applied kind. In that same spirit, we end by pointing out some desiderata in Caribbean variation studies. We consider these to lie in two broad areas. First, variationist work is needed to refine accounts of the syntactic variation which occurs in creole situations. As we improve on our models of the linguistic systems of creole communities, we will continue to confront complex sociolinguistic realities composed of multiple varieties. Future work can help us produce models that express the systemic relationships of alllects in a continuum situation without discounting the existence of separate varieties and the manifest social functions of their use. While variationist creolists of the early days, following DeCamp. insisted that the different varieties, or lects, encountered in continuum situations like Jamaica add up to a mono systemic whole - a claim that resonates most readily with generativist perspectives on syntax5 - a view of performance as based on distinct varieties makes sense at least for some steps of the analysis. Native speakers in creole communities, who perform the linguistic data that we aim to explain, typically think of themselves as having separate varieties at their disposal that can be used in discourse work. Several of the papers in this volume remind us of the distinctions in social functions between varieties. e.g. Deuber, Gut, Youssef, Dray and Sebba, and Migge and Leglise; note especially how Walker and Sidnell's ethnographically informed quantitative analysis suggests a view of a continuum that features separate systems. Promising avenues for future research include taking greater interest in stylistic variation (Eckert and Rickford 2001) and indexicality (Silverstein 2003) in creole communities, as these have the potential of informing syntactic description in the area of the co-existence of varieties and systems in continuum situations. It may also be helpful to incorporate the idea of an "ethnolinguistic repertoire" into our 5· Note however that increasingly, current generativlst syntacticlans (e.g. Adger and Smith 2005; Trousdale and Adger 2007) are grappling with (micro-)variation, as opposed to earlier theorists, who tended to ignore it
Introduction
thinking about creole community grammars. 6 A better understanding of the way that the grammars of different varieties in the same sociolinguistic system are linked becomes more pressing as new domains of creole literacy continue to emerge, e.g. in computer-mediated communication. The second area for useful future work lies in the description of intra-varietal diversification. Irvine's important work on the Jamaican acrolect (2004, 2005, 2008) demonstrates and addresses the need to take empirical, variation-based approaches to the description oflects, and to abandon definitions ofbasilect and acrolect as each other's opposite, in cases where no further specification is provided. This necessarily includes a detailed look at variation along social categories. A recent special issue of Sargasso (Miihleisen and Walicek, eds., 2009) on gender and sexuality in creole linguistic settings makes an important contribution in this area. While we hope that the current volume provides a representative overview of what has been done on variation in the Caribbean since Hymes (ed., 1971), it does not signal a wrapping up of the work, but the unfolding of new directions. We hope it shows that though there are many things which sociolinguists working on Caribbean varieties still take for granted, there remain many unexplored and unmapped areas, and that there are still questions which need to be explored if not answered. Fortunately for contemporary linguists working on Caribbean varieties, these absences are coming to light in a period where the broader field of linguistics is actively seeking creole data. Typologists are demanding more and better descriptions of creoles, and they are becoming more willing to include creoles in their samples, as opposed to treating them as noise in the data (e.g. Haspelmath et. al. 2005). In essence, this volume represents our throwing down of the sociolinguistic gauntlet, and we invite as many scholars as possible to pick it up so that this joint project can move forward.
Acknowledgments The editors would like to thank Maryam Bakht, Dagmar Deuber, Hubert De vanish, Christian Mair, and Peter Patrick for helpful consultations during the preparation of this volume; the anonymous reviewers of the papers; series editors Miriam 6. Benor (2010) defines "ethnolingulstics repertoire" as "a fluid set oflinguistk resources that members of an ethnic group may use variably as they index their ethnic identities" (160). In creole settings, the notion of indexing ethnic identities may be more or less pertinent, depending on the setting (the Belize research at the heart of Acts ojlde11tity is an example of a case where inter-ethnic distinction is certainly relevant). The important aspect of this concept is that it provides a unified view of all varieties in a community as a repertoire, which allows for a notion of distinct varieties with different social functions.
7
8
Lars Hinrichs & Joseph T. Farquharson
Meyerhoff and Umberto Ansaldo for the guidance and encouragement, as well as patience and perseverance, that they provided during the preparation of the manuscript; Shawn Warner-Garcia for substantial help with proofreading and formatting; and the authors of these papers for participating and believing in this project. Every effOrt has been made to trace and contact copyright holders for quoted materials. The editors and publishers would be pleased to hear form any copyright holders not acknowledged in this volume, so that this may be amended at the earliest opportunity.
References Aceto, M. & Wllllams, J.P. (eds.). 2003. Contact Englishes of the Eastern Caribbean. Amsterdam: John Benjamlns. Adge~ D. & Smith, J. 2005. Variation and the minimalist programme. In Synta.t: and Variation: Reconciling the Biological and the Social, L. Cornips & K. Corrigan (eds.), 149-178. Amsterdam: John Benjamlns. Bayley, R. 2002. The quantitative paradigm. In The Handbook ofLa11guage Variation and Change, J. K. Chambers, P. Trudgill, & N. Scbilling-Estes (eds.), 117-141. Malden MA: Blackwell Benor, S. 2010. Ethnolinguistic repertoire: Shifting the analytic focus in language and ethnicity. Journal ofSociolinguistics 14(2): 159-183. Cedergren. H. & Sankoff, D. 1974. Variable rules: Performance as a statistical reflection of competence. Language 50(2): 333-355. Chomsky,. N. 1965. Aspects of the Theory ofSyntax. Cambridge MA: The MIT Press. DeCamp, D. 1971. Toward a generative analysis of a post-creole speech continuum. In Pidginizatlo1! and Creolizatio11 ofLanguage.s, D. Hymes (ed.), 349-370. Cambridge: CUP. Eckert, P. 2000. Linguistic Variation as Social Practice: 1he Linguistic Co11struction of Identity 111 Belten High. Malden, MA: Blackwell Eckert, P. & Rickford, J. R. (eds.). 2001. Style and Socioli11guistic Variation. Cambridge: CUP. Escure, G. 1997. Creole a11d Dialect Continua: Standard Acquisitio11 Processe.s in Belize and Chb1a (PRC). Amsterdam: John Benjamlns. Fought, C. 2003. Chicano English i11 Context. Houndsmills: Palgrave MacMillan. Haspelmath, M., Dryer, M.S., Gil, D. & Comrie, B. 2005. 1he World Atlas ofLanguage Structures. Oxford: OUP. Hinrichs, L. & White-Sustaita, J. Forthcoming. Global Englishes and the sociolinguistics of spelling: A study ofJamaican blog and email writing. English World-Wide 32(1). Hymes, D. 1971. Introduction to Part V, Variation and use: A range of English-linked cases. In D. Hymes (ed), 299-304. Hymes, D. (ed. ). 1971. Pidginlzation and Creollzatlon of Languages: Proceedings of a Conference Held at the University of the West Indies, Mona, Jamaica, April1968. Cambridge: CUP. Irvine, A. 2004. A good command ofthe English language: Phonological variation In the Jamaican acrolect Journal ofPidgin and Creole Languages 19: 41-76. Irvine, A. 2005. Defining Good English in Jamaica: Language Variation and Language Ideology in an Agency of the Jamaican State. PhD dissertation, University of the West Indies Mona
Introduction Irvine, A. 2008. Contrast and convergence in Standard Jamaican English: The phonological architecture of the standard in an ideologically bidialectal community. World Englishes 27(1): 9-25.
Johnson, D.E. 2009. Getting off the GoldVarb standard: Introducing Rbrul for mixed-effects variable rule analysis. Language and Linguistics Compass 3( 1): 359-383. Kouwenberg, S. & Singler, J. V. 2008. Introduction. In The Handbook ofPidgin and Creole Studies, S. Kouwenberg & J.V. Singler (eds.), 1-16. Malden MA: IDack:well. Le Page, R.B. & Tabouret-Keller, A. 1985. Acts ofIdentity: Creole-Based Approaches to Language and Ethnicity. Cambridge: CUP. Mair, C. 1992. Problems in the compilation of a corpus of standard Caribbean English: A pilot study. In New Directions in English Language Corpora: Methodology, Results, Software Developments, G. Leitner (ed.), 75-96. Berlin: Mouton de Gruyter. Mair, C. 2003a Acts of identity. Interaction-based sociolinguistics and cultural studies: An introduction. AAA Arbeite11 aus Anglistik und Amerikanistik 28(2): 195-199. Mair, C. 2003b. Language, code, and symbol: The changing roles of Jamaican Creole in diaspora communities. AAA Arbeiten aus Anglistik und Amerikanistik 28(2): 231-248. Mair, C. 2009. Cmpus linguistics meets sociolinguistics: studying educated spoken usage in Jamaica on the basis of the International Cmpus of English. In World Englishes - Problems, Properties a11d Prospects: Selected Papers from the 13th IAWE Conforence, T. Hoffmann & L. Siebers (eds.), 39-60. Amsterdam: John Benjamins. Mendoza-Denton, N. 2008. Homegirls: Language a11d Cultural Practice among Latina Youth Gangs. Malden, MA: Blackwell Miihleisen, S. 2002. Creole Discourse: Exploring Prestige Formation and Change across Caribbean English-Lexicon Creoles. Amsterdam: John Benjamins. Miihleisen, S. & Walicek, D.E. (eds.). 2009. Sargasso- Joun1al of Caribbean Literature, Language and Culture 2008-2009. Special issue: Linguistic Explorations of Gender and Sexuality. Patrick. P.L. 2008. Pidgins, creoles, and variation. In The Handbook ofPidgin and Creole Studies, S. Kouwenberg & J. V. Singler (eds.), 461-487. Malden MA: Blackwell Rickford, J.R. 1987. Dimensions of a Creole Conti11uum: History, Texts & Li11guistic Analysis of Guyanese Creole. Palo Alto CA: Stanford University Press. Rickford, J.R 1988. Connections between sociolinguistics and pidgin-creole studies. lnterJ~a tionalJournal of the SociologyofLaJ~guage 71: 51-58. Rickford, J. R. (ed.). 1988. b1ternational Journal of the Sociology of Language 71. (Special issue: Socioli11guistics and Pidgb1-Creole Studie.s). Sebba, M. 1993. London Jamaican: Language Systems in b1teraction. London: Longman. Silverstein, M. 2003. Indexical order and the dialectics of sociolinguistic life. Language and Communication 23: 193-229. Tagliamonte, S. 2006. Analysing Sociolinguistic Variation. Cambridge: CUP. Trousdale, G. & Adger, D. (eds.). 2007. Journal ofEnglish Language and Linguistics 11(2). (Special issue: Theoretical Accounts of Dialect Variation). Zhang, Q. 2005. A Chinese yuppie in Beijing: Phonological variation and the construction of a new professional identity. Language in Society 34(3): 431-466. Zhang, Q. 2008. Rhotactzation and the 'Beijing Smooth Operator': The social meaning of a linguistic variable. Journal ofSociolinguistics 12(2): 201-222.
9
PART I
Variation and linguistic systems
Revisiting variation between sa and o in Sranan* Donald Winford Sranan employs two markers, sa and o, which both express some kind of futurity, but the precise nature of the difference between them has been difficult to pin down. Some researchers claim that sa and o are both primarily temporal in nature, but others treat sa as primarily modal, and o as primarily temporal, though they disagree on how far its meaning overlaps with that of o. This paper employs data from historical texts as well as examples and judgments elicited from five native speakers of Sranan in order to determine the precise nature of the differences in meaning and use between the two markers. I build on the insights of researchers like Voorhoeve and Seuren to demonstrate that sa really conveys a range of modal meanings that include futurity, strong expectation, and probability. This can be ascribed to the influence of Dutch zullen. On the other hand, o is primarily temporal in nature, expressing future tense. The analysis suggests that sa and o are not really in competition in contemporary Sranan.
1.
Introduction
Sranan employs two markers, sa and o, which both express some kind of futurity, as illustrated in the following example:
pramisi yu taki mi o/sa kon tamara. lsg IMPF promise 2sg caMP lsg FUT/MOD come tomorrow 'I promise you I'll come tomorrow:
(1) Mi e
*
The research on which this paper Is based was funded by NSF Grants #SBR-930636 and #BCS-0103826, for which I wish to express my great appreciation. I also wish to thank our consultants in Suriname, especially Robby Morroy, Hertoch Linger, Margie McBean, Renata de Bies, and Yolanda Boldewyn for spending so many long hours patiently explaining the uses ofsa and other TMA markers. Special thanks are due to my co-PI, Bettina Mlgge, for her detailed work on Pam.aka and Gbe and for providing data and commentary on sa in Pamaka My thanks also to the two anonymous referees whose comments helped improve this paper. All shortcomings that remain are my responsibility.
14
Donald Winford
However, there has been a great deal of controversy over the precise meanings and uses of the two auxiliaries. In particular, researchers disagree about whether sa and o are in competition with each other or convey quite different meanings. Earlier views on this issue fall into two categories. First, there are studies such as Simons (1954) and Donicie (1954, 1956), which claim that sa and o are both primarily ternporal in nature, but there are slight differences due to the kind of attitude speakers have toward the event. Second, there are studies such as Voorhoeve (1957, 1962), Seuren (1981), and Wendelaar and Koefoed (1988) which treat sa. as primarily modal and o as primarily temporal, though they disagree on how the meaning of sa overlaps with that of o. Despite their different interpretations, all of these researchers suggest that sa and o compete with each other for the expression of futurity. This competition, however, is not a case of variation between two variants that express precisely the same meaning, so it does not fall under the ambit of the sociolinguistic variable as it has traditionally been described. So far, no one has oftered a complete and plausible explanation of the precise relationship between sa and o. In this paper, I re-examine this issue by analyzing uses of sa and o in contemporary Sranan, based on elicitations and recordings made with native speakers during 2002. I argue that the conflicting views on this issue can be reconciled if we take into account the different histories of sa and o in Sranan and the differences in the ways they are used in particular discourse contexts. In Section 2, I first draw on evidence from the early Sranan texts to trace the historical developments that seem to have led to the difterences between the two auxiliaries. I examine the uses of sa and o in earlier Sranan and argue that sa was established as the marker of futurity before o emerged as its competitor in the early 18th century, leading to competition between the two. In Section 3, I examine and critique previous analyses of the meanings and uses of sa and o in modern Sranan. Then, in Section 4, I present my own analysis of the relationship between the two auxiliaries. Based on evidence from elicited data, I argue that sa is in fact a modal that conveys strong expectation or probability, while o is primarily temporal in nature, expressing future tense. I also suggest that there has been continued and growing influence of Dutch zullen on the interpretation of sa in contemporary Sranan, which has led to further differentiation between sa and o.
tar
2.
Historical excursus
The variation and apparent competition between sa and o in contemporary Sranan has its roots in the historical past, particularly in the course of the 18th century, when the TMA system of early Sranan was still undergoing development. The earliest known example of a future marker in early Sranan is wil (< English will?),
Revisiting variation between sa and o ln Sranan
which occurs in the Herlein fragment of 1718 (Van den Berg 2007: 190). However, in court records of 1745 (Van den Berg 2000: 91), the attested future marker is sal, which appears to have its source in Dutch za.l (see below). In VanDyk (1765), zal (also spelled zel, za., sa.l, sa, sa.a, etc.), appears to be the sole auxiliary used to express futurity (Van den Berg 2007: 188). The following examples illustrate this point. (2) Pikien morre wi za drinki koffi. little more we fut drink coffee 'We'll drink coffee in a little while: (VanDyk c1765, in Arends and Perl1995: 150) (3) Mi fon zomma. morre den za. ziki kwetti. 1sg beat person more 3pl FUT sick quite 'If I beat them more, they will get really ill: (VanDyk, in Arends and Perl1995: 176)
In these early texts, za.l, now increasingly spelt sa, is also used in a more modal sense, for instance to convey a hypothetical meaning, as in the following example.
mi sa jerri? Mi jerri wan gon piki toe trom. (4) No, hoe fa no, how manner 1sg FUT hear 1sg hear one gun speak two time 'No, how should I have heard? I heard a gun shoot twice~ (VanDyk, cited in Arends and Perl1995: 205) These uses of sa are confirmed for later 18th century Sranan by Schumann (1783: 146), who analyzes the auxiliary as follows: Sa: will (future tense, i.e., to point something out that is in the future; not: become or develop); also: shall [sollen]. (Van den Berg 2007: 188 [her translation])
In VanDyk (1765), however, we find evidence that sa. has a new competitor, namely the combination progressive de +go, which appears to express futurity in some contexts. (5) Zili de go didde. Anno gray hessi ba.kke. seed PROG go die it-NEG grow hastily back 'The seeds will die and won't sprout that easily again: (VanDyk, cited in Arends and Perl1995: 210) Schumann (1783: 50) also points out that "very frequently, go is used to indicate future tense instead of sa" (Van den Berg 2007: 191) [her translation]), as in the following example. (6)
Dem go
killi mi.
3pl FUT? kill 1sg 'They'll kill me:
(Schumann 1783: 50)
15
16
Donald Winford
Most of Schumann's examples, however, involve the combination of de +go, as in: (7) Mi de go jam. 1sg PROG go eat 'I'm going to eat:
(Schumann 1783: 50)
Texts of the late 18th century such as Weygandt (1798) provide further evidence of these uses of de +go. Van den Berg (2007: 194) points out that go in most of these cases may still be interpreted as a main verb, hence the combination de +go is not necessarily grammaticalized as a marker of futurity at this time. Over the course of the 19th to mid-20th centuries, de + go appears to have undergone a gradual process of grammaticalization, evolving into the combination of progressive e + go, which in turn eventually yielded the future tense auxiliary o. Nickel and Wilner (1984: 58, note 18) observed that "current speakers of Sranan Tango understand o as e +go." On the other hand, Seuren (1983: 224) claims that use of (main verb) go to express futurity "reflects an older stage of the language... [and is] virtually reduced, nowadays, to the religious services of the Moravian Church:' This brief historical overview will have to suffice for our present purposes, though it would be useful for future research to trace the evolution of future o and its growing competition with sa during the 19th and earlier 20th centuries. This would require a great deal of exhaustive historical research, which is beyond the scope of the present paper. At any rate, as late as the mid-20th century, various researchers still viewed sa and o as being in competition for the expression of futurity. Voorhoeve (1957: 379) claimed that futurity could still be expressed either bye +go or by sa. Voorhoeve and Kramp (1982: 8) suggested that "the modal sa may be in the process of being replaced by the purely tense marker o~ The following section examines these earlier views on the competition between the two auxiliaries.
3·
Previous analyses of sa versus o
All previous studies of Sranan agree that o functions as a marker of future tense. They differ in their view of whether sa. is primarily modal or not. Simons ( 1954) suggests that sa and o are both temporal in nature, but that o is used for an immediate future, while sa is used for other kinds of futurity. Donicie (1956) also treats both as temporal markers, but argues that o is used for events that are mentally closer. Other researchers distinguish more clearly between the two. Voorhoeve (1957) was among the first to draw a clear distinction between sa and e +go, which was apparently the common form of the marker of futurity at the time.
Revisiting variation between sa and o ln Sranan
He claims that "it is possible to render the occidental future in Sranan by means of sa + verb and by means of e -go + verb," but the latter "is a verbal group consisting of the auxiliary e plus main verb go, and hence cannot be described as a category within the system of verbal forms (formed by prefix+ verb)" (1957: 379).1 While acknowledging that his informants "were originally of the opinion that the sa- forms and the verbal groupe- go+ verb might be used interchangeably with only a slight difference in meaning (p. 380), he found several instances where the two markers could not be interchanged. He ascribes this to the fact that sa. has a "hypothetical" character not found in e-go, and describes the various discourse uses of sa as follows. First, there are cases where sa is used to express "a wish, intention, or expectation of the subject" (p. 380), as in the following examples (glosses are mine; free translations are Voorhoeve's).2 (8) Mi sa teki Streeteresser limonade. (In reply to "What will you drink?") 1sg MOD take Streeteresserlemonade 'I'll take Steeteresser lemonade: (9) Mi sa. taki nang a bakra. 1sg MOD talk with DET Whiteman 'I shall talk with the white man: In the above cases, Voorhoeve (p. 381) notes that sa might be interpreted as conveying "future tense," but goes on to illustrate other uses of"sa-forms which are impossible to consider as indicative of the future." These uses include the following. To convey a "dubitative" sense (i.e., of incredulity) in certain questions expressing indignation on the part of the speaker. (10) I sa go taigi a man dati? 2sg MOD go tell the man that 'You've really been so stupid as to tell this to that person?' To expresses an "optative" sense, as with the verb winsi 'wish: (11) Mi sa winsi a e kon tamam. 1sg MOD wish 3sg IMPF come tomorrow 'I only hope he comes tomorrow: To express the sense of can, that is, a "potential" sense.
Note that Donicie (1956: 183) represents the future marker as (de) go, indicating that de was optional.
1.
2. Though Voorhoeve doeslft specify which sentence expresses a wish or which an intention, I assume that sentence (8) expresses a wish, while (9) expresses an intention.
17
18
Donald Winford
Here, Voorhoeve simply repeats a claim by Donicie (1956: 186), who compares this use of sa with the use of sa 'can' in Saamaka, but gives no examples. I have found no examples of this use of sa in contemporary SN, though it is quite common in the Maroon creoles (Migge and Winford 2009). Voorhoeve sums up his account by noting that "our future tense is actually reproduced occasionally by means of the verbal form prefixed by sa-, but more often by the verbal group go +verb (1957: 382). However, he continues, it is impossible to replace the verbal form with the prefix sa- by means of the verbal group go + verb when the action is seen as hypothetical. The verbal form prefixed by sa- is used moreover as a dubitative, a potential, an optative, and an irrealis. (1957: 382-383)
For this reason, Voorhoeve describes sa. "not as indicative of the future, but as indicative of the modus non-realis" (p. 383). Voorhoeve (1962) very briefly summarizes the earlier analysis, describing sa as conveying "irrealis," and noting the form with prefix sa- usually indicates the future, as does the auxiliary go with main verb. However, the verbal form sa- with verb has more possibilities, which cannot all be interpreted as indicative of the future. Between sa- + verb and e-go or o + verb there is only a slight difference in meaning, the sa- + verb form having a more hypothetic (sic) character. (1962: 40)
On the whole, Voorhoeve's account still falls short of explaining the precise nature of the similarities between sa and o (which allow them to alternate in some cases), or the differences (which prevent them from doing so). Moreover, while he offers some good discussion of some of the discourse uses of sa, his characterization of its modal meaning as "irrealis" is too broad to explain either its basic sense or its uses in discourse. Other markers such as lean 'can: musu 'must: and indeed future o itself also express "irrealis" meaning, so the precise nature of the meaning of sa is left unclear. As we shall see, some of the uses of sa described by Voorhoeve can be found in contemporary Sranan. I will discuss these in more detail, and with more illustration than Voorhoeve supplies. I will also offer an alternative analysis of the modal meanings of sa. that explains these uses in a principled way. There are also some noticeable differences between Voorhoeve's account of the use of sa in the more conservative varieties of Sranan of the 1950s and my own account of its uses in contemporary Sranan. In particular, there is no evidence in my data for the so-called "potential" use of sa. to express the sense of'can: Moreover, Voorhoeve says nothing about the deontic and epistemic uses of sa that I discuss later. Seuren (1981) also attempts to distinguish between the meanings of sa and o, describing the former as a modal auxiliary and the latter as a future tense marker. This accords with my own analysis. His analysis in fact contains more insight than
Revisiting variation between sa and o in Sranan
Voorhoeve's. Like Voorhoeve, Seuren points out that sa has a variety of functions, including that of indicating "future or temporal posteriority relative to a defined moment in time" (1981: 1054). Seuren acknowledges that he has "found it very difficult to formulate generalizations concerning the correct temporal usage of sa and o" (ibid.). However, he offers the suggestion that
sa is the proper form for neutral predictions, but also for future events or situations resulting from somebody's (the speaker's) insistence, order, wish or promise. 0 on the other hand, indicates a future event or situation resulting from some pre-established plan or from natural causes already present at the time of speaking. (1981: 1054) Seuren's view of the temporal uses of sa closely resembles that of Voorhoeve, though, unlike Voorhoeve, he provides no examples. As we will see, this account of the "ternporal" uses of sa. agrees well with my own analysis of the use of sa to convey what I shall refer to as dynamic modality. On the other hand, I analyze o as a marker of future tense, which can convey prediction, intention, or simply future time reference. In addition, Seuren notes that sa has modal functions that "are more salient than the temporal meanings" (ibid.). He describes these functions as follows, providing the respective examples (glosses are mine, translations are Seuren's except where indicated). To indicate the speaker's wish or order. (12)
Yu sa
go.
2sg MOD go 'You shall go~ To indicate the speaker's hypothesis. (13) A sa ben dringi doti watm. 3sg MOD PAST drink dirty water 'He will/must have drunk dirty water: To "express the speaker's disbelief or doubt as to the truth of what is said in the following infinitive.: In this case, according to Seuren, sa "tends to occur in past tense, ben sa, and does what the subjunctive does in so many other languages" (ibid.).
ben du en. (14) A taki dati mi ben sa 3sg say that 1sg PAST MOD PAST do it. 'He said that I <subjunctive> had done iC ['He said that I might/must have done it: - DW]
19
20
Donald Winford
Seuren's account of these modal uses of sa accords more closely with my own analysis than Voorhoeve's does. I argue below that what Seuren refers to as the use of sa to indicate the speaker's wish or order (a) represents the deontic use of sa.. And what he refers to as the use of sa to express the speaker's hypothesis (b) and the speaker's doubt as to the truth of a proposition (c) represent the epistemic uses of sa. Wendelaar and Koefoed (1988) argue that sa is competing with o in two senses: first, the two are mutually substitutable, with slightly different meanings; second, o is gradually replacing sa. in certain contexts. They attempt to distinguish sa from o in terms of two parameters: the degree of certainty of the event and whether the event in question is "realizable" or not. They define a realizable future event as one that can be realized without intervention from other agents or factors, by which they appear to mean an event under the control of the speaker. They provide the following example (p. 68) (glosses are mine, translations by Marinus Stephan). 3 (15) Tamara mi o/sa bai a buku gi ju. tomorrow 1sg PUT/MOD buy DET book for you 'Tomorrow I will/shall buy the book for you.' A non-realizable future event, on the other hand, is one that is either literally impossible, or can only be realized by intervention of other factors, such as a third party or a factor beyond the speaker's control (p. 68). The following example illustrates.
kon. (16) Tamen·a. alen o!sa tomorrow rain PUT/MOD come 'It will/shall rain tomorrow.' Another factor which they claim influences the choice of sa versus o is the speaker's degree of certainty about the occurrence of the future event. Combining the factors of realizability and certainty yields complex future meanings like the following. Realizable but either certain or uncertain. (17) Mi o/sa yepi yu. 1sg PUT/MOD help you 'I wilV shall help you.'
3· I wish to express my thanks to Marinus Stephan for translating Wendelaar and Koefoed's paper, including the examples, from Dutch into English for me. I modified his English translations of the sentences so as to make a distinction between 'will' and 'shall: corresponding to o and sa respectively.
Revisiting variation between sa and o ln Sranan
Non-realizable, but either certain or uncertain. (18)
Yere, a
o/sa
ferteri wan tori.
hear, 3sg FUT/MOD tell a story. 'Listen, he will/shall tell a story.' Wendelaar and Koefoed presented a number of test sentences in both Dutch and Sranan to two informants, asking them to indicate whether o or sa would be appropriate to use in each case. Their goal was to test four hypotheses, as follows (1988: 69). la. lb. 2a. 2b.
o is used for the (relatively) certain realizable future. o is used for the (absolute) certain non-realizable future. sa is used for the uncertain (doubtful) realizable future. sa is used for the neutral and uncertain non-realizable future.
Their results show that prediction 1b is confirmed without exception, while hypothesis 2b is confirmed in 20 of 22 instances. However, they found that informants chose sa 12 times out of a possible 46 instances in the case of hypothesis 1a (where o was predicted), and chose o 9 times out of a possible 25 instances for hypothesis 2a (where sa was predicted). Clearly, these results fail to identify precisely why speakers chose sa or o. Despite the inconsistency of their results, Wendelaar and Koefoed claim that their hypothesis concerning the differences in meaning between sa and o have been fairly well confirmed. They suggest that future research may find answers to the exceptions to their hypothesis. As we will see, several aspects of their findings are borne out by my own data. For instance, they are certainly right in suggesting that degree of certainty and degree of control by the agent (not necessarily the speaker) are important factors in determining which marker is chosen. They are also right that various dimensions of meaning need to be taken into account if we are to explain the subtleties of the uses of sa and o, especially the former. One of the major problems confronting us is to determine what is "realizable" or "uncertain" from the perspective of the speaker. This is not something that can be determined a priori but requires input and intuitions from native speakers. Such intuitions, as we shall see, often depend on the discourse context, which typically determines whether sa or o can be used and, if so, in what interpretations. In the following section, I introduce the methodology that was employed in obtaining data and intuitive judgments from my informants, as well as the framework I used for the analysis of the meanings and uses of sa and o.
:n
22
Donald Winford
4· Methodology
The different analyses of sa and o that have been offered in the literature reflect the fact that these auxiliaries can convey a fairly broad range ofinterpretations in context This is particularly true of sa, a fact that makes its core meaning quite difficult to pin down. If we are to clarify its relationship to o, we need a more explicit descriptive framework as well as an appropriate methodology for collecting and analyzing data. It should be obvious by now that the alternation between sa and o in modern Sranan does not lend itself to the traditional tools of quantitative sociolinguistic analysis. We are not dealing here with variation between two variants that express the same meaning, but rather with a choice between two auxiliaries that both express some kind of futurity but can still convey quite different interpretations depending on the context in which they are used. The framework used here is modeled after Dahl's (1985) typological study of TMA systems in a broad sample of the world's languages. A key assumption of that framework, which is most relevant to the present study, is that every TMA category has a dominant meaning and often has other secondary meanings as well. In general, the dominant meaning of a category is represented in its primary or prototypical uses, while secondary meanings are contextually-determined interpretations that arise from more peripheral uses of the category. A full analysis therefore requires emphasis on the crucial role of the discourse context in deciding the interpretation ofTMA categories. With this in mind, I employed methods of data collection designed to capture as much discourse context as possible, so as to determine how each auxiliary could be used. The data for this study come from elicitations conducted with five native speakers ofSranan Tongo in Paramaribo in the summer of 2002. They included one male and one female over the age of fifty, and 3 informants - two females and one male - under the age of forty. The questionnaire used for the elicitations was a modified version of Dahfs (1985) questionnaire. It consisted of a number of sentences and short connected texts in English, which were offered for translation by informants. They were given clear indications, with additional explanation where necessary, of the contexts in which they were to envisage the sentences being uttered. For example, to elicit a verb with future time reference, the prompts illustrated in (19) to (21) were among those used. (19) [Q: What are you planning to do right now?] I WRITE letters. (20) [The boy is expecting a sum of money] When the boy GET the money, he BUY a present for the girl. (21) [The boy thinks that he will perhaps get a sum of money] If the boy GET the money, he BUY a present for the girl.
Revisiting variation between sa and o ln Sranan
Material enclosed in square brackets is meant to provide a context for the utterance to be elicited (the translation of which is outside the brackets). Verbs are offered in bare form (capitalized in the text) so as to minimize the possibility of interference from English when translating. The elicitations were conducted with a view of obtaining not just translations of the prompt sentences, but more importantly the informants' judgments as to whether sa or o would be more acceptable in particular contexts. In cases where both were deemed acceptable, informants were asked to explain the degree of similarity or difference in the meaning they conveyed. Finally, the informants were invited to create their own discourse contexts to illustrate their judgments of the appropriate uses of the two auxiliaries. In the next section, I will examine the various discourse contexts in which sa. and o are used, and attempt to clarify the degree to which they are substitutable in such contexts.
5·
The uses of sa and o in contemporary Sranan
In this section, I put to the test previous views, including those of Voorhoeve, Seuren, and Wendelaar and Koefoed, that certain contexts allow alternation between sa and o, and that sa has a kind of "temporal" meaning in these cases. I begin the analysis by considering my informants' judgments concerning when o is appropriate to use, but not sa. First, there are situations in which the speaker is the agent of the future action, such as the following test sentence from Dahl's questionnaire: (22) [Mother to child:] If you not STOP PLAY with that ball I TAKE it away.
Efu yu no tapufu prei nanga a ba.l, if 2sg NEG stop for play with DET ball mi o (*sa) puru en na yu anu. 1sg FUT pull it LO c 2sg hand 'If you don't stop playing with that ball, I'll take it away from you' But there are also cases in which the speaker is not the agent of the future action, such as in the following test sentence, adapted from Dahl. (23) [Q: What HAPPEN if you PUT a stone in this bag?] It TEAR.
Efu yu poti wa.n ston ini a saka. disi, a saka o (*sa) priti. if 2sg put ART stone in DET bag DEM, DET bag FUT tear 'If you put a stone in this bag, it will tear=
23
24
Donald Winford
There were four test sentences of this type, containing open conditional clauses, and informants were unanimous that o was more appropriate than sa in all cases. One informant commented that use of sa would "sound stiff, like a foreigner trying to speak Sranan." The general consensus among informants was that, in the above cases, o conveys a sense of relative certainty on the speaker's part. Sa is pragmatically inappropriate in these cases, since it implies uncertainty, or perhaps more accurately, speculation. Note that it matters little whether the future action is in the control of the speaker, so Wendelaar and Koefoed's distinction between "realizable" and "non-realizable" future does not apply here. It is clear from examples like these that o is a marker of future tense, which conveys a strong sense of certainty and prediction. On the other hand, sa lends itself to a fairly broad range of interpretations, which makes it very difficult to pinpoint its core meaning. Despite this, there are several contexts in which alternation between sa and o is possible. In some cases, my informants perceive subtle differences between the two, while in other contexts the distinction is neutralized to some extent. There were a number of test sentences in which all informants agreed that either sa or o could be used, though their meanings differed slightly. These include cases where some uncertainty is already expressed in the sentence, for example by use of the adverb ka.nde 'perhaps~ (24) Kande mi sa/o kon tide neti, efu mi abi ten. perhaps lsg MOD/FUT come today night if lsg have time 'Maybe I shalVwill come tonight, ifi have time~ One informant noted that sa is used in a case like this when you're not really willing, or when youCl do something reluctantly. Another noted that only elderly people tend to use sa in such sentences. It would seem then that informants still recognize a subtle difference in meaning between o and sa, which is almost neutralized in contexts like these. This is also true of questions, where sa and o alternate with subtly different meanings. For instance, one informant notes a contrast between the following requests. (25) a.
Yu o man leni mi moni? 2sg FUT can lend 1sg money 'Will you be able to lend me some money?'
b. Yu sa man leni mi moni? 2sg MOD can lend lsg money 'Would you be able to lend me some money?' This informant notes that he wouldn't use sa, but other people use it. For him, sa implies that you don't need the money. This may be due to the fact that sa has a
Revisiting variation between sa and o ln Sranan
more speculative character than o. This may also explain why informants generally thought that sa was more appropriate than o when asking questions, as in the following example:
mi? (26) Efu mi aksi yu wan sa.ni, yu sa. piki if lsg ask 2sg a thing 2sg MOD answer me 'Ifl ask you something, would you answer me?' This brings us to other cases where sa and o alternate, but informants perceive a sharper distinction between them both in meaning and in the kind of discourse context they require. Let us first consider cases where the speaker is the agent of the future action, as in the following. (27)
[Q: What are you planning to do right now?] I WRITE letters.
Mi o/sa
skrifi brifi.
lsg PUT/MOD write letter 'I will/shall write letters: The first choice of all informants in this case was o, which all agreed conveys a sense ofintention on the part of the speaker. They also unanimously agreed that sa could only be used to indicate the sense that the speaker has nothing else to do, so that writing letters is like a last resort. For them, sa conveys the sense that one might as well do something, because one has no other option. This interpretation ofsa seems to reflect one of the most common uses of this auxiliary to indicate a choice or option, and will be discussed further below. Note that this interpretation differs from both Seuren's and Voorhoeve's view that sa expresses simple intention. In cases where someone other than the speaker is the agent of the future action, informants have somewhat different intuitions regarding the interpretation of sa, though they agree on o. Consider the following. (28)
[The boy is expecting a sum of money.] When the boy GET the money, he BUY a present for the girl.
Te
a
boi kisi a moni, a o/sa bai when DET boy get DET money 3sg FUT/MOD buy wa.n kado gi a umapikin. ART present for DET woman-child 'When the boy gets the money, he wilVshall buy a present for the girl: Informants agreed without exception that o is used here to indicate the speaker's certainty about the future event. But they also judged sa. to be appropriate, though in somewhat different interpretations. Three informants (two older and one
25
26
Donald Winford
younger) indicated that sa would be appropriate in a context like the following, where the boy is not certain to get the money. (29) Efu a hoi kisi a moni a sa bai wan kado if DET boy get DET money 3sg MOD buy ART present
gi a
umapikin.
for DET girl 'If the boy gets the money, he shall buy a present for the gire For these informants, sa conveys a degree of uncertainty, in keeping with the fact that the boy is not sure to get the money. The two other younger informants suggested that sa is appropriate in a context like the following. (30) [Context: It's his girlfriend's birthday, but the boy hasn't bought a present for her. The speaker thinks he should have bought a present but knows he didn't have the money.]
abi
mon~
A
boi no
DET
boy NEG have money but
a
sa
bai wan kado
te
ma
a
kisimoni
when 3sg get money
gi a
umapikin.
3sg MOD buy ART present for DET girl 'The boy doesn't have money, but when he gets money, he'll buy a present for the gire In this case, sa conveys the speaker's strong expectation that the boy will buy the present As one informant puts it, "sa here is like shall, it implies that I shouldn't worry, that the boy will be able to buy the present=' There seems to be some difference between the intuitions of these two groups of informants. A similar difference in judgment emerges from their comments on the use of sa and o in sentences like the following. (31) [What your brother DO if you don't got to see him tonight, do you think?] He COME tomorrow (to see me).
A
o/sa
kon tamara
fu luku mi.
3sg FUT/MOD come tomorrow for see me 'He will/shall come tomorrow to see me~ Again, all informants agreed that both o and sa could be used in this case, and that o conveys certainty that the future event will happen. But again, the first group indicates that use of sa conveys the sense that the speaker is not certain that the brother will come. On the other hand, the other two younger informants again interpret sa as expressing a sense of strong expectation. As one of them puts it, sa
Revisiting variation between sa and o ln Sranan
indicates that "he must come, because one way or another, we must see each other=' There is no necessary incompatibility between the different judgments of the informants with regard to the way sa can be interpreted in these contexts. I will argue below that both the sense of uncertainty and the sense of expectation that informants perceive in sa are related to the fact that this modal functions as a marker of probability, and has shades of both deontic and epistemic meaning. 4 Finally, my informants also confirm Voorhoeve's (1957: 382) account of the use of sa to express a sense of indignation and incredulity (which he refers to as a "dubitative" sense), as in the following example. It would not be appropriate to use o in this case. (32)
[Context: someone tells you he thinks you told lies about him.]
Mi sa forteri tori fu yu? 1sg MOD tell storyofyou 'Would I tell lies about you?' As Robby Morray (pers. comm., 2002) notes, this implies "I would never do such a thing. How could you think I would do such a thing?" Both sa and o can combine with the past marker ben, producing the complex forms ben sa and bo (< bio < ben o), which both convey the sense of hypothetical or counterfactual past or present Again, there have been somewhat different views expressed about these combinations. Voorhoeve (1957: 376) points out, rather succinctly, that "the verbal form with ben sa indicates a past future," and also that it "generally indicates an irrealis" (p. 382), providing the following examples. (I have amended his examples. His glosses follow mine. The free translations are mine.)
mi ben sa aksi en dati. (33) Efu a ben kon esde, If 3sg PAST come yesterday, lsg PAST MOD ask 3sg that if he been-come yesterday, I would-have-asked him that 'If he had come yesterday, I would have asked him that: (34) Efu oto no ben de, mi ben sa e waka if car NEG PAST COP, 1sg PAST MOD IMPF walk
te mi no abi futu. until1sg NEG have foot If automobiles not existed, I would-walk until I not have feet 'If cars did not exist, I would be walking until I had no feet:
4 Still, it is possible that the differences ln judgment reflect the fact that older speakers tend to use sa to express more uncertainty, whlle younger ones use it to express stronger expectation. It is worth noting that the younger informant who expressed judgments simllar to those of the two older informants learned Sranan from her grandmother.
27
28
Donald Winford
Voorhoeve is correct to note that ben sa can express both present and past counterfactual meaning. However, he says nothing about ben o. Seuren does distinguish between ben sa and ben o, but does not elaborate much on their meanings, except to say that ben sa "does what the subjunctive does in so many languages (1981: 1054). He also claims. without further elaboration, that ben sa "occurs regularly in the main clause of past counterfactuals," while ben o is preferred in present counterfactuals (p. 1055). He provides the following examples. (Glosses are mine; translations are his, though fve amended the second one slightly.) (35) Efu a ben wroko mom tranga, a ben sa kisi mom moni. if 3sg PAST work more hard, 3sg PAST MOD get more money 'If he had worked harder, he would have earned more money;' (36) Efu a man ben o meki bari no~ a fowruben o freygwe. if DET man PAST PUT make shout now, DET bird PAST PUT fly go-away 'If the man were shouting now, the bird would fly away;' ['If the man were to shout now - DW] In fact, as we will see, ben sa and ben o (bo) can both occur in past and present counterfactuals, but with slightly different meanings. Wendelaar and Koefoed, for their part, treat the distinction between ben sa and ben o as identical to the one they propose for sa and o, that is, as a distinction between uncertain and certain future. Their argument therefore suffurs from the same flaws as their analysis of sa and o. It seems fair to say; therefore, that none of the earlier studies has provided a clear or adequate explanation for the difference between ben sa and bo. I will argue here that the distinction between ben sa. and bo parallels the distinction I have made between sa and o. Informants were presented with six test sentences involving counterfactual conditions, and asked to determine whether bo or ben sa would be more appropriate in each case. Three sentences involved counterfactual situations in the present, such as the following. (37) Efu mi ben abi moni mi bo!ben sa bai wa.n wagi. if 1sg PAST have money 1sg PAST.PUT/PAST MOD buy ART car 'If I had money, I would buy a car: The other three sentences involved counterfactual situations in the past, such as: (38) Efu mi ben sabi taki a abi yepi fanowdu, if 1sg PAST know COMP 3sg have help need mi bo!ben sa yepi en. 1sg PAST.PUT/PAST MOD help 3sg 'If fd known she needed help, I would have helped her:
Revisiting variation between sa and o ln Sranan
All but one informant expressed a clear preference for bo in all six sentences. The exception was an older female who preferred ben sa in all but 2 cases. The four informants who accepted both ben sa and bo, however, offered somewhat different views of how they differed in meaning. One informant. a younger female, saw little difference between the two, but claimed that elderly people were more likely to use ben sa, and that she wouldn't use it herself. For a younger male informant, ben sa expresses more of a wish in some contexts, such as (37) above, but is totally unacceptable in some contexts, for instance with verbs like abi 'have' and 'sabi' know (thus *ben sa abi/sabi). In other cases where the speaker is the agent of the hypothetical act, he accepts both ben sa and bo, but notes that the former expresses a sense of certainty, while the latter suggests a duty that must be carried out, perhaps with reluctance (the speaker is not sure s/he would have done it). Two other informants agree with these judgments. The following example illustrates this further. (39) Efu mi ben sabi taki a de dya, if lsg PAST know caMP 3sg be here
mi bo!ben sa
go luku en.
lsg PAST.FUT/PAST MOD go see 3sg 'Iffd known she was here, I would have gone to see hd As Robby Morray notes, ben sa in this case implies "I have the option to go, I decide." He compares the use of bo vs. ben sa in the context of having to go to two funerals, one willingly (a good friend has died) where bois appropriate (since it expresses more certainty), and one reluctantlywhere ben sa is appropriate (because it implies a social obligation). This distinction is quite similar to that between o and sa, as described earlier by the same informants. In general, however, the trend in modern Sranan is that the distinction between ben sa. and bo seems to be neutralized in most contexts. As Robby Morray (pers. comm., 2002) puts it, "there is very little difference between bo and ben sa, so people tend to use bo all the time. Sa conveys a more subtle meaning, so in generalit appears less frequently. Older people use (ben) sa more often, and distinguish it from (ben) o." Similar opinions were expressed by all other informants. However, there still remains a fairly clear distinction between sa and o, as we saw above. In the following section, I explore the meaning and uses of sa in more detail to clarify how and why it differs from o.
6.
The meanings and uses of sa in Sranan
The discussion so far has provided further evidence to support the views of previous researchers that sa is used to express various modal meanings that cannot be
29
30
Donald Winford
expressed by future o. The uses of sa surveyed in the previous section may all be interpreted as cases of dynamic modality and represent the most frequent use of this modal in my data, which is to express some kind of futurity or perhaps more accurately the speaker's expectation that some event will occur. 5 In addition to this, sa can also express epistemic and deontic modality. which might be regarded as secondary uses. Neither of these uses is possible with future o. In all cases, the apparent ditrerences in meaning expressed by sa are contextually determined. In the following subsections, I will first provide support for these interpretations of sa and then argue that they are influenced by Dutch zullen. 6.1
Cases where sa expresses dynamic modality
In its most typical use, sa expresses what has been referred to as dynamic modality. which Perkins (1982: 252) explains "is concerned with the disposition of certain empirical circumstances with regard to the occurrence of some event' Use of sa in this sense can yield interpretations such as commitment on the part of the speaker as agent of the action or the speaker's expectations concerning actions for which s/ he is not the agent. My analysis here is based primarily on the detailed commentary of my informants on what they perceive to be the meaning of sa in various contexts which they themselves created in order to explain their intuitions. First, let us consider in more detail situations in which the speaker is the agent of the future action. 6.1.1 Cases where sa is used to express a future act of the speaker In this use, as Voorhoeve (1957: 380) pointed out, sa typically expresses a wish, intention, or expectation of the speaker. The common element in all these interpretations is that the speaker exercises some choice or option available to him. This may have the force of a promise, as the following examples show.
(40) [Context: You owe someone money; and she demands payment.]
Mi sa gi yu a sani, yere. lsg MOD give you ART thing hear 'I shall give you the money, okay?' (41) [Context: I'm renting your house and you want me to leave. After repeated requests, I finally decide:]
Te mi kisi a moni na a kba fu a mun, when 1sg get ART money we ART end of ART month 5· It's not dear whether the use of sa to express incredulity and indignation can best be described as a case of dynamic modality. It is certain! y neither epistemic nor deontic. The sense of incredulity/indignation is really a product of the discourse context, but its basic sense seems to be dynamic.
Revisiting variation between sa and o ln Sranan
mi sa yuru 1van tra oso. lsg MOD rent ART other house 'When I get the money at the end of the month, I shall rent another house: In other contexts, sa can convey the sense of a threat.
moni, mi sa Jon yu. (42) Efu yu no gi mi a if 2sg NEG give lsg DET money lsg MOD beat 2sg 'If you don't give me the money, I shall beat you: In other cases, sa conveys the sense that the choice of an option is forced on the speaker by circumstances. This may involve a duty that must be fulfilled, as in the following example, provided by an informant with his commentary. (43)
[Context: A relative you never liked dies. He had stolen money from you. You decide:]
Mi sa go beri mi omu. lsg MOD go bury lsg uncle 'I shall attend my uncle's funeral: Informant's commentary: "This suggests I've been playing with the idea, and I come to the conclusion that although I don't want to, it's my decision. Sa implies it's not my choice, and indicates doubt. I'm not sure I should, but I will (reluctantly). It would be reasonable not to go (since he stole from me), but out of moral obligation I should go:' (Robby Morray, pers. comm., 2002) A similar example is provided by another informant. (44)
[Context: I keep telephoning the city authorities asking them to fix the road, but they keep ignoring me. I decide:]
Mi no o bel moro. Mi sa skr~(i. wan brifi. lsg NEG PUT call more lsg MOD write ART letter 'I won't call anymore. I shall write a letter: Informant's commentary: "Sa here is stronger than o. Because of the response I got, I have no choice but to write. There is no other option=' Informants generally agreed that sa expresses a sense of reluctant choice in these cases. 6.1.2
Cases where someone other than the speaker is the agent of the future action
In cases where the speaker refers to an action or situation under the control of another agent, he indicates an expectation based on his judgment of the circumstances. Use of sa in these cases conveys a sense of strong expectation that the event will occur.
31
32
Donald Winford
(45) [Context: John's father has died in Holland, and John is in Suriname. He hates flying, but still:]
Fu di na en p'pa dede, a sa go na p'tata. for that Foe his father die 3sg MOD go we potato 'Since it's his father that died, he shall go to Holland.' As Robby Morray (pers. comm., 2002) puts it: "Sa does not convey possibility here. It implies he has other options, and he concluded that going to Holland was the best choice." The above examples clearly illustrate that sa does not primarily convey a sense of uncertainty, as Wendelaar and Koefoed (1988) suggested. Rather, the views expressed by Voorhoeve and Seuren are closer to the mark. In all these cases, the sense conveyed is rather that of strong likelihood or expectation, based on the speaker's assessment of available options or choices available to the agent involved. This of course involves a certain degree of doubt, which is what lends sa. its uncertain or speculative character. This sense of weighing options can also be observed in cases where sa is used in a more epistemic sense to express probability, as discussed in the following section. 6.2
Epistemic uses of sa
Sa can also be used in epistemic senses which cannot be conveyed by o, particularly when it occurs with imperfective e or the copula de. (46) Den pikin sa (*o)e sribi now(naw). the.pl child MOD IMPFV sleep now 'The children should be sleeping now.'
In this use, sa conveys the sense that the speaker has made an inference or deduction based on the evidence available to him. This is in tact quite close to the epistemic use of English must, that is, the sense of probability. Informants were quite clear that sa does not convey any sense of futurity in these uses, but refers only to present probability. It would be impossible to use o in this sense. Consider also the following.
(47) Jan sa
de na oso. Di djonsro mi sii a pasa. John MOD COP LOC house just now lsg see 3sg pass 'John should be at home. I just saw him go by.'
Robby Morray (pers. comm., 2002) suggests that sa here (as in other cases) implies a certain doubt and weighing of options. In the above example, the speaker provides the reason for inferring that John is at home. Or there might be a situation
Revisiting variation between sa and o ln Sranan
where the speaker concludes that John should be at home, since he isn't at the office or playing football or something else. In such a case, "a logical choice is he's at home, but there are other options" (Robby Morray, pers. comm., 2002). Also related to these epistemic uses are cases like the following, provided by one of the older informants, where sa refers to a past event and cannot be replaced by o. John sa go sribi kaba; en oso dungru. John MOD go sleep COMPL 3sg house dark 'John must have gone to sleep already; his house is dark~ Na den pikin sa musu nyan a kuku. Foe the-pl chile MOD must eat DET cake 'It's the children that must have eaten the cake~ Like the earlier examples, these epistemic uses of sa convey the speaker's assessment of the probability of a situation being true based on his assumptions, expectations, and weighing of options in relation to the situation in question. This is what really distinguishes epistemic sa from o, making it impossible for them to substitute for each other in this use. 6.3
Deontic uses of sa
Sa can also be used in a deontic sense to express a sort of command, where the deontic source is the authority of the speaker or some other person or institution that imposes the obligation. This confirms Seuren's observation that sa can express the speaker's "wish or order:' One such example is the following. (48) Alen sa kon tide neti. rain MOD come today night 'It shall rain tonight.'
One informant commented that only God can utter a sentence like this, since sa conveys the sense of a promise or command that it will rain. Another example of the use of sa with the force of a command is the following. (49) A
hoi sa tan dya tide neti. boy MOD stay here today night 'The boy shall stay here tonight: DET
An older informant notes that this conveys the sense of a strong promise, meaning that "the boy shall stay; fll see to it that he stays." Again, o cannot be used in this deontic sense. On the other hand, this strong sense of sa is weakened in a case where someone expresses a wish, as in the following, where o is also possible.
33
34
Donald Winford
(50) [Context: Speaker A says her plants are drying up, and laments the fact that no rain has fallen. Speaker B replies:]
Kande alen sa/o kon tide neti. perhaps rain MOD/PUT come today night 'Perhaps it shall rain tonight: In the following section, I argue that the three uses of sa in contemporary Sranan as described above are due primarily to the influence of Dutch zullen. 6.4
Parallels between sa and Dutch zullen
Recent research (Migge and Goury 2008; Migge and Winford 2009) argues that sa. derives from Dutch zal, the first- and third-person singular form of the Dutch auxiliary zullen 'shall= Others (e.g. Van den Berg 2007: 188) have suggested that its source was English shall, but there is no other English-lexicon creole that derives a marker of futurity or modality from shall. Moreover, as Migge and Winford (2009: 143, fn 8) point out, the only other creoles outside of Suriname that employ sa as a modal auxiliary are Dutch-lexicon Negerhollands and Berbice Dutch and English-lexicon basilectal Guyanese creole as spoken in Berbice. The first two ereales clearly got it from Dutch, and it is very probable that Guyanese creole borrowed it from Berbice Dutch (Robertson 1983: 13). All of this seems to offer strong circumstantial evidence for the claim that sa in the Surinamese creoles derives from Dutch zal. Indeed in the early Sranan texts, it is spelled variously as zal, zel, za, sal, sa, etc. (Van den Berg 2007: 199). Also, early texts such as Van Dyk (c. 1765) consistently translate sa with forms of zullen. Our discussion so tar has shown that sa in modern Sranan Tango can convey various interpretations in discourse, which include dynamic, epistemic, and deantic modalities. This range of interpretations is quite similar to that of Dutch zullen both in the early texts and in modern Dutch. As discussed in Section 2 above, we find several cases in the early texts where sa is used in a dynamic sense to express futurity as well as notions like a promise or a threat. The following are examples of the latter two uses from VanDyk (c1765: 40) with his Dutch translations. (The first example is cited in Van den Berg (2007); glosses and English translations are hers.) (51) Tamara manmanten na tien uri mi za kom na joe. tomorrow morning LO c ten hour 1sg PUT come LO c 2S ['Morgen Ogtend te tien Uuren zal ik by u koomen:] 'Tomorrow morning at ten o'clock I will visit you: (Van den Berg 2007: 189)
Revisiting variation between sa and o ln Sranan
(52) A doe wan trom morre mi za kiele da homan. she do one time more lsg MOD kill the woman ['Als zy het nag reis doed zal ik haar Kapot maaken~] 'If she does it one more time, I shall kill the woman= (VanDyk c1765 in Arends and Perl1995: 179) As noted in Section 2, we also find uses of sa where it expresses the sense of'should' or a 'hypothetical' meaning (see example 4). Van den Berg (2007: 191) also points out that early Sranan sa is sometimes used to describe hypothetical or uncertain events. There do not seem to be any clear examples of the epistemic or deontic uses of sa in the early texts. It would appear, then, that only some of the modal senses that sa expresses in contemporary Sranan can also be found in the 18th-century texts. Continuing influence from Dutch is likely to have reinforced these uses of sa and extended them further, for example into epistemic and deontic domains. Indeed, all of the modal meanings associated with sa.- dynamic, deontic, and epistemic- closely parallel those associated with zullen in contemporary Dutch. In its most typical use, zullen (like sa) conveys a kind of futurity that can be interpreted as a case of dynamic modality. 6 When the speaker is the agent of the future action, zullen conveys the sense of an intention or commitment to act in a certain way, as in the following? (53) Morgen zal ik dat boek wel kopen. tomorrow shall I that book ADV buy 'Tomorrow I will buy that book (but not today): (Wendelaar and Koefoed 1988: 68) Similar to sa, the intent to act can be construed as a promise or threat, given the right context. (54) Als je vanavond komt zal ik je het geld betalen. when you tonight come shall I you the money pay 'When you come tonight, I shall pay you the money= (55) Als je me niet betaalt zal ik je in elkaar slaan. when you me not pay shall I you in each.other beat 'If you don't pay me, I shall beat you up= 6. According to Adrienne Bruyn (pers. comm., August 2007), the future use of zullen is (currently) more typical of written and formal speech. 7· I am grateful to Adrienne Bruyn. Margot van den Berg, and Marleen van de Vate for providing translations of Dutch sentences as well as examples of the various uses ofzullen and their intuitions concerning these uses.
35
36
Donald Winford
Moreover, like sa, zullen conveys an expectation on the speaker's part that someone else will act in a certain way, as in the following (gloss and translation by Margot van den Berg). (56) Mijn vadet· zal je helpen. my father shall you help 'My father will help you=
(Wendelaar and Koefoed 1988: 68)
Zullen can also be used in an epistemic sense, to express probability. (57) De kinderen zullen nu wel slapen. the children shall now ADV sleep 'The children should be sleeping now' (=it's very likely they are).
zijn. (58) ]an zal nu wel thuis John shall now ADV at.home be 'John should be at home now: Finally, in certain contexts, zullen can have deontic force, conveying the sense of a command, as in the following. (59) De jongen zal vanacht hier blijven. the boy shall tonight here stay 'The boy shall stay here tonight (60) Het zal vanavond regenen. it shall tonight rain 'It shall rain tonight From a historical perspective, the use of sa to express dynamic modality has always been its primary function, as we saw in our discussion of the early Sranan texts above. This function seems to have derived directly from that of zullen. The other uses of sa in contemporary Sranan -to express epistemic and deontic modality - appear to have developed more recently under the influence of zullen (see Migge and Winford 2009 for more discussion). I do not, of course, claim that influence from zullen fully explains the path of change that sa seems to have been following in modern Sranan. Further research is needed to ascertain how far the current semantics and uses of sa can be attributed to such influence, as opposed to the dynamics of internal language change. But there is evidence that Dutch has been exerting considerable influence on the grammar of Sranan over the last several decades as more and more speakers of the language have become bilingual in Dutch.
Revisiting variation between sa and o in Sranan
7·
Conclusion
We've seen that sa and o are both possible in many contexts but express clear distinctions of meaning and hence cannot be said to be in competition. In other contexts, such distinctions appear to be fading, for instance in complements to verbs like 'promise, think' and the like or in combination with adverbs like kande 'perhaps,' where choice of one or the other seems to make for little difference in meaning. The combination ben sa. seems to be gradually giving way to bo for similar reasons. Apart from these contexts, however, the distinction between sa and o seems to be alive and well in contemporary Sranan. Historically, the distinction between the two began to emerge only in the course of the 18th century, when de go gradually became grammaticalized as a marker of future tense, yielding the current form o. Given the observations by Voorhoeve and Donicie regarding (d)e go as the source of o and Nickel and Wilner's observation that speakers in the 1980s still understood o as ego, this process of grammaticalization may have continued we 11 into the 20th century, though further research is necessary to confirm this. In addition, it seems likely that the meanings and uses of sa became increasingly subject to influence from Dutch zullen, which conveys epistemic and deontic senses similar to those conveyed by sa in modern Sranan. As a result of the emergence of o and the influence of zullen, sa seems to have become less associated with futurity and more associated with modal meanings and uses in modern Sranan. In short, o is not competing with sa in any significant way, since o cannot express the kinds of modal meanings that sa can.
References Arends, J. & Perl, M. 1995. Ear~v Creole Texts: A Collection of 18th-century Sranan and Sara maccan Documents. Frankfurt: Vervuert. Dahl, 0. 1985. Thnse and Aspect Systems. Oxford: Blackwell Donide, A. 1954. De Creolentaal van Suriname. Spraakkunst. Paramaribo: Radhakishun. Donide, A. 1956. De partikels sa en (de) go in de creolentaal van Suriname. West-Indische Gids 36: 183-191.
Mlgge. B. & Goury, L. 2008. Between contact and internal development: Towards a multi-layered explanation for the development of the TMA system in the creoles of Suriname. In Roots of Creole Structure, S. Michaelis (ed), 301-331. Amsterdam: John Benjamins. Mlgge. B. & Winford, D. 2009. The origin and development of possibllity ln the creoles of Suriname. In Gradual Creollzatlon: Studies Celebrating Jacques Arends, R Selbach, H. C. Cardoso & M. van den Berg (eds.), 129-153. Amsterdam: John Benjamins. Nickel, M. & Wllner, J. 1984. Papers on Sranan Tbngo. Paramaribo: Instituut voor Taalwetenschap (Summer Institute of Linguistics). Perkins, M. R 1982 The core meanings of the English modals. Journal ofLinguistics 18: 245-273.
37
38
Donald Winford Robertson, I. 1983. The Dutch linguistic legacy in Guyana: Berbice and Skepi Dutch. Carib 3: 11-23. Schumann, C.L. 1783. Neger-Engl.isches Worterbuch. In Early Creole Lexicography: A Study of C. L. Schumaruis Manuscript Dictionary of Sranan (1983), A. Kramp, 44-305. PhD dissertation, University ofLeide.IL Seuren, P. 1981. Tense and aspect in Sranan. Linguistics 19: 1043-1076. Seuren, P. 1983. The auxiliary system in Srana.n. In Linguistic Categories: Auxiliaries and Rilated Puzzles, Vol. 2, F. Heny & B. Richards (eds.), 219-251. Dordrecht: floris. Simons, R. D. 1954. Het partikel sa en het Surinaams. West-Indische Glds 35: 166-170. van den Berg, M. 2000. "Mi no sal tron tongo:" Early Sranan in court records, 1667-1767. MA Thesis, Radboud University ofNijmegen. van den Berg, M. 2007. A Grammar of Early Sranan. PhD dissertation, University of Nij megen van Dyk. n.d [c1765] Nleuwe en nooit bevoorens gezlene onderwyzinge In het Bastert Engels, of Neeger Engels, zoo als het zelve in de Hollandsze Colo11ien gebruikt word. Amsterdam: Jacobus van Egmont. Voorhoeve, J. 1957. The verbal system ofSranan. Lingua 6: 374-396. Voorhoeve, J. 1962. Srana11 Syntax. Amsterdam: Publications of the Bureau for Linguistic Research in Surinam. Voorhoeve, J. & Kramp, A. 1982. Syntactic developments in Sranan. Paper presented at the 4th Biennial Conference of the Society for Caribbean Linguistics, Paramaribo, Suriname. Wendelaar, W. & Koefoed, G. 1988. 'Sa' en 'd in het Sranan. OSO 7: 63-7 5. Weygandt, G. C. 1798. Gemeenzame leerwijze, om det basterd of Neger-Engelsch op een gemakkelyke wyze te leeren verstaa11 en spreeken, etc. Paramaribo: W.W. Beeldsnyder.
Inherent variability and coexistent systems Negation in Beq uia James A. Walker & Jack Sidnell The robust linguistic variation observed on the island ofBequia (St. Vincent and the Grenadines) raises the question of whether there is a single variable linguistic system or multiple systems. We examine the distribution and conditioning of variable negation (na, ain't or not/-n't) in three communities distinguished by ethnicity and socioeconomic history using variable rule analysis. The variant na is restricted to one community and is conditioned by co-occurrence constraints, while ain't is conditioned by auxiliary context and temporal reference, which serve to distinguish between communities. Thus, although there is variation between communities and speakers, there is also evidence for overlapping but discrete and coexistent systems. These results underline the importance of examining the distribution and conditioning of all features implicated in the linguistic system.
1.
Introduction1
Bequia is the northernmost of the Grenadines, located just south of St. Vincent (see Figure 1). Despite its small size Oust 7 square miles), this island features a surprising degree of linguistic diversity, apparent even to the casual observer. An array of apparently quite distinct linguistic varieties is spoken on Bequia, ranging from a relatively standard Caribbean English, through what is best described as a nonstandard variety of English (Williams 1987), to a more restructured variety showing affinity to other English-based creoles (Winford 1993; Aceto and Williams 2003), all of which are characterized by linguistic features found elsewhere in the Eastern Caribbean. 1. The Bequia project is generously supported by the United States National Science Foundation. the Social Sciences and Hwnanities Research Councll ofCanada and the British Academy. We thank Miriam Meyerhoff; the local research assistants and the residents of Bequia and St Vmcent who have welcomed and helped us there. Earlier versions of this paper received helpful comments from audiences at Sociolinguistic Symposium 16 (UniversityofLimerick) and the UniversityofEdinburgh. We thank two anonymous reviewers for helping to clarify the analysis and argumentation.
40
James A Walker & Jack Sidnell
..
..,_ ••
'
I)
BE QUIA
~ Q
EASTERN CARIBBEAN
If:'
Mount Pleasant
.. 0
\) ~
a
()
[j
I
'·
~j]
.;::?
SOUTH AMERICA
LaPompe
Paget Farm 'Southside'
Figure 1. Bequia, St Vincent, and the Grenadines
Such linguistic diversity presents a challenge to our models of the speech community and the linguistic system. A basic question that has continually confronted us in our research on Bequia is whether we can model the observed variation as a single, highly variable linguistic system or whether there are multiple coexistent systems. Indeed, this question has long beset the analysis of linguistic variation, especially in the context of language (and dialect) contact and creole-speaking communities. An early response in creole studies was the idea of'polylectal' grammars (e.g. DeCamp 1971; C.-}. Bailey 1973; Bickerton 1975), in which apparent variation is viewed as an artifact of conflating different levels of a (post-)creole continuum: that is, the community is in the process of shifting from a basilectal creole system to a more acrolectal English system via several mesolectal stages. Although more recent work has abandoned the diachronic assumption of the continuum (e.g. Rickford 1988), the view that variation reflects a mixture of different levels of the continuum is still widespread. Yet our research has shown us that we cannot escape the fact of variation, even at the level of the individual speaker in the same
Inherent varlabllity and coexistent systems
social and linguistic contexts. This leads to the question: can we maintain a model of multiple linguistic systems while acknowledging the fact of inherent variability? In the context of a similar discussion about African American Vernacular English (AAVE), Labov (1998: 139) proposes four conditions that would favour the recognition of coexistent linguistic systems, two of which are relevant to our research question. The first, segregation ofvariants, states that: The variants oflinguistic variables are not evenly distributed across texts or situations, but concentrated in long runs of the same value, so that extended stretches of speech show one value rather than the other. (ibid)
The fourth condition, strict co-occurrence, states that "rules show strict co-occurrence, so that one never applies without the other" (ibid.). Taken together, these conditions suggest that, wherever we find 'conspiracies' or co-occurring features that can be attributed to one linguistic system, and not to another, we may infer the presence of coexistent systems. In our contribution to this volume, we make use of Labov's conditions in a quantitative analysis of verbal negation on Bequia to propose a partial methodological solution to the problem of studying linguistic variation in (potentially) polylectal communities. In brief, our argument is that, although we may observe considerable variation aero ss communi ties and even individual speakers, we can nevertheless isolate partially overlapping yet discrete and coexistent systems localized in different communities on the island. We begin by reviewing the literature on negation in English-based creoles, nonstandard varieties of English, and African American English, with a view to identifying linguistic contexts which might be operationalized as factors to isolate different linguistic systems. Before detailing our analysis of variable negation and the factors conditioning the choice of negation, we provide some sociohistorical background on Bequia and our ongoing research there.
2.
Negation in English, English-based creoles, and African American Vernacular English
Variability in the expression of verbal negation is a longstanding feature of English. Postverbal not, which originated from an adverbial began to compete with preverbal clitic ne in the period between Old and Middle English, gradually winning out by the end of the Middle English period (Strang 1970: 312; Mitchell and Robinson 1992: 101-2). In Early Modern English, not began to vary with its contracted form -n't (the first attestation of contracted not is given in (la)), and ain't (spelled in different ways) made its first appearance as a contracted form of have and be (an early attestation is provided in (1b )).
41
42
James A Walker & Jack Sidnell
(1) a. But mayn't I Bar points, being the Challenged? ( 1652: John Tatham, The Scotch Figgaries, m· a Knot of Kna.ves IV, i;
Oxford English Dictionary) b. wee'l play heads or tails, who goes first, that's fair now, e'nt it? (The Mock-Tempest IV.ii; Duffett 1674) In modern standard varieties of English, ain't has all but disappeared, and not varies only with -n't. In nonstandard varieties, ain't survives as a variant in contexts of negated have and be (Feagin 1979; Cheshire 1982; Christian, Wolfram, and Dube 1988), though it occurs more frequently in have contexts (Cheshire 1991: 57). A related consideration is the grammatical person of the subject (Wolfram 1973), though this may have to do with the irregular paradigm of auxiliary verbs (cf. Hudson 2000). Variability in verbal negation is also a well-known feature of the English-based creoles (EBCs) of the Caribbean. The prototypical creole negator is an invariant form (usually no or na) placed before the verb and any tense-mood-aspect markers (Bickerton 1975: 43; Holm 1988: 171; Rickford 1992; McWhorter 1996: 276; Schneider 1997: 1057; Greene 1999; Parkvall1999: 38; Patrick 1999; Kephart 2003: 235; McPhee 2003: 34; but cf. Van Herk 2003). Another frequently cited negator is ain't (which occurs in various phonetic forms, such as ernt, en or eh; Rickford 1974; Bickerton 1975: 99; Winford 1983; Schneider 1997: 1057; Van Herk 2003: 255), though unlike the English form, it is said to occur across a wider range of auxiliary and tense contexts (Bickerton 1975: 99; Kortmann and Szmrecsanyi 2004: 1190) and in some cases its function resembles that of no/na (Schneider 1997: 1057). Other forms of negation include neva, doon(t)!duon, di(d)n(t) and the suffix -n, all obviously derived from English never, don't, didn't and -n't, respectively, though again these forms are said to differ in distribution and function in some EBCs (Bailey 1966: 90; Bickerton 1975: 95-6, 98, 101, 1996: 314; Holm 1988: 172-4; Schneider 1997: 1056; Greene 1999:86, 88; Patrick 1999: 201; McPhee 2003). Although variation among these different forms of negation is generally noted (Bickerton 1975; Winford 1983; Schneider 1997; Cutler 2003: 74; Kephart 2003; McPhee 2003), accounts difter as to the source of the variation. Some claim difterences in meaning or selectional restrictions for the distribution of each form. For example, McPhee (2003: 38) notes that anterior marker birn and future marker gtJ can only be negated with na/no, whereas did and WAZ are negated with a suffixed -n. More commonly, each variant is assumed to represent a different lector level of the continuum. Thus, Bickerton (1975, 1996) and Winford (1983) argue that, as speakers shift from the basilect to the acrolect, forms are first relexified, taking on the properties of the replaced form, before being reanalyzed according to the English system. Under this view, no is first replaced by ain't or didn't, which, unlike
Inherent varlabllity and coexistent systems
the English forms, are unanalyzed (i.e. monomorphemic) negators, before being reanalyzed according to the English system (i.e. bimorphemic). Others are more agnostic about the diachronic nature of the variation, simply noting that speakers alternate between forms of negation from different levels of the continuum (e.g. Patrick 1999: 199; Cutler 2003: 74; Kephart 2003). Despite the acknowledgment of this variability, there are very few quantitative studies of negation in EBCs. Winford's (1983) examination of eh (
3· 3.1
Variation in negation on Bequia Sociohistorical background
Except for occasional visits by the Caribs, Bequia was uninhabited before being acquired under the Treaty of Paris (1763) by the British, who developed sugar plantations and imported slaves from West Africa and elsewhere in the Caribbean. During the initial period of development, there was also a steady influx of British
43
44
James A Walker & Jack Sidnell
(mostly Scottish) settlers from Britain or Barbados and St Vincent. By the mid19th century, with the collapse of the sugar plantation economy and the abolition of slavery, the island's remaining inhabitants shifted to small-scale subsistence farming and maritime trade, such as fishing and (after 1875) whaling. Most of the current population of approximately 5,000 descends from freed slaves who had worked on the plantations, though a small minority descends from British Caribbeans from Barbados who were re-settled throughout the Grenadines in the 1850s, as well as from French and Portuguese people who settled during the mid-19th century (for a more detailed discussion of Bequia's history, see Meyerhoff and Walker to appear). Our research project was initially designed to examine social factors contributing to the maintenance of an ethnolinguistic boundary between 'black' and 'white' varieties, but our first few weeks of field research in 2003 convinced us that such ethnic categorization is problematic on Bequia and that people are more appropriately categorized according to where they come from on the island. Our fieldwork, conducted between 2003 and 2005, concentrated on three communities (see Figure 1): Hamilton, a predominantly African-descent community originating from a large plantation; Mount Pleasant, the traditional home of the 'Scottish' population; and Southside, a local term referring to a contiguous group of fishing and whaling communities: Paget Farm, La Pompe and Friendship. These communities are distinguished not only by difterent racial/ethnic mixes but also by different socioeconomic histories and circumstances (see Price 1988). Our fieldwork combined ethnographic observation with sociolinguistic interviews (Labov 1984) conducted and recorded by trained local research assistants. The interview corpus comprises over 100 hours of recordings with 62 speakers aged between 41 and 100 years old (see Meyerhoff, Sidnell, and Walker in preparation). We draw on this corpus for the analysis presented in this paper, focusing on a sub-sample of18 speakers, six from each ofHamilton, Mount Pleasant and Paget Farm,2 evenly balanced for sex. 3.2
Defining the variable context
Our linguistic focus in this chapter is the variable realization of verbal negation, which on Bequia may be expressed by a preverbal negator na (2a), ain't (2b ), or postverbal not (2c) and its contracted form -n't (2d). 3 2. The Southside speakers in our interview corpus come from Paget Farm and La Pompe. In this chapter, we focus on speakers from Paget Farm because their interviews were among the first to be transcribed in at least two passes.
3· Our speakers also use 11ever, but a preliminary analysis of its distribution suggests that it behaves more like an adverb than a verbal negator.
Inherent varlabllity and coexistent systems
(2) a. b. c. d.
All them we na bin cussing. You a.in't know all the hardship. And when he is old, he iz not depart. The dead don't have no hope.
The range of variation seen here is similar to that found elsewhere in the Caribbean. The variant na resembles the invariant preverbal negator found in other EBCs. As noted above, while aint occurs both in EBCs and in nonstandard English, its distribution is different Therefore it is not the mere presence of this feature which concerns us, but its linguistic conditioning. Although we extracted the first 200 sentences with negative marking, wherever possible, from each of the speakers in the sub sample, in this chapter we narrow the focus of our analysis to verbal negation. We excluded from the analysis several contexts that do not admit variation, such as imperatives, modals and semi-modals such as used to, which almost always occur with forms of not or -n't. In addition, other variable processes make it difficult if not impossible to distinguish between not or na in some contexts. For example, the variable occurrence of zero copula and the following phonological context together make the sentence in (3) ambiguous between preverbal na and not with a zero copula. (3) That not!na true.
(H003: 352)
All such tokens were excluded from analysis. This yielded a dataset of 1,720 tokens of verbal negation, distributed according to community as shown in the final column of Table 1. Table 1. Distribution of verbal negation in three communities on Bequia
Hamilton Mount Pleasant Paget Farm
%na
%ain't
%not!-n't
12 0 [N=2] 0 [N=2]
29 40 16
59 60 84
Total:
TotalN 626 518
576 1,720
4· All examples are reproduced verbatim from recorded interviews and are identified by the community (H = Hamilton; M = Mount Pleasant; P = Paget Farm), speaker number, and line number in the transcription. 5·
In this example, iz represents a reduced form ofhabitual doz.
45
46
James A Walker & Jack Sidnell
3·3
Factor groups coded
In addition to noting the form of verbal negation (na, ain't or not/-n't), we coded each token for a number of factors that have been proposed to affect the choice of negator in English-based creoles, African American English, and nonstandard English. As noted above, a primary consideration in the analysis of negation, at least for ain't, is the auxiliary context. We therefore distinguish among be contexts (4a; that is, ain't varies with am + not, are + not, is + not, 1vas + not, were + not), do contexts (4b; a.in't varies with does+ not, do+ not, did+ not), and have contexts (4c; ain't varies with have + not).6 (4) a. be (ain'tvs. am/are/is/was/were+ not) Though the money right now ain't flowing. The seed is not growing the same day. That time no boat airlt working with engine or nothing. Rum wasn't flourishing like now.
(P023: 368) (P014:285) (H006:490) (H005: 547)
b. do (ain't vs. do/does/did+ not) Me ain't want a whole entire argument. Some of them, they don't want a man. But them days long time you ain't have no help. And long time they didn't have the preacher.
(M301: 715) (M301: 161) (H012: 168) (H012: 336)
c. have (ain't vs. have/has/had + not) I ain't been to a doctor, you know. I haven't seen the person up to now.
(M302: 588) (P014:071)
If an English-like system is operative, we expect ain't to be preferred in contexts of present be and have, as in other nonstandard varieties of English. If a more creolelike system is operative, we expect the auxiliary context to be irrelevant. We tested for the influence of tense and aspect on the choice of negation in several ways. First. we noted whether each token referred to a past (Sa) or present (5b) event or state (tokens with ambiguous temporal reference were not retained for this factor group). (5) a. I didn't school here, you know. b. Your lifestyle na suit me.
(P024: 147) (M301: 714)
6. There were very few had+ not contexts. In fuel:, differentiating between have + not and did + not contexts is made difficult by other variable processes such as (t/d)-deletion and morphological past-marking, which render the lexical verb bare.
Inherent varlabllity and coexistent systems
Following on previous work (Walker 2000, 2001, 2005, 201 Ob ), stativity was coded on the basis of the lexical verb, regardless of the aspectual nature of the sentence in which it occurred. We coded whether the verb indicated a state (6a) or activity (6b). (6) a. They a.irlt experience nothing in the world. b. Anglican don't baptise, them doz sprinkle.
(P014:630) (M302: 953)
Finally, we coded for the aspect of the sentence in which each token occurred, distinguishing among habitual (7a-b), continuous or durative (7c), and punctual (7 d) aspect. (7) a. We not telling you because if we tell you, you wouldn't come back to see us. (H027: 1249)1 b. I don't go to Carnival. (HOOl: 158) c. But long time it wasn't so. (P023: 1179) d. They airlt catch it in time. (MlOl: 109) If a creole-like system is operative, we expect considerations of tense and aspect to be irrelevant to the choice of negator. In line with Labov's criteria, we coded several factor groups to address the question of co-occurring constraints discussed above. First, we hypothesize that the presence of creole tense-aspect markers, such as anterior bin, indicates acreole system. (8) Them na bin have no engine.
(HOOl: 199)
We also hypothesize that the use of object pronouns in subject position indicates a more creole-like system.8 (9) Me ain't got no time with that.
(M301: 370)
However, since we found that this variation occurs only in 1st person singular (I vs. me) and 3rd person plural (they vs. them), we consider only these grammatical contexts. Finally, since negative concord is another feature highly associated with creole varieties (though of course it also occurs in many nonstandard varieties of English; cf. Kortmann and Szmrecsanyi 2004: 1186), we coded negative 7· Verbal aspect was coded on the basis of the discourse context rather than the form of the verb. For example, in this sentence, although the progressive is used, the wider discourse context makes it dear that the speaker is referring to a habitual event. 8. AB an anonymous reviewer points out, the use of object pronouns in subject positions is attested in three other (demonstrably non-creolized) varieties of English: Newfoundland and the North and Southwest of England (Kortmann & Szmrecsanyi 2004: 1185). While the ultimate source of this feature may thus be nonstandard English, invariant object pronouns are commonly considered a basllectal feature in the Caribbean context (Holm 1988: 201 ).
47
48
James A Walker & Jack Sidnell
concord as present if there was another negated element in the same clause (lOa) and absent if there were other elements in the clause which could have been negated but were not (lOb). Tokens in which there were no other elements in the clause that could be negated (10c) were notincluded in this factor group. (10)
a. We didn't have no shoes. (M303: 011) (P014: 139) b. I don't go any place. c. When the wind na blow- remember, bin wind, no engine. (H005: 335)
All of these factors were considered individually and together using GoldVarb X (Sankoff. Tagliamonte, and Smith 2005).
4·
Results
Table 1 above shows the overall distribution of forms by community. Note that, while all three communities feature robust use of ain't and forms of not, na is virtually restricted to Hamilton. This result confirms our impression of Hamilton as the most creole-like variety on the island. On the basis of this restricted distribution, we restrict our analysis of na to Hamilton, before we proceed to an analysis of the linguistic conditioning of a.irlt. 4.1
The use of na.
An examination of individual speaker patterns of use within Hamilton revealed that, while all speakers used nato various degrees, two speakers ahnost never used it. Because including such speakers in the analysis of na might skew the results, they were excluded from subsequent analysis of this variant. The remaining data was subjected to multivariate analysis, the results of which are shown in Table 2. 9 Because auxiliary context showed interaction with stativity and has not been claimed to be relevant to the distribution of na, this factor group was not included in the analysis.
9· In Tables 2 and 3, each set of three columns shows one variable-rule analysis using GoldVarb X (Sankoff, Tagliamonte, and Smith 2005), with factor weights, percentages and total number of tokens for each factor. The factor weight represents the probability that each factor contributes to the occurrence of 11a (Table 2) or ain't (Table 3): closer to 0, less likely; closer to 1, more likely. The 'Input' indicates overall probability. The multiple-regression feature of GoldVarb (binomial step-up/step-down) considers all factors simultaneously and assesses the statistical significance and relative contribution of each (see Guy 1993; Young and Bayley 1996; Paolillo 2002, Walker 2010a).
Inherent varlabllity and coexistent systems Table2. Factors contributing to the occurrence of na (vs. ain't and not!-nt) in Hamllton, Bequia*
TotalN: Input:
429 .068
%
N
60 56 16 2
20 18 49 113
84 14
19 420
[.SO]
18 17
267 162
Negative Concord Present Absent
[.55] [.47]
21 15
66 104
Temporal Reference Past Present
[.52] [.48]
19 16
175 251
Aspect Habitual Durative Punctual
[.42] [.54] [.53]
12 18 23
137 217 73
Form of Pronoun me
I
.96 .92 .76 .20
Range:
76
Tense-Aspect Marker Present Absent
.96 .46
Range:
50
them they
Lexical Stativity Stative Nonstative
[.50]*~
*Excludes two speakers (n=197) with low use of na. Factor weights :2: .50 favour na, while factor weights <.50 disfavour. **[ ] = Not selected as significant.
Note that none of the factor groups intended to test the effects of tense and aspect (temporal reference, stativity, and aspect) are selected as significant, suggesting that na does indeed represent a tense- and aspect-neutral negator like that found in other EBCs. Only two fuctor groups are selected as significant: the presence of a tense-aspect marker, which favours na, and the form of the pronoun, with object forms and they favouring na (stated conversely, with subject I disfavouring). In light of
49
50
James A Walker & Jack Sidnell
Labovs (1998) conditions of segregation of variants and strict co-occurrence, discussed above, the highly favouring effuct of these two creole-like features on na. suggests that at least some of the observed variation is organized into small systems that we may call a 'lect'. 4.2
The use of ain't
We turn now to a consideration of ain't, which occurs at robust rates across all three communities and therefore lends itself to cross-variety comparison. As Table 1 shows, there are quite ditrerent rates of ain't in the three communities: most in Mount Pleasant, less in Hamilton (though here na also figures), least in Paget Farm. However, rates of use can fluctuate for a number of (non-linguistic) reasons. More important is the hierarchy of factors conditioning its occurrence. A problem confronting anyone who studies ain't is determining its variable context. If the interpretation of decreolization is correct, and ain't constitutes a relexification of na, it should occur indifferently across all tense-aspect contexts. If it represents part of an English system, we must restrict the analysis to presenttense contexts of have and be. Rather than biasing the analysis by assuming one system or the other, we allow the hierarchy of conditioning factors to indicate which underlying system is operative, analyzing ain't against all other forms of not/-n't (excluding na). If English-like restrictions on the choice of auxiliary are relevant, they should emerge from the analysis. While we acknowledge that this decision itself may bias the analysis in other ways, we believe that it minimizes bias more than an approach that assumes one system or the other a priori. As noted above, the type of auxiliary is potentially a paramount restriction on the occurrence of ain't, but it interacts with stativity. For this reason, in the analysis of ain't, we include auxiliary context but do not include statitivity. Table 3 shows the results of multivariate analysis of the contribution of these factors to the occurrence of ain't. Most factor groups selected as significant do not distinguish among communities. In all three communities, object pronouns in subject position and tense-aspect markers both tend to correlate with airlt (though the latter factor group is either not selected as significant (Paget Farm) or contains knockouts (i.e. occurrences of 100%) and therefore cannot be included in multivariate analysis (Hamilton, Mt Pleasant)). These effects parallel those of na shown in Table 2 and suggest that ain't may function in some respects like na, a universal negator. The presence of negative concord also favours airlt across all communities, a finding reported in studies of AAVE (Wolfram 1974; Weldon 1994; Walker 2005).
Inherent varlabllity and coexistent systems
Table3. Factors contributing to the occurrence of ain't (vs. not!-n't) in three comml.lllities on Bequia
Hamilton 449 .516
TotalN:
Input:
%
N
74 44 49 42
19 9 57 132
Mount Pleasant
Paget Farm
516 .417
.133
542
%
n
80 100 41 27
5 2 90 133
%
n
67 63 14 14
12 8 81 179
Form of Pronoun
me
I
.70 .49 .51 .47
Range:
23
them they
.90 KO .56 .44
.91 .92 .45 .45 47
46
Tense-Aspect Marker Present Absent
KO**
100 40
4 445
KO
100 39
6 510
[.83]** [.50]
50 16
4 538
60 35 20
99 339 5
.75 .41 KO
62 31 100
131 370 2
[.51] [.50] [.46]
16 17 13
144 380 8
34 42
182 333
[.48] [.51]
16 16
175 366
44 36 50
160 326 30
.53 .46 .72
19 14 29
151 356 35
44 10
45 218
Auxiliary Context
be do
have Range:
.72 .44 .27 45
34
Temporal Reference Past Present
[.55] [.46]
46 36
198 249
Range:
.38 .57 34
Aspect Habitual Durative Punctual
.46 .47 .67
Ra1!ge:
21
36 38 56
149 221 77
.53 .46 .78 32
26
Negative Concord Present Absent
.70 .36
Range:
34
~
61 28
77 109
.70 .36
58 34
34
Excludes tokens of na and two speakers with no use of ain't. =Not selected as significant; KO =Knockout (100%).
~* ~ ]
98 141
.85 .41 44
51
52
James A Walker & Jack Sidnell
The main differences between the communities emerge in auxiliary context and temporal reference. Both Hamilton and Mount Pleasant show a clear distinction between be, which favours airft, and do, which disfavours (the small numbers for have make it difficult to make claims about this context), while in Paget Farm, auxiliary context is not selected as significant.10 Temporal reference is selected as significant only in Mount Pleasant. 11 Taken together, the results for Mount Pleasant suggest that ain't is favoured most in present be contexts, likely the legacy of a nonstandard English system. In contrast, the temporal and auxiliary constraints are less restrictive in the other communities and ain't appears to behave more like a tense-neutral negator, likely a creole legacy.
5·
Conclusion
Our analysis of the variable expression of negation on Bequia suggests differences between communities. The creole-like preverbal negator na is virtually restricted to Hamilton and is conditioned by the presence of tense-aspect markers and the use of object pronouns in subject position. The use of ain't is conditioned by a number offactors, two of which (auxiliary context and temporal reference) appear to distinguish between communities. We suggest that these patterns result from the historical development of the difterent communities on Bequia. The people of (primarily) British ancestry living in and around Mount Pleasant inherited a nonstandard English system in which na was absent and ain't functioned as a tensed negator in be and have contexts, a pattern that clearly persists until the present day. On the other hand, the people of Hamilton and Paget Farm appear to have incorporated airft into a grammar in which verbs are negated by an invariant, tenseneutral marker. Thus, the use of ain't in past contexts may reflect a process of extension or relexification, in which ain'talternateswith na. These results demonstrate not only the complexity of the system of negation on Bequia, but also the necessity of considering the conditioning of variation rather than simply the presence of a particular form. to. Not only is this factor group not selected as significant, but the factor weights and the percentages for be and do are also virtually identicaL n. In Hamilton, although the factor weights are dose to . 5, the range of percentages is greater than that in Mount Pleasant. However, the direction of results is opposite to that of Mount Pleasant and to patterns observed in nonstandard English. In Paget Farm, the factor weights for temporal reference are very close to .5 and the percentages are virtually identkal. It might be argued that the overall low input value for ain't in Paget Farm (.133) is responsible for temporal reference not achieving significance, but this does not prevent other factor groups from being selected as significant.
Inherent varlabllity and coexistent systems
Finally; what do these results say about the question posed at the beginning of this chapter regarding the nature of multiple linguistic systems and inherent variability? There seem to be two systems operative on Bequia: a more creole-like system and a more English-like system. However, with the possible exception of na, these systems are not distributed categorically or uniformly across the island, but are reflected in differential conditioning of variation within each community. We suggest, then, that there is a complex set of constraints that work together to aggregate features into something we could call a lect This emerges in the collocational effects in which a number of features co-occur: the presence of tense-aspect markers, the use of object pronouns in subject position and the use of the negator na. While one of these features on its own would suggest the presence of a creolelike system, the co-occurrence or collocation of these features strengthens such a conclusion. These results constitute a quantitative illustration ofLabovs criteria of co-occurrence, providing strong evidence for the presence of coexistent systems. They also call into question our ability to isolate discrete lects on the basis of categorical distribution, and suggest instead that we must examine the distribution and conditioning of all features implicated in the linguistic system. As such, the research reported here provides a direction for the study of variation in polylectal speech communities.
References Aceto, M. & Willlams, J.P. (eds.). 2003. Contact Englishes of the Eastern Caribbean. Amsterdam: John Benjamins. Bailey, B. L. 1966. Jamaican Creole ~vntax: A 'ITansformational Approach. Cambridge: CUP. Bailey, C.-J. 1973. Variation and Linguistic Theory. Arlington VA:. Center for Applied Linguistics. Bickerton, D. 1975. Dynamics of a Creole System. Cambridge: CUP. Bickerton, D. 1996. The origins ofvariation in Guyanese. In Towards a Social Science ofLanguage:
Papers in Honor ofWilliam Labov, Vol1: Variation and ChangE in Language and Society, G. Guy, C. Feagin, D. Scb.lffrin & J. Baugh (eds.), 311-327. Amsterdam: John Benjamins. Bolton, W. 1982. A Living Language: The History and Structure ofEnglish. New York NY: Random House. Cheshire, J. 1982. Variation in an English Dialect: A Sociolinguistic Study. Cambridge: CUP. Cheshire, J. 1991. Variation in the use of ain't in an urban British English dialect. In Dialects of English: Studies in Grammatical Variation, P. Trudgill & ]. Chambers (eds.), 54-73. Singapore: Longman Singapore. Christian, D., Wolfram, W. & Dube, N. 1988. Variation and ChangE in Geographically Isolated Communities: Appalachian English and Ozark English. Thscaloosa AL: University of Alabama Press. Cutler, C. 2003. English in the Thrks and Caicos Islands: A look at Grand Turk. In Contact Englishesofthe Eastern Caribbean, M. Aceto & J.P. Williams (eds.), 51-80. Amsterdam: John Benjamins.
53
54
James A Walker & Jack Sidnell DeBose, C.E. 1994. A note on ain't vs. didn't negation in African American Vernacula.t Journal of Pidgin and Creole Language.s9(1): 127-130. DeBose, C.E. & Faradas, N. 1993. An Africanist approach to the linguistic study of Black English: Getting to the roots of the tense-aspect-modality and copula systems in Afro-American. In A.fricanisms in Afro-American Language Varieties, S.S. Mufwene (ed.), 364-387. Athens GA: University of Georgia Press. DeCamp, D. 1971. Toward a generative analysis of a post-creole speech continuum. In Pidginizatio1! and Creolizatio11 ofLaJ~guages, D. Hymes (ed..), 349-370. Cambridge: CUP. Fasold, R. & Wolfram, W. 1970. Some linguistic features of Negro dialect In Teaching Standard English in the Im1er City, R. Fasold & R. Shuy (eds.), 41-86. Washington DC: Center for Applied Linguistics. Feagin, C. 1979. Variation and Change in Alabama English: A Sociolinguistic Study of the White Community. Washington DC: Georgetown University Press. Greene, LA 1999. A Grammar ofBelizean Creole. Bern: Peter Lang. Guy, G.R 1993. The quantitative analysis of linguistic variation. In American Dialect Research, D.R. Preston (ed.), 223-249. Amsterdam: John Benjamins. Holm, J. 1988. Pidgins and Creole.s, Vol 1: Theory a11d Structure. Cambridge: CUP. Howe, D.M. 1997. Negation and the history of African American English. La11guage Variatio11 and Change 9(2): 267-294. Kephart, R. 2003. Creole English on Carriacou: A sketch and some implications. In Contact Englishes of the Eastern Caribbean. M. Aceto & J.P. Williams (eds.), 227-239. Amsterdam: John Benjamins. Kortmann. B. & Szmrecsanyi, B. 2004. Global synopsis: Morphological and syntactic variation in English. In A Handbook of Varieties ofEnglish, Vol. 2: Morphology and Syntax, B. Kortmann, K. Burridge, R. Mesthrie & E.W. Schneider (eds.), 1142-1202. Berlin: Mouton de Gruyter. Labov, W. 1972. Language in the Inner City. Philadelphia PA: University of Pennsylvania Press. Labov, W. 1984. Field methods of the project on linguistic change and variation. In Language In Use: Readings In Sociolinguistics, J. Baugh & ]. Sherzer (eds.), 28-53. Englewood Cliffs NJ: Prentice Hall Labov, W. 1998. Co-existent systems in African-American Vernacular English. In AfricanAmerican English: Structure, History, and Use, S.S. Mufwene, J.R. Rickford, G. Bailey & ]. Baugh (eds.), 110-153. London: Routledge. Labov, W., Cohen, P., Robins, C. & Lewis, J. 1968. A Study of the Non-standard English of Negro and Puerto Rican Speakers in New York City [Co-operative Research Report 3288, Vol I]. Philadelphia PA: US Regional Survey. McPhee, H. 2003. The grammatical features ofTMA auxiliaries in Bahamian Creole. In Contact Englishes of the Eastern Caribbean, M. Aceto & ]. P. Williams (eds.), 29-49. Amsterdam: John Benjamins. Me Whorter, J. H. 1996. The diachrony of predication in Saramaccan creole: Synchronic and typological implications. Studies in Language 20(2): 275-301. Meyerho:tt; M., Sidnell, J. & Walker, J.A In preparation. A sodolinguistic overview ofB equia (St Vincent and the Grenadines). Ms., University of Auckland/University of Toronto/York University. Meyerho:tt; M. & Walker, J.A To appear. Bequia English: Language in St. Vincent and the Gre1!adines. London: Battlebridge. Mitchell, B. & Robinson, F. C. 1992. A Guide to Old English. Oxford: IDackwelL
Inherent varlabllity and coexistent systems Parkvall, M. 1999. Feature selection and genetic relationships among Atlantic Creoles. In Spreading the Word: The Issue of D!tfusion ai1Wng the Atla11tic Creoles, M. Huber & M. Parkvall (eds.), 29-66. London: University of Westminster Press. Paolillo, J.C. 2002. Analyzing Linguistic Variation. Stanford CA: CSLI. Patrick, P.L 1999. Urban Jamaica11 Creole: Variation 111 the Mesolect. Amsterdam: John Benjamins. Price, N. 1988. Behind the Planter:S Back: Lower Class Responses to Marginality in Bequia Island, St. Vi11cent. Hong Kong: Macmillan. Rickford. J.R 1974. The insights of the mesolect. In Pidgi,IS and Creoles: Curre11t Trends a11d Perspectives, D. DeCamp & I. Hancock (eds.), 92-117. Washington DC: Georgetown UP. Rickford. J.R 1988. Dimensions ofa Creole Continuum. Stanford CA: Stanford University Press. Rickford. J.R 1992. The creole residue in Barbados. In Old English and New: Essays i11 La11guage a11dLinguistics in Honor ofFrederic G. Cassidy, J. Hall, N. Doane, & D. Ringler (eds.), 183201. New York: Garland. Sankoff, D., Tagl.iamonte, S. & Smith, E. 2005. GoldVarb X: A Multivariate A11arysis Application. Department of Linguistics, University of Toronto and Department of Mathematics, University of Ottawa Schneider, Edgar W. 1992. Negation patterns and the dine of creoleness in English-oriented varieties ofthe Caribbean. In Studies in Caribbean La11guage II: Papers from the Ninth Bie11nial Qmjerence of the Society for Caribbean Linguistics, 204-227. St. Augustine: Society for Caribbean Linguistics. Schneider, E. W. 1997. The dine of creoleness in negation patterns of Caribbean English Creoles. In Language History and Linguistic Modeling: A Festschrift.for Jacek Fisiak on His 6oth Birthday, R Hickey& S. Puppel (eds.), 1055-1067. Berlin: Mouton de Gruyter. Strang, B. 1970. A History of English. New York NY: Methuen. Van Herk, G. 2003. Barbadian lects: Beyond meso. In Contact Engllshes ofthe Eastern Caribbean, M. Aceto & J.P. Williams (eds.), 241-264. Amsterdam: John Benjamins. Walker, J. A. 2000. Present Accounted for: Prosody and Aspect in Early African American English. PhD dissertation, University of Ottawa Walker, J.A. 2001. Using the past to explain the present: Tense and temporal reference in Early African American English. Language Variation and Change 13(1 ): 1-35. Walker, J. A. 2005. The ain't constraint: Not-contraction in early African American English. Language Variation and Change 17(1): 1-17. Walker, J.A. 2010a Variation In Linguistic Systems. London: Routledge. Walker, J.A. (ed.). 2010b. Aspect In Grammatical Variation. Amsterdam: John Benjamins. Weldon, T. 1994. Variability in negation in African American Vernacular English. Language Variation and Change 6(3): 359-397. Williams, J. P. 1987. Anglo-Caribbean English: A Study of its Sociolinguistic History and the Development of its Aspectual Markers. PhD dissertation, University of Texas at Austin. Winford. D. 1983. A sociolinguistic analysis of negation in Trinidadian English. In Studies In Caribbean Language, L.D. Carrington (ed.), 203-210. St. Augustine: Society for Caribbean Linguistics. Winford, D. 1993. Predication in Caribbean English Creole.s. Amsterdam: JohnBenjamins. Wolfram, W. 1974. The relationship of White Southern speech to Vernacular Black English. Language 50(3): 498-527. Young, R & Bayley. R 1996. VARBRUL analysis for second language acquisition research. In Second Language Acquisition and Li11guistic Variation, R Bayley & D.R Preston (eds.), 253-306. Amsterdam: John Benjamins.
55
Putting individuals back in contact Accommodation strategies by Barbadians in Ipswich* Michelle C. Braiia-Straw
Researchers have commented that linguistic complexity and the level of social integration can constrain adults' ability to acquire phonological changes (Chambers 1992; Kerswill1996; Evans 2004). This preliminary study explores the extent to which first generation Barbadian immigrants acquire Ipswich English (IpsE) or retain their own phonology (Island Barbadian English, or IBE). Initial results indicate that Barbadians employ four different strategies for individual vowels: (1) maintenance, (2) partial accommodation, (3) full accommodation, and (4) phonetic divergence. One outcome is the emergence among Barbadian females of an interdialect form with the creation of homophones not present in IBE or IpsE. Finally, the paper raises the possibility of new dialect formation in subsequent generations.
1.
Introduction
Contact between speakers of different dialects and languages is a major source of language change and contributor to new dialect formation. Many of the contact situations researched have involved adult migration in the initial stages of contact We know that changes can be acquired across the lifespan of an individual (Trudgill 1986; Kerswil11996; Mees and Collins 1999; Nahkola and Sanilahti 2004; Sankoff and Blondeau 2007). In some cases, this will lead to an increase in style-shifting, in others to more permanent changes or to 'long term accommodation: a term used
*
This research is supported by the Economic and Social Research Council (grant no. PTA-
030-2002-0 1347).
58
Michelle C. Brai!.a-Straw
in Trudgill (1986: 11, adapted from Giles et al. 1973). 1 Few studies have discussed the extent to which adults contribute to language change in their community. This research is a preliminary investigation of the changes made by first-generation immigrants in adulthood and the nature of the input they provide to subsequent generations. In studying in a creole diaspora community, this research helps to expand the field of creole studies, which has tended to focus on the historical outcomes oflanguage contact Furthermore, by focussing on synchronic vowel change, the research begins to redress the balance in a field that has typically concentrated on the most basilectal speakers and on grammar, thereby presenting a fossilised view of creole societies. Creole societies are characterised by migration and change, and more studies are needed to research all aspects oflanguage in Caribbean communities whether in the Caribbean itself or in the diaspora. The Barbadians in this study first arrived in Ipswich, East Anglia, during the 1950s and 1960s. In pursuing the question of possible outcomes of contact, the Barbadians may have begun a process of convergence to IpsE having lived in Britain for over 40 years. Convergence then is the first possible outcome. This study considers the degree to which Barbadians have accommodated to IpsE as a starting point to determine the direction of change by subsequent generations of Barbadians. The main research questions directly addressed by this paper are: 1. What learning strategies are employed by first-generation Barbadian adults? 2. Do speakers differ in the degree to which they approximate Anglo patterns? 3. What role do linguistic and social factors play in individual speakers' learning strategies?
Divergence from IpsE by converging to Jamaican English is also possible because Barbadians may instead accommodate to the other main Caribbean variety in Ipswich.2 Just over half of all Caribbean immigrants to the UK during the 1950s and 1960s were Jamaican, whereas Barbadians represented just 8% of immigrants (Peach 1996: 28). It is understandable then that the main focus of research in the UK has been on Jamaicans with the presumption that speakers from other islands have adopted Jamaican features (Sutdifte 1982; Edwards 1986; Sebba 1993). However in Ipswich, Barbadians are almost as numerous as Jamaicans and more 'visible' in the community, 1. In this paper, I discuss the process of language change by adults as 'accommodation; acknowledging potential linguistic and social constraints faced by adults learning a new variety. However, researchers also use the term 'acquisition' with a focus more on the outcome rather than the process (Evans 2004). I employ both terms more or less used Interchangeably, but use 'accommodation' to emphasize process and 'acquisition' to emphasize outcome. 2. For the purposes of this study, I use the term divergence to refer to phonetic divergence in instances where Barbadians' vowels are more distant to the Anglo vowel position than the source dialect, IBE.
Putting individuals back in contact
especially in the public domain. For instance, when I first entered the community in 2000, I assumed that most Caribbeans in Ipswich were of Jamaican origin and was surprised by the cool reception I first received when I contacted the Ipswich Caribbean Centre to ask if I could talk to people of Jamaican background.3 I discovered that many people at the centre were Barbadian and that Barbadians and Jamaicans had maintained island rivalries in Ipswich. Bessy; one of the Barbadian informants in this study expresses her displeasure when outsiders assume she is Jamaican. (1)
Bessy ...because of our visibility and our nationality people would point us out, you know, "wunna Jamaica-~ everybody was a Jamaican or American because although we suppose so illiterate, they didn't know anything about us but we knew quite a bit about them.
Maintenance is possible in the event that Barbadians wish to assert their own identity as distinct from both Ipswich Anglos and Jamaicans. Where convergence or divergence is partial, then first generation Barbadians may provide the type of input (partial retention and incomplete assimilation) that characterizes the new dialect formation scenario discussed in Patrick (2004). To sum up, Barbadians may engage in one or more of the following processes:
1. Convergence to IpsE. 2. Convergence to Jamaican phonology present in the Ipswich Caribbean diaspora. 3. Maintenance of IBE phonology to emphasize Barbadian identity as distinct from both Anglo and Jamaican identities. 4. Partial maintenance of IBE or incomplete processes of assimilation to IpsE or Jamaican English.
2.
Individual language change in adulthood
Scholars agree that linguistic complexity constrains the types of features that can be acquired by adults (Trudgill1986; Chambers 1992; Kerswill1996). I am interested in adults' ability to alter their phonology. Chambers asserts that simple phonological rules can be acquired at any age but complex rules are not generally acquired beyond the age of puberty. Chambers defines simple and complex phonological rules this way: [Simple rules are] automatic processes that admit no exceptions, [and complex rules have] opaque outputs, that is, they have exceptions or variant forms, or- a type of complexity that comes up especially in dialect acquisition - they have in their output a new or additional phoneme. Cllambers ( 1992: 682) 3·
Names of speakers and organisations are pseudonyms.
59
6o
Michelle C. Brai!.a-Straw
A number of studies have reported the degree to which adults can acquire new phonological features in adulthood. Sankoff and Blondeau (2007) report that some speakers in Montreal acquire a new pronunciation of /r/ after the age of 15. They examined a discrete variable that involved the change from apical [r] to dorsal [R] in the same phonological environments. resulting in a simple change. At the other end of the scale, Payne ( 1980) reports that the 'short a' pattern in Philadelphia was highly complex and had to be learnt lexically. Not even children born in Philadelphia were able to learn the 'short d pattern unless their parents had also been born in Philadelphia. In the UK, Wells (1973: 34-35, 108-11 0) found that adult Jamaican immigrants in London could not acquire new phonological oppositions, such as hear/hair and beer/bare. In addition to linguistic constraints, social factors also help to determine whether speakers acquire new phonological features during adulthood. Evans (2004) explored the extent to which Appalachian migrants in Michigan acquired a Northern Cities Shift feature (low-front raising of/ae/). She found that social network and gender were better predictors of acquisition than age. Following these studies, I investigate both social and linguistic factors that might constrain adult Barbadians' acquisition ofipsE.
3·
Methodology
In Britain during the 1950s and 1960s, most Caribbean migrants settled in larger cities, but this study examines the original migrants to Ipswich, a small regional town with a strong East Anglian identity. The Barbadians in this study came directly to Ipswich and formed part of a larger Afro-Caribbean community in the town. The two largest groups were Jamaicans and Barbadians (34% and 21% of Mro-Caribbeans respectively. Census 1991). In total, Mro-Caribbeans represent just 2.7% of the Ipswich population. Table 1. Speakers' characteristics Pseudonym
Berty Buster Bessy Betty Scott Simon Sarah Susan
Age
Sex
64 66
M M
72
F
72
F
74 74 76 71
M M F F
Age of arrival
Year of arrival
Year of birth
Time in
21 21 29 29 Born Born Born Born
1960 1955 1960 1960
1939 1934 1931 1931 1929 1929 1927 1932
43yrs 45yrs 43yrs 43yrs
UK
Putting individuals back in contact
Table 1 displays speakers' characteristics. The study comprises data from four Barbadians and four Anglos. split equally between males and females, aged 64 to 76 years. The Barbadians were selected because they arrived in the first main wave of immigrants from the Caribbean to Britain and because they arrived in Ipswich as adults in their 20s. They are compared to Anglo speakers who are selected to match the Barbadians as much as possible. Anglos lived in the same mainly working class neighbourhoods as the Barbadians and had similar levels of education. I employed the 'friend-of-a-friend' technique and participant observation to locate participants in both communities (see Milroy and Gordon 2003: 75 for further discussion). During the fieldwork phase, I volunteered at the Ipswich Caribbean centre frequented by the Barbadians and was able to interact and observe their interactions with other members of the Caribbean community. Following my discovery of the rivalries between Barbadians and Jamaicans in Ipswich, I was careful to downplay my Jamaican heritage in favour of a Caribbean one, interested in the experience of Ipswich Barbadians. I had originally planned to study Jamaicans following previous studies in the UK, but had trouble finding Jamaicans in Ipswich. The discovery of the prominence of Barbadians in Ipswich provided the opportunity to broaden the scope of existing research with its focus on Jamaicans. Sociolinguistic interviews, reading passages, and word lists were recorded using a Marantz CDR 300 transportable recorder and AKG C420 headset condenser microphone. Eight vowels were analysed for this study: FLEECE, KIT, GOOSE, FOOT, THOUGHT, LOT, BATH, and TRAP. For each vowel, the same three lexical items per speaker were selected as far as possible from word lists to minimise the effects of phonetic conditioning. 4 Barbadians pride themselves on their high levels of literacy, and informants experienced no problems in reading the passages and word lists. I presume that speakers will pay most attention to their speech when reading word lists and will be motivated to display their most British English pronunciation following Labovs (2001a) attention to speech methodology. Furthermore, studies have shown that minority ethnic groups are under pressure to assimilate to Anglo norms and that immigrant varieties have traditionally been highly stigmatised accents in the UK (Giles 1970; Fought 2002). Vowel formant values were measured and analysed using the Kay Elemetrics Computer Speech Lab model4500. I measured and plotted the first (F1) and second formant (F2) values of the eight vowels, broadly corresponding to vowel height and the degree of advancement, respectively. F1 and F2 measurements were taken at the temporal midpoint of the vowel. Boberg (ms.) has found that temporal midpoint 4 Bessy and Buster were recorded for a previous study and several of their lexical items differ. This introduces an element of bias into the data, requiring a cautious approach to interpreting results. A larger study is planned that will aim for a more balanced dataset in terms of token selection.
61
62
Michelle C. Brai!.a-Straw
often coincides with the central tendency of a vowel for long vowels. In the related pilot study. I conducted a more detailed set of measurements for long vowels that confirmed Boberg's observations (Straw 2006: 130-135).5 The formant values were normalised for differences in speakers' vocal tract length, using the LobanovZ-Score transformation (see Lobanov 1971). Vowel formant charts displayed the position of the vowels in perceptual space, the first formant corresponding to vowel height and the second formant corresponding to vowel backness and lip rounding.
4· Analysis Complex linguistic changes are unpredictable in nature. At their most complex, speakers must learn them lexically. In the context of vowel change, vowels exist in continuous space so that change in one vowel potentially impacts another. The following section discusses the extent of difference between IpsE and IBE andreveals that some changes can be considered simple, i.e. where all lexical items of a vowel move to a vacant position in the speaker's system. GOOSE is the clearest example of a simple change. More complex changes arise from vowels that move into the vowel space of an existing vowel. In this case, Barbadians must make multiple simultaneous adjustments to their vowel systems and the extent to which they make a change in one vowel impacts another. Partial accommodation or divergence strategies are likely to have the kind of unpredictability of outcome commented upon by Sankoff and Blondeau (2007: 583). However, I propose that a fine-grained instrumental analysis of the processes of vowel change can account for at least some of this apparent unpredictability. 4.1
Linguistic factors
As there are no published data on vowel variation in Barbados, the description in Wells (1982) is taken to represent Barbadians' first dialect 6 Figure 1 shows the position oflsland Barbadian English (IBE) which is marked by the lexical sets. The arrows mark the changes required for IBE to match IpsE. The endpoint of each
5· In the related pUot study in Straw (2006: 132-135), three Barbadians and three Anglo females were analysed using two different sets of measurements for vowel nuclei and compared at 25ms from vowel onset and at vowel midpoint). The results confirmed that patterns of interspeaker and intraspeaker variation were replicated and that midpoint measurements oflong and short vowels were sufficiently reliable. 6.
Wells's (1982) sources are Collymore (1970), Hancock (1971), and Haynes (1973).
Putting individuals back in contact
FLEECE
+-------GOOSE FOOT
KIT
~,~~/w~.7i BATH
THOUGHT=NORTH
*IBB source diale..-r. according to Wells (1982: 585)
Figure 1. Island Barbadian source dialect and target Anglo dialect* arrow marks the position of the IpsE vowels. FLEECE, KIT, and FOOT have similar vowel qualities in both varieties. Barbadians must front GOOSE to acquire IpsE, a simple phonetic change (Chambers 1992).7 Other changes are more complex in that IpsE has mergers for BATH, START, and PALM as well as THOUGHT, NORTH, and FORCE, whereas the vowel distribution of IBE is radically different (Wells 1982: 585). In order to examine the potential changes needed to IBE phonology, one token belonging to the BATH lexical set (bath) and two from the START lexical set (start, cart) have been analysed for all speakers. Similarly, one each of the THOUGHT, NORTH, and FORCE lexical sets are analysed to identify the direction of changes made by Barbadians and to provide the basis for a more in -depth investigation with more token numbers and speakers. Barbadians need to adjust TRAP and BATH, currently low central vowels, by moving BATH to low back position and fronting TRAP. Although in IBE, TRAP and BATH are both low central vowels, they are still phonemically distinctive by vowel length. For the Anglos, BATH, START, and PALM are merged, particularly because IpsE is non-rhotic. BATH comprises longer vowels than TRAP, but vowel length is not phonemically distinctive in IpsE, though Wells (1982: 585) claims it is for Barbados. 8 For the Barbadians, rhoticity plays a role because Barbadians are fully rhotic. Therefore, START is distinguishable from BATH and TRAP 7· The reader should note that when discussing changes to Barbadians' vowel systems, this paper does not discuss observable change but rather infers change by comparison to Barbadians current systems with IBE or presumed source dialect and with IpsE, the target dialect Therefore, terms such as "front" and "raise" are inferred changes rather than actual observable movements. 8. No tokens from the PALM lexical set were analysed for this study. I use the BATH lexical set therefore to incorporate START where they are fully merged. I distinguish BATH from START in the discussion of the Barbadians vowels because they are distinguishable by rhoticity.
63
64
Michelle C. Brai!.a-Straw
(Wells 1982: 585). Whereas for BATH, Barbadians need only back this vowel to match Anglos, they must cease to pronounce /r/ in postvocalic environments of the START lexical set, e.g. pm·t, sta.rt, cart. Barbadians must also separate LOT from THOUGHT/NORTH, the back vowels. Again, this might be possible because LOT is phonemically shorter than THOUGHT according to the available data on IBE (Wells 1982: 585). Barbadians need to raise LOT to low-mid back position. They also need to merge the nucleus of FORCE [o:r] with the nuclei of NORTH and THOUGHT [a~v], which involvesraisingTHOUGHTandNORTH up to FORCE in mid-high position and optionally raising all three to high-back position, a merger which is theoretically acquirable at any age (Kerswill1996: 200). The Anglos are non-rhotic, therefore Barbadians will have to lose their pronunciation of /r/ post-vocalically for NORTH and FORCE lexical sets. Rhoticity is a strong stereotypical feature of IBE and recognised as a stereotype by other Afro-Caribbeans in the Caribbean. It is likely to be highly salient both for Anglos and Afro-Caribbeans and, according to Chambers (1992), should be among the first features to be lost. Phonotactic constraints in the speaker's own variety may hinder the process of learning a new variety. The BATH lexical set difters for Barbadians and Anglos. Ipswich Anglos have fully merged BATH and START lexical sets (Straw 2006: 155). In the Barbadians' first dialect, START is distinguished from BATH by rhoticity: the vowel is followed by /r/. Barbadians also have overlap in TRAP and BATH/ START vowel quality, whereas for Ipswich Anglos, TRAP is a front vowel and BATH/START are back vowels. Barbadians must separate TRAP from BATH/ START to acquire IpsE vowel quality. The distinction between BATH and START because of rhoticity may interfere with the process of accommodation to IpsE. Additionally, where pronunciation changes lead to conflicts with existing vowels in a speaker's own dialect, the speaker may not accommodate to the new variable. The potential for homonymic clash might hinder Barbadians in accommodating to the Anglo system. Barbadians already have a back vowel (THOUGHT/NORTH) in the position where they would need to move BATH/START. In separating TRAP and BATH/START, Barbadians may be constrained by the need to avoid homonymic clash between the START vowels and NORTH vowels. If rhoticity is maintained there is a high potential for homonymic clash, producing homonyms with some highly frequent lexical items such as far-for and are-or. If Barbadians lose rhoticity, there is a much lower potential for homonyms with THOUGHT: taught-tart and caught-cart.
4.2
Social factors
In addition to the linguistic constraints on accommodation discussed above, the degree to which Barbadians can access Anglo networks is likely to affect their ability
Putting individuals back in contact
to identify and accommodate to the IpsE vowel system. Access to Anglo networks allows Barbadians to identify the Anglo target variety (see LePage and TabouretKeller 2006: 182 for a discussion of the constraints on individuals' ability to change their speech). Close-knit communities that are norm-enforcing, like the networks among Barbadians and other Afro-Caribbeans, have been found to hinder language change (see for example Milroys 1987 study of Belfast communities). Barbados has maintained close ties with Britain throughout its colonial history, firstly through slave ownership and later through government and education, up until Barbados became independent in 1966. Links have since continued due to the rise in tourism in the last 40 years. Today. Barbados is one of the most popular destinations for British tourists. At the time that the Barbadians in this study were growing up in Barbados, the legacy of colonialism had given rise to a racialised society where Whites had the highest socio-economic status, followed by Coloureds (Barbadians of mixed European andAfricandescent) and lastly Blacks(Le Page and Tabouret-Keller 2006: 48). Barbados was also a gender-divided society. Race and gender would have determined the types of social networks in which the Barbadians operated, influencing their Ianguage and the extent to which they used Barbadian creole grammar rather than standard Barbadian English grammar taught in the education system (Blake 1997; Van Herk 2003). Their social networks would also have determined the level of exposure to speakers of other varieties, including British English, but we do not know what impact this exposure would have on Barbadians' phonology. However, Wells (1982: 585) provides a description of Barbadian vowels with some indication of variation. It is clear from a comparison of Wells and descriptions of Suffulk English that the GOOSE, LOT, THOUGHT and BATH lexical sets are different in the two varieties. Holmes (1993) notes the tendency for first generation male immigrants in Australia to acquire the host language more successfully than females because males have more opportunity than females to develop wide social networks in the new country. The Barbadians in this sample came to England prior to Barbadian Independence and would therefore have grown up in a Barbados which was subject to British rule and a British education system. They all followed other family members, a typical pattern of migration from the Caribbean. All except one were keen to improve their education and employment prospects. Berty stands out from the other Barbadians because he was considered a Coloured person in Barbados, and as such, had more opportunities and access to different groups of people than is usual for Blacks. His parents were able to pay for higher education and he had higher job aspirations than the other Barbadians - to be a dentist. In contrast, Betty was raised in a more rural location than the other speakers and did not have aspirations to improve her situation. Rather, she moved to England under pressure from other family members and would have preferred to stay in Barbados. The females' social networks were demonstrably more restricted in Barbados than the
65
66
Michelle C. Brai!.a-Straw males. Female networks involved family or friends of family. while the males had networks including members drawn from outside of the family. Wider social networks in Barbados are likely to have determined the extent to which Barbadians might adopt more standard speech. Berty is a more standard speaker than the other Barbadians in the sample due to his higher social position and wider social networks in Barbados. We can see from an excerpt that Berty uses predominantly standard grammar. (2) Berty It was fun growing up in Barbados, as one would expect from, you know... a tropical island, urn, and so on. School was ...I always say to myself~ and a number of people has agreed with me as well, the education system in Barbados was one of the best in the world, without a doubt, urn ... As a matter of fact, if I could have afforded it, I would have sent all my kids back to Barbados to be educated. The only non-standard grammatical feature is lack of agreement between the subject people and the verb has. Berty is also more likely to have standard pronunciations of consonantal variables that are represented in standard orthography. In this excerpt, the is pronounced using standard pronunciation rather than non-standard [d] typical in Barbadian and other Creoles. Blake (1997: 40) observes that the most creole-like features are likely to be heard in the speech of rural Barbadians. Betty's social situation would have afforded her little opportunity to access British varieties and her restricted social position and rural upbringing indicates that she came to Britain speaking a more creole-like variety. The excerpt below contains several non-standard grammatical features. (3) Betty
He ra gone round the back. Come! Shut back and come, because I know who it is, I just see him pass across there. You can go th1"0ugh the back and tell him fm busy. [Interviewer] Do you want him to come back later or... ? I don't know, he ain't coming back. Just open the door, can't remem if it ra lock.
In this excerpt, Betty has two instances of copula deletion (he gone, it lock) as well as non-standard pronunciations for thmugh, the [tru, di] and non-marking of past tense (I just see him, it ra lock). Barbadians need to negotiate their social position in the new society. and the extent to which they accommodate to Anglo norms is likely to depend on their degree of integration into Anglo networks but also the degree to which they are able to identify the IpsE vowels (LePage and Tabouret Keller 2006: 182). Wider networks and more exposure to a range of different varieties in Barbados is likely
Putting individuals back in contact to increase Berty's awareness of differences between varieties in general and facilitate the process of accommodation to new varieties. Migration to Ipswich has resulted in changing social networks for the Barbadians. Initially, Barbadians' closest networks were with family and other AfroCaribbeans. The Barbadians' networks widened to the Anglo community through work, and in Berty and Bessy's case, through education. Berty again stands out as having had most access to Anglo networks. He had one year in further education in Ipswich and initially worked in a factory. After 10 years in Britain, he established his own business and bought a house in a traditionally white residential area. Berty's wider networks with Anglos through his political activities led to him to become a councillor and eventually, mayor. He has also developed networks with the Anglo and Afro-Caribbean communities through sustained community work. The other Barbadians were also involved with many similar activities, but the factors which set Berty apart from the other Barbadians are his higher-status employment and his residence in a middle-class area. In the discussion that follows, I argue that Ipswich Anglos have local pronunciations for several vowels and Barbadians would probably not have encountered these pronunciations before arriving in Ipswich.
5·
Results
5.1
GOOSE vowel
Table 2 shows each speaker's mean F2 scores for GOOSE, indicating the degree of vowel fronting. Speakers' vowel systems have been normalised to take account of vocal tract differences between males and females. The range of a vowel is based on a larger sample of 12 Anglos and 12 Barbadians and is between zero and 2.5 for a front vowel and between zero and -2 for a back vowel. The scale is between zero and -2 for a high vowel and between zero to 3 for a low vowel. 9 Therefore, a high front
Table 2. Mean F2 values for GOOSE Anglos
GOOSE Mean
Sarah Scott Simon Susan
0.7567 1.4364 0.7114 0.6245
9·
Barbadians
Berty Bessy Betty Buster
GOOSE Mean -0.2099 -0.4315 -0.7565 -1.0117
These values represent the transformed Hertz values according to the Lobanov scale.
67
68
Michelle C. Brai!.a-Straw
vowel would have a score close to Fl = -2 and F2 = 2.5, and a high back vowel would have values close to Fl = -2 and F2 = -2. Positive values greater than zero indicate a front voweL and negative values greater than zero indicate a back vowel. For the Anglos, GOOSE is a central-to-front vowel and individual tokens can front as far as KIT and FLEECE. Barbadians must front GOOSE and they vary in their implementation of this strategy, a simple phonetic change for this word class. While Barbadians have not acquired the level of fronting shown by Anglos, the range of variation suggests small changes. Berty stands out because his GOOSE is almost a central voweL indicating partial accommodation, while the other male, Buster, has a back vowel indicating maintenance. For the females, GOOSE is a more front vowel than Buster's but less front than Berty's. 5.2
TRAP and BATH vowels
Figures 2 and 3 plot the TRAP and BATH vowels for Anglos and Barbadians respectively. TRAP is marked by squares on the graphs. Each speaker's vowels are plotted in different shades. The vertical line on the graph indicates the position of a central vowel with respect to the degree of tongue advancement. The horizontal line marks a mid-vowel position in terms of tongue height. Name-shade
Sarah- white
Simon - black Susan - grey shadow
Scott- grey 2.5
2.0
1.5
1.0
0.5
F2 0.0
-0.5
-1.0
-1.5
-2.0
-2
TRAP
D
BATH
6
-1.5 -1
-0.5 0
•
......
•
•
• I D
llbo
.
,t"" ""
•• ... • • •
Figure 2. TRAP and BATH vowels- Anglos
0.5
1.5 2
F1 2.5
Putting individuals back in contact
Name-shade
2.5
Berty- white
Betty- black
Bessy- grey
Buster- grey shadow
2.0
1.5
1.0
F2
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2
TRAP
D
BATH
D
-1.5 -1
-0.5
•
0 ~
• • D
....
ll
~
•
....
.... ~
D
•
0.5
ll
\.M
1.5
•
2 F1 2.5
Figure 3. TRAP and BATH vowels- Barbadians
The main point to note is that there is no overlap in vowel space between TRAP and BATH for Anglo speakers. TRAP is a low, front vowel and there is little variation between speakers with the exception of Scott, who has produced one midfront token (grey square). BATH has a greater vowel height (Fl) range than TRAP and is a low back voweL positioned to the right of the vertical central line. From Figure 3, the substantial overlap in vowel quality between TRAP and BATH is immediately observable for Barbadians, unlike Anglos. Three Barbadians have a front TRAP vowel. Betty, however, has a very diffuse TRAP vowel (black squares). 10 Buster is the only Barbadian to maintain a central IBE TRAP vowel (grey shadow squares), whereas the other three Barbadians have acquired the front IpsE vowel pattern for TRAP. Betty and Bessy have front tokens of BATH in the vowel space of TRAP, partially retaining the IBE overlap in vowel quality. They have not accommodated to the Anglo pattern by backing BATH but rather have diverged by fronting BATH Only two tokens are visible on the graph because her third token was an extremely low outlier F 1 (5.2) and has been excluded. None of the speakers in the larger sample in Straw (2006) had a value below F3=3. 10.
69
70
Michelle C. Brai!.a-Straw
along with TRAP perhaps. Vowel length does not seem to be operating as a distinctive feature (see discussion in Section 6 below). Furthermore, Betty has a very diffuse vowel for BATH (as with TRAP), indicating that she has changed her pronunciation of the vowels but that she has no recognised target. Buster shows little evidence of accommodating to Anglo patterns, retaining the IBE system for both vowels. Berty is the only Barbadian to have had some success. In fronting TRAR he has managed to separate it from BATH, even though he has not altered his pronunciation of BATH from the IBE pattern. 5·3
LOT and THOUGHT vowels
Figure 4 displays the vowel formant charts of LOT and THOUGHT for Anglos and Figure 5 for the Barbadians. For all the Anglos, LOT is located in the lower right-hand quadrant of the vowel chart, a back vowel ranging in height from low to mid. There is no overlap with speakers' THOUGHT vowel which is located in the top right-hand quadrant, a mid- to high-back vowel. Additionally, each speaker has a high degree of separation between LOT and THOUGHT vowels. For example, Sarah's THOUGHT vowel (white squares) is highly separated from her LOT vowel (white triangles). Name - shade
Sarah -white Soott - grey
Simon -black !Wan- grey shadow F2
2.5
2.0
THOUGHT
15
1.0
05
0.0
-O.S
-1.0
-15
-2.0 -2.0
•
• 'i••
D
6
lltl•
•
. l.
1/1 ....
~~~
.. ..
"'
•
-0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0
Figure4. LOT and THOUGHT vowels- Anglos
Putting individuals back in contact
Name-shade Berty- white Betty - black Besay - grey Buster - grey shadow
2.5
2.0
1.5
1.0
0.5
F2 0.0
-0.5
-1.0
-1.5
-2.0 -2.0
THOUGHT
D
LOT
~
-1.5 -1.0
. •• "' • .ll. • D
AD
~l/ll
""'•
•
•
-0.5 0.0 0.5 1.0 1.5 2.0 F1 2.5 3.0
Figure 5. LOT and THOUGHT vowels- Barbadians
Barbadians' LOT vowel is higher than the low-back position of IBE and higher than the low to low-mid position oflps.E, indicating overshoot. Most tokens are in mid position. Berty (white triangles) and Bessy (grey triangles) most approximate the position of the Anglo system. Like the Anglos, their LOT vowel lies in the lower right-hand quadrant. Buster (grey shadow triangles) and Betty (black triangles) have a higher LOT vowel than Anglos. They show evidence of overgeneralisation by having raised LOT with THOUGHT. Overgeneralization was also present for BATH. For the THOUGHT voweL Barbadians have raised THOUGHT from the low-back position ofiBE but vary in the degree to which they have raised this vowel and have not reached the Anglo target. Berty is the only Barbadian who has a higher THOUGHT vowel than LOT and most resembles the Anglo pattern. Figure 6 displays Barbadian speakers' vowel formant charts for BATH and THOUGHT. Barbadians have partially accommodated towards the Anglo pattern for THOUGHT, resulting in an interdialectal form in which BATH now partially overlaps with THOUGHT, a feature that is present in neither IBE nor IpsE. This feature is found in both females' vowel systems (Betty and Bessy) but not in male Barbadians' systems.
71
72
Michelle C. Brai!.a-Straw
Name-shade
Berty- white Bessy- grey
2.5
2.0
Betty- black Buster- grey shadow
1.5
1.0
F2
0.5
0.0
-0.5
-1.0
-1.5
-2.0 -2
THOUGHT BATH
D
-1.5
6
-1 D
•
-0.5
D
..
[).
....
~
....
....
&"..
[).
•••
•
0 0.5
1.5 2
Fl 2.5
Figure 6. BATH and THOUGHT vowels- Barbadians
6.
Summary of results and discussion
The Ipswich Anglo pronunciation of GOOSE, LOT, and THOUGHT are traditional Suffolk features that have been traced back as far as the late 19th century (Ellis 1889, Kokeritz 1932). IBE speakers may be more or less standard speakers in their use of grammar, but there is no evidence of these vowel forms existing in Barbados. Any approximation to Anglo vowels can reasonably be presumed to have taken place in Ipswich. Table 3 provides a summary of the changes in vowel quality inferred for each Barbadian speaker. All speakers have accommodated in their speech. Berty, in particular, stands out as having approximated the most to IpsE. He acquires a similar vowel quality to Anglos in three of the five vowels (GOOSE, TRAP, and LOT), followed by Bessy who acquires two vowels (TRAP and LOT). According to Chambers (1992), GOOSE, TRAP, and LOT involve a straightforward gradual movement of the lexical sets into positions previously unoccupied in the Barbadians' vowel system. Presumably if TRAP is phonemically distinguished from BATH by vowel length, then Barbadians should be able to make the necessary change in vowel quality.
Putting individuals back in contact Table3. Summary of accommodation strategies employed by Barbadians Name
Buster and Betty difter from Anglos in all five vowels. Berty>s social position might account for his greater similarity to Anglo patterns. His social networks and social status in Barbados and Ipswich difrer from the other Barbadians. He had access to a greater range oflooser and less dense social networks with Anglos which could have facilitated his language change. Milroy's (1987) study finds that language change occurred most in individuals who had loose, uniplex social networks. In the US, Evans (2004) investigated the language change of Southern Appalachian migrants to Michigan in the North. She found that social networks were a greater predictor of acquisition of the northern dialect than other factors, such as age. In Barbados, Berty had a higher level of education and his higher racial status as a Coloured person might have afforded him access to a wider range of social groups, including British English speakers. In Ipswich, he owned his own customer-oriented business which would have provided him with greater opportunities to interact with Anglos. Furthermore, he lived in a traditionally White residential area and owned his own house. It is likely that Bertfs access to Anglo networks in
73
74
Michelle C. Brai!.a-Straw
Ipswich would have provided him with the ability to identify the Anglo variety, an essential precursor for accommodation (see also Le Page and Tabouret-Keller 2006: 182 discussion of individual agency in language change). Table 3 also reveals four different strategies used by individuals. 1.
No accommodation/maintenance. Both the male Barbadian speakers show
maintenance, with Buster showing the most maintenance. The females show no evidence of maintenance. This study is too small to make claims about gender patterns. Further study will investigate the role of gender in acquisition and the generalisation that females tend to lead males in adopting changes in progress (Labov 2001b: 261-293). In this instance, we are considering whether female migrants lead males in adoption of features in the context of language acquisition. 2. Partial accommodation - undershoot and overshoot. All speakers have evidence of undershoot where vowels have moved in the direction of the Anglo variety but have not reached the target. For example, all speakers show undershoot in their THOUGHT vowel, which is higher than in IBE but not as high as the Anglo pattern. Buster and Betty show evidence of overshoot in their LOT vowel, which seems to have risen along with THOUGHT, partially maintaining the overlap in vowel quality with THOUGHT. Recall that overlap between THOUGHT and LOT is a feature ofiBE. Wells (1982: 585) states that although THOUGHT and LOT have the same vowel quality in IBE, they are phonemicallydistinguished byvowellength. However, Straw (2006: 174-175) observes that the differences in vowel length between LOT and THOUGHT for the Barbadian speakers is highly variable and is substantially less than that required for phonemic distinctiveness (1.3: 1 for the Barbadians when the minimum ratio for distinctiveness is considered to be 1.6: 1 - see Beckford Wassink 1999: 189). Additionally, Barbadians seem to have maintained distinctive vowel quality in their THOUGHT, NORTH, and FORCE tokens. There are too few tokens for a robust generalisation, but Figure 5 shows that Bessy in particular has a diffuse vowel for THOUGHT. Barbadians continue to distinguish the THOUGHT lexical set from the NORTH and FORCE sets because they retain rhoticity in the latter two sets. One reason for Barbadians maintaining rhoticity may result from the observation that first-generation AfroCaribbeans maintain island affiliations after migration to Britain (Western 1992; LePage and Tabouret-Keller 2006: 47; Braiia-Straw 2007). The two most distinctive groups for Ipswich Afro-Caribbeans were Barbadians and Jamaicans. The Barbadians in this study all self-identified as Barbadian and were keen to maintain their Barbadian identity in the Afro-Caribbean community, often in contrast to Jamaicans. They were also identified as Barbadian from
Putting individuals back in contact
their speech by other Afro-Caribbeans in Ipswich. Full rhoticity has been noted as a stereotypical feature of Barbadian speech compared to other Anglophone Caribbean varieties (Wells 1982: 584; Van Herk 2003; Blake 2004), and this feature might have been retained by Barbadians in Ipswich to signal their island Barbadian affiliation in the wider Ipswich Afro-Caribbean community. 3. Phonetic divergence. In production, this is the observed outcome of two difterent strategies. Betty, in her TRAP and BATH vowels, shows a kind of divergence because her vowels are so diffuse that they are no closer to Anglo patterns than to IBE. It may be that Betty has not identified the Anglo vowels as a target. Another form of divergence is evident in Bessy's BATH vowel, which is front along with TRAP, maintaining a partial overlap of the lexical sets. Her BATH vowel is further away from the Anglo back vowel than it would otherwise have been, which is technically divergence. Lack of distinctive vowel length between TRAP and BATH for Barbadians (ratio 1.4: 1, Straw 2006: 175) has resulted in BATH and TRAP possessing the same vowel quality as in IBE. Future research is needed to understand how speakers orient to these differences. 4. Full accommodation/acquisition. Two Barbadians, Berty and Bessy, show evidence of acquiring Anglo patterns. supporting the notion that simple phonetic changes are possible for adults. In this study, the short vowels TRAP and LOT are acquired by both speakers. These changes involve movement to positions that are unoccupied by other vowels in speakers' systems. However, BATH and THOUGHT involve an order of complexity that makes them harder to acquire for Barbadians.
7·
Conclusion
Mobility and change are integral to creole societies, both in the Caribbean and in the diaspora. This study, with its focus on British Barbadians, extends existing research beyond language change in island creole communities to that of the diaspora in Britain. Diaspora studies in Britain have predominantly concentrated on the grammar of Jamaicans, the largest of the Caribbean groups to settle in Britain. However, there exist a number of areas in the UK where other groups predominate, for example, the prominent Barbadian communities in Reading and Ipswich. This study has shown that Barbadians have very different identities as compared to Jamaicans, and although they do participate in the wider Afro-Caribbean community, it is important for them to distinguish themselves from Jamaicans, at least in the first generation. All Barbadians in this study show evidence of language learning and vowel change in adulthood. Studies such as Holmes (1993) highlight the importance of
75
76
Michelle C. Brai!.a-Straw immigrants' social networks in their new country. but preliminary results suggest that the Barbadians' social position and networks prior to migration and afterward affect the degree to which they approximate Anglo patterns. A broadening of networks into the Anglo community clearly facilitates change, and a broader set of networks in their native country increases exposure and ability to identify different varieties. Phonotactic differences between varieties (i.e. rhoticity and differences in vowel quality) do seem to hinder acquisition in adults and have interacted with social identity, particularly island affiliation, to hinder accommodation. Speakers' vowels change within a continuous space such that change in one vowel has an impact on other vowels in the system. Many of the studies of adult language change have been of discrete variables or have been treated as such (Evans 2004; Sankoff and Blondeau 2007). One of the side effects of partial accommodation for Barbadian females is the emergence of an interdialectal form resulting in START and NORTH/FORCE partially sharing the same vowel space, creating new homophones such as pm·t-port and fa.rce-fm·ce. This interdialectal form resulting from accommodation processes is not in evidence for males. Patrick (2004) summarises the main outcomes for minority immigrant communities. They either assimilate to become native speakers of the host community's dialect or they maintain their first dialect to function as a marker of in-group ethnic identity. This study forms part of a larger three-generational study of British Barbadians which has the overarching aim of exploring a rare third scenario: new dialect formation resulting from strategies that fall somewhere between retention and assimilation (Straw 2006). The findings from this study show that the process of change begins with the first-generation adults who employ a range of strategies that make new dialect formation possible. Moreover, Straw (2006: 166) notes that BATH tokens overlap in vowel space with THOUGHT tokens in second-generation Barbadian females. More data is required to determine whether the interdialectal form evident in first- and second-generation females might be an innovation that establishes itself in the Barbadian community as a social identity marker. Research is underway to investigate adult accommodation in the wider Afro-Caribbean community in Ipswich and to discover the role of island affiliation, gender, and social networks in second dialect acquisition.
References Beckford Wassink, A 1999. A Soclophonetic Analysis of Jamaican Vowels. PhD dissertation, University of Michigan. Blake, R. 1997. ~ll d we is one?' Race, Class, and Language in a Barbados Community. PhD dissertation, Stanford University.
Putting individuals back in contact make, R. 2004. Bajan: Phonology. In A Handbook of Varieties ofEnglish, Vo11, E. W. Schneider, K. Burridge, B. Kortmann, R. Mesthrie & C. Upton (eds.), 501-507. Berlin: Mouton de Gruyte.t: Boberg, C. The acoustic measurement of English vowels. Ms., McGill University. Braila-Straw, M.C. 2007. Maintenance or assimilation? Phonological variation and change in the realization of It/ by British Barbadians. In Synchronic and Diachronic Perspectives on Contact Languages, M. Huber & V. Velupillai (eds.), 3-22. Amsterdam: John Benjamins. Census. 1991. County report Suffolk [Ptl, Table 7]. OPCS. Chambers, J. K. 1992. Dialect acquisition. Language 68(4): 673-705. Collymore, F. A. 1970. Notes for a glossary ofwords and phrases of Barbadian dialect. Bridgetown, Barbados: Advocate Co/Barbados National Trust (Cover title is just Barbadian Dialect). Edwards, V. 1986. Language in a Black Community. Clevedon: Multilingual Matters. Ellis, A.J. 1889. On Early English Pronunciation, with Especial Reference to Shakespeare and Chaucer (Part V reprinted 1999). Suffolk: Early English Text Society. Evans, B. 2004. The role of social network in the acquisition oflocal dialect norms by Appalachian migrants in Ypsilanti, Michigan. Language Variatlo11 and Change 16(2): 153-167. Fought, C. 2002. Ethnicity. In The Handbook ofLaJ~guage Variation and Cha11ge, J.K. Chambers, P. Trudgill & N. Schilling-Estes (eds.), 444-472. Malden MA: mackwell. GUes, H. 1970. Evaluative reactions to accents. Educational Review 22(3): 211-227. GUes, H., Taylor, D., & Bourhis, R 1973. Towards a theory of interpersonal accommodation through speech: Some Canadian data. Language in Society 2(2): 177-192. Hancock, I. 1971. A surveyofthepidgins and creoles of the world. In D. Hymes (ed..), 509-525. Pidginization and Creolization ofLanguages. Cambridge: CUP. Haynes, L. 1973. Language in Barbados and Guyana. Attitudes, behaviours, and comparisons. PhD dissertation, Stanford University. Holmes, J. 1993. Immigrant women and language maintenance in Australia and New Zealand. b1ten1ational Joun1al of Applied Linguistics 3(2): 159-179. Kerswill, P.E. 1996. ChUdren, adolescents, and language change. Language Variatio11 and Change 8(2): 177-202. KOkeritz, H. 1932. The Phonology ofthe Suffolk Dialect: De.scriptlve a11d Historical Inaugural Disser tatlon. Uppsala: .Appelbergs Boktryckeri Aktiebolag. Labov, W. 2001a. The anatomy of style-shifting. In Style and Sociolinguistic Variation, P. Eckert & J.R. Rickford (eds.), 85-108. Cambridge: CUP. Labov, W. 2001 b. Principles ofLi11guistlc Change: Social Factors. Oxford: Blackwell. LePage, R B. & Tabouret-Keller, A. 2006. Acts ofIdentity: Creole-Based Approaches to Language and Ethnicity, 2nd edn. Fernelmont: E.M.E. Lobanov, B. M. 1971. Classification of Russian vowels spoken by different speakers. Journal of Acoustical Society ofAmerica 49(2B): 606-608. Mees, I. M. & Collins, B. 1999. Cardiff: A real-time study of glottalization. In Urban Voices: Accent Studies in the British Isles, P. Foulkes & G. Docherty (eds.), 185-202. London: Arnold. Milroy, L. 1987. Language and Social Networks, 2nd edn. Oxford: Blackwell. Milroy, L. & Gordon, M. 2003. Sociolinguistics: Method and Interpretation. Oxford: Blackwell. Nahkola, K. & SaanUahti., M. 2004. Mapping language changes In real time: A panel study on Finnish. Language Variation and Change 16(2): 75-92. Patrick, P.L 2004. British Creole. InA Handbook ofVarietlesofEnglWz, Vo11, E.W. Schneider, K. Burridge, B. Kortmann, R. Mesthrie &C. Upton (eds.), 231-243. Berlin: MoutondeGruyter.
77
78
Michelle C. Brai!.a-Straw Payne, A. 1980. Factors controlling the acquisition of the Philadelphia dialect by out-of-state children. In Locating Language in Time and space, W. Labov (ed), 143-178. New York NY: Academic Press. Peach, C. 1996. Ethnicity in the 1991 Census, Vol 2. London: HMSO. Sankoff, G. & Blondeau, H. 2007. Language change across the lifespan: /r/ in Montreal French. Language 83(3): 560-588. Sebba, M. 1993. London Jamaican: Language Systems In Interaction. London: Longman. Straw, M. 2006. Dialect Acquisition and Ethnic Boundary Maintenance. PhD dissertation, University of Essex. Sutcliffe, D. 1982. British Black English. Oxford: IDackwell. Trudgill, P. 1986. Dialects In Contact. Oxford: Blackwell. Van Herk, G. 2003. Barbadianlects: Beyond meso. In Contact Engllshes ofthe Eastern Caribbean, M. Aceto & ]. P. Williams (eds.), 241-264. Amsterdam: John Benjamins. Wells, J.C. 1973. jamaican Pronu11clation i11 Lo11don. Oxford: Blackwell Wells, J.C. 1982. .AJ;cents of English, VoL 3. Cambridge: CUP. Western, J. 1992. A Passage to England: Barbadian Londo11ers Speak of Home. Minneapolis MN: University of Minnesota Press.
Relative markers in spoken Standard Jamaican English Ulrike Gut
The aim of the study is to test the claim that Jamaican Creole is the dominant shaping influence on the emerging spoken Standard Jamaican English (SJE) and the opposing claim that its structural properties are the result of contact with standard English in schooling and literacy. Focussing on relativization strategies, a total of 4,287 relative clauses with overt relative markers were analysed, drawn from the Jamaican component of the International Corpus of English (Greenbaum 1996). No direct or indirect influence from Creole was found in spoken S]E. Rather, there is ample evidence that the relativization strategies of spoken S]E reflect influences from the written English standard.
1.
Introduction
The complex linguistic situation of}amaica can be described as a lectal continuum that ranges from the basilect Jamaican Creole (Patois) to the acrolect Jamaican English, two separate linguistic systems which are linked by intermediate mesolectal varieties (e.g. DeCamp 1971). It appears that the two poles of the language continuum are not genetically related: the acrolect is assumed to have emerged from 17th and 18th century input of British dialects, whereas the basilect is regarded as a product of creolization caused by earlier language contact between British dialects and Niger-Congo languages that resulted in extensive structural mixing (Patrick 2004). Jamaican Creole is the dominant language of the majority of the population and is used in everyday spoken communication. The choice between Creole and acrolectal English, for those speakers who have command of both, is determined by socio-functional factors. English is appropriate in writing and formal spoken communication while Creole is used in informal spoken communication. When used in a formal context, Creole usually has an "anti-formal effect" (Allsopp 1996). Creole is very rarely used in written communication, although increasingly so in new informal text types such as e-mails (Hinrichs 2006).
so
Ulrike Gut
It has been variously suggested that in Jamaica, a local standard variety of English- Standard Jamaican English (SJE) -is emerging both in spoken and written communication (Shields 1989: 42; Mair 2002: 31; Irvine 2004, 2008). This conviction is based on sociolinguistic and linguistic evidence. As described by Schneider (2003), the emergence of a local standard English in postcolonial societies is closely connected with cultural and societal decolonisation events and processes, such as political independence and the development of a national identity. Typically, a common linguistic identity evolves when the local variety of English gains prestige over external norms, which in turn is often expressed in the creation of local variety dictionaries and grammars. The recent emergence of a local Jamaican standard with normative function is reflected in official documents such as the Revised Primary Curriculum published by the Ministry of Education in 1999, which states that the major objective of the language programme at school is to assist pupils "to acquire the target language Standard Jamaican English" (p. 14) (cited in Irvine 2004: 45f). Only 10 years before that, a Senior Education Officer had specified as the target of education in Jamaica "that our students develop proficiency in reading and writing Standard English" (cited in Shields 1989: 44). Shields (1989), Sand (1999), and Irvine (2004) describe the emergent SJE as an adoptive English spoken by influential Jamaicans of high repute such as politicians, businesspeople, educators and professionals in the media. SJE is assumed to be a home language for only a small section of the population and is acquired at school as a second language, after Creole, by the majority ofJamaicans (Patrick 2004: 408). Scholarly descriptions of spoken and written SJE show that it diverges systematically both from other world-wide standard English varieties and the neighbouring Caribbean varieties of English in terms of morpho syntax as well as phonology (Wells 1982; Shields 1989; Mair 2002; Irvine 2004). This variation has been speculated to derive from influence of Jamaican Creole. Mair (2002: 41) distinguishes between 'direct' influence, where a Creole form or rule is taken over into English without alteration, and 'indirect' influence, where Creole language structures are employed to explain a structure or word in the English utterance without themselves appearing in a clearly identifiable form. He finds some indirect Creole influence on written SJE: the overall frequency of definite and indefinite articles is lower in written SJE than other varieties of English and subject-verb agreement seems to be looser in written SJE than in standard varieties of English. Mair, however, notes that, on the whole, Creole plays a surprisingly small role in written SJE. Yet, he claims that borrowing and strong indirect influence from Creole is "clearly the dominant shaping influence on spoken English in Jamaica': and predicts that phonology as well as syntax and the lexicon will be affected (p. 36). Support for this hypothesis comes from an investigation of the phonology of spoken SJE by Irvine (2004). She found that all of the Jamaican Creole phonological features such
Relative markers in spoken Standard Jamaican Engllsh as /hi-dropping, palatalization of initial!k, glbefore/a, a:/and replacement of the alveopalatal affricate by palatalised stops also occurred in SJE but at a lower frequency than in Creole. SJE thus is not free of any of these Creole features, but all of them are present at a lower overall rate compared to Creole. Conversely, Christie (2003) assumes that relatively few structural differences between spoken SJE and standard English varieties reflect direct Creole influence. She cites cases such as lack of subject-verb concord and the lack of plural marking on nouns, which she considers to reflect change in progress in SJE rather than influence from Creole. Similarly, Shields (1989) claims that the structural properties of the emergent spoken SJE cannot be explained with influence from Creole but rather are the result of contact with standard English in schooling and literacy. She underlines this argumentation by showing that SJE exhibits features such as spelling pronunciation, stilted phrasing and a "careful, somewhat measured articulation" (p. 46) as well as the usage of a single style for both formal and informal conversation. Grammatical constructions listed by Shields as typical ofSJE, which reflect the influence of written standard English, include the particular usage of partitive constructions, questions embedded in the NP. and the differentiation between present and perfect tense. This study aims to test these conflicting claims with the example of relative clauses. Relativization strategies in Jamaican Creole were analysed by Christie (1996) and were shown to difter significantly from standard varieties of English. Relative clauses in SJE, which to date have not been investigated systematically, thus constitute an ideal field for testing possible influences ofJamaican Creole. Similarly, differences between relativization strategies in spoken and written standard English have been studied extensively (e.g. Sigley 1997) so that the possible influence of written English on spoken SJE can be analysed. The paper is organised as follows: sections 2 and 3 compare relativization in various standard varieties of English and in Jamaican Creole and SJE. Section4 describes the corpus data on which the analysis is based. Section 5 presents the results, which are discussed in Section 6.
2.
Relativization in spoken standard English varieties
Relativization in spoken English is a well researched area that covers several standard varieties (for Standard British English see Quirk 1957; Aarts 1993; Tottie 1997; for Standard American English see Guy and Bayley 1995; Ball1996; Tottie 1997; for Standard New Zealand English: Sigley 1997). In general, English speakers have a choice between the overt relative markers that, who, whom, whose and which as well as zero (0). In many contexts, speakers may choose among these options as illustrated in (1).
81
82
Ulrike Gut (1) a. The bike that she saw b. The bike which she saw c. The bike 0 she saw Previous research has shown that the choice of relative markers is restricted by several interrelated factors, among which are: the semantic properties of the relative clause (restrictive vs. non-restrictive) the semantic properties of the antecedent NP to which the relative marker refers the syntactic function of the relative marker (subject, object, or other) the position of the preposition in prepositional verbs in the relative clause The concept of restrictiveness of a relative clause refers to the effect the relative clause has on the referential scope of the noun antecedent. Non-restrictive relative clauses give supplementary, non-defining information about the antecedent since its identity is either previously identified or unique. The relative clause in (2) is non-restrictive since my mother is already unambiguously identified even without the additional information supplied by the relative clause. (2) My mother, who lives in Singapore, has come to see me. A restrictive relative clause, in contrast, supplies information about the antecedent that results in limiting its referential scope. Thus, the referential scope of my friend is restricted by the relative clause in (3). (3) My friend who lives in Singapore has come to see me.
In written language, non-restrictive relative clauses are usually marked by commas. The identification of restrictive relative clauses in spoken language is from straightforward. Many grammar texts claim that non-restrictive relative clauses in speech are typically delimited by pauses (Huddleston, Pullum, and Peterson 2002) or a combination of pausing and terminal intonation contour (Quirk et al. 1985: 1258; Biber et al. 1999: 280). Several studies have taken prosody as a cue for restrictiveness. Mesthrie (1991), for example, identifies 'near-relative' constructions in his South African Indian data where the place of the relative marker is filled by a personal pronoun but whose intonation pattern and placement in the matrix clause suggest a relative clause construction. Levey (2006: 54) reports on the difficulty of identifying restrictive relative clauses but relies on the intonation contour for his classification. Conversely, other studies have failed to identify a clear correlation between prosodic cues and type of relative clause. Fox and Thompson (1990), in their study on relative clauses in conversational data, did not differentiate between restrictive and non-restrictive relative clauses. They were not able to find "on intonational grounds[ ... ] clear cases of non-restrictive relative clauses[ ... ] Several cases[ ... ] were indeterminate" (footnote 2, p. 297f).
tar
Relative markers in spoken Standard Jamaican Engllsh
On the whole, non-restrictive relative clauses have attracted much less research than restrictive ones. All studies concerned with both types of relative clauses observed that non-restrictive ones are by far rarer than restrictive ones in spoken English. Their rate of frequency varies from 11.2% for New Zealand English (Sigley 1997) to 13.4% in educated British English of the 1950s (Quirk 1957) and 25.1% in the 1990s English represented in the British component of the International Corpus of English (ICE-GB, Hoffmann 2005). 1 Many studies report that the relative marker that and zero (0) do not occur in non-restrictive relative clauses. This holds for spoken British English (Biber et al. 1999), New Zealand English (Sigley 1997) as well as American English (Guy and Bayley 1995, who do not give any exact numbers, however). In addition to the distinction between restrictive vs. non-restrictive relative clauses, animacy of the antecedent noun also plays an important role in determining the presence and type of relative marker2. The relative markers who, whom and whose are restricted to refer exclusively to human antecedents in all standard varieties of English (British English: Tottie 1997; American English: Tottie 1997; New Zealand English: Sigley 1997). By contrast, the relative marker which occurs nearly exclusively with non-human antecedents (Tottie 1997; Sigley 1997). Only in British regional dialects does which occur with human antecedent nouns in up to 5% of all cases (Herrmann 2005; Peitsara 2002). The relative pronoun that can refer to both human and non-human antecedents. In educated spoken British English that refers to a human antecedent in 9.4% of all cases and to a non-human antecedent in 90.6% of all cases (Quirk 1957). The relative distribution of the overt relative markers with the semantic properties of the antecedent varies between spoken standard varieties of English. The percentage of human antecedents that have who is given as 80.5% for educated British English (Quirk 1957). In the analyses of parts of the Santa Barbara Corpus and parts of the BNC, both representing a wider range of American and British speakers, respectively, lower percentages of who were found: from 72.8% to 60.5% (Tottie 1997). Percentages of that occurring with human antecedents vary accordingly from 7% to 9% in educated British English (Quirk 1957; Tottie 1997) to 34.6% in the excerpts from the Santa Barbara Corpus and 25% in the sample of conversations from the BN C (Tottie 1997). The distribution of overt relative markers is equally as varied for non-human antecedents as for human antecedents. Non-human antecedents have which in 35.4% of all cases, that in 39.8% of all cases and zero in 24.8% in educated British English (Quirk 1957). Tottie (1997) found 1.
See Greenbaum (1996) for details on the ICE project.
2. Since most of the previous studies exclude non-restrictive relative clauses from their analysls, the following overview is based on restrictive relative clauses only.
83
84
Ulrike Gut
10% which and 46.1% that in the samples from the BNC. In the parts of the Santa Barbara Corpus of spoken American English analysed by her, which does not occur at all with non-human antecedents, tha.t occurs in 66.9% and zero in 32.2% of all such relative clauses. A third factor determining the speakers' choice of relative marker is its syntactic function in the relative clause. In (4), the relative marker that has a subject function, whereas in (5), its function is that of a direct object. (4) The woman that was leaving ... (5) The woman that he left ... A series of studies have shown that this factor is interrelated with the semantic properties of the antecedent noun (see Table 1). For human antecedents, the predominant relative marker in subject function is who in all varieties of standard English. Ball (1996) quotes 95% for American English, and the percentage for British English is 91% (Quirk 1957; Tottie 1997). Lower percentages for who in this position were found in the samples from the BNC with 67.1% and the parts of the Santa Barbara Corpus with 64% (Tottie 1997). That occurs in this position with a percentage between 7% and 38% (Tottie 1997; Sigley 1997), the zero marker occurs in up to 4% of all cases (Tottie 1997). The relative markers what and which are extremely rare in subject position when referring to a human antecedent noun. The most frequent relative markers with human antecedents in object function are that and zero in both British and American English (Quirk 1957; Aarts 1993; Tottie 1997) as well as in New Zealand English (Sigley 1997). Whom occurs in 0% to 25% of all possible cases in British English (Quirk 1957; Aarts 1993; Tottie 1997), while who and which are rare, as is illustrated in Table 1. The most frequent relative markers in subject position with non-human antecedents are that and which with an almost equal percentage of occurrence in educated standard British English (Quirk 1957) and in some samples from the London-Lund corpus of British English (Tottie 1997). Conversely, the American English studied by Ball (1996), the American English contained in the Santa Barbara Corpus, the British English of the BNC, and New Zealand English have a higher frequency of that than which in this position (Tottie 1997; Sigley 1997). Like with human antecedents, zero occurs very rarely in subject position. The most frequent relative marker with non-human antecedents and in object position is that, which occurs nearly as often as zero in all standard English varieties. A frequency of 20.9% for which is reported by Quirk (1957) for educated British English, but was not found for either American or New Zealand English or in analyses of parts of the LLC or the BNC (Tottie 1997; Sigley 1997), where the percentage lies between 0% and 16%. Table 1 illustrates the systematic variation between relative markers with human and with non-human antecedent and their syntactic role, which was found to be highly significant by Ball (1996: 233).
Relative markers in spoken Standard Jamaican Engllsh Table 1. Relative distribution of who, whom, that, zero, what and which with human and non-human antecedent in subject and in object position in spoken standard English (British [Quirk 1957; Aarts 1993; Tottie 1997], American [Ball1996; Tottie 1997], New Zealand [Sigley 1997]) human antecedent subject
who whom that zero
what which
object
non-human antecedent subject
object
64-95%
4.3-9.4%
7-38%
0-25% 21-60%
49-96%
32-48%
34-71%
0.3-5%
40-63%
0-7%
2-51%
0-21%
0-4% 0-0.3% 0-2%
Relative markers with possessive function are very rare in the data analysed in previous studies on spoken English. Sigley (1997) reports the exclusive usage of whose with human antecedents (n = 5) and of that with nonhuman antecedents (n 1). A fourth factor influencing the choice of the relative marker that has been described in previous studies is the position of the preposition in prepositional verbs. English speakers have the alternatives of positioning the preposition as a "stranded preposition, at the end of the relative clause as in (6a) or to produce it before the relative marker, which is referred to as "pied-piping, as in (6b). Option (6c) illustrates an utterance with the omitted preposition in. (6) a. The woman who(m) I trust in. b. The woman in who(m) I trust. c. The place I wanted to live. Diverging average rates of pied-piping are reported in the various studies on standard English varieties. Quirk (1957) observed 44.6% pied-piped constructions and 55.4% stranded prepositions in restrictive relative clauses produced by educated British English speakers. The occurrence of omitted prepositions is not mentioned at all. Conversely, Biber et al. (1999) claim that pied-piping is considered hyper-formal or even incorrect in British English conversation. Sigley (1997) reports a frequency of 2.8% of pied-piped constructions as opposed to 97.8% of stranded prepositions for restrictive relative clauses in New Zealand English. Hoffmann (2005) analysed ICE-GB and showed that pied-piping is less frequent in private dialogue (9%) than in public dialogue and speeches (71 %) and almost ubiquitous in scripted monologues (94%). The position of the preposition is systematically correlated with the choice of relative marker. Pied-piping only occurs with which, never with tha.t or zero (British English: Quirk 1957; New Zealand
85
86
Ulrike Gut
English: Sigley 1997). In restrictive relative clauses with stranded prepositions, that and zero occur nearly equally often: 38.6% and 47.7%, respectively, in British English (Quirk 1957) and 50% and 46.9% in New Zealand English (Sigley 1997). In the case of human antecedents, stranded or omitted prepositions occur more often with zero (74.3%) than with the relative marker that (25.6%, Sigley 1997).
3·
Relativization in Jamaican Creole and in SJE
Relative clauses in spoken Jamaican Creole show distinct divergences from standard English. Based on an analysis of spontaneous Creole conversations and stories, Christie (1996) reports a very frequent co-occurrence of relativization and focussing strategies. She found relative clauses often co-occur with left dislocation, a focussing strategy in which a main -clause preceding NP is produced that is coreferential with an anaphoric pro form in that main clause. This is illustrated in (7). (7) Di liedi we ben gi mi di tuu die shi sel di plies The lady who had given me the two days' (work), she sold the place (Christie 1996: 50) Jamaican Creole has the relative markers a, we(h), wa(t), huufa, dat and huu as well as zero (Christie 1996; Patrick 2004). Wa(t) developed from the interrogative pronoun what, and we(h) has been hypothesised to originate from either where (Christie 1996: 55) or Northwest England wha (Patrick 2004: 426). Huu stems from English who and huuja derives from who+ for. The relative markers a and dat originate from the deictic term that and, in contrast to wa(t), do not have a form that varies with some feature of the co-referential NP. As yet, little is known about the exact distribution of Jamaican Creole relative markers. Huu and huufa are restricted to a human antecedent, where huufa, an innovation that is basilectal according to Christie (1996: 56), represents relativized possessive nouns. Huu, conversely, is an acrolectal form. There is an overlap in the distribution of wa(t) and we(h), with the exception of relativized NPs with a locative meaning, where only we(h) can occur. Dat receives no attention in the literature and is probably restricted to formal contexts (Christie 1996: 56). Three types of relativization strategies in Jamaican Creole have been identified. In the first type, an overt relative marker introduces a clause with a structural gap, which is the result of wh-fronting. This type of relative clause can be introduced by either we(h), wa(t), huufa or huu. Utterance (8) is an example of such a relative clause. For the second type of relative clause, Christie (1996: 54) proposes a deletion of the co-referential NP within the relative clause. This type of relative clause typically has no relativizer (zero) or is introduced by dat or purposive fi. It occurs
Relative markers in spoken Standard Jamaican Engllsh most often when the relativized NP is part of a nominal complement comprising an indefinite NP and simultaneously represents the subject of the clause. The most frequent construction of this type is an existential clause as shown in (9). (8) Mi rispek ar tu di dart we shi waak pan _, Mada I respect her to the ground that she walks on__, Mother (Patrick 2004: 426) (9)
Dia· woz a liedi liv wid tuu children There was a lady (who) lived with two children
(Christie 1996: 55)
In the third type of relative clause, resumptive pronouns occur within the relative clause as in utterance (10). This is rare in combination with a zero relativizer but especially frequent when the relativized NP is a possessor (Christie 1996: 58). In these cases, huu as a relative marker is not allowed. Prepositional verbs only have stranded prepositions in relative clauses in Jamaican Creole; pied-piped constructions do not occur. (10) Di uman we dem tiifar biebi gaan a stieshan The woman that they stole her baby has gone to the station (Christie 1996: 58) Some features of relativization in the emerging spoken SJE have been described by Christie (2003). She claims that the relative marker whom is overused in inappropriate contexts such as in the sentence "The man whom the woman said had stolen her purse" (p. 18) in both spoken and written SJE, which she analyses as an instance of hypercorrection. Further, relative clauses in SJE tend to be separated from their antecedents as in "The woman came into the room who was very attractive" (p. 17), which Christie does not interpret as evidence for Creole influence but as an example of a world-wide change in progress. Both Christie (1996: 59) and Patrick (2004: 427) report that resumptive pronouns are common in spoken SJE relative clauses such as in (11). (11) That is a drink that everybody would like it at Christmas (Christie 1996: 59) No comprehensive investigation of relativization in spoken SJE exists yet. It is the object of this paper to investigate whether there is- indirect or direct- influence of the Creole system of relativization on spoken SJE as suggested by Mair (2002) or whether spoken SJE rather reflects features of written standard English (Shields 1989). The major differences between Jamaican Creole and Standard English varieties lie in:
87
88
Ulrike Gut
a. the inventory of relative markers b. the occurrence of pied-piped constructions c. the occurrence of resumptive pronouns Specifically, Jamaican Creole has the relative markers a, we(h), wa(t) and huufa. which do not occur in standard English varieties. Further, pied-piped constructions, where the preposition of a prepositional verb precedes the relative marker in the relative clause, exist in standard English but not in Jamaican Creole. Conversely, Jamaican Creole has resumptive pronouns in relative clauses, which do not occur in standard English varieties.
4· Data and analysis
The data presented in this paper is drawn from a pre-release of the spoken part of ICE Jamaica (ICE-JA). 3 In September 2006 it consisted of 276 orthographically transcribed recordings comprising private dialogues such as conversations, and phone calls (text code sla), public dialogues such as classroom lessons, broadcast interviews and discussions. parliamentary debates, legal cross-examinations and business transactions (text code s2a), unscripted monologues such as spontaneous commentaries, unscripted speeches, demonstrations and legal presentations (text code sl b) as well as scripted monologues such as broadcast talks, broadcast news and non-broadcast speeches (text code s2b). ICE-JA is intended to be a representative sample of educated Standard Jamaican English and includes the speech of male and female business people, academics, judges. politicians. professionals in radio and television as well as national icons in sports and music. All the speakers in the corpus are adults and were born and raised in Jamaica. Transcriptions of the recordings consist of roughly 2000 words each. Several transcribers carried out the orthographic transcriptions, most of whom were German students of English without any prior familiarity with Jamaican English. Unfortunately, no clear guidelines exist concerning the transcription of possible Creole elements in the recordings. It is thus probable that some Creole relative markers will not have been marked accordingly. Relative clauses in the corpus were located with computer assistance for the overt relative markers. For the analysis of relative clauses that are introduced by zero, a quarter of the corpus was read (see Section 5.4) and numbers were extrapolated. In order to make the analysis comparable to other quantitative investigations several relative constructions were omitted from analysis. These include:
3·
See for information about the corpus.
Relative markers in spoken Standard Jamaican Engllsh
a.
sentential relative clauses, as in (12): (12) As teachers however I know that the most significant reward is the joy in a child's face when he succeeds at his lessons or when his hard work puts him on top of his class or when he tries so hard and gets a hug and words of encouragement from his teachers which I hope is being done every day (s1b-060)
b. adverbial relative constructions with the markers when, why and where c.
interrupted relative clauses where it does not become clear which syntactic function the relative marker has.
The total number of remaining relative clauses with overt relative marker in the corpus that were analysed in this study is 4,287.
5·
5.1
Results Inventory of overt relative markers
Eight different overt relative markers4 occur in ICE-JA: who, whom, whose, which, that, wa, dat and we(h). Table 2 illustrates the absolute frequency of each of the relative markers. That is the most frequent (42.34%) relative marker, followed by who (31.1 %) and which (25%). Whom and whose occur seldom with a frequency of 0.8% and 0.7%, respectively. Dat, a and weh occur once each (0.02%). No instances of wa or huufa as relative markers were found in the corpus. The total occurrence of the Creole relative markers is thus lower than 1%. The overall frequencies of the relative markers in spoken SJE differ from those reported for educated spoken British English by Quirk (1957). In his data, who occurs less frequently with a rate of 23.9%, which is much more frequent with 41.3%, and tha.t is less frequent with 34.8%. The relative clauses with Creole relative markers found in ICE-JA are given in (13), (14) and (15). Table 2. Absolute and relative frequency of the overt relative ma:rkers in ICE-JA*
who n %
whom
whose
which
that
1,335
33
28
1,073
1,815
31.1
0.8
0.7
25
42.34
dat
CJ
we(h)
0.02
0.02
O.Q2
* Relative markers wa and huufa not attested
4
I am using this term to refer to both relative markers and relative pronouns.
89
90
Ulrike Gut
(13) her pride and joy was this little one pikney dat finally reach university (s2a-037) (14) what was known as the ten cent fund weh was for for the students (s1a-013) (15) the man a call her she run gone (s1a-23) Due to the fast speaking rate in utterance (13), it is not unambiguously decidable whether the relative marker is dat or that. It is likely to be dat because the speaker talks about a remark of her grandmother's whom she had quoted with a Creole utterance earlier in the talk. Likewise, utterance (15), using relativizer a, is a conscious imitation of a Creole utterance on the part of the speaker. 5.2
Restrictive vs. non-restrictive relative clauses
The 4,287 relative clauses analysed in this study were divided into restrictive and non-restrictive ones. The difference between restrictive and non-restrictive relative clauses was determined on semantic grounds. Relative clauses were only classified as non-restrictive when the antecedent was clearly identified or unique as in the utterances (16) and (17): (16) Dixon who gets a header on to Whitmore who heads it uh farther on to Cornel (s2a-009) (17) better than the Millenium Dome which was uh celebrated (s2a-002) A total of 335 relative clauses that had an unambiguous non-restrictive function occur in the corpus. This constitutes 7.8% of all analysed relative clauses. Of these, 185 were introduced by who, 119 were introduced by which, 19 had that as a relative marker, six were introduced by whom and six by whose. Table 3 shows the relative distribution of the relative markers who, which, that, whom and whose in restrictive and non-restrictive relative clauses. Who is the most frequent relative marker in non-restrictive relative clauses with 55.2%, followed by which (35.5%). Conversely, in restrictive relative clauses, that is the most frequent relative marker (45.5%) and who and which occur to a nearly equal percentage (29.1% and 24.1 %, respectively). Since most previous studies focussed exclusively on restrictive relative clauses, only those types of relative clauses in spoken SJE can be compared with other varieties of English concerning the overall relative distribution of the relative markers. Table 4 presents the distribution of the relative markers who, which, that, whom and whose in spoken SJE and in spoken standard British English (Tottie 1997).
Relative markers in spoken Standard Jamaican Engllsh Table 3. Distribution of the relative markers who, which, that, whom and whose in nonrestrictive and restrictive relative clauses non-restrictive relative clauses N 185 119 19 6 6
who which that whom whose n
% 55.2 35.5
restrictive relative clauses N 1,150
% 29.1
5.7
954 1,796
1.8
27
0.7
1.8
22
0.6
335
24.1 45.5
3,949
Table 4. Distribution of the relative markers who, which, that, whom and whose in restrictive relative clauses in spoken SJE and spoken standard British English (BNC, Tottie 1997)
SJE
who which that whom whose
BrEng
29.1
18.5
24.1
13.1
45.5 0.7
68.4
0.6
The most striking difference between spoken SJE and spoken British English in the BNC is the high percentage of the relativizer which (24.1 %) in the former. Conversely, that occurs more often (68.4% vs. 45.5%) in spoken standard British English compared to spoken SJE. 5·3
Human vs. non-human antecedent
The 4,287 relative clauses were divided into those with a human and those with a nonhuman antecedent noun. Table 5 shows that non -human antecedents are nearly twice as frequent as human antecedents. It further illustrates that the distribution of the relative markers in spoken SJE depends crucially on the semantic properties of the antecedent The relative marker whom occurs exclusively with human antecedents, who and whose refer to human nouns in the overwhelming majority of cases (98.7% and 92.8%, respectively). Cases in which the relative marker who occurs with non-human antecedents include the nouns herbs, plants and country and instances of who being used to refer to groups of people as in ministries, sub-committee, Western Society, Soviet
Union, one segment
af the population, BNS + NBC, Water·house, CUC, island tours,
91
92
Ulrike Gut
Tivoli Ga.rdens, faculty and our investigative media. The relative marker whose is used to refer to non-human antecedents in the case of the noun phrases the Ministry ofFinance and Planning and international institutions. This restriction of who, whose and whom to human antecedents mirrors findings for standard varieties of English (see Section 2). Which is used with non-human antecedents in 99.6% of all cases in ICE-JA. It occurs with a human antecedent in four cases where the nouns are people, artists, person and subjects. Like who, which is also used to refer to companies (three times in total), to countries (10 times) and to groups of people (21 times). Again, therestriction of which to non-human antecedents corresponds to standard English usage (compare Section 2). Table 5 shows that the relative marker that is the only one that is systematically used for both types of antecedents, although it refers to a human antecedent in only 11.2% of all cases. This figure is equal to the one reported for educated British English (Quirk 1957) (see Section 2). The absolute frequencies of the relative markers dat, a and weh are too small to be evaluated usefully. Table 6 presents the distribution of the relative markers who, which, that, whom and whose with human and non-human antecedents in both restrictive and non-restrictive relative clauses. It can be seen that the relative marker that referring to human antecedents occurs almost exclusively in restrictive relative clauses. In non-restrictive relative clauses, a human antecedent restricts the relative marker choice to who, whom and whose. The function of the relative clause also influences the choice of relative marker referring to non-human antecedents. While which is predominant in such cases in non-restrictive relative clauses, that is preferred in restrictive relative clauses with non-human antecedent. This last observation is similar to the findings by Quirk (1957) on educated British English, who reports 99.3% usage of which with Table 5. Relative and absolute frequency of each of the observed relative markers with human and non-human antecedents non-human
hwnan
who whom whose which that
%
n
%
98.7
1,318
1.3
17
100
33
92.8
26
7.2
0.4
4
99.6
2 1,069
11.2
183
88.8
1,632
dat
100
a
100
weh n
n
100 1,566
2,721
Relative markers in spoken Standard Jamaican Engllsh Table 6. Distribution of the relative markers who, which, that, whom and whose with human and non-human antecedents non-restrictive relative clauses human antecedent
who which that whom whose
180
n
192
restrictive relative clauses human antecedent
non-human antecedent
5
1,138
119 18
4
12 950 1,614
non-human antecedent
6
182 27
5
143
21 1,372
2,577
non-human antecedents in non-restrictive clauses, but a balance of 47.1% which and 52.9% that in the same context in restrictive relative clauses. 5·4
Syntactic function of relative marker
Tables 7 and 8 illustrate the syntactic functions of the relative markers who, whom, whose, which and that occur in restrictive and non-restrictive relative clauses. In both types of relative clauses, who occurs predominately in subject position. In 18 cases, it occurs in an oblique position such as in utterance (18). (18) the people who we are going to be sharing this info with (s2b-029) A total of24 cases in the restrictive and non-restrictive relative clauses can be classified as whom-avoidance when who appears in the direct object position. The relative marker whom occurs either in direct object position or in oblique position. Considering all 68 relative markers with human antecedent and in direct object function (Tables 7 and 8), it is only in six cases (8.8%) that whom is chosen in spoken SJE. This figure is similar to those reported by Quirk (1957) and Aarts (1993) for spoken standard British English. Only in one case did a speaker use whom in subject position (19). (19) NN and Kevin Gilzene both of East Kirkland Heights in the area whom they believe can assist with their investigation (s2b-005) The relative marker whose occurs always in oblique (possessive) position in the relative clause. Both which and tha.t occur almost equally frequently in subject position (43.7% vs. 56.3%, respectively) and in oblique position (48.8% vs. 51.2%). This was also reported for standard British English (Quirk 1957; Tottie 1997) but not for standard American English, where that predominates (compare Table 1). In spoken SJE,
93
94
Ulrike Gut Table 7. Distribution of the relative markers who, whom, whose, which and that with human and non-human antecedent in subject position. object position. and oblique position in restrictive relative clauses non-human antecedent
human antecedent
who whom whose which that
subject
direct object
oblique
subject
1,098
22 3
12
132
38
18 23 21 2 12
688 878
direct object
oblique
136 620
126 116
direct object position that (81.9%) tends to be produced rather than which (10.1 %). This was also found for standard British and standard American English (Table 1). No difference between restrictive and non-restrictive relative clauses in terms of relative marker choice and syntactic function of the relative marker exists in spoken SJE. The overall low rate of that in non-restrictive relative clauses was already pointed out in Table 3. In order to determine the relative frequency of the zero relative marker (0) in object function, 25% of the corpus (66.45 texts) were searched manually for the occurrence of 0. This yielded a total of74 0 in object function, of which 10 had a human antecedent, and two 0 in subject function, both with human antecedents. Extrapolating this figure to the entire corpus, Table 9 shows the relative frequency of the zero relative marker in spoken SJE. The relative distribution of the various relative markers referring to human antecedents shows one important divergence from the figures reported by Aarts Table 8. Distribution ofthe relative markers who, whom, whose, which and that with human and non-human antecedent in subject position, object position and oblique position in non-restrictive relative clauses human antecedent
who whom whose which that
subject
direct object
178
2 3
non-human antecedent oblique
subject
direct object
oblique
10 4
5
5 3 5
104 11
3
Relative markers in spoken Standard Jamaican Engllsh
Table 9. Distribution (in%) of the relative markers who, whom, 0. which and that with human and non-human antecedent in object position across all relative clauses hwnan antecedent
non-hwnan antecedent
who whom
22.2
0
37
25.9
which that
35.2
60.8
5.6 13.3
(1993) for educated standard British English: who is much rarer there with 4.5%. Conversely, 0 in object position in relative clauses referring to non-human nouns is far more frequent in standard English varieties than in spoken SJE (Sigley 1997 reports 58.2% for New Zealand English). 5·5
Prepositions in the relative clause
It was further investigated whether relative marker choice in spoken SJE varies with the position of the preposition of prepositional verbs in the relative clause. A total of 387 relative clauses in the corpus contain prepositional verbs. Of these, 61.5% have pied-piping, 35.4% have stranded prepositions and in 3.1% of all cases the preposition is deleted. This last figure is higher than that reported by Quirk (1957) for British English and much higher than that observed for New Zealand English (Sigley 1997). The tendency to have relatively more pied-piping in nonrestrictive relative clauses than in restrictive ones, which was found for spoken British English (Quirk 1957), also exists in spoken SJE, although the overall numbers in this study are very small. Table 10 shows that who occurs almost exclusivelywith a stranded preposition such as in utterance (20). There is only one instance of preposition fronting (pied-piping) with who (21) in ICE-JA. (20) ... six lecturers who can draw on who you can go to ... (s1a-058) (21) appoint people who .. in who they have no confidence ... (s1b-022) Conversely, the relative marker whom occurs almost exclusively with pied-piped constructions as the one illustrated in (22). One of the two instances of whom with a stranded preposition is illustrated in utterance (23). (22) Doctor Cooper from whom you will hear later in the proceedings (s2a-022) (23) this accused is the man whom the complainant spoke of to the police (s2a-063)
95
96
Ulrike Gut Table 10. Distribution of the relative markers who, whom, whose, which and that with pied-piping, stranded, and deleted prepositions in restrictive relative clauses (numbers for non-restrictive relative clauses in brackets) pied-piping
who whom whose which that
stranded preposition
1 (0)
19 (0)
11 (3)
2 (0)
prepositions deleted
2 222 (5)
7 (0)
1 (0)
2 (0)
109 (2)
11 (0)
The two relative markers whose that are produced in combination with a preposition occur in a relative clause with pied-piping. The relative markers which and that show opposite behaviour in the way that which is used predominantly with pied-piped constructions and that with stranded prepositions. This is similar to standard British and New Zealand English, where pied-piping only occurs with which (Quirk 1957; Sigley 1997). Eleven of the 12 instances where a preposition was deleted in the relative clause occur with the relative marker that in utterances such as (24). Five of the instances involve the verb to go as in (25). Only once is there a missing preposition in combination with which in the spoken part ofiCE-JA (Example 26). (24) something that you listen (s2a-35) (25) places that we go (s1a-68) (26) several of the proposals which the committee worked assiduously in putting forward (s1b-56) In order to test the claim about spoken SJE being monostylistic (Shields 1989), the position of the preposition in relative clauses was compared in two different speaking styles. Table 11 lists the relative occurrence of pied-piped constructions and Table 11. Distribution of the relative markers who, whom, whose, which and that with piedpiping, stranded and deleted prepositions in private dialogue and scripted public monologue private dialogue pied-piping
who whom which that n
stranded preposition
scripted public monologue pied-piping
stranded preposition
14
2 8
56 49
9 (12.5%)
63 (87.5%)
6 58 (87.9%)
8 (12.1%)
Relative markers in spoken Standard Jamaican Engllsh stranded prepositions in private dialogues (text category s1a) and scripted public monologues (text category s2b) in spoken SJE. It can be seen that, like in spoken British English (Hoffmann 2005), stranded prepositions occur predominantly (87.5%) in private dialogue whereas pied-piping is nearly the rule (87.9%) in scripted public dialogue. 5.6
Relative clause constructions
It was next analysed how many of the 4,287 relative clauses in the spoken part of ICE-JA co-occurred with the focussing strategy left dislocation. In total, 13 such constructions occurred in the corpus, three of them with the relativizer who, eight with that, and two with which. Examples for these constructions are given in (27), (28) and (29). This means that the vast majority (99.7%) of all relative clauses is embedded in other syntactic constructions in spoken SJE. (27) The persons who are hosting events they want to have (s2a-51) (28) The two customers that were in your shop sir they were from that area (s1b-069) (29) That right leg which is the very next stride of his that's not where it should be (s2a-05) In the entire corpus, only one instance of a resumptive pronoun in a relative clause was found (see utterance 30): (30) I've met some who they seem (s1a-55)
6.
Summary and discussion
This study was concerned with relativization in spoken SJE. Specifically, its aim was to test both the claim that spoken SJ E shows features ofJamaican Creole influence as suggested by Mair (2002) and the opposing claim that it rather reflects an influence of written standard English as predicted by Shields (1989). 4,287 relative clauses with overt relative markers taken from the spoken part of ICE-JA were analysed according to the factors constraining their choice. The first finding is that the inventory of relative markers in spoken SJE is largely the same as in standard English. The total occurrence of Creole relative markers is three (0.06%) in the corpus, two of which are used in Creole quotations. However, it needs to be repeated here that it is possible that some Creole relative markers were not transcribed and thus missed in the automatic corpus analysis.
97
98
Ulrike Gut
Since the function of the relative clause has been established to constrain relative marker choice in various standard varieties of English, the relative clauses contained in ICE-JA were divided into restrictive and non-restrictive ones. Nonrestrictive relative clauses in spoken SJE are very rare with a percentage of 7.8%. This figure is lower than that reported by Quirk (1957) for educated British English (13.4%), which might however simply reflect the difficulty of identifying nonrestrictive relative clauses in spoken language and the methodological differences in determining such clauses between the various studies. The overall distribution of the overt relative markers in restrictive relative clauses in spoken SJE seems to diverge from that in standard British English in two important ways: which occurs less frequently in SJE, while that is used more often. However, a reliable comparison of the relative distribution of relative markers is made difficult by the fact that the corpora do not contain the same text types. Quirk's data only consists of public and private dialogues whereas ICE-JA also includes unscripted and scripted monologues. The second line of investigation concerned the variation of relative marker choice with the semantic properties of the antecedent. Like in the standard varieties of English, in spoken SJE, who/whom/whose are restricted to human antecedents whereas which occurs exclusively with non-human antecedents. The relative marker that can refer to both human and non-human antecedents, and is used for human antecedents equally rarely in SJE (about 11% of all relative clauses with human antecedent) as in educated British English (Quirk 1957). The semantic properties of the antecedent interact with the function of the relative clause in constraining the choice of relative markers in spoken SJE. That with a human antecedent occurs only in restrictive relative clauses. With a non-human antecedent noun, which is preferred over that in non-restrictive relative clauses, whereas the preference is the opposite in restrictive relative clauses. This replicates findings by Quirk (1957) on educated spoken British English. In general, in SJE, non-restrictive relative clauses have human antecedents equally often as non-human antecedents. In restrictive relative clauses, conversely, non-human antecedents are significantly more frequent than human ones. This finding stands in contrast to observations made by Quirk (1957) for educated British English and Sigley (1997) for New Zealand English. They report that only between 17.8% and 25.5% of all non-restrictive relative clauses have human antecedents whereas between 28.9% and 44.6% of all restrictive clauses have human antecedents. Spoken SJE thus has a more pronounced tendency to have non-restrictive relative clauses with human antecedents. Again, the heterogeneous methods of determining non-restrictive relative clauses in the various studies may contribute to this divergence. The third factor constraining the variation of relative marker choice in spoken SJE, which is interrelated with the function of the relative clause and
Relative markers in spoken Standard Jamaican Engllsh 99
the semantic properties of the antecedent, is the syntactic function of the relative marker in the relative clause. With human antecedents, who is used predominantly in subject function, who and whom occur in direct object and oblique position and whose occurs only in oblique position. Of all direct object positions, only 8.8% are filled by whom in SJE, which is similar to findings for standard British English (Quirk 1957; Aarts 1993; Tottie 1997). One divergence of relative marker choice in SJE from standard English varieties lies in the underuse of 0 in object position in relative clauses referring to non-human antecedents. The rate of zero in such contexts is much higher in British and New Zealand English (Quirk 1957; Sigley 1997). The correlation of relative marker and the position of the preposition in prepositional verbs in spoken SJE is similar to that of standard British and New Zealand English (Quirk 1957; Sigley 1997): pied-piping occurs with which or whom, whereas stranded prepositions occur with that or who. Yet, the overall percentage of pied-piped constructions in spoken SJE is higher than that in standard British and New Zealand English (Quirk 1957; Sigley 1997). The distribution of piedpiping occurs predominantly in scripted public monologues, and stranded prepositions occur mostly in private dialogue, which was observed by Hoffmann (2005) for British English. The two last findings concerned the syntactic constructions that relative clauses occur in and the frequency of resumptive pronouns in the relative clause. Relative clauses that occur together with left dislocation are extremely rare (0.3%) in SJE and resumptive pronouns are virtually non-existent (n = 1). The results of this corpus-based analysis thus contradict the claims about differences between relativization in SJE and in standard English made by Christie (2003) and Patrick (2004). Resumptive pronouns in relative clauses are extremely infrequent with a total of one out of 4,287 relative clauses. Likewise, only one occurrence of whom used in subject function in the relative clause was found in ICE-JA. Christie's (2003) claim of an overuse of whom in subject position in the emerging SJE is thus not substantiated by this study. Equally, only one instance of a relative clause separated by its antecedent was observed in ICE-JA. This was utterance (29), where the relative clause does not follow the antecedent but is produced at the end of the utterance: (29) You mean everybody should get benefits who are in this schedule (s1b-73) The systematic differences between relative marker choice in spoken SJE and in standard English varieties can be summarized as follows: in general. which occurs less frequently in SJE, while that is used more often. 0 in object position in relative clauses referring to non-human antecedents is used less frequently in SJE than in standard English. SJE has more pied-piping than standard English.
100
Ulrike Gut
How can these differences be explained? The findings of this study clearly contradict the hypothesis that spoken SJE shows direct or indirect influence from Creole (Mair 2002). In Section 3, the major differences between relativization in Jamaican Creole and in standard English had been summarized as differences in the inventory of relative markers and in the occurrence of pied-piped constructions and resumptive pronouns. None of these Creole features occur in spoken SJE. Creole relative markers are used in only 0.06% of all cases and occur in Creole quotations or conscious imitations. Pied-piped constructions occur in SJE with an overall frequency of 61.5%, which is higher than that in standard English varieties (Quirk 1957; Sigley 1997). Resumptive pronouns are practically non-existent in spoken SJE. The results of this study thus do not confirm the claim that spoken SJE shows Creole influence. Conversely, it is in line with Mair's (2002) findings on written SJE: in the student essays he analysed, extremely few direct lexical loans occur and no direct Creole influence in the area of grammar was found. Those grammatical divergences he observed (e.g. overall lower number of articles, lack of subject-verb agreement) in written SJE reflect at best indirect influence from Creole but - since they can be observed in other postcolonial varieties of English (Mair 2002: 52) - might better be explained by other mechanisms. The observation that spoken SJE shows extremely little influence from Creole stands in contrast to Irvine's (2004) findings on SJE phonology. On the one hand, this might be due to the fact that onlyrelativization was investigated in the present paper. It is possible that relativization happens to be a grammatical area in which very little Creole influence occurs whereas other grammatical or phonological structures are more affected. Future research on other structures and variables can show whether and where direct or indirect Creole influence can be found in spoken SJE. On the other hand, however, another explanation for this discrepancy might be that Creole influence shows differently on different linguistic levels. It might be restricted to the area of phonology for a number of reasons: For one, English phonology is less emphasized in school than English grammar. In addition, phonology is the one aspect that cannot be drawn from written texts which introduce English to Jamaicans at school. Moreover, school teachers might simply not be aware of the phonological differences between spoken SJE and spoken standard American or British English. Creole structures in grammar are probably much easier to detect for them than phonological features influenced by Creole. This last hypothesis is backed up by Irvine's (2004) study. She reports that the Jamaican acrolectal speakers she investigated stated that they consciously avoid phonological features that are stigmatized as Creole features. Nevertheless, their speech showed typical Creole characteristics such as the production of a voiced alveolar stop at the beginning of they.
Relative markers in spoken Standard Jamaican Engllsh
The hypothesis that Creole influence is restricted to the phonology of spoken language suggests that the norm-orientation for English phonology and English grammar differs in Jamaica. It is likely that whereas a Creole form in grammar is treated as a mistake by teachers and educated speakers, American or British English pronunciation in the speech of a Jamaican is considered affected and inappropriate. This is the case in many other postcolonial countries, e.g. in Singapore where an accent that combines some American, some British and many local features is considered most appropriate (Poedsjosoedarmo 2002: 149ff), or in Nigeria where Nigerian speakers with a British accent are ridiculed for speaking in an affected way (Udofot 2003: 204), although standard British English serves as the model for grammatical structures. Is there any evidence that the properties of spoken SJE reflect influences from the written standard as proposed by Shields (1989)? All of the differences between SJE and standard English observed in this study point in that direction. First, in SJE, a higher percentage of which with non-human antecedents was observed, which according to Sigley (1997) is a feature of writing only. In spoken standard English as well as in informal writing that is preferred in this context. Second, the relative marker that is mainly used to refer to human antecedents in informal speech in standard English; in writing, who is preferred. Speakers in ICE-JA avoid that with human antecedents even in spoken language. The low occurrence of 0 in object position might reflect prescriptive rules in English grammar books that recommend the avoidance of the zero relative marker in written English (Sigley 1997). Finally, the high percentage of pied-piping in spoken SJE mirrors the general tendency of pied-piping to occur more frequently in written than in spoken standard English (Sigley 1997), especially when the writing has a high level of formality (Hoffmann 2005). Shield's (1989) explanation of SJE properties as reflecting the predominant contact of Jamaicans with written English and their acquisition of the language in a (prescriptive) school setting is thus supported by the data in this study. Yet, no evidence was found for her claim that SJE is monostylistic. Clear differences in pied-piping and preposition stranding were found between informal private dialogue and public scripted monologues. More research on other variables is needed to support the present finding that systematic stylistic differences exist in spoken SJE.
References Aarts, F. 1993. Who, whom, that, and 0 in two corpora of spoken Engllsh. English Today 9: 19-21. Allsopp. R 1996. Dictionary of Caribbean E11giish Usage. Oxford: OUP.
101
102.
Ulrike Gut Ball, C. 1996. A diachronic study of relative markers in spoken and written English. Language Variation and Change 8(2): 227-258. Biber, D., Johansson, S., Leech, G., Conrad, S. & F.lnegan, E. 1999. Longman Grammar of Spoken and Written English. London: Pearson. Christie, P. 1996. Jamaican relatives in focus. In Caribbean Language Issues Old and New, P. Christie (ed.), 48-60. Kingston, Jamaica: University of West Indies Press. Christie, P. 2003. Language In Jamaica. Kingston: Arawak Publications. DeCamp, D. 1971. Toward a generative analysis of a post-creole speech continuwn. In Pidginlzation and Creollzatlon ofLanguages, D. Hymes (ed.), 349-370. Cambridge: CUP. Fox, B. & Thompson, S. 1990. A discourse explanation of the grammar of relative clauses in Engl.lsh conversation. Language 66(2): 297-316. Greenbaum, S. 1996. Introducing ICE. In Comparing English Worldwide: The International Corpus ofEnglish, S. Greenbaum (ed. ), 3-12. Oxford: Clarendon Guy, G. R & Bayley, R 1995. On the choice of relative pronouns in English. American Speech 70: 148-162.
Herrmann, T. 2005. Relative clauses in Engl.lsh dialects of the Britlsh Isles. In A Comparative Grammar of British English Dialects, B. Kortmann, 't Hermann, L. Pietsch & S. Wagner (eds.), 21-123. Berlin: Mouton de Gruyter. Hinrichs, L. 2006. Codeswitchingon the Web: English andJamalca11 Creole in E-Mail Communication. Amsterdam: John Benjamins. Hoffmann, T. 2005. Variable vs. categorial effects: Preposition pied piping and stranding in British Engllsh relative clauses. Joun1al ofEnglish Linguistics 33(3): 257-297. Huddleston, R, Pullum, G. & Peterson, P. 2002. Relative constructions and unbound dependencies. In The Cambridge Grammar of the English Language, R Huddleston & G. Pullum (eds.), 1031-1096. Cambridge: CUP. Irvine, A. 2004. A good command ofthe English language: Phonological variation in the Jamaican acrolect Joun1al ofPidgin and Creole Languages 91(1): 41-76. Irvine, A. 2008. Contrast and convergence in Standard Jamaican Engl.lsh: The phonological architecture ofthe standard in an ideologically bidialectal community. World Englishes 27( 1): 9-25.
Levey, S. 2006. Visiting London relatives. English World-Wide 27(1): 45-70. Mair, C. 2002. Creolisms in an emerging standard English World-Wide 23( 1): 31-58. Mesthrie, R 1991. Syntactic variation in South African Indian English: The relative clause. In English around the World, J. Cheshire (ed), 462-473. Cambridge: CUP. Patrick, P. 2004. Jamaican Creole morphology and syntax. In A Ha11dbook of the Varietie.s of English, B. Kortmann & E. W. Schneider (eds.), 407-438. Berlin: Mouton de Gruyter. Peitsara, K. 2002. Relativizers in the Suffolk dialect. In Relatlvlsation on the North Sea Littoral, P. Poussa (ed.), 167-180. Munich: Lincom. Poedjosoedarmo, G. 2002. Student teachers' differing attitudes to accents of English. In The Teaching and Use ofStandard English, E.-L. Low & S. Teng (eds.), 146-155. Singapore: Singapore Association for Applied Linguistics. Quirk, R. 1957. Relative clauses in educated spoken English. English Studies 38: 97-109. Quirk, R, Greenbaum, S., Leech, G. & Svartvik, J. (eds.). 1985. A Comprehensive Grammar ofthe English Language. London: Longman. Sand, A. 1999. Linguistic Variation in Jamaica: A Corpus-Based Study of Radio and Newspaper Usage. Tiibingen: Narr.
Relative markers in spoken Standard Jamaican Engllsh 103 Schnelder, E. 2003. The dynamics of new Engllshes: From ldentity construction to dialect birth. Language 79:233-281. Shields, K. 1989. Standard English in Jamaica: A case of competing models. English World-Wide 10:41-53.
Sigley, R. 1997. The infiuence of formality and channel on relative pronoun choice in New Zealand English. English Language and Linguistics 1(2): 207-232. Thttie, G. 1997. Relatively speaking: Relative marker usage in the Britlsh National Corpus. In To Explain the Present: Studies in the Changing English Language In Honour of Matti Rissanen. T. Nevalainen & L. Kahlas-Tarkka (eds.), 465-481. Helsinki: Societe Neophilologique. Udofot, I. 2003. Stress and rhythm in the Nigerian accent ofEnglish. English World- Wide 24(2): 201-220.
Wells, J. 1982. Accents ofEnglish. Cambridge: CUP.
PART II
Variation and identity
"Flying at half-mast"? Voices, genres, and orthographies in Barbadian Creole* Janina Fenigsen Bajan, a Barbadian creole, has no standard orthography. It is written and published in a variety of spellings, and its readership and the generic scope remain limited. The reception of Bajan texts is regimented by the history of representational practices that used non-standard spelling as parody. Because within the cultures of script linguistic prestige depends on the existence of a robust range of written genres, this reception hinders the raising of the prestige of Bajan. The relationship between Bajan spelling and its prestige, then, comes around a full circle: the lack of standardized orthography reinforces the low prestige ofBajan that prevents its emergence. Because the low prestige restricts social mobility oflower-class Bajan speakers, it reinforces social hierarchies. Thus, the interpretive practices of the readers whose stance toward Bajan texts is not determined by spelling are particularly interesting. I consider such alternative readings and propose how to account for them.
If we keep condemning something about ourselves which cannot be removed, we will be always flying at half-mast[ ... ] I am begging you, then, to come to an objective view about our 'dialect,' one of our basic characteristics and resources. (Callender 1987: 23)
This research stemmed from my larger ethnographic study oflanguage relations in Barbados supported by grants from the Research Institute for the Study of Man (1993); Slgma Xi, The Scientific Research Society (1994); and from Latin Amertcan Studies Department at Brandeis University (1994 Jane's Travel Grant). I am indebted to the many Barbadians who made this work possible and especially to Vireen, Glorta, Maureen, Ertc, and Kim Ramsey from Arawak Hill. I use pseudonyms for persons who requested anonymity. Kim Ramsey gave me the permission to use her actual name. The name of the village in St. Peter's partsh where I did fieldwork (Summer 1992, March-August 1994, Swnmer 2002) is fictive. I am also grateful to Jeanette Layne-Clark who granted me a telephone interview (September 1996), to Norma Meek. and to two Barbadian journalists who preferred to remain anonymous. At earlier stages, the paper has benefited from crttical readings by Jane Goodman, Judith T. Irvine, and Alexandra Jaffe. I also wish to thank Lars Hinrtchs and two anonymous reviewers for their help in making it a better manuscript.
108
Janlna Fenigsen 1.
Introduction
In the settings that share Western histories and ideologies of linguistic standardization, the legitimacy oflanguage varieties is linked, among other things, to writing (Crowley 1989; Grillo 1989; Bourdieu 1991; Devonish 2000; Mi.ihleisen 2002: 183; Collins and Blot 2003). The efforts of various minority groups to substantiate their claims to political autonomy via linguistic distinctiveness show that the ideological re-evaluation and official legitimacy of vernaculars, including creole languages, is tied to their expansion into written registers (Eckert 1983; Urla 1988; Jaffe 1996; Roberts 1997; Romaine 2005). Yet, as I have argued in previous work, not all written genres spell out a linguistic endorsement of the vernacular (Fenigsen 1999). In the absence of a standardized orthography, as is the case with Bajan, writing involves decisions about spelling. These decisions are enmeshed with ideologies that enregister orthographic variation as indexical of stereotypical social personae, identities, and communicative contexts and genres (Preston 1982, 1985; Rickford and Traugott 1985; Mishler 1991; Schieffelin and Doucet 1998; Jaffe 2000; Jaffe and Walton 2000; Silverstein 2003; Agha 2005). The hegemony of orthographic standardization casts any divergence from the standard as marked: both a resource and burden. The ideological markedness of orthographic deviation is a semiotically powerful tool for explicitly political projects and for stylistic strategizing. Politically, Non-Standard (NS) spelling may serve as an assertive "anti-language" (Halliday 1978) for claiming linguistic and cultural difference (Joseph 1987; Sebba 2007). 1 Stylistically, it can be used to depict orality, outspokenness, and humor (Preston 1985; Fenigsen 1999; Jaffe and Walton 2000). But the same diacritical contrasts with the standard that create an ideological space to problematize its hegemony call up an interpretive framework that equates any divergence from it with intellectual inferiority, crudeness, and parochialism (Preston 1982, 1985; Callender 1987; Fenigsen 1999; Jaffe and Walton 2000). It is the entextualized voice of the folk liminally pitched between a Herderian authenticity and Rabelaisian carnivalesque. For Barbadians, the NS spelling summons yet another construct which coordinates its perception: the notion of the lack of systematicity in spoken Bajan that many Barbadians, including those who are Bajan-dominant, attribute to this variety. Common references to Bajan as "talking anyhow" and a "broken English" The assertive use ofNS spelling has been documented by Sebba for British Creole (1998) and other varieties (2007), Romaine for Hawai'i Pidgin (2005), Winer for Trinidadian Creole (1990), Hinrichs and White-Sustaita (forthcoming) for domestic and diasporic Jamaican Creole online, and Deuber and Hinrichs for Nigerian Pidgin and Jamaican Patwa in Computer Mediated Communication (2007).
1.
"Flying at half-mast"? Voices, genres, and orthographies in Barbadian Creole 109
devoid of grammar reflect this notion. A statement made by a Barbadian journalist who described Bajan to me as "essentially. a collection of linguistic errors" illustrates it quite well. Thus, as a vehicle of writing, the NS ofBajan is nonstandard (and problematic) in two related yet distinct senses. First, it is divergent from the Standard English orthography, an unquestioned paragon of scriptural technology in Barbadian ideologies of literacy. Second, it is nonstandard in not having been systematized, that is, in not being norm-governed. In the research on spelling, this second issue has been eclipsed by the first one. Yet, in the British thought, the treatment of systematic, unified language and its invariant orthography as a prerequisite of rationality. social good, and the welfare of the state dates back at least to the sixteenth century when- a century before John Locke's crusade for language purification in the name of governmentality (Bauman and Briggs 2003) -John Hart justified the need for orthographic reform in these very terms (Blake 1997). Within the pale of this British language ideological tradition, which extends to Barbados (Fenigsen 2007), linguistic systematicity has been long treated as a cornerstone of modernity. Although its actual impact may be hard to gauge, comments made by Barbadian readers I interviewed about Bajan writing point to yet another factor that weighs in negatively on the reception of NS spelling. That factor functions metapragmatically (Silverstein 1993) via the syntagmatic patterning of written text. Bajan is closely related to English, and its entextualizations take hybrid forms that draw together Standard English (SE) orthography and NS spelling. The diacritical contrast between the two varieties is usually accomplished by selective renditions of shibbolethic features of Bajan phonology through the use of characters that comprise SE alphabet. So, for example, the title of a poem discussed below is spelled "Lissun tuh de Bishop=' For a reader who is literate in SE and who processes writing in a linear mode, the SE-spelled segments of such a hybrid form set up an expectation as to the SE-patterned sequence of characters in the unfolding written word which becomes then disrupted by a divergent usage. So, while it may not be easy to separate the stigma related toNS spelling from the one related to Bajan whether spoken or written, I have argued above that the spelling itself carries baggage that stands in the way of its acceptance. 2 In literary cultures possessed of textual conventions for representing vernaculars such as 2. The abllity to better separate the ideological impact of NS spelling of Bajan entextualizations from that of Bajan entextual!zed through a standardized spelling would require experimental research. Because this paper is concerned with the ways in which readers have been able to suspend the generically regimenting interpretations ofNS spelling so as to see other textual dimensions of that writing, such research would be beyond its scope. I am grateful, however, to Lars Hinrichs and to an anonymous reviewer of this manuscript for pointing me in the direction of this analytic issue.
no Janina Fenigsen
Barbadian and our own, NS-spelled voices have been consistently lower-class, and the practices of literarization that regiment the entry of vernaculars into literary and journalistic texts (Pollock 1998) have been restricting their generic range to humor and parody (Rickford and Traugott 1985; Schieffelin and Doucet 1998: 293-294; Fenigsen 1999; Sebba 2007: 105; see also Bakhtin 1981: 363-364). The authors who would like to break away from these conventions have to contend with the expectations of readers and publishers (Fenigsen 2003). My earlier work on representations of Bajan in print shows that the association ofNS spelling with humor and parody reinforces the dominant ideology of Bajan as inferior and further hinders its expansion into a more robust range of print genres (Fenigsen 1999). A cartoon by Winston Jordan published in The Nation, one of the Barbadian daily newspapers, captures this division of spelling labor.
Figure 1. "Chango" cartoon. Winston and Winslow Jordan. The Nation, November 4, 2006,p. 6A
"Flying at half-mast"? Voices, genres, and orthographies in Barbadian Creole
In setting the newspaper column in the SE orthography and the exclamation of a distraught monkey in the NS spelling, the cartoon is a diagrammatic icon (Parmentier 1985) that represents the allocation of spelling strategies to textual genres within the newspaper while associating the NS spelling with (simian) orality. I have discussed these issues in some detail in previous work arguing that Bajan has been present in print primarily as a representational target of discourse hostile to it and so perceived by its speakers. My goal here is not to reiterate that analysis but to elaborate on it by attending to some of my expanded data that complicate the negative force of NS-spelled entextualizations of Bajan. In the present analysis, I will suggest that the regimenting role of spelling in allocating the text to a genre is not always straightforward for the reader but rather can be inflected by extratextual factors, by broader contextualizing frameworks, and by more nuanced interpretations of the ideological positioning of the author's voice toward the one represented through such a spelling, what Bakhtin (1981: 359) called the intentionality of the representing linguistic consciousness. I start by identifying distinctions between Bajan and Barbadian English phonology. Next, I situate the writing and reading of NS-spelled texts against the backdrop of the Barbadian sociolinguistic situation and consider the ways in which writers in Bajan situate themselves toward Bajan and NS spelling. Finally, I discuss the interpretations of NSspelled Bajan texts by four readers from Arawak Hill the village where I did my fieldwork, while attending to their generic allocation of these texts.
2.
Bajan and the Barbadian sociolinguistic situation
The Barbadian sociolinguistic situation reflects more than three centuries of British colonization. As did other creole languages in the New World, Bajan developed on plantations in the contact between the British colonizers (elites as well as indentured servants) and enslaved Africans. It has drawn its substance from the many languages spoken by the populations in contact (Handler and Lange 1978; Rickford and Handler 1994). While some Bajan syntactic structures may suggest an African influence (Burrowes and Allsopp 1983), nearly the entire lexicon was drawn from the regional Englishes spoken by the resident British colonizers (Niles 1980). Bajan differs from spoken Barbadian English (BE) in its phonological features, which I illustrate using Rickford's (1987) adaptation of Cassidy's phonemic system (see Appendix I). The written English form is in the parentheses. For consonants, these features notably include interdental stops. Bajan has non-fricative allophones of interdental voiced and voiceless fricative stops in BE: dhem => dem; thing=> ting. This stigmatized feature, highly salient to Barbadians, is (prescriptively) absent from BE. Other phonological features of Bajan include a tendency to reduce consonant
111
112
Janlna Fenigsen clusters, r-fullness, the use of glottal stop allophone [bAt] => [bA?] ('but'), and a preference for an open syllable structure often considered to be a carryover from West African languages (see also Alleyne 1980; Sutcliffe and Figueroa 1992). This preference sometimes results in deletion of the word-final syllable-final consonant (jlaag =>fie, 'flag') and sometimes in vowel epenthesis (bakl => bakele, 'buckle'). Other characteristic Bajan features include the presence of pure vowels in the distribution corresponding to diphthongs in BE. For example, in Bajan, BE ai => a ('f). Also, Bajan is characterized by the heightening of the first part of diphthongs where a diphthong obtains: mit ('right') is likely to be articulated as 1~it. Another feature is the nasalization of vowels in the environment of nasal consonants. Since some semantically salient phonetic features of spoken Bajan have no close English equivalent, their representation involves more than a stylistic choice. For example, in English the negation of the verb 'can' is achieved through the addition of the morpheme 'not,' whether in a full or contracted form. In Bajan, the negation of'can' is achieved through the lengthening and, at times, the nasalization of the vowel: kaan. Some articulations also include the palatalization of the initial stop. In writing, this form is usually represented as cahn, kahn, or kyahn. In this instance, the distinct Bajan phonology is not only a pragmatic index of a distinct phonology but also carries important morphemic function. In order to preserve the semantic integrity of that variety, the choice is between "respelling" it (Preston 1982: 309) and translating it into English, thus obliterating the distinctiveness of Bajan. Another Bajan lexical item is the pronoun for the second person plural usually spelled wunna, likely derived from West African languages (Holm 1988a). Although Bajan today is formally dose to its metropolitan lexifier, English (Cruickshank 1916; Reinecke 1938; LePage 1958; Niles 1980; Roy 1984; Holm 1988b; Rickford 1992; but see Blake 1997), the intervarietal distinctions matter to the speakers. Without fluency in the spoken Barbadian English (BE), distinct from British and American spoken standards in prosody and some lexical usage, it can be difficult to access better jobs, schools, and upper-class social networks.3 Like many other vernacular and creole languages (Rickford 1983; Rickford and Traugott 1985), Bajan is considered by its speakers to be natural, indicative of in-group solidarity, and superior to BE in its capacity for humor and intimacy. Yet its limited presence in print and the prescriptive diglossic ideology that stipulates BE as the language of officialdom, of business elites, of many churches, and of schools contributes to the prestige differential between the two varieties and their speakers. Only some Barbadians, in addition to their competence in Bajan, are comfortable speaking English. For many, their command of written English, acquired at school, is not matched by its spoken 3· My own research has been focused on Bajan. I am not familiar with any work devoted specifically to Barbadian English.
"Flying at half-mast"? Voices, genres, and orthographies in Barbadian Creole 113
command As a schoolteacher in the village where I did research observed, "We cannot speak the language we write, we cannot write the language we speak." Most Barbadians consider BE as mandatory in official discourse and in serious journalistic reporting. While Bajan may be an important emblem of the "grassroots" Barbadian identity, within domains characterized by the demand for English, its use is to be confined to carefully managed sidelines and to sanitized quotes, framed by a change in voice quality or by a distancing phrase ("as they say...") that detaches the authorial voice from a ventriloquated one and marks the deliberateness of creole usage (Fenigsen 1999; see also Bourdieu 1991: 69). The growth of jobs in education, government administration, tourism, offshore banking, and publishing (Freeman 1991) over the years has expanded the domains of linguistic censorship, increased pressures for English, and heightened linguistic insecurity (Bourdieu 1991:55, 69). A Barbadian acquaintance of mine from the village declined to talk to a Barbadian store manager because his potential interlocutor appeared to be a more authoritative speaker of English. The felt need to speak "good" English has been strengthened further by the ideology that identifies Barbadian nationhood with Westminster democracy and with Standard English, as well as by the long-standing tradition of economic emigration to Canada, Great Britain, and the States, with English as its linguistic vehicle. In many ways, the legitimacy of English in Barbados saturates "the whole process of living" (Williams 1977: 110). 2.1
Writing Bajan
Brought into sharp relief by its intrusion into writing where SE is de rigueur, Bajan allows writers a stylistic shortcut to the rich domains of meaning it summons: lower-class orality, outspokenness, and a grassroots Bajan identity. In Bakhtinian terms, the function of written Bajan is to quote "another's speech in another's language, serving to express authorial intentions ... in a refracted way" (1981: 324). These ideologically inflected depictions of local heteroglossia provide a flexible tool for charting the authorial voice and its positioning toward Bajan and its speaker. Yet, as I have suggested above, the use of Bajan in literary expression can be problematic for Barbadian writers and readers, influenced by a century and a half of mass education that made SE the tool of Barbadian literacy. In this paper, I will argue that, for the readers in this study, their perceived selfpositioning of the writer toward Bajan and Bajan voices is a key factor in their decisions of whether to treat an NS-spelled text as humorous and parodic. Because of this, I would first like to address the ways Barbadian writers situate themselves toward Bajan. In discussing these authorial stances, I draw on the distinction suggested by Jaffe and Walton (2000). In their experimental study of US readers' interpretations of non-standard orthographies in North American English texts,
114 Janlna Fenigsen
Jaffe and Walton (2000) make a distinction between "distancers'' and "nondistancers=' The distancers accepted SE spelling as a "linguistic and moral baseline" and identified NS spelling with a stigmatized social category. While the nondistancers too recognized social language hierarchies, they "did not always hear a stigmatized voice in the non-standard text" (2000: 571-572). Jaffe and Walton point out that these categories are not clear-cut; instead, they "are axes of relative orientation" of readers toward NS-spelled texts and the social identities they perceived as represented through those texts (2000: 571). Because, unlike Jaffe and Walton's readers, some Barbadian writers express clear ideological commitment to the dialect, I modify the distinction as a contrast between "self-distancing" and "self-identification=' Three of the writers whose views I discuss below (Jeanette Layne-Clarke, Norma Meek, and Austin Clarke) have authored the texts discussed by the readers and presented in the next section of this paper. While Timothy Callender's writings were not among the texts selected by the readers, I include his perspective on Bajan and NS spelling because of his explicit position on the issues that became well known and influential in the literary circles. In the ensuing analysis, I do not wish to advocate any particular stance toward written Bajan; rather, I would like to point out the complexity that characterizes the relationship that writers have with the medium of their craft. In her recent poem, "Dialeck in Crisis," Jeanette Layne-Clarke, a well-published and popular author, defends Bajan and Bajan writing as central to the Barbadian way of life. Dialeck in Crisis
But leh dem prance an' hop I still en gine stop puttin' Bajan dialeck 'pon show... As Sa.ndi say, duh could like it m· lump itFrom my pen it will always flow!
But let them prance and hopI am still not going to stop putting Bajan dialect on the show... As Sandi says, they can like it or lump itFrom my pen it will always flow!
(verses omitted)
(verses omitted)
Dialeck is a. element o' cultun:, Part o' we way o' life, An' to a.ttack a.nybody dat promotin' it Is to try to engender strife.
Dialect is an element of culture, Part of our way of life, And to attack anybody who is promoting it Is to try to engender strife.
Duh should realise dat we vernacular
They should realize that our vernacular
Like a tasty pot o' stew,
Like a tasty pot of stew,
"Flying at half-mast"? Voices, genres, and orthographies in Barbadian Creole 115
Dat de seasonin' we does lace it wid Does mek it sweet fuh true.
That the seasoning we lace it with Does make it truly sweet http://www.caribvoice.org/, 01/04/2008
Although through the referential layer of the text the poet delivers a spirited defense of Bajan, some aspects of the text seem to qualify if not problematize the degree to which the author identifies herself with Bajan. For one, the text refers to Bajan in a folklorized, homely light- a "tasty pot of stew," the imagery that evokes intimacy; informality; and quotidian conversations, that is, the contexts identified with the varieties that inhabit the Low end of diglossic distribution. Further, it seems significant that the line "From my pen it will always flow!" (which most directly represents the authorial voice) is in SE lexically and syntactically and is entirely spelled in the SE orthography, so that the iconic meaning contradicts the referential one. The author's self-distancing from this variety was clear in an interview she granted me many years ago, in which she emphasized that she did not grow up speaking Bajan, became interested in it by "hearing maids and other such people; and noted that she would never use Bajan in her serious writing. Another Barbadian poet, Norma Meek, whom I interviewed in 2002 (in person) and in 2004 (via e-mail), repeatedly expressed hope that Bajan and its speakers would eventually gain full acceptance. She viewed Bajan poetry as a goal in itselfas well as being instrumental to stimulating literaryinterests among Barbadians. Meek told me that she thought of Barbados as a bilingual country like Canada. Although she often found it difficult to publish through official venues and has been accused in the media for linguistically corrupting the Barbadian youth with her Bajan writing and by her use ofNS spelling, the digital technology has allowed her to publish her work herself (see also Mi.ihleisen 2002: 203-204). Meek considers both Bajan and English as central to her identity as a poet and a Barbadian. Unlike Meek, some writers who identify themselves with Bajan programmatically renounce NS spelling. In his writer's manifesto, Timothy Callender called for the recognition ofBajan as "a serious mode of national communication:' protested the "assumption that [Bajan] smacks oflow-life, immorality and ignorance" (1987: 25), and considered writing Bajan necessary for "recording the way of life of our people (1987: 28). Yet, he rejected its "phonetic spelling" arguing that: Dialect-writers spoil their work by the pwposeful mis-spelling of words. They take delight in personal, whimsical spelling [...]. They accept the view that dialect is strictly an uneducated way of talking, and they try to reflect the un -education of the dialect -speaker by the corruption of normal, standard spelling. (Callender 1987: 29)
Yet, in spite of his argument that the distinctiveness of Bajan can be adequately marked by "the structure or rhythm of the language rather than [Non-Standard]
116
Janlna Fenigsen spelling of the words" (1987: 29; see also Brathwaite 1984), in his own literary practice the writer frequently departed from that position by using the NS spelling. For example, in his short story, "Grandfather Willie and Mr. Hitler" (from the same collection as his programmatic essay and published under the same date), Callender writes, "'Man, don't tek it so serious. man ... Gimme a drink,' Gran'father Willie says, 'and I going tell yuh the story"' (1987: 10). Without an authorial commentary about the inconsistency between the writer's program of rejecting NS spelling and his writing practice, it is not clear how to account for it. The issue is even more intriguing because Callender's discussion of writing Bajan refers to this particular short story, and both texts have been published together. The writer's position that a divergence from SE orthography is a "corruption of normal, standard spelling" (1987: 27) suggests that the hegemony of the orthographic norm ingrained by schooling and other practices of literacy- rules even over those who aim to disrupt the Standard's monopoly. Callender's programmatic approach toNS spelling and the notion of conveying the distinctiveness of Bajan through "the structure or rhythm of the language" may have been influenced by Frank Collymore, a Barbadian poet, author of Barbadian Dialect (1992 [1955]), a founder and editor of the literary magazine Bim where Callender published his early work, and Callender's teacher and literary mentor when Callender attended Combermere, which is the oldest secondary school in Barbados. Collymore had left his mark on many of his students and young writers. including Austin Clarke. In an interview that appeared in Pagitica (November 1999), Austin Clarke, Barbadian writer who lives in Toronto, credited Collymore for informing his strategies for writing Bajan: [...] It began [when Barbadian novelist George Lamming] would be on a program that originated from the BBC [...].It came on Sunday nights. So in Barbados, on Sunday nights, we would hear stuff written by Barbadians, or Trinidadians, Jamaicans. And we leap up with, "Jesus Christ, they're talking 'bout me! From England It's got to be very important:' [So] that suggested legitimacy. That's when I thought I would try to reproduce, in my way, the way my Barbadian characters talk. Now, I had an intellectual problem with that; [by] which I mean, not many people outside of Barbados would understand what I was saying. So I, therefore, thought fd capture the essence of the language. [Poet and actor Frank Collymore] told me many years ago, "You don't have to write your language with all these apostrophes, and spell the words the way they sound in the dialect:' He meant, write the words in a traditional manner, like English. But you have got to understand the rhythm, so [ ...] you're dealing with the arrangement of the words to create a different flow. [Then] you can spell the words the same as they're spelled in English. Takes time. Austin Clarke, interview for Pagitica No.1. November 1999.
"Flying at half-mast"? Voices, genres, and orthographies in Barbadian Creole 117
The interview documents the impact of the British media in the legitimization of Bajan writing for young Barbadian writers in the late 1940s. It also highlights the trade-offs involved in the decisions to write Bajan, notably, the risk of alienating the potential readership. 2.2
Reading Bajan
As I suggested above, for Barbadian readers, the NS spelling is problematic for several reasons. First, as I learned through my fieldwork, even these Barbadians who declare interest in Bajan literature propose that reading it requires a special skill. This perception is rather common. For example, a contributor to a Barbadian discussion forum wrote recently, I remember at one point in my life when Standard English was the only thing I knew how to read and write, though I was proficient in speaking patois, Janet (sic) Layne-Clarke's column, Lickmout Lou was a real challenge to read, until I taught myself. Now I have a good grasp of many local dialects. including Jamaican, which I enjoy reading. Juju 101 BarbadoS@FlyBarbados, Nation News, 1/23/2007 http://classifieds.nationnews.com,7/1/2008
While as a native speaker of Polish, I am not in the best position to assess the difficulty that Barbadians literate in English face in reading NS-spelled Bajan texts, it seems that the stance also reflects a degree of self-distancing from these texts by their readers and, at least in the case of Juju 101, also from Bajan itself. This selfdistancing is implicit in the way Juju 101 refers to his or her own Bajan competence as being "proficient" in speaking it. usually a term of choice when referring to one's competence in a language other than one's own. 4 As I noted above, Barbadians often object to Bajan writing by citing the variation in NS spelling which offends their expectations toward printed text to be orderly and systematic. And some view the spelling variation as unavoidable because they perceive Bajan itself as inherently unsystematic. But, the variation in spelling practices has been also invoked as a technological problem with publishing Bajan texts. In an interview with me, an assistant editor of one of the main Barbadian daily newspapers, responsible for copy-editing, viewed working with Bajan texts as a particular challenge. While in line with the dominant Barbadian language ideology; he objected to publishing texts that "violate Standard English grammar and orthographic 4 I should note that in the absence of specific information about Juju 101 's national identity, and given Juju 101's use of the term "patois" in reference to Bajan (a label I never heard used by a Barbadian talking about Bajan), it is possible that Bajan indeed is not a native variety for this particular forum participant.
us Janlna Fenigsen usage [and] have no alternative standards to refer to=' He justified his position that Bajan texts should be taken off the newspaper columns primarily in terms of a technical concern. For him, Bajan should "stay away from printed page:' because "newspapers should disseminate correct language no matter what that language is," and in Bajan's case, "you can't get it spell-checked so how do we know it is correct?" Aside from its resistance to the technologies of reading and publishing, NS spelling offends two divergent though historically related sets of values. For the many Barbadian readers who view SEas the paragon of linguistic correctness and authority, the NS spelling is no more than its corruption. As one person said, "I can't get it read. Have no patience for this foolishness. If you take your time to sit down and write, you should be able to do better than that. Is giving bad example to the youngsters, that's all it is. If you not prepared to spell properly don't write at all." For those who identify with Bajan, NS spelling evokes a problematic generic tradition of at best humorous, at worst derogatory, representations of that variety. What kinds of interpretive frameworks, then, allow some Barbadian readers to move beyond these default readings?
3·
So, how funny is itf Orthographies, voices, and genres
Briggs and Bauman (1992) argued that in the study of genres, instead of assuming their formal fixity, we should consider the ideological aspects of social practices of production and reception of genres and recognize the fluid, open-ended nature of their classifications (Bauman 1986b; Hanks 2000: 135-144). While neither these scholars nor the influential theoretician of genres, Bakhtin (1981), consider the role of orthography in relation to genres, orthography plays a significant role in the allocation of Bajan texts to particular genres, as it does in many other cultures of writing that contrast standard and non-standard spellings. In particular, NS spelling invokes "low," facetious genres (Rickford and Traugott 1985). As one of my Barbadian interlocutors put it. to illustrate for me the association between the NS spelling and humor, "if you write somebody death certificate like this, people be thinking you's making sport~ And yet, notwithstanding these indexicalities of NS spelling, some Barbadians provide NS-spelled texts with readings more nuanced than those summoned by the dominant ideology and the sedimented traditions of writing. I analyze here spelling choices in four Bajan texts: two poems, a newspaper essay, and the lyrics for a religious song written for a Bible study group, while discussing their interpretations by the readers for whom the NS spelling did not fully determine the generic attribution of texts. The textual and interpretive data were collected in the course of my ethnographic fieldwork in Barbados in 1992, 1994,
"Flying at half-mast"? Voices, genres, and orthographies in Barbadian Creole 119
and 2002. I also conducted some follow-up interviews over the phone in 1996 and via e-mail in 2004 and 2005. I participated in the daily life in a village I call here Arawak Hill, recorded a range of communicative events, and did 68 "spontaneous or free conversation" interviews (Wolfram and Fasold 1974: 48) with the residents. I found out that Bajan- although prevalent in the discursive practices in the village, in Bridgetown, and in the hallways of the Cave Hill campus of the University of the West Indies - aside from a political cartoon or an occasional satirical column or poem, was nearly absent from Barbadian newspapers. I incorporated the theme of Bajan in print in my informal interviews asking 68 adults in the village whether they read Bajan texts. Of the 49 regular newspaper readers, all were familiar with Bajan writing but only four acknowledged reading it: Vireen, Gloria, Maureen, and Eric. They offered to save Bajan texts for me, and it is these texts that comprise the textual data for this analysis. At the time of the study, Vireen was in her late 20s. She was a teacher in the local primary school finishing her M.A. degree at the University of the West Indies in Cave Hill. In addition to her teaching responsibilities, Vireen was very involved in the lite of the school, presiding over the Parent Teacher Organization, organizing fund-raisers. and tutoring students. Her commitment to the school and her students and her nearly completed advanced degree would make her a likely candidate for an assistant principal ("Head Mistress). Some of her colleagues as well as some of the parents thought her promotion unlikely, however, due to her "deep" Bajan accent. Vireen herself acknowledged what she humorously referred to as her "phonological handicap:' but said she did not worry about it, pointing to the responsibilities that a promotion would entail without a proportional remuneration. The issue became moot, when shortly after the completion of my fieldwork in 1994, the tiny village school closed down and its staff and students were transferred to a larger school a few miles away. The discussions of the texts were impromptu. A few weeks after we first talked about Bajan writing, I met Vireen on the road and she invited me to her house. She just got a slim volume of Bajan poems by Norma Meek and oftered to show it to me. She also saved for me Jeannette Layne-Clark's satirical poem, "I gine to Illaro Court" Citing the statement that an NS-spelled obituary would be interpreted as a joke, I asked Vireen whether she agreed with the view that NS spelling marked text as humorous. She observed that many Barbadians looked at these texts that way, but suggested that the interpretation of NS-spelled texts was more complex. My discussion of the spelling of each of the two texts is followed by its interpretation by Vireen. The first text is a poem written by Jeanette Layne-Clark that appeared in a Barbadian newspaper.
120
Janina Fenigsen
Figure 2. "I gine to Illaro Court:' Jeanette Layne-Clark. 1he Sunday Advocate, March 13, 1994,p.8
In a subsequent interview with me in 1996, the author noted that in her spelling she aimed to achieve a close phonological representation of the way Barbadians actually speak. However, the examination of the text shows an over-representation of creole features. This over-representation is most tangible in the writer's categorical avoidance of 'th' clusters. There are 36 tokens of d, t, and f, taking up all of the 36 slots in which 'th' would appear in SE orthography - as in, "Oat I had wid de hierarchy." This is quite unlike the Bajan spoken practice where Bajan speakers alternate SE interdental fricative articulations with Bajan labio-dental fricatives and dental stops. The following excerpt from my recordings of Bajan conversations illustrates the alternating use of these articulations, each marked in bold print, where /th/ represents the interdental fricative: "Mostly those people must be wearing a tie, de person in government will be wearing a tie:' The radical erasure of 'th' clusters in the poem's spelling has to be seen in the light of the dominant linguistic ideology that treats the Bajan articulations as highly stigmatizing (see also Bauman 1986a: x). The poem is also characterized by the lavish application of a phonologically superfluous apostrophe (59 count). As I suggested elsewhere (Fenigsen 1999), the only representational function of the apostrophe is to index the absentee markers ofSE, as in the items li'l or 'cardin', the latter of which is framed with an apostrophe in the word-initial and word-final positions. The apostrophe does not cue the reader for any particular articulation. Instead, it serves a diacritic function of
"Flying at half-mast"? Voices, genres, and orthographies in Barbadian Creole t:n
specifying the difterence between the print representation of Bajan and Standard English spelling, the strategy that represents Bajan not in its own terms but rather in terms of its perceived shortcomings. It also insinuates an intertextuallink between the poem and other texts accumulated over the long history of derogatory representations of creole varieties that have drawn on the same formal resources. A similar use of the apostrophe is illustrated by the title of one such derogatory text, published in the Barbadian Times in 1877, "Old Peter's Address on 'Larnin"' and 'Spellin"' (Fenigsen 1999). Some other aspects of the poem's spelling merely suggest oral "allegro" speech quality (Preston 1985: 329) of "poor li'l Cinderilla's" language, as in the item hummuch. The NS rendition of some SE glides as pure vowels (as in tek and nuh), on the other hand, accurately represents their articulation by some Bajan speakers. In her generic attribution of the poem, Vireen paid attention to the relationship between NS spelling and the phonological properties of Bajan discourse and to the degree and kind of divergence of NS spelling from SE spelling. Further, she considered other, content-related, properties of the text. Vireen noted that the NS-spelling of the text was very "dense:' that is, characterized by a high number of words depicted inNS spelling, in comparison with the "plain-looking" (SE-spelled) words, and noted that the author was "leaving those 'th's all out." The poem includes 128 NS-spelled lexical items out of the total of 511 -that is, about 25 percent. While Vireen considered this spelling strategy as suggestive of"making sport," and of"ridiculous" (satirical) writing, she was very clear that her classification of the text as satirical was ultimately decided by her perception of the author's derogatory self-positioning toward the represented character and her language. While the poem is written in the first person, Vireen had no doubts that the voice of the feverish social striver, a "poor-bracket body" who wants "to get into dat inner-circle" and "eat cake by de chunk an' get pissin' -drunk" was ventriloquated rather than claimed by the author as her own. While the orthographic density of the voice assigned to the "poor li'l Cinderilla" contributed to creating the ideological distance of the author from the character, that distance for Vireen depended on the author's descriptive strategies. Taken together, these strategies make the poem an exemplar of an "objectivized word; fully dominated by the authorial voice, which makes the "Bajan" voice serve the author's ideology (Bakhtin 1973: 164). A destructive parody of Bajan and its speaker (Bakhtin 1981: 364), the text bears out Callender's point that NS spelling is used to reflect the "un-education" of the dialect speaker and the assumption that Bajan "smacks of immorality, lowlite and ignorance." Vireen's other choice was Norma Meek's poem, "Lissun tuh de Bishop."
122
Janlna Fenigsen Listun tuh de Bishop Duh put one a we people at de top a de wite people big expensive church! Duh in nuh higher office.
Do you believe that God has called you ... ? I believe... Ten times a nite we hear it pon t.v. duh offring it fuh sale on viddytape Like roll yuh bamsy, wuk-up And behave rude? De Bishop waun no, too.
so all a we no dat.
So now we culture cud come in de Church? Tuk ban an lanship? Calypso? I waiting now to see ... Buh wait. .. We culture does include dem tings He talk bout de Crop Over Festival and lash out at de vulgar spectacle of young and old who did wuk up fuh so and did get drunk and did behave real crude.
If. .. what I saw is Barbadian culture As Ch1·istians we can do without those aspects. Bishop Brume, uh cud shake yuh hand! Guh long an talk yuh talk! Uh glad to hear yuh. Wunna Bajans, Pay attenshun to dis man! Lis sun, nuh? (Meek 1993: 30-31) Though also dominated by dense NS spelling (49 out of 143 words, that is 34%, comprising the Bajan verses of the poem) and by a categorical absence of'th' dusters in these verses, Vireen considered it a serious political commentary. The poem code-switches and voice-switches so as to express the poet's reaction to the widely publicized nomination of the first Black Barbadian as an Anglican bishop in "de white people big expensive church" and to the bishop's ensuing disappointing criticism of "vulgar" folk forms of Barbadian culture. Three formally and ideologically distinct voices are engaged in a rich argument about cultural loyalty and betrayal: the SE voice of the Anglican Church and its white metropolitan culture, the contemplative Bajan voice of the author, and an ironically cast voice of the bishop's Bajan accomplice. The poem's ideological drama is set up as a disenchantment of the author's hope that "now we culture cud come in de Church," first disrupted by the bishop's authoritative voice and finally destroyed by another Bajan voice that caters to the censorial bishop in the poem's coda.
"Flying at half-mast"? Voices, genres, and orthographies in Barbadian Creole 123
The formal distinction between the authoritative voice of the Bishop and his Church and the Bajan folk expression is sharply demarcated by the contrast between the Standard English of that authoritative discourse and the NS spelling and creole syntactic and lexical forms of the two Bajan voices. The discursive excerpts representing the bishop's voice are further set off as alien by the poet's use of italicized font (see also Kress and Leeuwen 1996). The formal distinction between the Bajan voice of the authorial reflection and that of the bishop's Bajan interlocutor and ideological accomplice in the coda is parsimoniously achieved by the orthographic contrast between theSE-spelled authorial I in "I waiting now to see" with the Bajan-accented uh in, "Bishop Brume, uh cud shake yuh hand!" The way in which Vireen read the poem aloud and her subsequent analysis supported the salience of those formal distinctions. In her animation of the text, Vireen differentiated between the three formally and ideologically marked voices and performed their conversational interaction (Hill1995). Vireen read the authorial-voice stretches (stanzas 1 and 3-6) as "her own" - at a fast pace, with a relatively level prosody, and in the voice quality that usually characterize her unagitated, conversational speech. She set off the italicized SE stretches of the text by prosody and timing evocative of a sermon held in the local Anglican church. Some more notable features of that rendition included the holding (in the musical sense) of word-final consonants, a slower pace, and noticeable breaks between words. In reading the poem's coda (stanza 8), Vireen assumed an exaggerated, obsequious voice quality that she later labeled a "yes sir" speech. She attributed that way of speaking to some "old-timers" attempting to please and impress someone in a position of authority, archetypically, a plantation owner or manager. Vireen reflected that it was the distinct spelling of 'uh' in "Bishop Brume, uh Cud shake yuh hand" that first alerted her that the coda belonged to a voice different from the authorial one, also spelled inNS at times. Although the poem is characterized by numerous instances ofNS spelling, including the invariant representation ofBajan values of'th' clusters, a highly stigmatizing feature in Barbadian oral discourse, due to her perception of the poet's sincere engagement with important issues of Bajan culture, Vireen privileged its serious reading. The two other readers, Gloria and Maureen, were sisters. Gloria was a nurse who recently returned to Arawak Hill from Canada where she worked in a hospital for several years. Maureen worked as a maid in one of the hotels on the West Shore. At the time, Gloria was in her early 40s; Maureen was in her early 30s. Gloria and Maureen told me that while they grew up "with Bajan," mainly through their work experience they became "comfortable" moving between Bajan and English. Austin "Tom" Clarke's text "We Is All Murderers!" was produced when I visited the sisters at their home. When I was about to leave, Gloria remembered having saved Clarke's text for me. We sat down again and talked. As I did in my
124 Janina Fenigsen
•, '
FLY/ NOWWPAY LATER ITH
I
Figure 3. "We is all murderers!" Austin "Tom" Clarke. Sunday Sun, June 5, 1994, p. 21 A
conversation with Vireen, I asked them whether they considered any NS-spelled text as "making sport:' They, too, suggested that it was not just about the spelling. The essay is one of Clarke's regular contributions to his humorous, chatty political newspaper column marked by the frequent use of Bajan. In its spelling and ideological casting of Bajan voices, the text differs from the first two. In the text, Clarke, a resident of Toronto, addresses a robbery and murder committed in a Toronto cafe by four young Jamaicans - so identified by the media based on their "complexion" and "accent:' The culprits escaped, sending the Canadian police on a search and leaving Canadian public opinion shaken and vocally anti-West Indian. The entire West Indian community found itself under suspicion and scrutiny. That scrutiny, in Clarke's account, made those in the hostile limelight scramble to assert their non-Jamaican Caribbean identities by reclaiming their home-island accents. The accented bids for Bajan identity, quoted in the text as, "Gorblummuh! Lemme tell you something, see!" and "Tom, man, as one Bajan to a next, this place getting rough as so-and-so, yuh!" made by the author's friends who "in thirty years, does-'nunciate in pure Canadian twang" plead for readmission to the linguistically indexed Bajan-ness. The discourse of Clarke's text, marked by sparsely applied NS spelling (68 NS-spelled items out of nearly 1000 words) as well as by
"Flying at half-mast"? Voices, genres, and orthographies in Barbadian Creole 125
elements of creole syntax and lexicon, works in a synergy with the ideological fusion between the authorial! and the 1st person of the narrator to establish the writer as the Creole's emigrant keeper. Both Gloria and Maureen acknowledged humorous aspects of Clarke's writing. Yet as Vireen did in reading Meek's poem, they also pointed out Clarke's genuine engagement with serious issues. Referring to her Canadian experience, Gloria commented, "He really writes it like it is, man, with humor but on the serious side of it." With regard to Clarke's formal strategy, these two readers commended his near avoidance of NS-spelled substitutions for SE 'th' (only two instances in the entire text), which they both find particularly obnoxious; they thought this strategy to be particularly conducive to a serious reading of the text. Yet they primarily attributed their own interpretation to the sincerity of Clark's authorial voice perceived in its fusion with that of the narrator. The last text, "Come Into De Workshop," was written by Kim Ramsay, a journalism student at the UWI and an active member of the Pentecostal church in the village. Come Into De Workshop Refrain: Come into de workshop, Come into de workshop, Jesus Christ de Carpenter, He gine work pun you. Saw out de envy, Chisel out yuh pride, He gine work pun you. Nail up de greed, Hammer out de hate, Build up a solid gate, He gine work pun you.
(Refrain) Yuh, kant escape de Master's hand, If evuh you gih He a chance, He gine push you, saw you, chisel you too, He gine work pun you. Till you become wha He want you too, A vessel o' noble worth, So come into de workshop, He gine work pun you. (Refrain)
(Refrain) Rule out de self, Plane down yuh roughness, He gine work pun you. Straighten out yuh feelings, Square up yuh ego, He gine work pun you. (Ramsay 1994)
12.6
Janlna Fenigsen The person who first mentioned the text to me was Eric. In his 50s, Eric drove a taxicab on the coast and was a devout member of the Pentecostal congregation in the village. He was an excellent raconteur and told me that he could not imagine telling his stories and anecdotes in English. They simply would not "come off" right. But in his interview with me, Eric said that he did not read NS-spelled Bajan texts, considering them a waste of time. "We talk one way and write another," he laughed. We talked often and during one of these conversations, Eric pulled out a sheet of paper with Kim's hymn lyrics and said, "This Bajan writing don't waste your time:' Unlike the other readers, Eric did not offer a detailed interpretation of the text. He only said that in this case, NS spelling did not matter and qualified the text as serious because of its religious content. I talked about the text with Kim. She wrote the lyrics for a Bible-study group that she and her friends organized for young people in the district. At the time, Kim was in her mid-20s. I met her through her mother who was a plantation worker. Although Kim was very aware of the stigma attached to Bajan and of the reluctance of Barbadians toward Bajan writing, she chose to use Bajan rendered in NS spelling because of her belief that Bajan, the language in which Barbadians "think and feel." should be the medium of Barbadian religiosity. Kim telt that for Bajan speakers, the use of English in worship and other religious communication hindered their understanding and their emotional spontaneity, central to the Pentecostal dogma. For Kim, the use of English in church was a "detrimental legacy of the colonial past:' Unlike my other village interlocutors, Kim thought that Bajan could and should be standardized so as to enter the official repertoire, although she doubted that this would happen in a predictable future. In her spelling strategy, Kim decided on a compromise. While she noted that she would prefer a "fully phonological spelling," she felt that "ifyou do that, people will not take you seriously; they're used to 'Lickmout' Lou' (a humorous newspaper column authored by Jeanette Layne-Clarke and characterized by dense NS spelling) and other such texts=' While drawing on NS-spelled forms, Kim settled for their sparser application. Further, Kim wanted to deflect not only a humorous reading but also to avoid the impression that Bajan is "uneducated=' To this end, Kim used a morphologically English form/kant/rather than a widely used Bajan/kyaan/which she considered as "broad, broad" dialect Instead, she decided to assimilate it to Bajan through NS spelling. Her song became a permanent feature of the study group's meetings.
4· Conclusions These local readings suggest that although the relationship between the spelling of the text and its generic classification is salient to Barbadian readers, the impact of
"Flying at half-mast"? Voices, genres, and orthographies in Barbadian Creole 127
NS spelling may be less definitive than one might expect. The first text, characterized by relatively dense NS spelling (25% NS-spelled words), was classified as parodic. The second text, however, characterized by an even higher ratio of NSspelled words (34%), was read as serious, as was the last one, represented in a lighter brand of NS spelling. The allocation of those texts to either humorous or serious genre has been significantly influenced by the reader's perceptions of the ideological self-positioning and intentionality of the authorial voice, suggesting that in interpretive practices neither the identification of genres nor the reception of texts are fully determined by their formal properties (Briggs and Bauman 1992). NS spelling or not, texts can lend themselves to complex interpretations and "responsive understanding[s]" (Volosinov 1986: 41). Even though the stigmatizing potential ofNS spelling colors the writers' strategies and the readers' interpretations. that spelling alone does not fully determine the text's generic allocation. For those willing to engage with NS-spelled texts, the spelling gives up some of its regimenting power. The existence of such alternative interpretive spaces suggests other complex considerations that weigh on the classification of"Bajan" texts as humorous, parodic, serious, or in-between. These undisciplined readings indicate that once an NS-spelled text has made its first cut with the reader, its allocation to either humorous or (more) serious genres is defined not only by spelling and other markers of the Bajan voice but also by the reader's perception of the writer's self-positioning toward that voice. However, while NS spelling may not fully imply that a text will dissolve into humor and parody, the association remains. While readings such as the ones discussed in this paper suggest a potential for turning Bajan into a legitimate, "serious mode of national communication" (Callender 1987: 23), for many Barbadians an NS spelling and Bajan writing carry powerful connotations of transgression and parody. Unless the persuasiveness of that ideological framework becomes undone, Bajan writing will continue to "fly at half-mast~
References Agha, A. 2005. Voice, footing, enregisterment Journal ofLinguistic Anthropology, 15(1 ): 38-59. Alleyne, M.C.1980. Comparative Afro-American. Ann Arbor MI: Karoma. Bakhtin, M.M. 1973. Problems ofDostoevsky's Poetics. Ann Arbor MI: Ardiston Bakhtin, M.M. 1981. 1he Dialogic Imagination: Four Essays by M.M. BakhNn. Austin TX: University of Texas Press. Bauman, R 1986a. Story, Performance and Event: Contextual Studie.s of Oral Narrative. Cambridge: CUP.
128
Ja.nlna Fenigsen Bawnan. R. 1986b. Contextuali.zation. tradition and the dialogue of genres: Icelandic legends of the KraftasM!d Paper Presented at the Annual Meetings of the American Anthropological Association. Philadelphia PA Bawnan. R. & Briggs, C.L. 2003. Voices of Modernity: Language Ideologies and the Politics of Inequality. Cambridge: CUP. Blake, R. 1997. All o'we is one? Race, Class and Language in a Barbados Community. PhD dissertation. Stanford University. Bourdieu, P. 1991. Language and Symbolic Power. Cambridge MA: Harvard University Press. Brathwaite, E.K. 1984. History of the Voice: The Development of Nation Language in Anglophone Caribbean Poetry. London: New Beacon Books. Briggs, C. & Bauman. R 1992. Genres, lntertextuality, and social power. Journal of Linguistic Anthropology 2(2): 131-172. Burrowes, A & Allsopp, R. 1983. Barbadian Creole: A note on its social history and structure. In Studies in Caribbean La11guage, L. D. Craig et al. (eds.), 38-45. St Augustin: Society for Caribbean Linguistics. Callender, T. 1987. Independence and Freedom: A Story, an Essay and a Drama. Monroe Village, St. George, Barbados: Realization Studios. Clarke, A T. 1994. We is all murderers. Sunday Sun, June 5, p. 21 A Clarke, A T. 1999. Pagltica. No. 1. November. Crowley, T. 1989. Standard English and the Politics ofLanguage. Urbana IL: University of Illinois Press. Collins, J. & Blot, R.K. 2003. Literacy and Literacies: Texts, Power, and Ide11tity. Cambridge: CUP. Collymore, F.A 1992[1955]. Barbadian Dialect: Notes for a Glossary of Words and Phrase.s of Barbadian Dialect. Barbados National Trust. Cruickshank, J.G. 1916. Black Talk: Being Notes on Negro Dialect in British Guiana, with (I11evitably) a Chapter on Barbados. Demerara: Argosy. Deuber, D. & Hinrichs, L. 2007. Dynamics of orthographic standardization in Jamaican Creole and Nigerian Pidgin World Englishes 26(1): 22-47. Devonish, I. 2000. Language variation in Barbados. In Language, Blacks and Gypsies: La11guages Without a Written 'Iraditio11 a11d Their Role in Education. T. Acton & M. Dalphinis (eds.), 85-105. London: Whiting and Birch. Eckert, P. 1983. The paradox of national language movements. Joun1al ofMultili11gual and Multicultural Developme11t 4: 289-300. Fenigsen, J. 1999. "A broke-up mirror": Representing Bajan in print. Cultural Anthropology 14(1): 61-87.
Fenigsen, J. 2003. Language ideologies in Barbados: Processes and paradigms. Pragmatics 13(4): 457-482. (Special issue: Misrecognltlon, Linguistic Awareness, and Linguistic Ideologies: Ethnographies and Approaches). Fenigsen, J. 2007. From apartheid to incorporation: The emergence and transformations of modern language community in Barbados, West Indies. Pragmatics 17(2): 231-261. Freeman. C. 1991. From tourism to typing: Barbadian women in the global network. Listen Real Loud 11(1): 12-13. Grillo, RD. 1989. Dominant Languages: Language and Hierarchy in Britain and France. Cambridge: CUP. Halliday, M.AK. 1978. Language as Social Semiotic: The Social Interpretation of Language and Meaning. London: Edward Arnold
"Flying at half-mast"? Voices, genres, and orthographies in Barbadian Creole 129
Handler, J.S. & Lange F. W. 1978. Plantation Slavery in Barbados: An Archaeological and Historical Investigation. Cambridge MA: Harvard University Press. Hanks, W.F. 2000. Inter texts, Writings on Language, Utterance and Context. Denver CO: Rowman and Littlefield Hill, J. 1995. Voices of Don Gabrlel: Responsibility and self in a modern Mexicano narrative. In The Dialogic Emergence of Culture, D. Tedlock & B. Mannheim (eds.), 97-147. Urbana IL: University of Illinois Press. Hinrichs, L. & White-Sustafta, J. Forthcoming. Global Englishes and the sociolinguistics of spell~ A study ofJamaican blog and email writing. English World-Wide 32(1). Holm, J. 1988a Pidgins and Creoles, Vol 1: Theory and Structure. Cambridge: CUP. Holm, J. 1988b. Pidgins and Creoles Vol 2: Reference Survey. Cambridge: CUP. Jaffe, A. 1996. The second annual spelling contest: Orthography and ideology. American Ethnologist 23(4): 816-835. Jaffe, A. 2000. Introduction: Non-standard orthography and non-standard speech. Joun1al of Socioli11gulstics 4( 4): 497-513. Jaffe, A. & Walton, S. 2000. The voices people read: 0 rthography and the representation of nonstandard speech. Journal of Sociolinguistics 4( 4): 561-587. Joseph, J.E. 1987. Eloquence and Power: The mse ofLanguage Standards and Standard Languages. London: Frances Pinter. Juju 101 Barbados@FlyBarbados, Nation News, 23/01/2007 (7 January 2008). Kress, G. & van Leeuwen, T. 1996. Reading Image.s. London: Routledge. Layne-Clarke, J. 1994. I gine to Illaro Court Sunday Advocate, March 13, p. 8. Layne-Clarke, J. Dialeck in danger. Carib Voice. ( 1 January 2008). Le Page, R.B. 1958. General outlines of Creole English dialects in the British Caribbean. Orbis 7(1): 54-64. Meek, N. 1993. Lissun tuh de Bishop. In In My Small Con1er, N. Meek, 30-31. St. Thomas, Barbados: Norma Meek. Mishler, E. G. 1991. Representing discourse: The rhetoric of transcription. Journal of Narrative and Life History 1(4): 255-280. Miihleisen, S. 2002. Creole Discourse: Exploring Prestige Formation and Change across Caribbean English-Lexicon Creoles. Amsterdam: John Benjamins. Nlles, N. A. 1980. Provincial English Dialects and Barbadian English. PhD dissertation, University of Michigan. Parmentier, R.J. 1985. Diagrammatic icons and the historical process in Belau. America11 A11thropologlst 87: 840-852. Pollock. S. 1998. India in the vernacular mlllennium: Literary culture and polity. Daedalus 127: 1000-1500. Preston, D.R. 1982. 'Riti.n' fowklower daun 'rong: Folklorists' failures in phonology. Journal of American Folklore 95(377): 304-336. Preston, D.R. 1985. The li'l Abner syndrome: Written representations of speech. American Speech 60(4): 328-336. Ramsay, K. 1994. Come into de workshops. Reinecke, J.E. 1938 [1964]. Trade jargons and creole dialects as marginal languages. Social Forces 17, 107-118. [Reprinted in Hymes, D. (ed). 1964. Language in Culture and Society, 543546. New York: Harper and Row].
130
Ja.nlna Fenigsen Rickford,J .R. 1983. Standard and non-standard language attitudes in a creole continuum. Society .for Caribbean Linguistics, Occasional Paper No.6. Rickford, J. R. 1987. Dimensions of a Creole Continuum: History, Texts, and lJnguistlc Analysis of Guyanese Creole. Stanford CA: Stanford University Press. Rickford, J.R. 1992. The Creole residue in Barbados. In Old English and New: Essays In Language and IJngulsNcs in Honor of Frederick G. Cassidy, N. Doane, J. Hall & D. Ringler (eds.), 183-201. New York NY: Garland. Rickford, J.R. & Handler, J.S. 1994. Textual evidence on the nature of early Barbadian speech.
Journal ofPidgin and Creole Languages9(2): 1676-1835. Rickford, J. R. & Closs Traugott, E. 1985. Symbol of powerlessness and degeneracy, or symbol of solidarity and truth? Paradoxical attitudes toward pidgins and creoles. In The English Language Today, S. Greenbaum (ed.), 252-261. Oxford: Pergamon Roberts, PA. 1997. From Oral to Literate Culture: Colonial Experience in the English West Indies. Kingston, Jamaica: University of the West Indies Press. Romaine, S. 2005. Orthographic practices in the standardization of pidgins and creoles: Pidgin in Hawai'i as anti-language and anti-standard. Journal of Pidgin and Creole Languages 20: 101-140.
Roy,. J.D. 1984. An Investigation of the Processes of Language Variation and Change in a Speech Community in Barbados. PhD dissertation. New York, NY: Columbia University. Schieffelin, B.B. & Charlier Doucet, R. 1998. The 'real' Haitian creole: Ideology, metalinguistics, and orthographic choice. In Language Ideologies: Practice and Theory, B. B. Schieffelin, K.A. Woolard & P.V. Kroskrity (eds.), 285-316. Oxford: OUP. Sebba, M. 1998. Phonology meets ideology: The meaning of orthographic practices in British Creole. Language Problems and Language Pla,1ni11g 22(1 ): 19-47. Sebba, M. 2007. Spelli11g and Society: The Culture and Politics of Orthography around the World. Cambridge: CUP. Silverstein, M. 1993. Metapragmatic discourse and metapragmatic function. In REflexive La11guage: Reported Speech a11d Metapragmatlcs, J.A Lucy (ed.), 33-58. Cambridge: CUP. Silverstein, M. 2003. Indexical order and the dialectics of sodolinguistic life. Language and
Communication 23: 193-229. Sutcliffe. D. & Figueroa, J. 1992. System in Black Language. Clevedon: Multilingual Matters. Urla, J. 1988. Ethnic protest and social planning: A look at Basque language revival Cultural
A11thropology 3: 379-394. Volosinov, V.N. 1986 [1929]. Marxism a11d the Philosophy ofLaJ~guage. Cambridge MA: Harvard University Press. Williams, R. 1977. Marxism and Literature. Oxford: OUP. Wolfram, W. & Fasold, R. W. 1974. The Study ofSocial Dialects In American English. Englewood Cliffs NJ: Prentice Hall
"Flying at half-mast"? Voices, genres, and orthographies in Barbadian Creole 131
6. Appendix I Transcription Convention Consonants b voiced bilabial stop p voiceless bilabial stop d voiced alveolar stop t voiceless alveolar stop g voiced velar stop k voiceless velar stop gy voiced palatal stop ky voiceless palatal stop m bilabial nasal n alveolar nasal ng velar nasal ny palatal nasal v voiced labio-dental fricative f voiceless labio-dental fricative dh voiced interdental fricative th voiceless interdental fricative z voiced alveolar fricative s voiceless alveolar fricative zh voiced alveopalatal fricative sh voiceless alveopalatal fricative j voiced alveopalatal affricate ch voiceless alveopalatal affricate 1 alveolar lateral w labia-velar approximant/semi-vowel r alveolar approximant/semi-vowel, and rhotic y palatal approximant/semi-vowel h voiceless glottal fricative 'l glottal stop
Vowels high, tense, front unrounded lower-high, lax, front unrounded ee mid, tense, front unrounded e lower-mid, lax, front unrounded a low/open, short, central unrounded aa low/open, long, central unrounded
ii
132 Ja.nlna Fenigsen
ai
oh 0 Oi o ou oo u uu
falling diphthong lower-mid, long, back rounded lower-mid, short, back rounded falling diphthong short, central unrounded, unstressed, or short, back unrounded, frequently stressed falling diphthong long, mid, back rounded lax, lower-high, back rounded tense, high, back rounded
As John Rickford (1987: 9) notes, the system is phonemic, rather than phonetic or morphemic.
The creole continuum and individual agency Approaches to stylistic variation in Jamaica Dagmar Deuber 1his paper analyses creolisms in a set of conversations among educated Jamaicans. It focuses on morphology and syntax but le.xis is also considered Two
different approaches are applied: a quantitative approach in the framework of the creole continuum and a qualitative, interaction-based approach. The quantitative approach is useful to locate the set of data within the continuum - between the upper mesolect and the high acrolect - and to determine the relative "creoleness" of different features. The qualitative approach focuses on the fine details of variation within the data such as code-switching and helps to explain speakers' choices. The paper concludes that the creole continuum and individual agency complement each other as approaches to stylistic variation in Jamaican speech.
1.
Introduction1
Since DeCamp's (1971) pioneering work, approaches to linguistic variation which rank speech samples along a continuum from the deepest creole (basilect) through an intermediate range (mesolect) to Standard English (acrolect), based on quantitative findings, have been a major paradigm in the study of language use in creole speech communities such as those of Jamaica and Guyana (e.g. Bickerton 1975, Patrick 1999).2 From the beginning it was understood that there were correlations between speakers' linguistic performance and social rank (cf. DeCamp 1971: 385). While researchers have disagreed over the actual relevance of social factors in analyses oflinguistic variation in creole speech communities (e.g. Bickerton 1975: 6-7), Peter Patrick has fruitfully applied the methodology of quantitative sociolinguistics
Data transcription and analysis for this paper was supported by a grant from the Deutsche Forschungsgemeinschaft (DFG MA 1652/4) for the research group "Educated Spoken English in Jamaica" directed by Christian Malr, University of Freiburg. 2. In a recent study ofliterary dialect ln Michael Thelwell's The Harder They Come, Schneider and Wagner (2006) have also applied this methodology, complemented, however, by some qualitative analyses of style-shifting.
1.
134
Dagmar Deuber
in his detailed study of mesolectal Jamaican data from the Kingston neighbourhood ofVeeton (1999), and he affirms the importance of social aspects: Social stratification in Jamaica is crucial to understanding the extreme variability of contemporary Jamaican speech. The complex linguistic situation can be related to an equally intricate web of social relations, using the model of the creole continuum. (Patrick 2004: 408)
It was also recognized early on that speakers generally command a span of the continuum and may vary their stylistic level (ct~ DeCamp 1971: 350, 354). However, style has received less attention than social differentiation in studies of linguistic variation in Jamaica and other anglophone Caribbean territories. Quantitative studies done in the 1970s and 1980s indicated a close correlation between thecasual- formal and the creole- English dimensions (Winford 1991: 575-576), which suggests that stylistic variation can easily be accommodated in the framework of a creole - English continuum. However, more research needs to be done, especially, as Winford (1991: 576) already noted, in the case of"higher status groups=' It is one of the aims of the present paper to make a contribution in this area. Higher status speakers deserve particular attention in connection with stylistic variation as they tend to have a broad range of competence usually including both a form of the creole and the acrolect (Winford 1991: 576; cf. also Patrick 1999: 269-273). The present paper focuses precisely on this type of speakers, i.e. Jamaicans with a high level of formal education. The database is taken from the Jamaican component of the International Corpus of English (ICE-JA) and consists of forty texts - each text having about 2,000 words - from the category conversations (text codes S1A 001-040). 3 These represent the speech of university students, lecturers and other, mostly university-educated professionals in a range of relatively informal contexts. 4 The paper looks at creolisms in these texts, a creolism being defined here as any feature which can plausibly be attributed to the influence of Jamaican Creole (Jam C), although in some cases there may also be alternative explanations such as the common tendencies in the New Englishes studied by Sand (2005). 5 The primary focus will be on syntax and morphology but some lexical aspects will also be considered. 6 3· For details of ICE and the composition ofthe corpus, see Greenbaum (1996) or . 4· Most of the conversations were recorded on the campus ofthe University of the West Indies at Mona between 1999 and 2004. 5· Cf. also Mair (2002) on creolisms in written English and other fuctors shaping the emerging Jamaican standard 6. An analysis of phonological variation would be likely to yield very interesting insights as well (c£ also Irvine 2004), but this is beyond the scope of the present paper.
The creole continuum and lndivldual agency 135
A study of the language use of highly educated Jamaicans in the context of computer-mediated communication (CMC), i.e. e-mails and postings to internet discussion forums, by Hinrichs (2006) has actually questioned the usefulness of a continuum-based approach in accounting for the type of variation encountered in the data: [I]n the qualitative analysis of CS [code-switching] in naturally occurring language data, it is less helpful to describe a range of various intermediate lects on the continuum than to take language use to be the result of the strategies with which speakers and writers draw on the basic resources that they have at their disposaland Jamaicans see their resources as binary, i.e. at any given point during language production they are using either Jam C [Jamaican Creole] or JamE [Jamaican English] according to their own classification. (Hinrichs 2006: 11)
Thus, a qualitative approach looking at the interactional contexts of the use of JamC as opposed to JamE is more suited to these new written genres. As Hinrichs (2006: 40) explains, CMC has "generally not preserved the ordered and small transitions of the spoken continuum, but replaced them with two separate codes which are in principle easy to distinguish~ However, it is not dear whether this is due to the special nature of the genres in question or whether the gradual transitions from creole to English which have emerged in sociolinguistic studies generally do not apply to the way educated Jamaicans manipulate the varieties they command. Beyond the purely descriptive, the present study seeks to address this question and, furthermore, to discuss the respective merits of a quantitative approach in the continuum framework and a qualitative, interaction-based approach to stylistic variation. To this end, it presents two alternative ways of looking at the same selected data set: The first (Section 2) is a feature-based approach which focuses on analyses of the incidence of creolisms in relative, or, in the case of very low frequency items, absolute terms. 7 The second approach (Section 3) looks primarily at the contexts in which creolisms occur in the conversations and poses the question of why speakers vary. In this way. the paper also responds to a general trend in the study of stylistic variation. In the past fifteen to twenty years, that field has moved beyond variationist investigations in the Labovian framework correlating speakers' stylistic choices with predefined situational categories towards speaker agency approaches which try to shed light on the meanings of these choices in the context of speakers' interactions and identity negotiations (cf. the discussion in Schilling-Estes 2002). In order to enhance the understanding oflinguistic variation in Jamaica that can be derived from the conversations analysed here, the results will be compared 7· Quantitative analysis is limited to frequencies as creole-type constraints on the use of forms such as anteriority or stativity as determinants of the use of past tense forms are not expected to apply at the level of language use investigated here (c£ Patrick 1999: 265).
136
Dagmar Deuber
to previous authors' findings for other data sets. In the quantitative section, an important basis for comparison is provided by Sand's (1999) findings from a corpus of Jamaican radio texts comprising ca. 60,000 words, of which half is made up of news and formal talks and the other half of discussions and interviews. These data represent the more formal spoken language use ofJamaicans of a comparable social status to those in the conversations. 8 Where available, quantitative data for other social groups will also be taken into consideration. Hinrichs' (2006) study of the "corpus of Jamaican e-mail and other CMC" (COJEC, about 100,000 words) will be especially relevant for comparison in the qualitative section but a few quantitative findings from this corpus, mainly taken from Deuber and Hinrichs' ( 2007) work on orthography, will be used as well.
2.
2.1
Quantitative analyses Syntax and morphology
Tables 1-4 present quantitative results for the occurrence of selected syntactic and morphological creolisms in ICE-JA S1A 001-040. The comparative data used have also for the most part been integrated into the tables. The discussion below will first focus separately on the results for each area of syntax and morphology analysed, starting with questions and deft sentences and then moving on to copula, verb, and pronoun forms, and will then summarize the main findings. Wh-questions without do-support or inversion (see Table 1) are a syntactic creolism also identified by Sand (1999) in the broadcast discussions and interviews in her corpus. While she found this feature to be "clearly stylistically marked and restricted to certain contexts" (p. 142), such questions are one of the most widespread creolisms in the present data. Although Standard English syntax is the preferred choice overall they occur in many of the conversations and even in texts which show few other creole features. In addition, defied wh-questions introduced by is (though not the more basilectal a), a JamC feature which does not occur in Sand's radio data, 9 are attested in ICE-JA S1 A 001-040. These are marginal, though, as the figures in Tables 1-4 show. However, in declarative sentences, where Standard English has a parallel cleft construction with it followed by a form of be, clefting
8.
Most of these data have meanwhile been incorporated in ICE-JA
9· Sand does not spedfically mention this, but all her examples of is-defts (1999: 145) are declarative sentences, and in an additional search of all broadcast Interviews and discussions In the current version ofiCE-JA (texts SIB 021-050), no wh-question introduced by deft is was found.
The creole continuum and lndivldual agency 137
with is is very common in the present data. 10 In such a case, a is also attested. Sand (1999: 145) reports only is-defts, arguing further that this should be considered an "informal" or even "markedly mesolectal" feature in her radio corpus. For copula variability (see Table 2), 11 comparative JamC data are available from Rickford (1999). He analyses an interview with two informants who represent more or less the opposite end of the social spectrum from the speakers in ICE-JA: "retired Jamaicans in their seventies[ ... ] who live off the land in the rural and relatively isolated northeastern village ofWoodside" (Rickford 1999: 144). The comparison shows that there is indeed a big gap between their speech and the conversations from ICE-JA analysed here. Rickford's informants not only show a high rate of copula absence before Ving, but a (or its past equivalent bina, not attested in ICE-JA S1A 001-040) followed by the verb stem- supposedly the basilectal equivalent of V ing but also used "at all levels of the mesolect" (Patrick 2004: 420) -is actually more frequent in their speech than Ving. Before going (to)- or, in their speech, rather gwain (tu) - copula absence is almost categorical and it is also high before adjectives, while in the case oflocative predicates, de is a common variant. In the present data, by contrast, the highest rate of copula absence in any environment is only 14%, in the case of going(to)!gonna, and the overt JamC copula forms a and de are very rare. The comparison with Sand's (1999) results places the present data much closer to her radio corpus, as expected, but there are both similarities and differences. Zero copula before Ving and going (to)!gonna seems to occur at a similar rate in the conversations and Sand's interviews/discussions: the combined figure for these two categories in the conversations is 10%, and Sand's (1999: 114) combined figure of 6% for the radio corpus includes the news and talks in which there are no instances of zero copula. However, zero copula before adjectives is clearly rarer in Sand's data: she reports only two instances, and this is also among the features which she classifies as "mesolectal, (1999: 147-148). Zero copula before locative predicates, which has a similar rate of occurrence in the present data as zero copula before adjectives, is not reported at all by Sand (1999), nor are the overt JamC copula forms a and de.
In comparison to the corresponding texts in the British component ofiCE, it is worth notingthat deft sentences are far more frequent overall in the Jamaican material (a total of only 16 cleft sentences was found in ICE-GB S1A 001-040).
10.
n. I follow the tradition in variationlst studies of the copula to include auxiliary be in progressives and going-to-futures under the term copula.
138
Dagmar Deuber Table 1. Quantitative results for syntactic and morphological creolisms I: questions and cleft sentences Creolism
Examples from ICE-JA S1A 001-040
ICE-JA Radio data S1A 001-040• (Sand 1999) AVAILABLE number [total
JamC data
NOT
number]
percent Questions wh-questions without do-support/ inversion wh-questions introduced by Is CJeft sentences introduced by Is
Which school you went to
62 [282]
(SlA 029)
22%
So what youre doing.for summer (SlA 032) Is who was telling me
2
(SlA023)
Is only production must go on in the afternoon
15.. [37] 41%
(SlA 018)
introduced by a
(some occurrences, in discussions/ interviews only)
A lie you are telling (SlA 013)
(some occurrences, in discussions/ interviews only)
1 [37] 3%
• Co Wits in this and subsequent tables exclude occurrences in repetitions, hesitations and false starts; uncertain parts of transcriptions; quoted or imitative speech; and neutralization contexts. •• Includes one instance of a JamC predicate cleft construction not possible in EngUsh (is drive you drive come up here, SlA 034).
In the area of verb morphology (see Table 3), the comparison of the overall results from Patrick's (1999) analysis of past marking in his mesolectal data (interview as well as spontaneous speech) with the findings from the present data 12 suggests a similar gap between JamC and }arnE as has become evident in the comparison of copula forms. In Patrick's data, inflected verbs account for only 35% of the total of past reference verbs, whereas in the present data, English inflectional past marking predominates. However, if we consider that Patrick's informants in fact represent the whole range of the mesolect, from a "low" group with a past inflection rate of only 10% to a "high" group with a rate of over 70% (cf. Patrick 1999: 201), it becomes clear that the transition from mesolect to acrolect is a gradual one. In fact,
This should be taken as a rough comparison only as the variable is not defined in quite the same way here as in Patrick's (1999) analysis. Specifically,. Patrick (1999) includes past particlples whereas they are excluded here because they are very rarely unmarked (cf also Rickford 1986: 388). 12.
The creole continuum and lndivldual agency 139
Table2. Quantitative results for syntactic and morphological creolisms II: copula forms Creolism
Examples from ICE-JA S1A 001-040
JamCdata: ICE-JA Radio data Rickford S1A001-040 (Sand 1999) (1999)* number [total number/percent number] percent
percent CopuJa forms**
zero copula
we leaving now (SlA 033)
beforeVing zero copula before going
65 [693]
58%
9%
if we going communicate
20 [143]
with these people (SlA 003) 93%
14%
(to)lgonna zero copula before pred. adjectivest
the university Inefficient (SlA020)
65%
4%
zero copula before locative predicates
She here for three years' learning (SlA 026)
28%
5%
aV
(frequent) Me a talk dropouts from grades four (SlA 011) me naa tanya another year
naa*V
copuJade before locative predicates
32 [865]
6%
{
(all occurrences In discussions/ Interviews)
2 (both in discussions/ Interviews)
10 [221]
(S1A015) ('I'm not going to stay here another year') everybody de ya (SlA 025) ('everybody is here') 31%
11 8
2 [221]
1%
• Percentages are taken from Rickford's (1999) Table 6 (p. 151); for a (which Rickford collapses with bina) cf. his Table 4 (p. 149). •• Specific data exclusions in the analysis of copula forms in ICE-JA S1A 001-040 follow Rickford ( 1999: 146). t Includes sta:tive participial predicates. t This is the negated equivalent of a. Most occurrences in ICE-JA S1A 001-040 have future reference (cf. Bailey 1966:46; Patrick2004:419).
the present data fall rather neatly between the upper mesolect and the "high acrolect" (cf. Patrick 2004: 410 for this term). The latter is well represented by Sand's radio data - especially. of course, the news and talks - in which there are few tokens of unmarked past reference verbs. (Note also that the JamC past marker did, an attested but rare feature in the present data and absent from Sand's, is mostly
140
Dagmar Deuber
Table 3. Quantitative results for syntactic and morphological creolisms III: verb forms Creolism
Examples from ICE-JA S1A 001-040
JamCdata: Patrick (1999)' number [total number]
percent Verb forms unmarked past you decide you know I ca11 tense do library after all because
..
1279 [2125]
149 [1036]
60%
14%
you came to UWI and see some young people (S1A 027) unemphatic did A man him daughter did die 100 [2125] as past marker unmarked 3rd pers. present (regular verbs)
ICE-JA Radio data S1A001-040 (Sand 1999) number [total number] percent number
gt [1036]
(S1A025)
5%
1%
All she need to do 11ow is to go and do some albums for
NOT
98 [455]
AVAILABLE
22%
herself(S1A 032)
27, most in discussions/ interviews
(some occurrences, especially in discussions/ interviews)
• The data are taken from Patrick's Table 6.4 (1999: 206). '"The analysis of past marking in ICE-JA Sl A 001-040 extends to unmarked past reference verbs (including those with a possible perfect interpretation) and verbs inflected for simple past tense. It excludes all forms of be; modals; auxiliary and pro-predicate do (as well as structures where auxiliary do is absent but would be required in Standard English); verbs in clauses containing 11ever, which can function as a negative past marker; habituals with used to; and verbs in irrealis contexts. t Did+ V only. A further two instances of did occur with participial predicates.
associated with those of Patrick's speakers who have the lowest inflection rates, cf. Patrick 1999: 206.) The status of the present data with respect to inflection for the third person singular present is probably similar to what has been observed for past marking, i.e. it is intermediate between the upper mesolect and high acrolect. As can be seen in Table 3, the rate of unmarking is somewhat above 20% for regular verbs. Quantitative JamC data are not available in this case but third person singular -s is a feature highly associated with English in the Jamaican context, which Patrick (2004: 415) has described as absent from "all forms of JamC:' In the "high acrolect," it is not likely to be omitted except for instances of notional or proximity concord and the occasional slip (cf. the findings presented by Sand 1999: 134; also Mair 2002:49, 52-53). Although overt JamC forms are mostly very rare in the present data, there are a few more common ones, including certain pronoun forms. Table 4 shows the frequency of all JamC pronoun forms which occur more than once in ICE-JA S1A 001-040. As one can see in the table, the relatively frequent forms are them as third person singular subject and possessive pronoun and especially him (with the initial
The creole continuum and lndivldual agency 141
Table 4. Quantitative results for syntactic and morphological creolisms IV: pronoun forms Creolism
Examples from ICE-JA S1A 001-040
JamCdata
NOT AVAILABLE
ICE-JA Radio data S1A001-040 (Sand 1999) number [total number]
percent Personal and poss. pronoun forms me subject
M£ can't bother (Sl A 012)
59 [3263]
2% meposs.
him (masc.) subject
In Portmore back at me mummy (SlA 037) him just motivate me to want to learn more (SlA
5 [355]
1% 77 [475]
16%
021)
him (masc.) poss.
you can't see no part of him body (SlA 035)
them subject themposs.
11 [76]
14% 57 [705]
them have a lot of bills on them hand (SZA 016)
8% 7 [96]
7%
h variably absent in pronunciation) as third person singular subject and possessive pronoun. Me as first person singular subject or possessive pronoun is less frequent. In comparison with COJEC, it is instructive to note that the JamC second person plural pronoun unu -which, as a non-English derived form, will be highly marked in predominantly English speech or writing - is attested 43 times in that corpus (in various spellings; cf. Deuber and Hinrichs 2007: 31) but only once in ICE-JA S1A 001-040. Other forms attested only once in ICE-}A S1A 001-040 are we as object and as possessive pronoun. On a more general level, the findings which can be gleaned from Table 1 and the discussion so far are:
1. The conversational data in ICE-JA S1A 001-040 reach deeper into JamC than the radio data analysed by Sand (1999), but overall they remain within the range that can be described as acrolectal. 2. To the extent that comparative JamC data are available, these clearly show the differences between JamC and the data analysed here, but they also suggest that the transition from the acrolect to the mesolect is a gradual one, supporting Patrick's ( 2004: 41 0) observation about the "lack of a sharp upper boundary"
142
Dagmar Deuber
ofJamC. It is necessary, however, to fill the gaps in quantitative JamC data, e.g. by analyses of copula or pronoun variability in the mesolect. 3. There seems to be a hierarchy among syntactic and morphological creolisms such that some are relatively easily integrated into informal JamE whereas others are marked. Overall zero forms are more likely to occur in informal JamE than overt creolisms but the situation is complex; there is a hierarchy both among zero forms and overt creolisms and some zero forms (e.g. zero copula before adjectives) are actually less likely to occur than certain overt forms (cf. him as subject or possessive pronoun). The hierarchy of creolisms which has emerged from the quantitative analysis ofiCE-JA S1A 001-040 seems to correspond in many ways to a ranking of features in the framework of a sociolinguistic continuum, but this needs to be researched in more detail. 4. Although the present picture ofJamaican language variability has many gaps, it supports the notion that JamC and JamE represent opposite poles of a continuum rather than two strictly demarcated varieties, as in a bilingual or diglossic situation. Social factors are important in accounting for variability in Jamaican speech but so is stylistic difterentiation, especially within the acrolectal range. 2.2
Lexis
In addition to the above-mentioned unu, another non-English-derived word analysed quantitatively by Deuber and Hinrichs (2007: 31) is pikni. This JamC lexical item (consistently spelled pickney by the CMC writers) occurs 45 times in COJEC. A search of COJEC for the English equivalent child(ren) revealed 86 occurrences (Lars Hinrichs, pers. comm.), which means that fully one third of the total for this lexical variable are instances of pikni. This is in marked contrast to ICE-JA S1A 001-040, where there are 54 instances of child(ren) but only one of pikni (see Table 5). Overall lexical creolisms are rare in ICE-JA S1A 001-040. The evidence remains scarce even if the scope of the analysis is broadened to include items listed in Allsopp's Dictionary of Caribbean English Usage (2003), which may be considered Caribbeanisms rather than (Jamaican) Creolisms, such as the idiomatic expression and thing. Table 5 provides a fairly exhaustive list of Creolisms or Caribbeanisms in the data (except for quoted or metalinguistic speech and items that will be discussed in Section 3 below). As this table shows, apart from a few exclamations and idiomatic expressions, one finds the very occasional noun or adjective, and a somewhat higher number of verbs. Most of the latter are indirect creolisms (cf. Mair 2002) in the sense that a verb which also exists in English is used with a JamC meaning, or cases where the
Table 5. Quantitative results for lexical Creolisms or Caribbeanisms Creolism/Caribbeanism*
Examples from ICE-JA SIA 001-040
Number of occWTences in ICE-JA SIA 001-040
Exclamations
raas [indicating opposition or anger] raatid [indicating opposition or anger]
[speaker A] you are rural rural typical Jamaican educated [speaker C] Me raas (SlA 002) If somebody Is seeking Information asking questions or so don't cut across <#> Right<#> Because that would break the flow of everything <#> Raatid (S lA
2
023) Idiomatic expressions
and thing 'and so on head (Is) not good 'not right in the head'
with all the uprising and thing (SlA 017) [speaker B] Dahlia Is shorter [speaker A] [...]your head not good[ ... ]youre not taller than Dahlia (SlA 034)
8
~
Nouns
lick 'blow'
I used to get licks from my mother every day (SlA 024)
plkni 'child'
some little piknl looking thing
pocomania ("a cult mixing
Ole pocomania [exclamation] (SIA 013)
revivalism with ancestral-spirit-possession." Cassidy/LePage 1980) walk foot 'pedestrian'
I don't want no walk foot (S1A 034)
~
(S1A025)
Adjectives
gone 'past, lasf
last semester gone (SlA 025)
2 (both in S1A 013)
i a
8. ~
~ e. ~
.Q ~
t t;1
Creolism/Caribbeanism*
Examples from ICE-JA S1A 001-040
Number of occurrences in ICE-JA S1A 001-040
raatid 'damn, blasted'
half of the raatid people (SlA 003)
1**
business 'be concerned' carry 'take
they just seems to just despise each other and you know always like bad-talking each other (SlA 007) I don't really business (SlA 026) mummy did hafi carry her go hairdresser
4
somebody somewhere'
(S1A026)
cuss 'curse' juk 'stab' lick down 'knock down
you probably cuss if it don't work (SlA 007) juk the fork in It (SlA 026) That's how people get lick down all the time (SlA 034)
Verbs bad-talk 'speak maliciously'
2 2 (both in SlA 034)
look for 'visit'
When last you go look .for your mummy
2
(S1A036)
reach intr. 'arrive' reach to 'arrive at' wait on 'wait for'
Every time you go down to Western Union the money don't reach yet (SlA 034) we reach halfway to the Union right now (SlA 034) I'm waiting on him to call (SlA 031)
• All items appear either in Cassidy & Le Page ( 1980), Allsopp (2003), or both. •• There is also one occurrence of raatid as a noun (all of this raatid, SlA 002).
3
1 ~ §. !!:
The creole continuum and individual agency 145
di:ffurence lies in the complementation of the verb rather than the verb itself. However, not only do these verbs tend to co-occur with JamC morphosyntactic features, but their International English synonyms appear mostly to be preferred, though in such a small data set, the evidence is of course sparse. For example, visit (or see) seems to be preferred to look for in this sense and get there to 1-each without a complement (also, reach is more often complemented by a noun phrase than a prepositional phrase), and take (somebody somewhere) is more frequent (6 occurrences) than carry in this sense. The only apparent exception is wait on (wait fordoes not occur at all in ICE-JA SlA 001-040), an item which seems to be entering standard usage in the Caribbean (Allsopp 2003: s.v. wait)P Mair (2002, 2009) has argued that certain formal lexical items which are more commonly used in JamE than other varieties of English (especially natively spoken ones) may be considered a type of very indirect creolism in the sense that their less formal synonyms might be avoided because they are shared with JamC. 14 This would mean that they are "anti -informal" usages, the opposite of the "anti -formal" level of usage to which many direct creolisms are assigned by Allsopp (2003). A case in point is the plural form persons vs. people. A comparison of the use of these two items in ICE-JA S1A 001-040 as well as the broadcast texts in ICE-JA and the corresponding texts in the British component of ICE (Table 6) clearly shows the contrast between the two varieties of English. Within ICE-JA, persons is used especially in the formal news and talks, which lends some support to the hypothesis of hypercorrection or hyperformality. However, the figures for the broadcast dialogues and the private conversations do not di:trer greatly although the latter also feature morpho syntactic creolisms which are avoided in more formal speech. This suggests that while hypercorrection may be the ultimate source of the use of persons in Jam.E, 15 this has now reached an extent where the item has lost some of its formal connotations and has simply become a viable if somewhat rare synonym of people even at the lower acrolectallevel. Overall, the findings for lexis, limited though they are, are consistent with the findings for syntax and morphology. They support the classification of the present sample as largely within a lower acrolectal range. In addition to displaying certain
13. In the full pre-release version of ICE-JA that was used in this study (470 texts), the fre-
quency of wait on compared to wait for is only 10: 28, with a difference between the spoken component of the corpus (7: 10) and the written one (3: 18),however.
14. A similar argument of avoidance has been put forward by Irvine (2004) for some phonological variables. 15. As Mai.r (2009) shows on the basis of data from the Oxford English Dictionary quotation base, archaic usage is another possibility, but this leaves unexplained why the same usage of persons is not attested in other ex-colonial varieties.
146 Dagmar Deuber
Table 6. Frequency of people vs. persons in selected texts from ICE-JA and ICE-GB' conversations
people:persons, number (percent persons)
ICE-JA ICE-GB
broadcast news/talks
(SlA 001-040)
broadcast discussions/ interviews (SIB 020-050)
270: 35 (11 %)
270:45 (14%)
104: 48 (32%)
245:0
256: 1
104:0
(S2B 001-030)..
* Cowrts exclude occurrences in repetitions, hesitations and fu.lse starts; uncertain parts of transcriptions; and quoted or imitative speech. ** As the category broadcast talks in ICE-JA was not yet complete at the time of this writing. only the first ten texts in this category from both corpora were used
general peculiarities ofJamE (or Caribbean English), this allows the use of some rather informal creolisms, but very salient ones like pikni are extremely rare.
3·
Qualitative analyses
As this section examines the contexts of use ofcreolisms (syntactic, morphological or lexical), it presents extended quotations from the database with marking of creolisms (by boldface) added. In connection with the ICE textual mark-up symbols which occur in the quotations, it is particularly important to note that in the case of overlaps, marked by<[> ... for each string involved in the overlap and<{> for the whole set of overlapping strings as shown below, other speakers' overlapping strings- numbered if there are several -are always listed at the end of a complete speaker turn: <$A><#> text <{1> <[1> text text <{2> <[2> <,> text <$B> <#> <[1> text <$C> <#> <[2> text <$>is the speaker identification symbol, <#> indicates the beginning of a text unit roughly corresponding to a sentence in written English, and <,>marks a short pause. For full details of the mark-up, see Nelson (1996, 2002). The discussion will start out with a brief consideration of inter-individual variation in ICE-JA S1A 001-040, which will establish the necessary background for the analyses of intra-individual variation which will follow. 3.1
Inter-individual variation
The speakers in ICE-JA S1A 001-040 employ a considerable range of different speech styles, with their choice apparently being determined by a complex interplay
The creole continuum and lndivldual agency 147
of several factors, including aspects of the speech situation such as the participants and the relation between them, the setting and the topic, as well as the speakers' personal background and preferences. Examples (1) to (4) illustrate this range. The examples are representative either of the whole text from which they are taken or of what can be considered the baseline style of a speaker. All speakers in the texts in question are UWI students and all of them except speaker B in S1A 009 are female. Example (1) is from a text in which both of the informants use International Standard English grammar and lexis practically throughout the conversation: (1) [ICE-JA SlA 009, general style; topic: lecturer] <$B> <#>No he's a great teacher<#><}><-> I<=> I don't question how<{><[> what he teaches <$A><#><[> No I'm not talking about his being a great teacher <#> fm talking about people saying that he is so exceptionally hard <,> <#> But he isn't <.> <#> He is quite accommodating <,> <#> And I think it's the same thing with Mister [name] <#>It's just that we haven't been exposed to him as much as these people Example (2) illustrates a style which is fairly close to the Standard English end of the spectrum but which shows occasional absence of verbal or nominal inflections: (2) [ICE-!A SlA 032, general style; topic: boyfriend] No actually well I guess when we just started talking he was kind of<,> you know always wondering<#> He was a bit<,> you know sceptical about you know <,> who my friends are on campus and especially because he knew I was in a relationship with somebody on campus before and you know wondering<,> if<}> <->you're in <=>you're talking about their personal<,><#> I guess the whole thing go with<}><-> when <1-> <=> when you are on campus or a community the university and your boyfriend is not or your partner <,> <#> It's like you are around so many other of your peers where you see you know<}> <->people more <1-> <=>you're more likely to find people of your age <,> people who have the same goals like you and<,> all of that so I don't know if that's a part of the problem <#> I think that is it sometimes Examples (3) and (4) show not only zero copula in different contexts but also some overt creole forms, with a greater density of such features in (4), though even in that example there is alternation between the JamC and English variants of some of the relevant variables:
148 Dagmar Deuber
(3) [ICE-JA SlA 034, geneml style; topic: bayfriend] <$B> <#> He's kind of boring I realize <#> I have to find somebody new <{1> <[1> <,> <#> fllhave to go find somebody man but being that I'm always home I don't lrnow how that is going to go <#> Well I going to start back school as of next week <#> So more than likely I going to find somebody<#> I want a nice tall<#> Him don't have to be cute <{2> <[2> <,> <#>I don't want a cute guy<#> I want a nice tall dark guy<#> He has to be like five nine the least <$A><#> <[1> Mhm <$A> <#> <[2> Mhm <$A> <#> Mhm <#> That kind of short though <#> That's like my height (4) [ICE-/A SJA 021, speaker K.H., baseline style; topic: lecture1j Sometimes I think him teaching us like<}><-> we<=> we're basic school<,><#> Sometime I'm like clearly that we understand what he's saying but you see when him aks us a question and we don't communicate it in a way that him want to hear it him make it so it's like you don't understand <#> several words understand but what exactly he want us to say<#> But<}><-> that<=> that's the only problem like sometime when him aks us questions clearly the answer is so simple but we're thinking too deep 3.2
Intra-individual variation
With regard to intra-individual variation, both Sand (1999) and Hinrichs (2006) make a distinction between more abrupt changes, which are referred to as "codeswitches," and more subtle ones, for which the term "style-shifts" is used. In her analysis of the radio corpus, Sand (1999: 157) describes the news and talks as generally mono-stylistic, i.e. consistently formal. In the interviews and discussions, she found subtle shifts to be rather common, and she observes that generally "style-shifting into a more informal style with increased creole features is parallel to the emotionalinvolvementofthe speakers" (1999: 158). Code-switching is described by Sand (1999: 158) as rare throughout her corpus. In Hinrichs' (2006) CMC corpus, by contrast, code-switching is such a prevalent phenomenon that he made it the sole focus of his study. He argues that "in speech, only switching is a consciously employed stylistic strategy, and is therefore represented in writing" (p. 37; italics original). For the purposes of the present investigation, it is recognized on the one hand that there may be greater and lesser changes in the language of a speaker, given that the quantitative approach in the preceding section has corroborated the
The creole continuum and lndivldual agency 149
existence of a hierarchy among creole features. On the other hand, this very hierarchy from "less creole" to "more creole" of course makes it impossible to establish a clear boundary between style-shifting and code-switching. The discussion below will focus on the contexts in which intra-individual variation occurs in ICE-JA S1A 001-040 and on the purposes for which speakers exploit the different linguistic resources at their disposal. I will consider separately variation related to the topic, i.e. an aspect of the speech situation (cf. Blom and Gumperz's (1972) "situational code-switching"), and what Blom and Gumperz (1972) have termed "metaphorical" as well as identity-related variation. This corresponds to similar distinctions in several approaches to style-shifting and codeswitching. Bell (2001: 166) has provided a useful overview of these under the headings of"responsive" (situational code-switching and similar phenomena) and "initiative" (metaphorical code-switching etc.). He also suggests, however, that the responsive-initiative distinction is not a dichotomy but "that these may be two complementary and coexistent dimensions of style, which operate simultaneously in all speech events" (2001: 165). The present findings will support this. 3.2.1
Topic-1-elated variation
There is definitely no general correlation between any topic and a particular style, as the four examples in Section 3.1 above make clear: they show both a personal topic (boyfriend) and a nonpersonal one (lecturer) being discussed in both English- and creole-influenced styles. However, in several texts one can observe a cluster of creole features in a segment where a topic associated with a high degree of emotional involvement is being discussed. Below, two examples of this will be presented. The first example is from text S1A 037, an interaction between a fieldworker (speaker Bin (5)) and one informant, W M. (speaker A in (5)). W M. first talks at some length about his undergraduate studies in Computer Sciences as well as a technical support job he is doing at UWI in what can be considered his baseline style, which is somewhere in between the styles of examples (2) and (3) above. The conversation then turns to the topic of the military; which is of very high personal interest to him (cf. ll. 27-31 in (5)). This change in the topic coincides with a notable increase in creole features, as is evident in (5). In particular, in addition to such forms as going- V futures without a copula and verbs unmarked for past tense or third person singular present, ll. 40ff. show considerable use of the pronoun forms him and them in subject function. There is also an instance of a very rare creolism, them as a (redundant) pluralizer in the wicked people them (1. 39).16 There is onlyoneotherinstanceofthls form in the present dataset (teachers them,SlA 013). A further two instances of N them represent an associative plural (Mitko them 'Mitko and his assoctates: two occurrences in SlA 008).
16.
150
Dagmar Deuber
(5)
5
10
15
20
25
30
35
<$B> <#> Oh you were saying something about uhm the difference between what you're doing <$A> <#> Okay and uh <,> oh about undergrad stuff<#> I mean the courses in Com Sci and all of them<}><-> are<.> compu <1-> <=>are just like two compulsory courses in second year and you have to do four third-year courses and you can really choose around <#> If you like programming you can just do pure <}> <-><.>pro <1.> <1-> <=>programming courses and still <}> <-> <.>wou <1.> <1-> <=> would uh be like probably marketable <#> Well I mean unless of course you want to do programming <,> <#> People kind of think Com Sciences were a hands-on type of thing know how to use programs know how to fix computers <#> Com Sci don't teach that <$B> <#> Okay <#> So uhm you're going to continue in this job now <$A> <#> Uhm yes and no <#> Uh actually I have applied to Digicell to <.> like in the I T department to do some similar to what fm doing here but means to an end <#> I don't really want do that for like too long cos ultimately <,> what I want do is like commercially <#> So I mean I don't know how long I going be doing that for <#> And of course I have applied to the J F as well so should I be called in I have to leave in to go to that <$B> <#> Tell me more about that wanting to become a pilot ultimately <SA> <#>Yeah <#> And the military aviation is like the real gateway to do it <#> That's what I really prefer to do <$B> <#>So why that interest <SA> <#> Why <#> I think I've word very very very probably unusual interest in the military<#> I think I really really love it like every single thing about it and coupled with flying <#> So I mean military aviation seem to be my thing<{><[><,><#> Alright that's what I love anyway <$B> <#> <[> Mhm <SB> <#>Okay<#> So you haven't any family members in the military or anything like that <$A> <#> Yeah I have a brother that is in the Gulf War now <&> interruption of recording <#> Well he is in Kuwait<#> He was actually down there before he came back up because of a minor operation he did in early April <#> <} > <-> Stayed in <= > stayed in Jamaica for about a month and the wicked people them
The creole continuum and lndivldual agency
40 send him back down already <#>And him don't even finish heal yet <#> So he's in Kuwait now and he might soon be going back over into Iraq <#> He's not really a frontline person so I'm not like a hundred per cent worried yet because<}><-> him doing like<=> he's like a I don't know one of them warehouse type persons 45 doing some quartermaster duties <#> So I think him say him control a warehouse type of thing <#> You know take picture with them humvee and thing yes Uncle Sam vehicle <,> <#>Yeah <} > <-> him<=> him do him tour already I don't know why him don't leave the stupid army now but then you see people ask me 50 if I going go fight war too <#> Jamaica army not going fight no war <#> The only war we going have is Tivoli and them place and even so I words be going down there because I'll be going in the air wings <,> so I'd still be okay free from all of them stress <$B> <#> Okay<#> So you'll be training with J D F 55 <$A> <#>Hopefully hopefully <$B> <#>And they do all<}><-> the <=>the air what do you call it <$A> <#> word surveillance air sea rescue then the whole heap of stuff whole heap of stuff <#> I think the coolest 60 thing they do is how they land their chopper right over the hospital <0> imitates sounds <10> <#> Oh boy I don't even remember what them call that operation but <0> clicks tongue <10> <#> You know when them carry people from hospitals like in the country come up to sick people thing because of admissions<#> Yeah 65 man it's like one little small piece of grass them just land it <#> It look even better in the night <#>You really wonder how them find it <#> So it's them thing I want to do <#> Go England fi train and come back and<,> living on the J D F for a while (ICE-JA SlA 037) This example is also of interest in connection with the use of persons vs. people discussed above (relevant NPs are underlined): there are four clear instances of people in the passage, one in the part that illustrates W. M~s baseline style (1. 11) and three in the segment about the military (ll. 39, 49, 63). The only instance of persons (1. 44) occurs in the latter segment and is part of a noun phrase which also contains the JamC demonstrative them. 17 As an "anti-informal" usage would be incongruous in this context, the example supports the supposition made above This demonstrative occurs four more times (1145, 49, 51, 64); in all four cases, the following noun is unmarked for plurality and one of these NPs also contains the expression a11d thing mentioned in 2.2 above, which Allsopp (2003: s.v. and 4. 9, s. v. thing 3.1) classifies as "anti- formal" 17.
151
152
Dagmar Deuber
that in Jamaican speech, persons is to some extent simply a synonym of people without special connotations. Example (6) below is from the speech ofK. H., who was already quoted in (4) above. This speaker appears in three texts, S1A 021-023, which are all from the same recording. As example (4) has shown, her baseline style already has anumber of creole features (in fact, she has one of the most creole baseline styles in the sample). However, there are two segments in the three texts, one in S1A 022 and one in S1A 023, in which such features are particularly dense and in which she also employs JamC features such as progressive a and complementizer fi which do not occur at all elsewhere in the three texts. In both of these segments she vividly recounts episodes from a linguistics field trip to Guyana in which she took part. Example (6) shows the relevant segment in S1A 022. In this segment she relates how she got extremely annoyed at one of her male fellow students, who refused to help when the bus broke down and had to be pushed. Here it is particularly evident how she is getting carried away by the story, as at one point she even remembers that the conversation is supposed to be in English and apologizes to the fieldworker (11. 9-10), but then carries on in essentially the same style. (6)
<$A> <#>Me and I think Indrah were sitting at the front and two other people <#> So we come out and we had to push <} > <-> the <1- > <=>the van for it to get started <#>People think me dainty and whatever you know but me reach ina the back and me a push 5 and ago on <#>So I ask<}> <->the man <1-> <=>the boys to come out and help us <#> I don't do bus <0> laughter <10> <#> <}> <->Me say me did <=> me bout fi get dark you know <#> several words <0> laughter <#> Sorry English <0> laughter <10> <#> 10 Ooo sorry<#> No so<}><-> me I I went <1-> <=>I went to Sir and say Sir can you believe this <#>This old jeyeseyaise boy <0> laughter <10> <#> Sorry <#> He started to well him don't want do bus <#> Also we it's four girls you know four girls get behind the bus and other people in the 15 bus as well but<}><-> them<=> it never make no sense them come out and push <#> Four of us four girls at the back of the bus pushing it<#> Can you imagine that<#> And we ask him to come and help us <#> Him don't do bus <#> I was most upset
In addition to several JamC grammatical features including the above-mentioned a (1. 4) andfi (1. 8), note the use of 1-each followed by the JamC preposition ina in 1. 4 and especially the word jeyesey-aise, which, according to Joseph Farquharson (pers. comm.), is an extremely striking creolism in the speech of a young person
The creole continuum and lndivldual agency 153
(it appears in the supplement of Cassidy/Le Page [ 1980] and is explained by means of a quotation which says "inferior, like a donkey with sore ears dressed with jeyes"18). These items mark the passage as rather "more creole" than the speakers' baseline style. From an overall perspective, the kind of variation described in this section seems to resemble what has been described as style-shifting by previous authors. It involves large segments of texts and, if these are considered as a whole, the changes appear rather subtle, consisting in an increase or decrease in the frequency and types of creole features that occur rather than an abrupt switch from one linguistic system to another when there is a change in topic. 3.2.2
Metaphorical and identity-1-elated variation
In metaphorical variation as described by Gumperz (1982), 19 "code, dialect and style switching processes" function as contextualization cues, i.e. "means by which speakers signal and listeners interpret what the activity is, how semantic content is to be understood and how each sentence relates to what precedes or follows" (p. 131; italics original). Such changes in the language used by a speaker may have a mere highlighting function, e.g. when a quotation is set off from the rest of the discourse by the use of a different language or language variety, which may or may not be the one that the speaker quoted had actually used. However, as Gumperz emphasizes (pp. 93-94), the direction of the change may be significant as well. especially in the case of the oppositions he summarizes under the heading "personalization versus objectivization" (p. 80), which can be seen as reflecting the opposition between "we-code" and "they-code:' Identity-related variation overlaps with such cases of metaphorical variation in that itis based on linguistic contrasts functioning as contextualization cues and in that it draws on the social values associated with different languages or language varieties. However, the symbolic potential of linguistic features may also be exploited in the creation and presentation of speaker identity in ways that are not only more direct but also freer and more creative than the allusions to the "we"/"they" -code opposition de scribed by Gumperz.20 As SchillingEstes (2002: 390) remarks, "people utilize stylistic resources, not only to indicate relatively longstanding group affiliations and personal attributes, but also to make temporary meanings in ongoing interaction." Examples below proceed from the 18. /eyes is an abbreviation of/eyes'fluid (see Cassidy/LePage 1980: s.v.jeye.s). 19. Gumperz (1982) in fact uses the term "conversational code-switching" but this seems to correspond to mom and Gumperz's (1972) "metaphorical code-switching~ 20. Cf. also the continuum of meanings of code-switching as a contextualization cue from "contrastive" to "inherent" ones presented by Hinrichs (2006), e.g. p. 135, who elaborates a proposal by Auer (1992).
154 Dagmar Deuber
metaphorical type of variation where linguistic contrasts serve mainly a highlighting function to the most strongly identity-related types of variation. Examples (7) and (8) illustrate variation in texts from ICE-JA S1A 001-040 which can be considered metaphorical in the sense of Gumperz (1982). (7) I tried out for a play once right<#> And I was supposed to play this roughneck Jamaican boy <,> right <#> And I read the lines with all the passion and what not and the director said thank you <{3> <[3> <,> yeah <#> So I went to him afterwards and I said uhm what wrong <#> He says boy you couldn't convince me and you can't convince nobody (ICE-JA S1A 002) (8) <$C> <#>If my husband dies and I remarry I'll think about having another one [child] <$D> <#> Mhm <0> laughter <$A> <#>My husband naa dead left me <#> We a dead together <#> <{> <[> <}><->You know what I've been<.> do <1.> <1-> <#><=>Can I tell you what I've been doing <#> Can I tell you how serious I take this<#><}><-> Me andhimnaa <.>de <1-> <=> himnaadead left me<#> And me naa dead left him words <0> laughter <10> <$C> <#> <[> <}><->There <1-> <=>there is a greater<.> possi <#>Aha <$C> <#>There is statistically<{> <[>my husband is very likely to diebefore me <$A><#><[> Me serious<#> You know what I'm doing now <$A><#><}><-> Him naa <=>My husbandnaa dead left me<#> <{1> <[1> We a dead together<#> You know what I'm doing now <#>I'm now going through all my will all my documents wherever they are and I'm putting a third party on<#> <{2> <[2> Two more persons on the part of my husband cos him naa go benefit from it cos we a dead together <0> laughter <#> No I've done it I've gone through <#>I've been doing it for the last about a month <{3> <[3> or so <$C> <#> <[1> Statistically <0> chuckles <10> <,><#>Why <$C> <#> <[2> Mhm <$C> <#> <[3> No statistically my husband is more likely to die before me <#>He's twenty-something years older than me (ICE-JA S1A 013) In example (7), creole features highlight quotations. In (8) the speaker keeps repeating a strongly creole phrase involving negated progressive a (with future reference) as well as dead 'die' and left 'leave' in a serial construction (i.e. 'he won't die
The creole continuum and lndivldual agency 155
leaving me' fi won't die leaving him') to emphasize her prediction that she and her husband will die at the same time, for which of course she has no objective basis and to which the other participants in the conversation react with laughter. Her subjective predictions contrast sharply with her matter-of-fact description of the arrangements she has actually been making for after death (probably under the impression of her brother's recent fatal car accident). A special type of variation which can be related to the "we"/"they -code opposition occurs in connection with metalinguistic comments, where variation in language use seems to highlight the opposition talked about. Interestingly, the "we"code in is Patois (9) - an informant's description of topic-related variation in educated Jamaican language use -but in (10) it is JarnE, which, as Christie (2003: 63) writes, "is a sign of Jamaican identity, but, unlike Creole [... ] is not a symbol of if' (9) I prefer to call it as a Jamaican relaxed language rather than refer to it as a Patois because we all speak it <,> uhm <,>when we talking to each other <,>while we are speaking proper English <#> <}><->It it <1-> <=> it just depends on what we are talking about as to whether we descend <,> <#> And I'm not two or three words <}> <->whether <1-> <=> whether <mention> descend is a correct word or whether we move across <.> into <,>Patois (ICE-JA S1A 002) (10) <#>Yeah so <0> laughter <10> <#>So uhm <,> <}><->the <1-> <=> <,>the English we speaking is of course<.><}><-> it's it's <=>it's just<}><-> it's not like<=> it's not like you know there is all Oxford English orB B C English (ICE-JA S1A 002) The next example arguably combines the use of creole features in the context of a quotation (or imitative speech) with an identity-related aspect. The speaker in question (Bin (11)) has a somewhat creolized baseline style, with occasional occurrences of verbs unmarked for past tense or third person singular present as well as zero copula and them as third person plural subject or possessive pronoun. In (11), the baseline style can be observed from 1. 12 onwards, where the phrase "but apart from that" signals a change in the topic. In the preceding lines, there is a duster of creole forms including highly marked ones such as a introducing a cleft sentence2 1 and naa as the negated equivalent of progressive a, and these forms are only partly accounted for by the quoted or imitative segment in 11. 9-10. It seems 21. A your friend that is an instance of the special type ofJamC cleft structure which occurs in sentences equating two NPs (c£ Bailey 1966: 87). It has not been counted in the analysis in Section 2.1 due to its being quoted or imitative.
156
Dagmar Deuber
rather that the speaker is using a more creole style in this context in order to emphasize her disidentification with her gossipy former fellow residents whom she connects with JamC speech. (11)
<$A> <#> So what was your experience like living on hall <} > <-> on
<1-> <=>on Irvine hall <$B> <#> It was good and bad because I wasn't used to having so many females around me who just wanted to know so much about 5 your business and it was annoying to tell you the truth because them just want know everything <#> If them see a guy come in your room them want know if he your boyfriend I mean <#>If them see a person visited you often <#> A your friend that <#>Where you know that girl there from <#> I mean everybody had a story 10 to tell you about somebody and you have to wonder if them naa tell somebody a story about you too when you tell them your business so <#>But apart from that <0> speaker A clears throat <10> the block I lived on was really cool because we would always go out together party together<}><-> we <1-> you know<=> we really had a 15 lot of parties together a lot ofbonding we talk a lot and<,> I think that even though everybody was difterent <}><->it was <1-> <=>there was still something special about the way that we lived over there (ICE-JA S1A 016) Whereas in (11) JamC features were apparently used by the speaker to distance herself from a particular social group - though she is a JamC speaker herself- the opposite seems to be the case in (12). In this example from a text which overall does not have many creole features, two overt creole forms are used in what may be described as a symbolic act of identity with creole speakers not proficient in English: (12) Cos I mean a thing like the Constitution<,><#> Even me no understand someone who knows English <#> Cos I mean there's a problem in terms of rights <#> A Creole speaker cannot understand so what about him (ICE-JA SlA 040) Example (13), another instance of variation that can be interpreted as identity-related, is from S1A 002, a text that has been quoted from repeatedly in this section. Overall, this text does not have very many creole features, but occasionally such features are used for special effects. In (13), the speaker jocularly takes on the persona of a country person: (13) <$C> <#>I must say that when I came to Kingston I was most shocked <,>to realise that<}><-> people did<.> peo <.!> <1-> <=>people
The creole continuum and lndivldual agency 157
would not go into a shop and say could I have some milk please <{><[>they'd say could I have some white milk <,>as opposed to some chocolate milk<,> cherry milk <$A><#><[> Please <$B> <#>We have always known that Kingston people stupid (ICE-JA S1A 002) A speaker who has a particular tendency to employ creole for special effects is the one already quoted above in (8). In another text from the same recording as the one quoted in (8), she uses creole phrases in several passages in which she casts herself as a person who does not take her studies at university overly seriously but who does not miss a single social occasion, and who rejects the attitude of those with a more ambitious orientation. This is shown in examples (14) to (16). (14) But the truth is that is nothing<#> Because no care how much Master's you have if you dead it naa no usell <#>So what's the point (ICE-JA S1A 012) (15) I remember this girl in my Master's programme who topped the class means straight Xs <#>Now frankly me just no want get<}><-> no<.> s <1.> no <1-> <=>no <1=><1}> pass but ... A<}><-> is never<=> was never my thing (ICE-JA S1A 012) (16) That's why I don't understand students <#> The little time you de a UWI you no come a every little thing them have (ICE-JA S1A 012)
In the last example, note in addition to the locative copula de - a very rare feature in the present data- the use of a as a preposition. According to Allsopp (2003: s.v. a7), "this is a strong basilectal particle in Anglophone Cr[eole]s" and "outside of wholesale Cr[ eole] contexts this prepositional usage usu[ ally] signals code-switching among educated speakers with a desired effect in mind." Generalizing, one can say that the types of variation which can be described as metaphorical or identity-related rely especially but not necessarily on the symbolic potential of salient overt creole features which may be considered part of code-switches. It is clear from the examples given in this section that the type of interactions represented in the conversational component of ICE-JA permit a freer use of creole features for such special effects than the more formal types of texts. However, it is also true that in the sample studied here, identity-related variation in particular is mainly associated with certain texts and speakers. There is no parallel in this sample to the "wealth and diversity of examples of identity-related code-switching which Hinrichs (2006: 134) found in COJEC.
22.
No care is a subordinating conjunction ('no matter'); naa is here a contraction of no + have
'doesn't havf!.
158
Dagmar Deuber Finally. it has to be emphasized that the difterent types of variation which have been discussed in separate sections here should not be considered as mutually exclusive. Both of the examples in the section on topic-related variation (3.2.1) in fact arguably have an identity-related element in addition to the observed shifts in topic and language use: W. M. in (5) is distancing himself from the "wicked" Americans who sent his brother back to war shortly after undergoing an operation in ll. 38-39, and in 1. 4 of example (6), K. H. is emphasizing that she is not the kind of dainty person that some people think she is. Also, at least some of the examples of metaphorical or identity-related variation quoted in the present section are certainly also connected with the emotions generated by a topic, such as annoyance in example (11). This once again shows the difficulty of separating subtle shifts triggered by a speaker's emotional involvement in a topic from presumably more conscious code-switches.
4· Conclusion
This paper has presented quantitative data on the occurrence of several syntactic and morphological creolisms, as well as selected lexical features, in a sample of private interactions from the Jamaican component ofiCE and compared the findings both to more formal types of spoken English represented in that corpus and to JamC data where available. It has also provided qualitative descriptions of variation within the sample and explored the ways speakers use JamC features in these conversations for a range of di:ffurent discourse strategies. Taken together, the different aspects dealt with have shown dearly that there are very finely graded transitions in educated Jamaican speech from more English to more creole ways of language use. Overall, the language of the conversations can be located in between the high acrolect and the upper mesolect, and may be described as informal JamE. However, this label should be understood as designating a range of language use rather than a clearly demarcated variety. The qualitative descriptions have given an indication of the subtlety of stylistic variation in this range as well as the complexity of its determinants. Although there are cases of what are undoubtedly codeswitches into JamC, the line between these and more subtle style-shifts can be hard to draw. All in all. then, the notion of a continuum from more English to more creole varieties seems to better capture the range of stylistic variation in educated Jamaican speech than a hi -varietal model such as diglossia. If the new informal types of written communication which have arisen in CMC show less evidence of the gradual transitions from English to creole which are characteristic of spoken usage, and more evidence of drastic contrasts between codes, this is most likely due to the nature of these new electronic genres, in which
The creole continuum and lndivldual agency 159
language use may be informal but is more planned and less spontaneous than in speech. As Hinrichs (2006: 21) emphasizes, "features of speech enter CMC through conscious stylization and imitation." It is probably for this reason that the deliberate use of features strongly associated with JamC within an overall English baseline code for complex discourse strategies such as the creation of social personae is apparently much more common in CMC than in speech. In broad terms, the present data and analyses certainly support previous findings on the nature of stylistic variation in a creole continuum in that there is a definite correlation between the casual - formal and the creole - English dimensions, and in that educated speakers have been shown to display a range of stylistic competence including the acrolect as well as a form of JamC as "a national vernacular shared across social boundaries, a resource and locus ofidentity for Jamaicans" (Patrick 1999: 272). However, it has also become clear that the casual- formal dimension is a complex one and that variation along this dimension is more fine-grained than what correlations between language use and pre-defined situational categories can reveal. Furthermore, the functions and meanings of JamC features in informal JamE discourse are manifold and cannot be fully understood by reference to JamC as a Jamaican "we-code" alone. This study has revealed some of the complexity of stylistic variation in educated Jamaican speech by combining a broad quantitative approach in the framework of a continuum of lects and qualitative investigations of the contexts of variation. In this way, it has also thrown light on the difterent strengths of both approaches. The quantitative approach is useful to locate a set of data within the larger spectrum of linguistic practices that occur in the Jamaican continuum, and it can help to establish a cline of features from "less creole" to "more creole" ones. A qualitative approach, on the other hand, makes it possible to differentiate between different contexts within a set of data and to determine the stylistic connotations of different features more precisely. It also helps to explain why speakers make certain choices, and, seen in this light, features which appear negligible in quantitative terms may assume great significance in the way speakers make meaning in interaction. Thus, although the present paper supports the notion of a continuum in Jamaican speech, a simple ranking of varieties or features along this continuum does not fully do justice to the data at hand. Of course, the traditional. quantitative approach to style and investigations in the speaker agency framework are not necessarily mutually exclusive. As Schilling-Estes (2002: 391) notes, "given the importance of the discourse contexts in which features occur in shaping their meanings, researchers are increasingly complementing investigations of aggregate levels for features[ ... ] with investigations of where in discourse stylistic resources are used" (italics original). The present paper has shown that in the Jamaican context, both a traditional approach in the
160
Dagmar Deuber
continuum framework which ranks varieties on the basis of quantitative findings and an investigation of how individuals use the linguistic resources at their disposal in interaction can be useful to describe and explain language variability. Thus, the creole continuum and individual agency complement each other as approaches to stylistic variation in Jamaican speech.
References Allsopp, R. 2003. Dictionary of Caribbean E11glish Usage. Kingston, Jamaica: University of the West Indies Press. (reprint; first published by OUP In 1996). Auer, P. 1992. Introduction: John Gumperz' approach to contextual!zation. In The CoJ~textual ization of Language, P. Auer & A. di Luzio (eds.), 1-37. Amsterdam: John Benjamins. Bell, A. 2001. Back In style: Reworking audience design. In Style and Socioli11gulstlc Variation, P. Eckert & J.R. Rickford (eds.), 139-169. Cambridge: CUP. IDom, J.-P. & Gumperz, J. J. 1972. Social meaning In linguistic structure: Code-switching In Norway. InDirections i11 Sociolinguistics: The Ethnography ofCommuJ~ication, J.J. Gumperz & D. Hymes (eds.), 407-434. New York NY: Holt, Rinehart, and Winston. Bailey, B.L. 1966. Jamaican Creole Syntax: A Transformational Approach. Cambridge: CUP. Bickerton, D. 1975. Dynamics of a Creole System. Cambridge: CUP. Cassidy, F. G. & LePage, R.B. (eds.). 1980. Dictionary ofJamaican English, 2nd edn. Cambridge: CUP. Christie, P. 2003. Language In Jamaica. Kingston: Arawak. DeCamp, D. 1971. Toward a generative analysis of a post-creole speech cont!nuwn. In PidginlzaNon and Creolizatlon of Languages, D. Hymes (ed), 349-370. Cambridge: CUP. Deuber, D. & Hinrichs, L 2007. Dynamics of orthographic standardization In Nigerian Pidgin and Jamaican Creole. World Engllshes 26: 22-47. Greenbaum, S. 1996. Introducing ICE. In Comparing English Worldwide: The International Corpus ofEngilsh, S. Greenbaum (ed), 3-12. Oxford: Clarendon. Gumperz, J. J. 1982. Discourse Strategies. Cambridge: CUP. Hinrichs, L. 2006. Codeswitchingon the Web: English and Jamaican Creole In E-Mail Communication. Amsterdam: John Benjamins. Irvine, A. 2004. A good command ofthe English language: Phonological variation In the Jamaican acrolect Journal ofPidgin and Creole Languages 19: 41-76. Mair, C. 2002. Creolisms In an emerging standard: Written English in Jamaica. English WorldWlde23: 31-58. Mair, C. 2009. Corpus linguistics meets sociolinguistics: Studying educated spoken wage In Jamaica on the basis of the International Corpus of English. In World Engiishes - Problems, Properties and Prospects: Selected Papers from the 13th IAWE Conference, T. Hoffmann & L. Siebers (eds.), 39-60. Amsterdam: John Benjamins. Nelson, G. 1996. The design of the corpus. In Comparing English Worldwide: The International Corpus ofEnglish, S. Greenbaum (ed), 27-35. Oxford: Clarendon. Nelson, G. 2002. Markup manual for spoken texts. . Patrick, P.L 1999. Urban Jamaica11 Creole: Var iatlon i11 the Mesolect. Amsterdam: John Benjamins.
The creole continuum and l..ndivldual agency Patrick. P. L 2004. Jamaican Creole: Morphology and syntax. In A Handbook of Varieties of English, Vol I I, B. Kortmann & E. W. Schneider (eds.), 407-438. Berlin: Mouton de Gruyter. Rickford, J. R 1986. Past- marking ln the Guyanese mesolect: A dose look at Bonnette. In Variation in Language: NWAV-XV at Stanford, K.M. Denning, S. Inkelas, P.C. McNair-Knox & J.R Rickford (eds.), 379-394. Stanford CA: Stanford University Department of Linguistics. Rickford, J.R 1999. Variation in the Jamaican Creole copula and its relation to the genesis of AAVE: New data and analysis. In Creole Genesis, Attitudes, and Discourse: Studies Cele11ratlng Charlene J. Sato, J.R Rickford & S. Romaine (eds.), 143-156. Amsterdam: John Benjamins. Sand, A. 1999. Linguistic Variation in Jamaica: A Corpus-Based Study of Radio and Newspaper Usage. 1Ubingen: Narr. Sand, A. 2005. Angloversals! Shared Morpho-syntactic Features ln Contact Varieties of English. Habilitation dissertation, University ofFreiburg. Schilling-Estes, N. 2002. Investlgatlng stylistic variation. In 1he Handbook of Language Variation and Change, J.K. Chambers, P. Trudgill & N. Schilling-Estes (eds.), 375-401. Oxford: Blackwell Schnelder, E. W. & Wagner, C. 2006. The variability of literary dialect ln Jamaican Creole: Thelwell's The Harder They Come. Journal ofPidgin a11d Creole La11guages 21: 45-95. Winford, D. 1991. The Caribbean. In English Around the World: Socioli11gulstlc Perspectives, J. Cheshire (ed), 565-584. Cambridge: CUP.
161
Language attitudes and linguistic awareness in Jamaican English Andrea Sand This paper presents results from a pilot study testing the acceptability of a number ofle.xical and morphosyntactic features which have been identified in previous corpus-based analyses as more frequent in educated Jamaican English usage than in other standard varieties of English. The results from the questionnaire survey show that there are considerable di:ffe rences between individual items on the questionnaire, with regard to their acceptability in writing as well as their regional background Such differences merit further investigation and must be taken into account in the codification of a standard Jamaican English.
1.
Introduction
Since the 1970s, the prevailing view of the linguistic situation in Jamaica has been the notion of the creole continuum as developed by DeCamp (1971) or Bickerton (1973, 1975). According to this theory; there is a basilectal or broad variety ofJamaican Creole at one end of the scale and an acrolectal variety of standard English at the other end. 1 For decades, the acrolectal variety was considered to be in line with the norms of British or international standard English at least with regard to morpho syntax and lexicon. Only much more recently do we find studies pointing out difterences between standard British English and educated Jamaican usage, e.g. by Christie ( 1989), Shields-Brodber (1989), Mair (1992), and Sand (1999) .2 These studies whether based on student essays, anecdotal evidence, or corpus data all suggest that educated Jamaican English is more than standard British or standard international English spoken with a Jamaican accent. However, if we conclude from the data analyzed so far that an endonormative Jamaican Standard 1.
For a more detailed discussion of the continuum model, see Sebba (1997: 210-224) or Sand
(1999: 50-61). 2. And even more recently, in-depth studies of the phonology of educated or acrolectal Jamaican English, e.g. Irvine (2004).
164 Andrea Sand
English is in the making, we need to know more about speakers' and writers' attitudes towards the features under analysis. In other words, are we dealing with new norms or with - conscious or unconscious - deviations from previously existing norms? And if they turn out to be conscious deviations from the norm, what is their discourse or social function? Are the speakers aware that the features under analysis are significantly more frequent in Jamaican English usage than, for example, in British or American English? A first step towards answering these questions is to supplement the corpus findings with native speaker acceptability judgments. For this purpose, a pilot study was undertaken in the form of a questionnaire survey.
2.
Methodology
In order to test the acceptability of forms frequently encountered in corpus data consisting of written and spoken formal Jamaican English, I devised a questionnaire which was distributed to 100 students and staff at the University of the West Indies, Mona. Seventy-two questionnaires were returned, of which six were incomplete. The majority of the respondents (59) 3 were born and raised in Jamaica and have not stayed abroad (54). Of those who have stayed abroad, only one stayed longer than six months in an English-spealdng country. Except for one native speaker of Spanish, whose questionnaire was not considered in the analysis, all respondents claimed English and/or (Jamaican) Creole as their first language. A differentiation of respondents in terms of gender was ruled out by the small number of male respondents: 57 are female, seven male, and eight did not indicate gender. With regard to age, 53 fall into the 18-25 age group, 12 in the 25-45 age group, and seven did not provide any information with regard to their age. The average respondent of the survey therefore is a young Jamaican woman who has not lived abroad and is presently receiving undergraduate university education. They thus represent the kind of"educated speaker" (ct~ Greenbaum 1996: 6) sampled for the Jamaican subcorpus of the Interna.tional Corpus ofEnglish (ICE) from which the examples in the questionnaire were taken. A larger sample, which is evenly balanced between male and female speakers, would be a desideratum for further research, as many previous sociolinguistic studies (e.g. classic studies like Labov 1966 or Trudgil11974) indicate a higher degree of adherence to standard norms for women. 3· Those who were not born in Jan~aica came from Trinidad (4), Dominica (1 ), St Maarten (1) and Venezuela (1). Six respondents did not complete the country of origin-question in the personal information section.
Language attitudes and linguistic awareness in Jamaican Engllsh 165 The participants were asked to judge the acceptability of two similar sentences and asked to assign the difterent forms to different varieties of English. if possible. The questions were designed to elicit information on the following phenomena: 1. Formal vs. informal lexical iterns: persons vs. people, assistance vs. help, freeness
vs.freebie 2.
Loan words from Jamaican Creole vs. international standard English lexemes: backative vs. support, mash up vs. destroy 3. American English vs. British English lexemes: truck vs. lorry, toward vs. towards 4. Pluralisation of mass nouns and concord with collective nouns: equipment(s), ·waste(s), police with singular and plural verb 5. Backshift in reported speech and past perfect in complex sentences The questionnaire was designed on the basis of data from newspaper texts and radio broadcasts included in ICE-Jamaica which had been found to show significant differences compared to British and American data (cf. Sand 1999). The original example was then paired with a British or American English version of the same sentence to provide uniform contexts for the target lexical item or construction. In the instructions, it was pointed out that the examples in the questionnaire were taken from written texts to indicate a maximum degree of formality. The respondents were asked whether they considered the examples appropriate for written English and which alternative they preferred in their own writing. In addition to that, they were asked whether they associated any of the examples with a particular region of the English-speaking world, such as Great Britain, the United States of America, or Jamaica. This was done in order to check whether the respondents recognized the features identified as frequent in Jamaican English writing and formal speech as typically Jamaican.
3· 3.1
Survey results Lexicon
In the first category, which contrasts formal with informallexemes, three forms were tested: persons (set off against the neutral term people),freeness (paired with the informal American freebie) and assL~ance (as opposed to neutral help). According toresearch on register (e.g. Biber 1995), lexical items in English which derive from Latin or French etyma are generally considered to be more formal than those of Germanic origin. In many publications on the New Englishes (e.g. Sand 1999: 101-110 and Roberts 1988: 29 on Jamaican English or Gorlach 1991 on the New Englishes in general), a tendency to prefer such Latinate lexical items especially in formal or written
166 Andrea Sand
usage is described as typical of these varieties. It is therefore of interest in the present context whether the respondents express a preference for the more formal items in writing, as they are not the preferred choice in present-day written British English. The case offreeness is slightly different from assistance and persons, because it is archaic in international Standard English and it is a hybrid form of a Germanic stem with a derivational suffix borrowed from French. Its Jamaican usage is based on a semantic innovation in Jamaican Creole, but since it was paired with the informal American term fi-eebie, it is interesting to compare it with the other two lexemes in this group. The following sample sentences were used: (1) a. b.
There are some people who believe that football is a boring game. There are some pe1·sons who believe tha.t football is a boring game.
(2) a. There are those who argue that the concept ofa free university education would encourage a freebie mentality among students. b. There are those who argue that the concept ofa free university education would encourage a freeness mentality among students. ( 3) a. b.
The organizers received generous help from volunteers. The organizers 1-eceived gene1"0us assistance from volunteers.
In all three questions, the respondents preferred the more formal forms (persons, assistance, freeness), but at the same time, the majority conceded that neither form should be considered bad English. The cases of assistance/help and persons/people are very similar indeed, as illustrated in Table 1 be low. The majority of respondents preferred persons and assistance and a sizable numher also consider them to be better English. However, over 80% also do not consider any of the four lexemes to be 'bad English' or 'wrong: thus people and help are also definitely not stigmatized but rather neutral. The comments on help/assistance suggest that speakers consider them synonymous (e.g. "Both words convey the same meaning," or "Same meaning, but expressed differently by different countries") but Table 1. Acceptability of assistance vs. help and persons vs. people (in % of 70 and 68 responses)
'bad English'
'prefer'
'better English'
assistance
54.4
38.8
5.9
help persons people
25.0
10.3
10.3
65.7
47.1
1.4
21.5
7.1
14.3
Language attitudes and linguistic awareness in Jamaican Engllsh 167
have a clear sense that assistance is more formal and thus preferable in writing (e.g. "(b) [i.e. assistance] more formal'). In the case of persons and people, some respondents appear to make a semantic distinction, as illustrated by the following comments: a.
"People could refer to a nation while persons are more likely to find football
b.
"People refers to a group. Persons -you can actually give a number to represent
boring=' those who believe the game is boring." There is also a striking similarity in the regional categorization of these four lexemes, as illustrated in Table 2. In the case of help, 29.4% considered it to be typically Jamaican, while only 2.9% thought the same of assistance. Even more pronounced is this difference for people (42.9%) and persons (8.6%). Therefore, it is not very surprising that the formal items persons and assistance were classified as 'British' by over 30% of the respondents. This is in stark contrast to the corpus data we have analyzed, which showed significantly higher frequencies of assistance and persons in the Jamaican corpus in comparison to its British, American, New Zealand and Kenyan equivalents (cf. Sand 1999: 104f). The Jamaican semantic innovation freeness (cf. Dictionary of!amaican English (DJE, Cassidy and Le Page 1967), Dictionary of Caribbean English Usage (DCEU, Allsopp (ed.), 1996), discussed in Mair 2003), with its concrete sense of'something given free, without cost,' was paired with its informal American near-synonym freebie to test whether the respondents would be able to identify the American term and whether they regarded freeness as equally informal because of its Creole origin. The results for this question reveal a definite preference for freeness over freebie. The preferred form also receives significantly higher ratings in the "better English" category (30.6% over 14.5%), but 70.3% of the respondents concede that neither form should be considered 'bad English or wrong' (see Table 3 on next page). However, there were also a number of respondents who consider both forms 'bad English or wrong: While only two respondents commented on the fact that they consider freeness to be informal or think it "should be put in quotations; all other comments make the point that freebie is regarded as "informal," "slang:' Table 2. Origin oflexemes assistance vs. help andpersonsvs.people (in% of70 and 68 responses)
asslstance help persons people
Jamaican English
British English
American English
other
2.9
33.8
16.2
1.5
29.4
8.8
13.2
3.0
8.6
35.7
21.4
1.4
42.9
11.4
11.4
2.9
168 Andrea Sand
Table 3. Acceptability offreebie vs.freetless (in% of 62 responses) 'prefer'
'better English'
'bad English'
freebie
25.8
14.5
27.4
freeness
45.2
30.6
14.5
neither
29.0
54.8
61.3
Table4. Origin oflexemesfreebie andfreetless (in% of 62 responses) Jamaican English
British English
American English
other
freebie
9.7
14.5
22.6
0.0
freeness
40.3
16.1
4.8
0.0
or "jargon=' Freeness is recognised as Jamaican by the majority (40.3%) of those respondents willing to categorise the word with regard to its regional distribution, while fewer respondents (22.6%) were aware of the American origin of freebie (cf. Table 4 above). Only one respondent comments on the Creole origin offi-eeness, while another considers freebie "a Jamaican slang word=' Freeness is the only item in the questionnaire which was recognized as a typically Jamaican form but still preferred over the alternative choice in writing. And although there are some comments on the informality of .freeness, it is significantly less stigmatized than other lexemes of Creole origin with only 14.5% of the answers in the 'bad English or wrong' category (as opposed to 27.4% for .freebie). The relatively neutral quality of freeness is supported by the results from the comparison of the second group oflexemes, backative and suppm·t, as well as mash up and destroy. The first item in each of these pairs is a direct loan from Jamaican Creole, while the second one is a standard English synonym. In our corpus data, the JC lexemes mash up and backative did occur in writing (cf Sand 1999: 79f.; Mair 2003: 84f.), but always 'flagged' as intentional deviations from the standard in quotation marks or within direct speech. They were included in the questionnaire to test whether this reflects a higher degree of stigmatisation as opposed to .freeness which repeatedly appears without quotation marks in our data. The following sample sentences were used: (4) a. b.
The students often do not have the financial backative to secure a loan. The students often do not have the fina.ncia.l support to secure a loa.n.
(5) a. A lot of houses were mashed up in the rain. b. A lot of houses were destroved in the rain.
Language attitudes and linguistic awareness in Jamaican Engllsh 169 The hypothesis was supported by the answers in the language attitude survey. While
support and destroy received similarly positive judgments, backative and mash up are somewhat different with regard to the acceptability ratings. While only 1.5% of the respondents actually prefer mash up to destroy, 10% prefer backative over support. On the other hand, 78.5% consider mash up 'bad English or wrong: while only 44.3% view backative negatively (cf. Table 5 below). The written comments, however, suggest that both lexemes are regarded as unacceptable in writing: a. b. c. d.
"Jamaicans would tend to use backative in oral speech but write support." "backative is a colloquial term" "(b) [i.e. backative] influence oflocallanguage" "(a) [i.e. mash up] Creole or colloquial term is used,
It does not conclude that were destroyed is not said or written in formal Jamaican contexts but that Creole interference is a major issue for standard form. Some people also commented on the hybridity of the form of mash up used in the example, namely the past participle mashed up. Obviously, they did not regard the lexeme as part of the standard English lexicon and thus found the standard English inflection odd: a. b.
"(a) [i.e. mashed up] is a mixture of Creole and standard English." "mash up is more Jamaican than mashed up"
Just as unanimous as the vote against the use of these Creole lexemes in writing was their regional classification (cf. Table 6 on next page). They are recognized as Jamaican by 80% of the respondents, by far the highest score in the regional-labels category in the whole questionnaire. Interestingly, the standard English lexemes suppm·t and destroy were often classified as both British and American, indicating their neutral status with regard to regional distribution. But when we take a closer look at the figures, the percentages for the 'British' category are significantly higher than those for the 'American' category. This could be an indication that for many Jamaicans, 'good English' or 'proper English' in writing are still associated with the former exonormative standard which is believed to be Standard British English. Table 5. Acceptability of mash(ed) up vs. destroy(ed) and backative vs. support (in% of 65 responses) 'prefer' mash(ed)up destroy(ed) neither backative support neither
'better English'
'bad English' 78.5
1.5
3.1
96.9
87.7
1.5
1.5
9.2
20.0 44.3
10.0
7.1
88.6
75.0
2.9
1.4
15.7
52.8
170
Andrea Sand Table 6. Origin oflexemes mash(ed) up, destroy(ed), backative, and support (in% of 65 responses)
Jamaican English mash(ed) up destroy(ed) backative support
British English
American English
other
80.0
1.5
0.0
1.5
3.1
56.9
29.2
0.0
80.0
1.4
4.3
0.0
10.0
62.9
42.9
0.0
But since there is a considerable influence of American English in Jamaica through the media, tourism, and Jamaican migrants to the US, two stylistically neutral American lexemes were also tested in the survey as a third category; namely the relatively well-known transparent lexical feature truck, as opposed to British lm-ry, and the rather opaque lexical feature toward, which is predominantly used in American English, as opposed to British towards. The sample sentences were: (6) a. It is still not known what caused the blaze that also destroyed a truck
that was parked nearby. b. It is still not known what ca.used the bla.ze that also destroyed
~
that was parked nearby. (7) a. Effm·ts should be made towm·ds constitutionalniorm. b. Efforts should be made toward constitutional reform. The respondents of the survey favor truck over lorry, which is in line with my previous corpus analysis based on newspaper and radio usage (Sand 1999: 80f.). However, both forms are neutral with regard to the categories 'better English' or 'bad English: so lorry is not stigmatized in any way (cf. Table 7 below). With regard to regional classification, lorry received the highest percentage (73.6%) in the whole questionnaire in the category'British English; so we can safely assume that it is recognised as distinctly British. This is quite remarkable, as truck is the preferred form. I can only assume that this surprisingly high degree of Table 7. Acceptability of truck vs. lorry (in % of 72 responses)
truck lorry neither
'prefer'
'better English'
'bad English'
72.2
15.3
4.2
12.5
16.7
4.2
15.3
68.1
91.7
Language attitudes and linguistic awareness in Jamaican Engllsh
linguistic awareness of a form from a different variety of English stems from that tact that the lexeme pair lorry/truck is one of the classic textbook cases illustrating lexical differences between British and American English. This awareness for regional differences is also illustrated by the following comments: a. b. c.
"From this it is seen that differences in culture largely affect vocabulary and how words are used=' "(b) [i.e. lorry] is used in British context and will not be necessarily understood by other society "It depends on the country one is born in that affects our speech:'
Interestingly. truck must have been sufficiently nativised to be classified as 'Jamaican' by 59.7% of the respondents, as opposed to only 34.7% in favor of the category 'American' (cf. Table 8 below). The case of toward versus tmvards is rather difterent, for a number of reasons. First of all, we are dealing with an opaque feature, because the difference between the two items is minimal and they belong to the closed-class category of prepositions, which are not as salient as open-class items such as nouns. Secondly, there is the added difficulty that these forms differ with regard to a word-final <s>, which bears close resemblance to the morpheme {-s}, indicating plural in standard English but not in Jamaican Creole. The respondents of the survey have all been taught to be aware of the standard English plural morpheme, especially in writing, and thus many comments focused on the presence or absence of this feature despite the fact that the question had been designed to test a regionally restricted lexical feature: a. b. c.
"It is a habit of some Jamaicans to mispronounce their words or not pluralise their words when speaking." "Since efforts is plural it follows that the verb made tmvm·d should be plurar' "A singular verb takes a singular noun."
This probably explains the slight edge of tmvards over tmvard (cf Table 9 on next page). But generally speaking, the respondents have no clear preferences with regard to either form, as illustrated by the following comment: "In truth, I'm a bit baffled by the appropriateness of one statement over the other:' Table 8. Origin ofle.x:emes lorry and truck (in% of72 responses) Jamaican English
British English
American English
other
truck
59.7
8.3
34.7
1.4
lorry
1.4
73.6
16.6
1.4
171
172
Andrea Sand Table 9. Acceptability of toward vs. towards (in% of 68 responses)
toward towards neither
'prefer'
'better English'
'bad English'
41.2
41.2
22.1
54.4
38.2
32.4
4.4
20.6
61.8
Table 10. Origin oflexemes toward and towards (in % of 68 responses)
toward towards
Jamaican English
British English
American English
other
13.2
22.1
13.2
0.0
26.5
23.5
14.7
2.9
With regard to regional categorization, it is thus surprising to find the highest percentage for tawards as Jamaican, and again we notice higher percentages for British English than for American English, but evenly distributed for both lexemes. The Jamaican respondents are obviously not aware of the actual regional distribution of these two lexemes. To summarize the lexical questions of the survey, we can state that the respondents always preferred the more formal item of each pair. The use of direct loans from Jamaican Creole is obviously stigmatized and can safely be assumed to represent an intentional rhetorical device along the lines of Allsopp's (1996: lvii) 'antiformal' usage. The fact that freeness is preferred over freebie is probably due to the example chosen in which it was paired with an informal American lexeme and the form freeness itself appears to be rather formal due to the derivational suffix -ness. The regional categorization of the items was not always in line with the findings from our corpus, especially in the case of persons and assistance. Generally, there is a tendency to assign the dispreferred forms to Jamaican usage and the preferred forms to British usage. With regard to the prepositions, the respondents could not assign regional labels, as was to be expected for a closed-class item. 3.2
Morphosyntax
A feature which is situated in the interface of lexicon and morphosyntax is the pluralisation of mass nouns, as in furnitures instead of furniture. This feature is commonly cited as typical of the New Englishes and is also attested in ICE-Jamaica, though not as frequently as in ICE-India or ICE-Singapore (e.g. Platt et al. 1984-52; Schmied 1991: 70; Sand 1999: 135, forthcoming 77 -89). The two lexemes chosen for the questionnaire were toxic waste(s) and equipment(s), both of which
Language attitudes and linguistic awareness in Jamaican Engllsh 173
are generally not pluralized in Standard British4 or American English, because they occurred more than once in a plural form in the corpus data. The sample sentences used in the questionnaire were:
(8) a. The athletes need modern equipment to succeed. b. The athletes need modern equipments to succeed.
We are concerned that this could be an attempt to turn the area into a destination for toxic waste. b. We are concerned that this could be an attempt to tum the area into a destination for toxic wastes.
(9) a.
The survey showed that the respondents are very much aware of the status of these two nouns as collective nouns. There is an overwhelming preference (85.1% for waste, 81.9% for equipment) for the international Standard English forms (see Table 11 below) and a large number of comments point out the ungrammaticality of plural -s for theses nouns. Interestingly, exactly those arguments brought forward by Schmied (1991: 70) to explain the common occurrence of these and similar forms in African Englishes, namely the plural semantics of the lexemes in question, are given as reasons why the respondents do not judge equipments or toxic wastes as acceptable English, as illustrated in the following comments: a. b.
"Equipment is already a plural word." "'Waste' does not need an to show it is a lot:'
's
With regard to regional labelling, again a majority of respondents (25.4% for waste, 38.9% for equipment) attribute the forms without plural -s to British English, but Table 11. Acceptability of wastes vs. waste and equipments vs. equipment (in % of 72 and 67 responses)
wastes waste neither equtpments equipment neither
'prefer'
'better English'
'bad English' 46.3
10.4
22.4
85.1
62.7
4.5
4.5
14.9
49.3
16.7
15.3
76.4
81.9
81.9
9.7
1.4
2.8
13.8
4 There ls one instance of equipments in ICE-GB, however, which may indicate a certain degree of variability in the use of these nouns.
174 Andrea Sand Table 12. Origin oflexemes waste(s) and equipment(s) (in% of72 and 67 responses)
Jamaican English wastes waste equipments equipment
British English
American English
other
9.0
11.9
11.9
0.0
17.9
25.4
11.9
0.0
22.2
5.6
12.5
5.6
18.0
38.9
16.6
1.4
there is no clear picture with regard to regional provenance, as is to be expected in the case ofless salient lexica-grammatical features (cf. Table 12 above). A number of respondents also comment on the difficulty of assigning aregionallabel to something they consider a mistake: a.
b.
"No specific culturallabel may be attached to the incorrect version (equipments) as wherever in the world you go, there are set rules governing pluralization in the English language, yet we find that many native speakers of English do not adhere to such rules~ "Even persons using the Queen's English make this mistake."5
The evidence makes clear that while the forms under analysis, as well as similar forms, are attested in corpus data, even in formal text types, they are by no means accepted by a majority ofJamaicans. Unlike in the case of Indian English, these features would definitely not be included in the codification of a Jamaican English standard Another feature for which corpus evidence suggests that Jamaican English, as well as other New Englishes, diverges from British or American usage is the case of concord with collective nouns, such as government or police (cf. Mair 1992: 82-84; Sand 1999: 133-140, forthcoming: 166-172). In many recent studies on concord with collective nouns in British, American and New Zealand English (e.g. Hundt 1998: 82; Levin 2001: 78-86), police has been found to be one of the strongholds of plural concord. This is not the case in our Jamaican corpus data. The sample sentence in the questionnaire contains both verbal and pronominal concord, which may have influenced the results, as pronominal concord tends towards the plural form, even for nouns which have a higher incidence of singular verbal concord (cf. Hundt 1998: 89; Levin 2001: 159-161): (10) a. b.
5·
The County Police says it has received no formal report. The County Police sav they have received no formal report.
Note the use ofpersons In this comment, which is in line with the findings from ICE-Jamaica,
cf. Section 3.1 above.
Language attitudes and linguistic awareness in Jamaican Engllsh 175 Table 13. Acceptability ofpolice... theyvs.police... it (in% of 65 responses)
police...they police... it neither
'prefer'
'better English'
58.5
52.3
16.9
30.8
21.5
38.5
10.8
26.1
47.7
'bad English'
The respondents of the survey showed a definite preference (58.5% over 30.8%) for the sample sentence with plural concord, but the singular form is also widely accepted (47.7% consider neither form 'bad English: cf. Table 13 above) and the comments indicate very little stigmatization. Respondents have very varied views of the acceptability of the two sentences, as illustrated in the following comments: a. b. c.
"Both are correct:' "The sentence is talking about one County Police therefore to say 'they' would be incorrect=' "In example (a) the Police are plural, and the word 'it' is singular, so the structure of the sentence is incorrect='
This also holds for the regional labelling (cf. Table 14 below), where we find relatively similar ascriptions of both forms, as is also expressed in one comment indicating that "both could be from all regions:' The only difference to the other items on the questionnaire is that both constructions are attributed in more or less an equal number of cases to Jamaican usage. The case of singular or plural concord with police appears to be undecided for the Jamaican respondents and further research is required to find out to which extent Jamaican English is following the world-wide trend towards singular concord with collective nouns. It would also have to be tested whether the text type plays a role here, as Hundt (1998) has found significant differences between various text types in the case of New Zealand English. The final feature under analysis is sequence of tenses, both in indirect speech and to maintain the chronological order of events in complex sentences. Previous research, whether based on anecdotal evidence or corpus data (e.g. Schmied 1991: 66 on African English; Sima Bobda 1998:9 and Sand 1999: 119, forthcoming 113-120 Table 14. Origin of police... they and police... it (in% of 65 responses)
police. ..they police. .. l:t
Jamaican English
British English
American English
other
16.9
15.4
9.2
1.5
18.5
10.8
7.7
1.5
176
Andrea Sand
on a number of varieties), suggests that New Englishes, such as Jamaican English, often differ from international standard English in the application of tense shift rules. The questionnaire therefore contained two instances of reported speech intraduced by a reporting verb in the past tense, namely, (11) a. She said the chlorine tablets were expensive and had to be used everyday.
b. She said the chlorine tablets are expensive and have to be used everyday. (12) a. The director said the project had operated since 1984. b. The director said the project has opera.ted since 1984. In order to draw the respondents' attention to the tense of the reporting verb, these were also underlined. In international standard English, a reporting verb in the past tense requires the verbs in the embedded clauses to be shifted from present tense to past tense and from present perfect and past tense to past perfect. Jamaican English corpus evidence shows that this rule is not consistently applied. Therefore, the questionnaire presents examples to test the respondents' awareness of the backshift rule. The results are quite different for the two sample sentences, as can be seen in Table 15 below. In the case of (11) above (item 3 on the questionnaire), 49.2% prefer the backshifted version with past tense as opposed to only 29.2% who prefer the present tense version without backshift. For ( 12) (also itern 12 on the questionnaire), 54.2% prefer the version without backshift with the present perfect, as opposed to only 29.6% of respondents preferring the backshifted version with the past perfect. This means more or less reversed results for the two samples in this category. The reasons for the different profiles for these two backshift examples, one involving the shift from present tense to past tense, the other the shift from present perfect to past perfect, can only be explained by looking at the comments. Many respondents who commented on these sample sentences express an awareness of the option not to apply the backshift rule if the reference of the original utterance still holds at the time of the reported utterance (cf. e.g. Quirk et al. 1985: 10261029), as illustrated by the following comments: Table 15. Acceptability of said+ PRES vs. said+ PAST (in% of71 and 65 responses)
said+ PRES said+ PAST neither said+ PREsP said+ PAsTP neither
'prefer'
'better English'
'bad English'
29.2
27.7
30.8
49.2
33.8
13.8
21.5
38.4
58.4
54.2
48.6
18.3
22.5
18.3
29.6
22.5
32.4
50.7
Language attitudes and linguistic awareness in Jamaican Engllsh 177
a. b. c.
"(a) [i.e. present tense] gives a sense of the past, that the chlorine tablets are no longer expensive or needed to be used every-day;" "(b) [i.e. present perfect] shows that the project is still in operation." "The version would depend on whether the project still operates."
However, it appears that the option formulated in the grammars is interpreted as an obligation which actively prevents the use ofbackshift in contexts in which the original utterance is still valid at the time of reporting. This understanding is reflected in the comments on sentence 12a above, which remarks that "Jamaicans and Americans tend to use incorrect tense as in (a):' The main reason why the two sample sentences are judged differently seems to be the presence of the adverbial since 1984 in sentence 12 above, as stated in a number of comments such as: a. b.
".. .'since 1984' suggests a continuance~ "Word 'since' implies (project) continuity, therefore 'has operated' should be used."
This line of reasoning is especially interesting, as corpus analysis reveals a tendency in Jamaican English, as well as other New Englishes, to combine since with present tense (cf. Mair 1992: 85; Sand 1999: 118-120, forthcoming 108-110). With regard to regional labelling (cf. Table 16 below), it is striking that the use of present perfect in sentence 12 above has received the largest number (22.5%) of ascriptions to British English. This is not in line with the actual application of the backshift rule in British English corpora, but might indicate a strong exonormative pressure to use present perfect in the context of since. The final item on the questionnaire is concerned with the use of past perfect to maintain the chronological order of events in complex sentences. The sample sentence used (13 below) was taken from a newspaper report about the discovery of a corpse: (13) a. The autopsy revealed that her body was cut into pieces by a machine saw. b. The autopsy revealed that her body had been cut into pieces by a ma-
chine saw. Table 16. Origin of said+ PREs, said+ PAST (in% of71 and 65 responses)
said+ PRBs said+ PAST said + PRBsP said+ PAsTP
Jamaican English
British English
American English
other
13.8 12.3
10.8
4.6
1.5
16.9
7.7
1.5
12.7
22.5
11.3
1.4
11.3
8.5
9.9
1.4
178 Andrea Sand
Obviously, the mutilation of the body predates the autopsy and is therefore generally encoded in the past perfect to indicate the chronological sequence of events. However, corpus data shows many instances of "missing" past perfects in such contexts of past-in-the-past in written Jamaican English (cf. Sand 1999: 119). The present survey, however, reveals a clear preference (64.6% over only 16.9%) for the use of the past perfect, as can be seen in Table 17 below. The use of past perfect is clearly preferred and none of the respondents were indecisive with regard to assigning the label "better English" to one of the forms. It is the only item on the questionnaire in which the question about the status of"better or more correct English" elicited no answer in the category "neither': Only 4.6% of the respondents regarded neither form as 'bad English: which also stands in stark contrast to some of the other items on the questionnaire. The comments for this question, which are fewer than for other items on the questionnaire, indicate a clear case of stigmatization of the past tense form in this context: a. "(a) [i.e. PAST] wrong tense." b. "Colloquialism is atrecting the Standard form in the case of (a) [i.e. PAST]." The small number of comments could also be interpreted to indicate a clear-cut case of preference. With regard to the regional labels, we see another majority for assigning the preferred form to British English (33.8%) and the dispreferred form to Jamaican English (35.4%) (see Table 18 below), which is also expressed by a comment on the use of the past tense form by "a majority of Jamaicans=' Table 17. Acceptability of PAst vs. PPERF in a temporal sequence (in% of 65 responses) 'prefer'
'better Englisli
'bad English' 21.5
was cut PAsT had been cut PPERF
16.9
13.8
64.6
56.9
0.0
neither
18.4
29.2
78.4
Table 18. Origin of PAST and PPERF in a temporal sequence (in% of 65 responses) Jamaican English
British English
American English
other
was cut PAsT
35.4
9.2
21.5
0.0
had been cut
9.2
33.8
20.0
0.0
PPERF
Language attitudes and linguistic awareness in Jamaican Engllsh 179
One could interpret these results to indicate that there is an awareness of the frequent use of past tense in past-in-the-past context in Jamaican English, but a strong normative pressure to regard this usage as incorrect. Tense choices involving past tense, present perfect and past perfect are quite complex, as the standard English system is not easily mapped onto the Jamaican Creole system with the anterior marker bin and the completive marker done as well as a number of mesolectal forms to refer to events in the past (cf. Patrick 1999; Sand 1999: 118-125; also Bickerton 1975: 122-132 on the situation in the Guyanese continuum). It must be stated at this point, that a certain variability in the use of these forms can be observed in the Jamaican corpus data but no clear-cut new norms have evolved as of yet. To summarize the results for the morphosyntactic items of the survey; the general trend is an endorsement of international standard English norms, with the exception of the backshift from present perfect to past perfect, due to the reasons given above. This is contrary to corpus evidence and to the results concerning lexical features and indicates a strong (prescriptive?) pressure with regard to morphosyntax, at least in writing.
4
Conclusions
Despite the relatively small size of the sample, the present study shows that there are definite ditrerences with regard to the acceptability oflexical and morphosyntactic features which differentiate educated Jamaican usage from standard international or British English. Corpus data is a good base for showing which kinds of features occur in which contexts and at which frequency, but it does not tell us whether these features are endorsed by the speech community or not. While it is certainly a well-known fact in sociolinguistics that speakers' actual usage often differs from their metalinguistic commentary on individual features, as documented in studies ranging from Labov (1972) to Johnstone et al. (2006), the gap between actual usage and its model must not be significant if we want to assign variety status. If we assume an endonormativisation process for Jamaican English, a number of criteria must be met (cf. Mallin 2007: 169-173). The first group of criteria is concerned with the use of the variety in certain domains, such as education, administration, the media, and creative writing. This is without doubt the case in Jamaica. The second group of criteria applies to the linguistic features of the variety. A number of studies, as already quoted in the introduction above, have found distinctively Jamaican features with regard to phonology; lexicon, syntax, and discourse style. The important point is, however, that Mallin (2007: 173) stresses the fact that these features must be systematic and communal rather than
180
Andrea Sand
idiosyncratic deviations from the norm. Of the features identified in previous research, only very few are preferred forms, i.e. used in a majority of cases by a majority of speakers and writers. It remains to be seen what the status of such variants may be in a process of institutionalisation - accepted alternatives or 'mistakes'? I would argue that this is where acceptability surveys such as the present study come into the picture. If the majority of respondents accept a form, we can assume that it is part of the evolving standard. If the majority of the respondents reject a certain form or construction, we must accept that despite its more or less frequent occurrence in actual usage, it must be considered non-standard.This evaluation process is part of the third group of criteria listed in Mallin (2007: 172-173), those concerned with institutionalisation. Beginning codification can be seen in the DCEU, for example. But the general acceptance of the local variety and its label as well as closing the gap between performance model and linguistic behaviour are those points which merit further investigation. The survey revealed that many respondents were well aware of the Jamaican origin of a number of items under analysis, but the question whether there is a "standard Jamaican English" was not addressed in the questionnaire (see appendix). There are without doubt many more features that could have been investigated, but for a pilot study, the questionnaire proved to be sufficient As it turned out, many respondents did not complete the questionnaire, therefore future investigations should concentrate on shorter questionnaires, also using difterent tasks such as correction or fill-in-the blanks, or interviews with a larger number of respondents. Furthermore, the results of the present study are only concerned with formal written usage. It remains to be seen whether certain features are more acceptable in the context of spoken language. Text-type specific analysis of corpus data has already revealed that some features, for example progressives without the copula (cf. Sand 1999: 114-118, forthcoming 131f.) are restricted to informal spoken language. However, their acceptability among educated speakers has not yet been investigated. From the results of previous research on the acrolectal end of the Jamaican continuum and the results of the present survey. it appears that the speakers of English in Jamaica are in the process of examining the norms of British English and developing their own. At this point, it is difficult to predict which features will eventually be part of a codified Jamaican Standard English.
References Allsopp, R. (ed). 1996. Dictio11ary ojCaribbea11 English Usage. Oxford: OUP. Bickerton. D. 1973. The structure of the creole continuum. Language 49: 640-669. Bickerton. D. 1975. Dynamics of a Creole System. Cambridge: CUP.
Language attitudes and linguistic awareness in Jamaican Engllsh Bib~ D. 1995. Dimensions of Register VarlaNon: A Cross-Linguistic Comparison. Cambridge: CUP. Cassidy, P.G. & LePage, R. 1967. Dictionary ofJamaican English. Cambridge: CUP. Christie, P. 1989. Questions of standards and lntraregional differences in Caribbean examlnations. In English Across Cultures, Cultures across English: A Reader in Cross-Cultural Communication, 0. Garcia & R. Otheguy (eds.), 243-262. Berlin: Mouton de Gruyter. DeCamp, D. 1971. Toward a generative analysis of a post-creole speech continuum. In Pldginizatlon and Creolization ofLanguages, D. Hymes (ed), 349-370. Cambridge: CUP. GOrlach, M. 1991. English as a world language: The state of the art. In Engllshes: Studies In Varieties of English 1984-1988, M. GOrlach (ed), 10-35. Amsterdam: John Benjamlns. Greenbaum, S. 1996. Introducing ICE. In Comparing English World-wide: The International CorpusofEnglish, S. Greenbaum (ed.), 3-12. Oxford: Clarendon Hundt, M. 1998. New Zealand English Grammar: Fact or Fiction? Amsterdam: John Benjamlns. Irvine, A. 2004. A good command of the English language: Phonological variation ln the Jamaican acrolect. Journal ofPidgi11 and Creole Languages 19: 41-76. Johnstone, B., Andrus, J. & Danielson, A. 2006. Mobility, indexicality and the enregisterment of "Pittsburghese." Joun1al ofEnglish Linguistics 34:77-104. Labov, W. 1966. The Social Stratification ofEnglish in New York City. Washington DC: Center for Applied Linguistics. Labov, W. 1972. Socioli11guistic Patten1s. Philadelphia PA: University of Pennsylvania Press. Levin, M. 2001. Agreement with Collective Nou,IS. Lund: Lund University Press. Mair, C. 1992. Problems in the compilation of a corpus of Standard Caribbean English: A pilot study. In New Directions in E11glish Language Corpora: Methodology, Re.sults, Software Developme11ts, G. Leitner (ed.), 75-96. Berlin: Mouton de Gruyter. Mair, C. 2003. Kreollsmen und verbales Identitiitsmanagement im geschriebenen jamaikanischen Englisch. In Zwischen Ausgrenzungu11d Hybridisieru11g, E. Vogel, A. Napp & W. Lutterer (eds.), 79-96. Wtirzburg: Ergon. Mollln, S. 2007. New variety or learner English? Criteria for variety status and the case of EuroEnglish. English World- Wide 28: 167-185. Patrick. P.L. 1999. Urban Jamaican Creole: Variatio11 in the Mesolect. Amsterdam: John Benjamlns. Platt, J. T., Weber, H. & Ho, M. L. 1984. The New Englishe.s. London: Routledge. Quirk, R., Greenbaum, S., Leech, G. & Svartvlk, J. 1985. A Comprehensive Grammar of the English Language. London: Longman. Roberts, P. 1988. West Indians a11d Their Language. Cambridge: CUP. Sand. A. 1999. Linguistic Variation in Jamaica: A Corpus-Based Study of Radio and Newspaper Usage. 'llibingen: Nar.r. Sand. A. Forthcoming. Angloversals? Shared Morpho-Syntactic Features In Contact Varieties of English. Amsterdam: John Benjamins. Schmied. J. 1991. English In Africa: An IntroducNon. London: Longman Sebba, M. 1997. Contact Languages: Pidgins and Creoles. London: Macmillan. Shields, K. 1989. Standard English in Jamaica: A case of competing models. English World-Wide
10:41-53.
Simo Bobda, A. 1998. The lndigenizaNon of English In Cameroon and New Englishisms. Essen: LAUD. Trudgill, P. 1974. The Social Dttferentlation ofEnglish In Norwich. Cambridge: CUP.
181
182
Andrea Sand
6. Appendix Questionnaire: Varieties of English The examples (printed in ita.lics) are taken from WRITTEN texts collected in various English-speaking countries. Focussing on the underlined sections, please answer the questions following each sentence pair. Thank you for your time! 1. a. It is still not known what caused the blaze that also destroyed a truck that was
parked nearby. b. It is still not known what caused the blaze that also destroyed a lorrv that was
parked nearby. -
Which version would you prefer in your own writing? D a) D b) D no preference
-
Which version do you consider "better English" or "more correct English"? Db) D neither D a)
-
Is there one version which is dearly "bad English" or "wrong"? Db) D neither D a)
-
Could you assign "regional labelS: such as "Jamaican~ "American" or "British~ to any of the versions? a) b) _ _ _ _ _ _ _ __
-
Do you wish to comment on any of the examples?
2. a. The athletes need modern equipment to succeed. b. The athletes need modern equipments to succeed. -
Which version would you prefer in your own writing? Db) D no preference D a)
-
Which version do you consider "better English" or "more correct Englisli? Db) D neither D a)
-
Is there one version which is clearly "bad English" or "wrong"? Db) D neither D a)
-
Could you assign "regional labelS: such as "Jamaican~ "American" or "Britisli, to any of the versions? a) b) _ _ _ _ _ _ _ __
Language attitudes and linguistic awareness in Jamaican Engllsh 183
3.
Do you wish to comment on any of the examples?
a. The directm· sa.id the project had opera.ted since 1984. b. The director said the project has operated since 1984. -
Which version would you prefer in your own writing? D a) Db) D no preference
- Which version do you consider "better English, or "more correct English,? D a) Db) D neither - Is there one version which is clearly "bad English" or "wrong"? D a) Db) D neither -
Could you assign "regional labels~ such as "Jamaican~ "American, or "British~ to any of the versions? ~ b) _________________
-
Do you wish to comment on any of the examples?
4. a. The students often do not have the financial backative to secure a loan. b. The students often do not have the financial suppo1·t to secure a loan. -
Which version would you prefer in your own writing? D a) Db) D no preference
-
Which version do you consider "better English" or "more correct English,? Db) D neither D a)
- Is there one version which is clearly "bad English, or "wrong"? D a) D b) D neither -
Could you assign "regional labels~ such as "Jamaican': "American, or "British~ to any of the versions? ~ b) _________________
-
Do you wish to comment on any of the examples?
5. a. Efforts should be made towards constitutional reform. b. Effm·ts should be made toward constitutiona-l reform. -
Which version would you prefer in your own writing? Db) D no preference D a)
184 Andrea Sand
-
Which version do you consider "better English" or "more correct English"? Db) D neither D a)
-
Is there one version which is clearly "bad English" or "wrong"? Db) D neither D a)
-
Could you assign "regional labels~ such as "Jamaican~ "American" or "British~ to any of the versions? b) _________________ a)
-
Do you wish to comment on any of the examples?
6. a. There are some people who believe that football is a boring game. b. There a.re some pe1-sons who believe that football is a boring game. -
Which version would you prefer in your own writing? Db) D no preference D a)
-
Which version do you consider "better English" or "more correct English"? Db) D neither D a)
-
Is there one version which is clearly "bad English" or "wrong"? Db) D neither D a)
-
Could you assign "regional labels': such as "Jamaican~ "American" or "British': to any of the versions? b) _________________ a)
-
Do you wish to comment on any of the examples?
7. a. We are concerned that this could be an attempt to turn the area into a desti-
nation fm· toxic waste. b. We are concerned that this could be an attempt to turn the area into a desti-
nation for toxic wastes. -
Which version would you prefer in your own writing? D a) D b) D no preference
-
Which version do you consider "better English" or "more correct English"? Db) D neither D a)
-
Is there one version which is clearly "bad English" or "wrong"? Db) D neither D a)
Language attitudes and linguistic awareness in Jamaican Engllsh 185
-
Could you assign "regional labelS: such as "Jamaican': "American'' or "British~ to any of the versions? ~ b) _________________
-
Do you wish to comment on any of the examples?
8. a. The organisers received generous help from volunteers. b. The organisers received generous assistance from volunteers. -
Which version would you prefer in your own writing? D a) Db) D no preference
-
Which version do you consider "better English" or "more correct English"? Db) D neither D a)
- Is there one version which is clearly "bad English" or "wrong"? D a) Db) D neither -
Could you assign "regional labelS: such as "Jamaican~ "American" or "British~ to any of the versions? ~ b) _________________
- Do you wish to comment on any of the examples? 9. a. A lot of houses were mashed up in the rain. b. A lot of houses were destroved in the rain.
- Which version would you prefer in your own writing? D a) Db) D no preference -
Which version do you consider "better English" or "more correct English"? Db) D neither D a)
-
Is there one version which is clearly "bad English" or "wrong"? D a) Db) D neither
-
Could you assign "regional labelS: such as "Jamaican': "American'' or "British~ to any of the versions? ~ b) _________________
-
Do you wish to comment on any of the examples?
186 Andrea Sand
10. a. The autopsy revealed that her body 1vas cut into pieces by a machine saw. b. The autopsy revealed that here body had been cut into pieces by a machine
saw. -
Which version would you prefer in your own writing? Db) D no preference D a)
-
Which version do you consider "better English" or "more correct English"? D a) Db) D neither
- Is there one version which is clearly "bad English" or "wrong"? Db) D neither D a) -
Could you assign "regional labels~ such as "Jamaican~ "American" or "Britisli, to any of the versions? a) b) _________________
-
Do you wish to comment on any of the examples?
11. a. There are those who argue that the concept of a free university education
would encourage a freebie mentality among students. b. There are those who a.rgue that the concept of a. free university education would encourage a freeness mentality among students. -
Which version would you prefer in your own writing? Db) D no preference D a)
-
Which version do you consider "better English" or "more correct Englisli? Db) D neither D a)
-
Is there one version which is clearly "bad English" or "wrong"? D a) Db) D neither
-
Could you assign "regional label~ such as "Jamaican~ "American" or "British': to any of the versions? a) b) _________________
-
Do you wish to comment on any ofthe examples?
12. a) She said chlorine tablets were expensive and had to be used everyday. b) She said chlorine tablets are expensive and have to be used everyday. -
Which version would you prefer in your own writing? Db) D no preference D a)
Language attitudes and linguistic awareness in Jamaican Engllsh 187
-
Which version do you consider "better English" or "more correct English"? D a) D b) D neither
-
Is there one version which is clearly "bad English" or "wrong"? Db) D neither D a)
-
Could you assign "regional labels~ such as "Jamaican~ "American" or "British~ to any of the versions? ~ b) _________________
-
Do you wish to comment on any of the examples?
13. a) The County Police says it has received no formal report. b) The County Police say they have 1-eceived no jormal1-eport. -
Which version would you prefer in your own writing? Db) D no preference D a)
-
Which version do you consider "better English" or "more correct English"? Db) D neither D a)
-
Is there one version which is clearly "bad English" or "wrong"? Db) D neither D a)
-
Could you assign "regional labels~ such as "Jamaican': "American" or "British~ to any of the versions? ~ b) _________________
-
Do you wish to comment on any of the examples?
To process your questionnaire properly, we need to know a few things about you. Gender: D M D F
Age: D 18-25
D 26-45
046-65
CountryofBirth: In Jamaica since when? _ _ _ _ _ __ First language spoken in your family? _ _ _ _ _ _ _ _ _ _ _ _ _ __ Have you ever lived abroad? If yes, where and how long?_ _ _ _ _ _ _ __ Thank you for participating in our survey!
PART III
Variation and the community
The varilingual repertoire of Tobagonian speakers Valerie Youssef This paper challenges the creole contimrum model as the most accurate descriptor oflanguage development in Caribbean sociolinguistic complexes for the present. The model was constrained by its time and by quantitative generalization which lost sight of the individual as the locus of change. Much has changed internally since the late 1960's to render models which predict continuous change towards the acrolect inapplicable. If we now focus on the balancing oflanguage varieties in the individual, we can extend our focus to relevant community groups, taking into consideration both socio-economic factors and social-psychological factors for the society as a whole as well as for the individuals concerned. What emerges is a mh~:ed but structured varilingual competence representative of a social-psychological reality which demands balancing codes rather than shifting in time from one to another. Further, language use cannot be effectively described without considering individual attitudes and community ideologies.
1.
Introduction
This paper challenges the creole continuum notion in the form in which it was originally conceived (DeCamp 1971, Bickerton 1975), as an accurate descriptor of language use and change in the twenty-first century Anglophone Caribbean. It is more useful to use Carrington's (1976, 1993) term: the Caribbean sociolinguistic complex, which allows for multidimensionality in interaction among language varieties as opposed to the strictures of unidimensionality and unilinear language interaction and change proposed by the creole continuum. The paper argues that the traditional notion of a continuum has clouded our perceptions and intervened in the accuracy of our descriptions of language use in Anglophone Caribbean societies. Not only was the model an attempt to provide a societal linguistic descriptor which, from the outset, lost the individual in the midst, but it was constrained by its time and the expectations then for societal
192
Valerie Youssef
development. Much has changed internally in the Anglophone Caribbean to render models which predict continuous change towards the English acrolect inapplicable. In particular, the acrolect has ceased to be a variety to aspire toward for its own sake. It may still be useful to speak of a continuum of varieties, but it is important to recognize the precise significance of each variety for its speakers and the functional need for more than one variety, as each is appropriate to specific contexts in the language situation in which it operates. Varieties in a continuum can undergo relative stabilization. If we acknowledge: 1. That the individual is the locus of language development and change, and 2. That the transition between the individuafs language and that of the community demands qualitative as well as quantitative treatment, then we also need to recognize: 3. That there is a capacity for the balancing of language varieties in the individual, in accordance with the sociolinguistic appropriateness of each to its particular fit in the community and the capacity of each to represent particular dimensions of the individual's identity. We may then extend this capacity to relevant community groups. taking into consideration both socio-economic factors and social-psychological factors for the society as a whole as well as for the individuals concerned. To quote one of our Tobagonian informants: "A good Bethel twang? ... Is a mix-up, a mix-up:' The mixed nature oflanguage competence in situations of intense contact among separate language varieties demonstrates a particular kind of language competence and reflects the needs of an entire community's social-psychological reality. In relation to the above, it becomes evident that language use cannot be accurately described without consideration of the shifting worldview of the individuals and communities it describes. The language variety or varieties that best represent(s) and is most appropriate to their worldview becomes the dominant variety(-ies).
2.
Background
The language complex of my focus is on the island of Tobago, the smaller island of the twin island state of Trinidad & Tobago, 'twin' being a misnomer for the quite distinct societies, people, and- unsurprisingly -language situations that exist despite the political union of the two islands. Winford James and I undertook a major
The varllingual repertoire of Tobago Dian speakers 193
language collection exercise throughout Tobago in the context of The University of the West Indies Tobago Language Project (1991-1998). 1 Tobago has had a far less complex language history than Trinidad and additionally has escaped the industrialization and urbanization which characterized Trinidad's development in the twentieth century. Tied administratively to Trinidad from 1898, Tobago has had little legislative or financial control of its destiny throughout the twentieth century. It remained profoundly underdeveloped, continuing to rely on "peasant agriculture, plantation cultivation of cocoa and coconuts, fishing, the plying of various trades, and provision of government and private services" (James 1997: 46). The basilectal Creole flourished in these societal conditions in Tobago, even as it failed entirely to characterize the linguistic development of its neighbour Trinidad. The English acrolect, on the other hand, flourished among the more educated but plummeted in ideological significance in the newly independent state created in 1963. In this context, the local Creole language came to be perceived as 'we owri and the acrolect that of remote colonial masters who, in the wake of Independence, commanded much less respect than formerly. It would be fair to say that this change in attitude to the acrolect was common to both islands. When Eric Williams came to power in Trinidad & Tobago in 1956 with a promise of free government secondary schooling for all, characterized by his famous utterance: "The future of the children lies in their schoolbags." This established a shift of interest away from working the land, but it did not bring either the intended universal secondary education or the infrastructural and economic development that would open a range of new career opportunities. The Tobago Development Plan Report, Number 4, as late as 1998 (Tobago House of Assembly), found young people still getting inadequate preparation and training for the work force, resulting in the development of a culture of dependence on seasonal government employment and sporadic tourism involvement. Today in Tobago there is a catastrophic level of unemployment, within which a large sub-class of young people have no means to advance within the establishment system. The majority leave school between fourteen and sixteen and the young men in particular enter into an existence of'liming on the blocK the local expression for idling away the time on street corners, dreaming of a way up and out, potentially through football, with Dwight Yorke as the significant name on their lips and in their hearts. That an international footballer should have come out of Tobago is a great source of pride to the young men of the island, such that Yorke has come to epitomize all their hopes and dreams. As we discussed their lives and prospects with 1. This work was supported by a small grant from the University and much persistence and labour on our part The data set we collected fuelled a book-length study Games and Youssef 2003) and a number of descriptive papers (e.g. Youssef 1996, 2001 ).
194 Valerie Youssef
them, they talked not only of their immediate and long-term goals but also of a clash of worlds (Youssef 2005). On the one hand, there was the pull of the larger external world beyond the Caribbean, whose 'reality' is represented by the media of the United States and a subset of tourists. On the other hand, there is an internal conflict between Tobago and Trinidad, as the government promises much but delivers little in terms of real opportunity for the smaller island. The young people's statements allow us to observe real conflicts in the value systems with which they are grappling. Both the content of their talk and the structuring of it provide a means of looking into their worldview and ideology. The examples below are telling. Informant 1: Young man in his 20s, responding to a question as to how he and his friends passed their time. (1) Well mostly limes on the docks like, you knaw fishin, fun kind iJ ting. Nothin
more really. It have a beach and you find most of the follas dem does try to get a little change by workin the seine. And then in the evenin time might go play a. little sports any ting on the football ground and so on. Most a the youngfollas that is the main thing in the village, playin sports. So you find you jus home, jus relaxin an playin sports. It eh have nothing else to do. Well, mostly just hanging out in the dock area, fishing and fun kinds of things. Nothing more than that really. There's a beach and you will find that most of the young men try to make a little money by helping pull in the fish nets. And then in the evenings they might go and play some sport or the other on the football ground. For most of the young men that is the important thing in the village, playing sports. So basically you will find you are just at home, relaxing and playing sports. There isn't anything else to do. In 2006, the hopes and dreams of both Trinidadians and Tobagonians became fixed on their countrys performance in the World Cup. But even in the 1990's, young Tobagonian men declared that football was the only thing that they wanted to talk about, the only thing that might represent forward movement for them: Informant 2: Young man in his 20s, unemployed. (2) Well the only thing I like to talk about that ever really happen to me - that I
will never forget then in life - playin football, yes, playin sports - That I like so much - I find iz a keep dawn right through . ... you just getting poorer. I want to move on. Everybody tellin me I got the talent, I got this, I 1-eady to go an)I"M;ay, but I eh getting the push. I went an train with all kinds of teams and play an, you know - like me eh know is like I just eh gettin the push. That is my main thing. to tell you the truth, me an Dwight Yorke use to play together. I play against all dem nationa.ls in T1"inidad. Latapy all a them, all iJ them. An I know to mihself mih standard up there wi them.
The varllingual repertoire of Tobago Dian speakers 195 Well, the only thing that I like to talk about-it's the only thing that ever really happened to me, that I can't forget about in my life dream, is playing football, yes, playing sports. That's what I like so much. I find that you are kept back all the way though -you are just getting poorer. I want to move on. Everybody is telling me that I have the talent-! have that; I'm ready to go anyway, but I'm not getting the push (that I need). I went and trained with all kinds of teams and played, and, you know, it's as if I'm just not getting the push. That's my main thing. To tell you the truth, Dwight Yorke and I used to play together. I used to play against all those Trinidad nationals-Latapy, all of them, all of them. And I know myself; my standard is up there with them. The way up and out is perceived as being through Trinidad. Infonnant 3: Young man in his 20s. also unemployed: (3) a. I myself,- iz Trinidad where I go go. I go be right up there in the business
Area Club... workin there - more serious thing- trainin hard you knowworkin to make the na.tional team. That is mih ma.in a.im. For me, it's Trinidad that I'm going to move to. I'm going to be right up there in the business Area Club, working there-it will be a more serious kind of thing-training hard, you know, working to make the national team. That is my chief aim. b. When I talk in I foel I go beat all the rest a the fellas. I doh even study dem
talk, dey, you know. Why I can't travel? Why I can't travel fu enlighten? When I am talking I feel that I am going to beat all the rest of the young men in my area. I don't even take on their talk, you know. Why can't I travel? Why can't I travel and learn the ways of the world? The worldview which emerges in these young people's statements of their aspirations can be linked, as for us all, to the language varieties which they make use of as will become clearer in the discussion below.
3· 3.1
The language study Informants and data collection
In seeking out a valid linguistic picture for Tobago, we first sought to sample all relevant groupings in the society who might perform differently linguistically. The population was first divided on an island-wide scale according to what we perceived as relevant social factors, namely: age, gender, geographical locality and education. An intensive series of interviews was conducted, both formal and
196 Valerie Youssef
informal. in a range of villages throughout the island. Differences from classical Labovian methodology (Labov 1972), or even Bickertonian (Bickerton 1975), would have been the extremeness of our efforts to set up naturalistic non-threatening interview situations, visiting the villages and homes and using different interviewers, including village members themselves in some cases, to elicit difterent kinds of speech. We used no word lists or reading passages, but did our best to tap naturalistically into the full range of our informants' communicative competence. The survey study was followed up by a small community study in the village of Bethel, whose socio-economic structure was researched in detail and our subgroupings then derived from the structure of that particular community (cf. Youssef 2001). It is data from this study which we use in this community-focused paper. We collected data from two age groups: 16-21 and over 70. Ten informants per age group and gender were sought yielding a sample population of 37 for a village population of 1,638 (Central Statistical Office 1997). Within each subgroup, both educational level and associated network structure were considered. Profession or past profession was considered for older speakers, and, associated with this, the extent of travel and stay outside Tobago. The life experience and professional development of the older group made some comparisons impossible with the younger group. We interviewed ten young women, four of whom had left school after what is termed post-primaryeducation, at age 14, and six ofwhom had continued in school until sixth-form. Those who had left school continued their education in a variety of small vocational institutions. There were nine young men interviewed, four of whom had left school at 14 and five of whom were still in school. Three of the four who had left school were unemployed and one was attending a vocational school. At this stage, they had all lived only in the village, and differences in network structure came through their different educational experiences and contacts. Of the ten older women interviewed, four had had no formal education and had lived and worked on the land their whole lives; two had attended school until their early teens and had lived as home-makers in Tobago ever since; two had achieved tertiary education abroad and had been teachers themselves; and two had attended secondary school and had spent considerable periods of time working in Trinidad. This range of experience and diversity of network structure in different phases of their lives could not be compared to the younger group, as noted earlier. Of the eight older men interviewed, two had achieved tertiary education and six post-primary. The two with tertiary education had achieved professionally to the levels oflawyer and headmaster. Among the remaining six. one had been a Spiritual Baptist minister and one a police officer, while the remaining four had lived and worked at home on the land.
The varllingual repertoire of Tobago Dian speakers 197 4
The data sample and data analysis
A prototypical subset of the language systems in contact is exemplified below through a tabular representation of the marking of a verb. It will give the reader some sense of the differences among the systems in contact. It is important to note that, while quantification of data was carried out for forms which were of high occurrence, it was recognized that the very production ofbasilectal features needed to be given considerable attention, although their incidence was low. Whatever measures were used to create basilectal-appropriate contexts, the interview context by its very nature as a speech collection exercise demanded English or a variety dose to it of any informant who commanded such. For this reason alone, basilectal utterances were unusual. Hence, when they occurred in naturalistic contexts, comparative tables for the presence of those features were set up across the groups under examination. Additionally, with such features it was necessary in some cases to distinguish between forms produced specifically as demonstrative samples of prototypical Tobagonian basilect use, and others which were produced in the natural flow of speech. With the more regular range of forms. the range and extent of an individual's production was noted and the association of particular forms with particular speakers was specifically addressed to examine both the directionality of change and speaker range according to context. The presence of the low-occurrence forms
Table 1. Pre-verbal tense-aspect marking in Tobagonian Creole (past and present contexts)
Basilect
Mesolect
Acrolect
Past
Present
Relative Past - (RP) bin* RP imperfective- bina~ Perfective - 0 Imperfective - a Relative Past - did Habitual- use to Continuous- was+ -i11(g) Perfect/Completive - cbn Perfective - 0 Past prefect - had + -en* Present perfect - have + V -en* Simple Past - SE-ed Continuous - was + - in(g) Habitual- use to
Perfective - 0 Imperfective - a
Habitual- does Continuous- 0 + -in(g) Perfective - 0
* Marginal forms to the contact systems as assessed from our (limited) data set.
198 Valerie Youssef
is a significant indicator of the survival ofbasilectal features and the low production ofacrolectal ones also makes a contribution to our overall analysis of the directionality of change. 4.1
Findings
What emerged was, in one significant aspect, similar to my earlier findings in Trinidad. This finding was that many persons exhibit what I have elsewhere called varilingual competence (e.g. Youssef 1996), balancing their use of the contact varieties according to stylistic features of the immediate situation in which they found themselves, as well as according to user-based factors. They produced some mesolectal Tobagonian Creole (Tob MC) and some acrolect (Tob S), or some basilect (Tob BC) and some mesolect in a proportional representation of the two varieties which was appropriate to their own language competence. Apart from the oldest speakers in the sample (over 75) without any education, all study participants produced either basilect and mesolect or mesolect and acrolect, which would be intensively varied within single utterances. To quote one of our more linguistically versatile informants: Informant 4: Young professional woman in her 20s: (4) You want a good Bethel twang? Is a mix-up a mix up.
If I know you have educated background I try my best to speak so up - but when I meet the average people then I does say: "Ehman, wha: go up on a road de:. They does say 'Eh man, eh man, wha: me a go for? So all these things we ha: to meet up."
Do you want a good Bethel accent? It's a mix-up, a mix-up. Ifi know you have an educated background, I will try my best to speak to you in an elevated way- but when I meet average people I say: 'Hey man, what's going on up the road'. They say: 'Hey, man, what would I be going for?' So all these things we have to meet up. An example of extreme and unusual mixing, moving directly from basilect to acrolect is given below. Informant 5: A retired teacher in her early 70s, discussing a particular politician: (5) No say mi nah like de the man hiself. No say mi nah like the man hiself.
His style of approaching people when he goes to talk to people. fm not saying I don't like the man himself. Not that I don't like the man himself. His style of approaching people when he goes to talk to people. Figure 1 following shows the presence of preverbal imperfective a across the full range of our age-graded sample involving men and women 70 and above and
The varllingual repertoire of TobagoDian speakers 199 100 90 80 70 ~ 60 50 B IS 40 Po. 30 20 10 0
c
Older men
Older women
Young men
•
Usage
•
Natural Usage
Young women
Figure 1. Presence of a for older and younger men and women
young people between the ages of 16 and 21 for natural and 'example' usage. It indicates a slightly higher rate of natural usage for the form among young women as opposed to older women, and a reduced level among young men as compared to older men. Examples of this form in use include the following. Infonnant 6: A retired teacher, female: (6) When we go down to school dere, and you worry the Principal, Pinky wud
sta:t to tell you. .. 'Yuh see wha: a go on de: wi dem piknee, wha: yuh a see... Allyuh nah a discipline dem in de proper wa~ so wha: awi teach fu do?. When we go down to the school and we worry the principal she would start to tell you: "You see what is going on with those children, what you are seeing? You are not disciplining them in the right way, so what are we teaching for?" Infonnant 7: Male sixth-former: (7) Well school a go oka~ but for now I ha: fu waitfu results come out... Me a de
a hope dat he come out high but work ....
if I eh get no passes I go ha: to look fu
Well, school is going ok, but for the time being I have to just wait for the results to come out. .. fm hoping that they come out high, but ifi don't get any passes I will have to look for work... Figure 2 indicates presence of bin(a), which was a rare form in the relative past contexts in which it would be expected. The sample is important only for indicating that these classical Tob BC forms are present among young people in similar measure with older people and in contradistinction to suggestions of decreolization in
2.00
Valerie Youssef 100 90 80 70
!'c
... ~
~
60 50 40 30 20
n.l
20
~- ~
10 1 -
11
L
0
0 Older men
Older women
J
Young men
I L
J-
Yowtgwomen
Speaker group
Figure 2. Presence of bin for older and younger men and women time. In noting this, we naturally take into consideration the limitations of agegrading as a measure of shift in time, in particular the potential for speaker output to change in real time with age. Examples of this form in use include the following: Informant 8: Retired teacher, male: (8) Mih teacher bina Barbadian man. And lifo slow. ..Mi bina monitor but mi
only get 60 cents. 1he headmaster bina get $18... the second master bina $16. 1he third one bina 14... My teacher was a Barbadian man and life was slow. I used to make 60 cents. The headmaster made 18 dollars. The second master made sixteen. The third one made fourteen. Informant 9: Young man, unemployed: (9) He bina tell awi wha does wha really go on an ting in Trinidad... wha: does
really take place... Hi bina show awi all kinda ting. He was telling us what really goes on in Trinidad. What really takes place. He showed us all kinds of things. The presence of acrolectal have+ V-en is indicated in Figure 3 for older and younger men and women and indicates that this form is in greater use among older rather than younger speakers. Notwithstanding the possibility that there is more acrolectal speech across the board today than say fifty years ago, this form appears to have almost fallen into disuse for the younger age group.
The varllingual repertoire of Tobagonian speakers 40
37.5 ~
35
???
1-
~
30 1 1-
1-
-
1-
11.1
10 1 -
5
1-
10 1-
0
Older men
Older women
Young men
Young women
Speaker group
Figure 3. Presence of have+ V-en for older and younger speakers
Moving on to the figures for production of forms, Figure 4a shows production of be+ V-in(g) for continuous marking as opposed to 0 + -V-in(g) for older speakers, and Figure 4b shows production of the same variable for younger speakers. Again, it indicates much greater usage of the acrolectal auxiliary be among older rather than younger speakers. 80
70
vv
69
r-
r-60 c--
j
50
ii
40 c--
~
30 c--
.. ~
c--
31 23
20
1---
r--
•otdermen
10 c-0
0
• Older women
0
Tertiary
Professions
Post primary
Speaker group by education
Figure 4a. Production of be + - V-in(g) as opposed to e + V-in(g) by older men and women
201
202
Valerie Youssef
8
7.5
7 6 5
-
2
-
• Youngmen 0
0
• Youngwomen
0
Secondary
Post primary
Speaker group by education
Figure 4b. Production of be + - V-in(g) as opposed to "' + V-in(g) by younger men and women
Infonnant 10: A young man of 19, employed: (10) Some of dem complainin they ha: nutten to do. What I tell yuh, right...
Mos 3 de young people get take up in sellin drugs an ting. Well, you feel like you mus find some way of livin, so yuh sellin drugs, an den police behind yuh. Most 3 them jus wastin away dere life den. Some of them are complaining that they have nothing to do. What I can tell you is that most of the young people get involved in selling drugs and things like that. Well, you feel you must find some way of making a living, so you end up selling drugs and then the police are after you. Most of them are basically just wasting their lives. Of interest also was the marked difference in usage (according to age) of overt TOB MC preverbal marker does for the habitual The older speakers generally eschewed it, the highest percentage of production being 40% for any subset of older speakers. In contrast, the same form was dominant in the younger speakers' output. Rates of production were 55.5% and 42% for more educated young women and men and 70.6% and 65.8% for those with only post-primary education (up to age fourteen) . Indeed, it was the most marked feature across the board distinguishing older and younger speakers above any other feature.
The varllingual repertoire of Tobago Dian speakers 203 Infonnant 11: Young man of 19: (11) I have pigeons. I does take care a them. Ducks and rabbits an all like
that...Most of the evenings I does be here an chill out, sometimes go by girlfrien an talk a little bit. I keep pigeons. I take care of them. Ducks and rabbits and things like that. Most evenings I'm right here, hanging out; sometimes I go to my girlfriend's home and talk for a while. The further into the study we progressed, the more it became clear that older speakers generally appeared to disdain the mesolect, several of them referring to it as a corruption of the Standard which they perceived as having been imported from Trinidad in the course of the twentieth century. The basilect and the acrolect are regarded by older speakers as full languages, while the mesolect, although they use at least its less overtly marked forms like rA + V-in(g), is regarded as less than a full language. The young people in contrast focused usage on this same mesolectal variety and favoured it across the board, above both the acrolect and basilect varieties. For them, it seemed to be a significant variety in its own right whose overt features were markers of identification.
5·
Summary and conclusions
To summarize, it is clear that young people made greater use of Tob MC than older speakers, favouring rA + V-in(g) more than 90% of the time above be+ V-in(g) and also favouring does above Tob S Simple Present for habitual marking (Youssef 2001: 40-41). As noted previously, this generation has positive values for the mesolect and shows some evidence of regarding the basilect as "ole people talk" though still using it in private contexts. The more marked features of the mesolect such as does served as a badge of identity for the young people, characterizing their speech across the different educational levels. This greater cleaving to the mesolect is the most marked difference between young and old. The acrolect was marginal for the younger speakers, including those in Sixth Form, and in contrast to the older educated speakers who used it as their preferred variety and with a greater range and fluency. It is important to recognize, however, that for those young people who would go on to tertiary education and the professions, it would undoubtedly develop. The contrast between varieties is, for that group, necessarily overstated. Indeed, while there was more diversity of speech among the older speakers than the younger taken as a whole, their greater life experience must be key here to their greater linguistic diversity. Features such as further education, extensive travel outside Tobago, and network structure
204
Valerie Youssef
made for much more distinctness among the subgroups. When the figures were scrutinized according to background factors, it emerged for example that the only older group to use does extensively was that group who had spent substantial time in Trinidad. A majority usage of one variety, and a favouring of it, does not in any way preclude the overall varilingual competence we are describing: the balancing of varieties goes on in relation to appropriateness needs as well as ideological or identificational ones. Each variety has a symbolic value which is useful to call on according to circumstance. The concern of this paper is not just the trends in themselves, interesting as they are, but also the need for some degree of modification of methodologies in assessing the directionality of change in Caribbean sociolinguistic complexes. It is clear that amplification of the focus on the individual, very precisely described and defined sociolinguistically, and ranged as a member of a sub-group in the community, makes for greater insight overall at both individual and societal levels. Further, new norms are emerging, which are subtler than simply a decreolizing shift. If the mesolect is the more preferred variety of the younger rather than the older speakers, this is not simply evidence of a decline in use of the basilect but also a decline in use of the acrolect as a pure form in the context of some re-definition of the Caribbean self, as well as association with Trinidad and the wider Caribbean region. The shift away from Standard may be perceived in this context as significant, registering an outward focus societally, but an outward focus redefined in Caribbean terms of advancement rather than by association with the former colonial other. Unless the acrolect can attain some real relevance in the context of the younger people's lives and identities this marginality is likely to continue, save in the case of those who attain tertiary education. In this context, it is also useful to reconsider individual and collective competence in terms of varilingualism, insofar as that model recognizes the capacity for individual competence to be at least hi-varietal given the co-existence of two or three societal codes, each with separate and distinct social-psychological and ideological values. The model sees bi-varietalism not as a transitional phase but as a long-term condition in a society where dual codes do not so much compete as collectively represent a complex, ever-evolving multidimensional identity.
References Bickerton. D. 1975. Dynamics of a Creole System. Cambridge: CUP. Carrington, L 1976. Determining language education policy in Caribbean sociolinguistic complexes. Internatio11al Journal of the Sociology ofLanguage 8: 27-44.
The varllingual repertoire of Tobago Dian speakers Carrington, L. 1993. Images of creole space. Journal of Pidgin and Creole Languages 8: 227-236. DeCamp D. 1971. Toward a generative analysis of a post-creole continuum. In Pldglnlzatlon and Creolization of Languages, D. Hymes (ed), 349-370. Cambridge: CUP. Labov, W. 1972. Sociolinguistic Patterns. Philadelphia PA: University of Pennsylvanla Press. James, W. 1997. Students' TAM Errors ln the Context ofthe Speech ofTobago. PhD dissertation, University of the West Indies, St Augustine, Trinidad & Tobago. James, W. & Youssef, V. 2003. The Languages of Tobago. Trinidad & Tobago: School of Continuing Studies. Tobago House of Assembly. 1998. Tobago Development Plan, 1998-2000, Report Number 4, Scarborough, Tobago. Youssef: V. 1996. Varllingualism: The competence underlying code-mixing in Trinidad & Tobago. Journal ofPidgin and Creole Languages 11(1): 1-22. Youssef; V. 2001. Age-grading ln the Anglophone Creole of Tobago. World Engllshes 20(1): 29-46. Youssef: V. 2005. Unmasking ideology through language. Paper presented at the Eighth Annual Eastern Caribbean Islands- in- Between Conference, Tobago, November 10-12, 2005.
205
On the emergence of new language varieties The case of the Eastern Maroon Creole in French Guiana Bettina Migge & Isabelle Leglise
Creoles are traditionally asswned to lack stylistic depth. Researchers recognize a basilectal variety and/or a mesolect, which is generally assumed to result from contact between the basilect and a European language. While the historical foundation of this model has been much called into question, the sociolinguistic premises have received relatively little attention. This paper addresses this issue by exploring CWTent sociolinguistic practices in the Eastern Maroon community based on data from participant observation, findings from a questionnairebased language survey, and an analysis of natural recordings in French Guiana. The paper demonstrates that the Eastern Maroon Creole has distinct regional and stylistic varieties that play an important role in the identity politics of the community. Due to ongoing social change, they are undergoing social and linguistic changes, and new styles of speaking are emerging. Change is by no means unidirectional and is conditioned by a range of social forces.
1.
Introduction
Historical linguistic research has traditionally assumed that new dialects emerge gradually as the result of the spread oflanguages due to the migration of part(s) of their speakers to new locations or due to the relative isolation of part of the population in relatively geographically inaccessible locations such as mountainous areas. Social dialectological work (Labov 1963; Britain 2002) has also identified factors such as negative stereotypes, local rivalry, and the absence of public transportation as causing or enhancing dialect divergence. Finally, the bulk of sociolinguistic research has strikingly demonstrated that social factors such as class or social group membership, age, ethnicity, etc. play an important role in constraining patterns of interaction and thereby contribute to the divergence of dialects/varieties.
2o8 Bettina Migge & Isabelle Leglise
While dialectological and traditional sociolinguistic research on dialects implicitly or explicitly maintains that linguistic differentiation comes about due to gradual language-internal processes of change, language contact is typically invoked as a prime factor in the emergence of Diaspora varieties of a language. Cases in point are Diaspora varieties of Hindi (e.g. Siegel1988, 1990, 1997; Mesthrie 1991) and (new) varieties of English (cf. Kortman and Schneider 2004). Siegel for instance, has identified the following kinds of processes besides independent language-internal development as having played a role in the emergence of new varieties of Hindi (e.g. Trinidad Hindi, Mauritian Hindi, Guyanese Hindi): Dialect mixing: mixing of features from different regional and local varieties. Formal simplicity: regularization and reduction of categories and loss of inflections. 3. Dialect levelling: loss of input dialect features due to selection of equivalent features from other varieties. 4. Focussing: stabilization of a new variety based on the input varieties; sometimes strongly based on a majority variety. 1.
2.
A range of other processes of contact-induced change such as borrowing, convergence, L2 acquisition, and substratum influence (cf. Thomason and Kaufman 1988; Winford 2003) have been invoked as having played an important role in the creation of (new) varieties of English that arose in bilingual and multilingual contact settings (e.g. Singlish English, Irish English). With respect to creoles, it is widely believed that they have few internal varieties. The linguistic literature generally refers to the Creole and does not narrowly define the variety of creole they are dealing with; at most, we find a brief characterization of the social and regional background of the speakers from whom the data were drawn. There is relatively little research on the sociolinguistic makeup of creoles. For some settings, e.g. Haiti (Fattier 1999) and Jamaica (DeCamp 1971), it has been demonstrated that there is regional variation, but little is still known about how speakers conceptualize this variation and what role it plays in everyday interactions and in local identity politics. Regional differences are generally argued to be due to (partial) differences in the nature of the linguistic input during Creole formation (e.g. relative importance of regional European varieties, African languages, and other Caribbean Creoles), the past and current makeup of the population, and the patterns of interaction between the different population groups (cf. Winford 1997). In terms of social and stylistic varieties, some Creoles are described as more or less mono-stylistic (e.g. Belize). Essentially. it is argued that the Creole, which is used in informal settings, is in contact with the official language, e.g. English, which is used in formal contexts. While it is acknowledged that there is a certain
On the emergence of new language varieties
amount of interaction between the Creole and English, this 'zone' of interaction is argued to not constitute a separate variety. Other settings (e.g. Guyana) are described as consisting of two main Creole varieties. the so-called basilect and mesolect. The basilect is the variety that is structurally clearly distinct from the European input language and typically associated with rural populations. In contrast, the mesolect is associated with urban populations and is described as being linguistically intermediate between the European input language and the basilect. The traditional view (Bickerton 1975) maintains that mesolects emerged when speakers of the basilect gained greater access to the official language in the post-emancipation context; they allegedly borrowed features from English and integrated them into their basilectal Creole variety. Socio-historical and historical linguistic research on Caribbean Creole communities has however challenged this view, arguing that Creole societies were never socially and linguistically homogeneous. From the beginning, different social groups of slaves spoke different varieties of Creole because of different patterns of interaction and different degrees of access to English, African languages, and other contact varieties. According to this view, modern (mesolectal or basilectal) varieties essentially emerged due to processes of contact and linguistic focusing from these earlier varieties (Winford 1997). Qualitative research on variation in Creole communities demonstrates that Creole speakers do not only recognize the two (Creole and official language) or three varieties (basilect, mesolect, and official language) posited by quantitative sociolinguistic research. The members of Creole communities strategically and creatively draw on these varieties in order to construct individual and group social identities and social relationships. This has led to the emergence of new varieties that are associated with partially distinct social entities (e.g. social groups, settings) (Reisman 1970; LePage and Tabouret-Keller 1985; Garrett 2000). It then appears that Creoles, like other languages, have considerable internal complexity. This complexity seems to have emerged due to different kinds of contact processes. However, to date, little is lmown about the sociolinguistic structure of any one Creole and the social and linguistic processes that played a role in its emergence and maintenance in the past or that affect its makeup in the present. The aim of this paper is to investigate these issues in relation to the Eastern Maroon Creoles (EMCs) of Suriname and French Guiana. The discussion suggests that contrary to common assumptions, the speakers of these Creoles traditionally recognize a range of social, regional, and ethnic varieties. In addition, new varieties and practices continue to emerge, most likely spurred by the social changes that have been affecting these communities in the last 30 years. The data for this study come from a range of sources. Part of the data come from long-term participant observation mainly in one of the three Eastern Maroon (EM) communities, semi-guided discussions with EMs, and recordings of natural
209
210
Bettina Migge & Isabelle Leglise
interactions among and with EMs. The other data come from semi-guided interviews and discussions with members of the multicultural French Guianese society, a survey of school children in French Guiana aimed at obtaining linguistic practices, attitudes, and recordings of natural interactions in different settings such as at the marketplace, at work, within the family, among friends, etc. (see Leglise 2005, 2007). The paper is organized as follows: Section 2 presents the current social and linguistic context of French Guiana. Section 3 discusses the traditional sociolinguistic structure of the Surinamese Creole linguistic space according to the main group of native speakers, EMs. Section 4 investigates the social and linguistic changes that have been affecting the EM linguistic space in recent years, and Section 5 discusses the findings and their implications.
2.
Social and linguistic situation in French Guiana
French Guiana, located in South America, is a highly multilingual overseas department of France. Besides French, the official and ex-coloniallanguage, about 30 languages are spoken. It is relatively difficult to obtain precise figures on the numher of speakers for each language, including French, because French censuses do not record ethnic and linguistic information. However, available information suggests that about 20 of these roughly 30 languages are spoken by between 1% and 30% of the total population (Leglise 2007). Officially, and for most researchers working in the region, languages are subdivided into so-called indigenous (e.g. Amerindian languages and some Creoles such as Creole Guyanais) and immigrant languages (e.g. Haitian Creole, Hmong). But for most languages, this distinction proves problematic (Leglise 2004) because the region has been subject to several waves of migration over the last tew centuries. The current situation is, however, of particular interest because the scale of migration has increased considerably over the last 30 years: the 1999 national census figures show that more than half of the population was born outside of French Guiana. The following kinds oflanguages are currently spoken in French Guiana: six Amerindian languages (e.g. Arawak, Kali'na) four French-based Creoles (e.g. Haitian Creole, Creole Guyanais), five English-based Creoles (e.g. Sranan Tango, Ndyuka/Nenge(e) and see below), Hmong (an Asian language), and other official languages of the Caribbean and Amazonian region such as Brazilian Portuguese, Dutch, English, and Spanish. In this paper, we will focus on the Western part of French Guiana and specifically on the situation of five related English-lexified Creoles that originally emerged on the plantations of Suriname (Migge 2003).
On the emergence of new language varieties
Aluku, Ndyuka, and Pamaka are associated with three independent Maroon communities of the same name. They are the first languages of the members of these communities who have either been residing in French Guiana for more than two centuries or are recent migrants from the interior of Suriname. Based on sociohistorical and comparative linguistic data, linguists argue that they are dialects of a common language called Nenge(e), Businengetongo (Goury and Migge 2003) or Eastern Maroon Creole (Migge 2003). The fourth variety, Saamaka, is usually described as an English-based Creole with a significant proportion of Portuguese lexical items (Queixal6s 2000). It is associated with an ethnic group by the same name. Finally, Sranan Tongo is the mother tongue of the descendants of slaves who did not flee the plantations of Suriname. It also serves as a lingua franca in multiethnic Suriname (Carlin 2001). Most linguists in French Guiana argue that itis not spoken natively in French Guiana and is not part of the linguistic landscape (Queixal6s 2000; Goury 2002). A sociolinguistic survey (Leglise 2004, 2005), carried out over the last seven years, provided further insights into the current situation of these Creoles in French Guiana. First, it revealed that Nenge(e), essentially the numerically dominant EM variety Ndyuka, is not only practiced as a native language but appears to also be learned as an L2 by school children who are not ethnically Ndyuka. The latter children employ it to interact with Ndyuka friends in the school yard (Leglise 2004, 2005). Attitudes towards Nenge(e) among the whole population are quite contradictory though. It is often described negatively as being "the language of migrants" and considered to lack prestige. However, the fact that it is widely used as an L2 suggests that it is becoming a regional lingua franca especially in the town of St. Laurent du Maroni and among school children (Leglise 2004). According to Leglise (2007), about 30% of the population speak Nenge(e) as L1 or L2. Second, the expression Saramaka! functions as an insult in the school yard. It is associated with backwardness. Speakers of Saamaka tend to disguise their ethnic and linguistic background by declaring to be speakers of Nenge(e) rather than Saamaka and by employing Nenge(e) as their main means of communication, especially in inter-ethnic settings (Leglise and Puren 2005). While the Saamaka are, according to Price (2002), numerically the largest Maroon group in French Guiana, their children only make up 5% of the school population in the western part of French Guiana (Leglise 2005). The main reason for this seems to be that most Saamaka reside in remote rural locations that are badly connected to the main urban areas where most of the schools are located. Another explanation is that, due to recent waves of migration after the civil war in Suriname, they do not constitute the largest group anymore. Finally, despite frequent claims to the contrary, the survey found that Sranan Tongo is spoken as a mother tongue in French Guiana by both so-called indigenous
211
212
Bettina Migge & Isabelle Leglise
and immigrant populations. It is the mother tongue and main community language of a small Amerindian group residing in St. Laurent, the Arawaks (Leglise and Puren 2005). They became speakers of Sranan Tango due to a process oflanguage shift that started roughly 60 years ago. Sranan Tango is also widely used as a means of inter-ethnic communication. Its vehicularization rate (see Table 1) is not very high among school children possibly because it is only learned in adolescence or adulthood (Leglise 2004), and most children in French Guiana may not be able to distinguish between the contemporary urban Maroon varieties and Sranan Tango (Leglise and Migge 2005). In fact, a "sort of Sranan Tango:' locally referred to as Takitaki, is widely practiced by non-Maroons in the Western part of French Guiana (Leglise and Migge 2006). From the findings of the sociolinguistic survey, it then appears that the contemporary structure of the Creoles of Suriname in French Guiana is much more complex than described in the literature. In the next section, we further investigate the sociolinguistic structure of these Creoles by exploring the (traditional) linguistic ideology of the main native speaker group, the EMs. Table 1 illustrates the vehicularization rates - meaning its importance as a vernacular or means of communication in a region - for different languages spoken in the biggest town of western French Guiana, St. Laurent du Maroni. For each language, the absolute vehicularization rate is calculated, following Dieu and
Table 1. The vehicularization rates of different languages spoken in St. Laurent duMaroni (Uglise 2004) Percentage of speakers (ll-L4)
Weight of the language asLl
Vehicularization Vehicularization absolute weighted rate rate/amount of speakers
Brazilian Portuguese
4.2%
2.5%
1.7
0.07
Haitian Creole Dutch Sranan Tongo Creole Guyanais Kalfna Ndyuka Swn of varieties of Nenge(er
5.4%
4.7%
1.1
0.06
5.4%
2.4%
2.2
0.12
19.6%
8.3%
2.3
0.46
13.7%
1.8%
5.7
0.78
5.3%
3.6%
1.5
0.08
57.1%
39.9%
1.4
82.1%
56.1%
1.5
0.81 1.27
* Swn of varieties of Nenge( e) means the figures for all three varieties (Aluku, Ndyuka. Pam aka) combined.
On the emergence of new language varieties
Renaud (1979), by dividing the percentage of speakers within a town or a country by the percentage of L1 speakers. The weighted vehicularization rate, following Calvet (1993), gives more precise insights. It is calculated by dividing the vehicularization rate by the total number ofL1 speakers. Table 1 shows that the Englishbased Creoles or varieties of Nenge(e) - Aluku, Ndyuka, Pamaka -and among them Ndyuka are most widely spoken in St. Laurent as they have the highest vehicularization rates. Creole Guyanais, the former main lingua franca of French Guiana, and Sranan Tonga, the lingua franca of Suriname, have a lower rate of usage. Other languages, e.g. the Amerindian language Kali'na and Brazilian Portuguese, are little used among school children in St. Laurent.
3·
The (traditional) native view of the EM linguistic space
Eastern Maroons employ a range of language terms that refer to difterent locally recognized varieties. Traditionally, they differentiate between varieties that are associated with local ethnic groups (i.e. regional varieties) and those that are associated with specific settings (i.e. registers) or social groups (i.e. social varieties). Below we discuss each in turn. 3.1
Ethnic or regional varieties
EMs make reference to five EM varieties, namely Aluku, Ndyuka, Pamaka, Kotika, and Saakiiki. The first three varieties are associated with independent EM communities by the same name. The members of the Kotika and Saakiiki communities are members of the ditrerent upriver Ndyuka lineages who came to settle along the lower Maroni/Maroweijne River and the Commoweijne River since the early part of the 20th century in search of what is often referred to as a 'better' life. Due to geographical separation from the upriver Ndyuka community and the significant size, particularly of the Kotika community; they have over the years come to be recognized as quasi-independent communities and their varieties as distinct from upriver varieties. However, to date, there are no studies that have systematically investigated the similarities and difference between upriver Ndyuka and Kotika varieties. The differences between the three main EMCs (Aluku, Pamaka, Ndyuka) are largely phonological and lexical in nature (e.g. Gouryand Migge 2003). 1 From a linguistic point ofview, they are relatively minor, but they function as important markers oflocal ethnic identities. For instance, sociallyvery significant phonological difterences There are also intonational differences between the three varieties, but they have not yet been investigated
1.
213
214
Bettina Migge & Isabelle Leglise between Aluk:u (AL) and Pam.aka (P) on the one hand and Ndyuka (ND) on the other hand involve the alternation between long and short word-final vowels in some lexical items (as seen in 1), the alternation in vowel height of word-final front vowels (as seen in 2) and the absence and presence of intervocalic liquids (as seen in 3). (1) word-final vowel length
PM/AL
ND
gloss
wata
wataa
nenge boso
nengee bosoo
water person/language brush
(2) word-final vowel height PM/AL ND gloss
meki teki
meke teke
make take
(3) Prealization of intervocalic liquids
PM/AL
ND
gloss
kali weli
kai wei
call wear
Beside the five EM varieties, EMs recognize four further ethnic varieties, namely Saamaka, Sranan Tonga, Matawai, and Kwinti. Saamaka, Matawai, and Kwinti are each associated with the three Maroon communities of the same name from the Western part of Suriname. Sranan Tango is the language of the urban population, particularly those of African descent often referred to as fotonenge 'the Blacks of Paramaribo (Foto )~ Although all Surinamese Creoles descend from the same plantation varieties (Migge 2003), they differ somewhat linguistically and are not fully mutually intelligible due to partially different linguistic developments. The varieties associated with Sranan Tango have been subject to relatively strong influence from Dutch while the varieties that developed into Saamaka (SM) and Matawai in particular were in close contact with Portuguese or Portuguese contact varieties during their emergence (Arends 1999; Smith 1999). The nature of the similarities and ditrerences between the Surinamese Creoles can be illustrated by considering the area of potential modality (Table 2). Briefly, Table 2 shows that in all the Creoles of Suriname, learned ability is expressed using the verb sa(bi). Differences are found with respect to the expression of other potential meanings, however. First, while the Maroon Creoles have a single potential category which is expressed using the marker sa, the urban Creole Sranan Tango distinguishes several different categories, namely physical ability, root and epistemic possibility, and permission, using distinct elements to convey them. Second, the EMCs ditrerentiate between positive and negative potential contexts while Saarnaka does not. Third, Aluk:u and Pamaka use man to convey negative potential modality while Ndyuka uses poy.
On the emergence of new language varieties 215
Table2. Future and possibility in the Creoles of Suriname (based on Migge 2006) Forms PM/ALND
Category SM
Meanings
SN*
Learned Ability
sabi
sabi
sa
sablfu
Ability or skllls acquired through learning or tralnl.ng. Potential
sa man
sa
poy
sa sa
man/ kan
Positive Negative
kan
mag
sa
sa
ka1!-
kan-
de
de
sa ka11 Positive kande ka11de Negative (sa)
Physical ability (Deontic) Abllity subject to physical or natural law. Deontic (root) possibility Ability/possibility subject to moral or social law, involving situations under the agent's control Permission Deontic possibility imposed by authority (social, legal, etc.). Epistemic possibility Possible situations, or situations to the certainty of which the speaker is not committed
Synchronic and diachronic linguistic and soda historical evidence suggest that the early plantation varieties were characterized by a relatively great amount of variation. With respect to the potential domain, it seems likely that at least several of the currently attested elements, e.g. sa, kan, man, and pay, coexisted in earlier varieties as means for expressing the same or partially overlapping potential modality meanings. Over time, each form became associated with different emerging social and ethnic groups. We could hypothesize that the distinction between sa and kan is related to a ruraVurban or Maroon/non-Maroon social differentiation since the Dutch-derived item kan continues, even today, to be strongly associated with an urban/town orientation. The distinction between man and pay conveys a socially pertinent inter-Maroon social distinction: the smaller Maroon groups were historically much dominated by the Ndyuka and therefore still like to differentiate themselves from it. 2 1he adoption of Dutch-derived elements such as mag is most likely due to the comparatively strong influence from Dutch on the varieties that developed into Sranan Tonga.
2.
Note also that poy has overtones of being 'deep; 'original,' and 'powerful.'
216
Bettina Migge & Isabelle Leglise
3.2
Social varieties
Traditionally, Eastern Maroons distinguish five broad social varieties:
1. 2. 3. 4. 5.
lesipeki taki 'respect speech' kowounu taki 'ordinary speech' basaa nenge(e) 'non-Maroon speech' keliki taki 'church speech' fositen nenge 'earlier talk'
Respect speech differs from ordinary speech both socially and linguistically. While ordinary speech is commonly identified with low status and everyday social interaction (e.g. informal chats among women and men, parent-child talk. talk accompanying subsistence work activities), respect speech is reserved for formal settings (e.g. talk among and with elders, socio-political meetings, delicate topics). Linguistically; respect talk is characterized by a range of negative politeness strategies, e.g. a special polite vocabulary to replace potential taboo terms, special status-indicating address terms, verbal indirectness, special turn-taking rules, etc. (Brown and Levinson 1987; Migge 2004). Ordinary speech is best described as conversational talk. It is characterized by positive politeness strategies including relative directness of expression, use of vulgar and taboo vocabulary and the relative absence of special turn-taking rules.
Basaa nenge is broadly associated with L2 varieties of the EMC. They can generally be differentiated from native talk on the basis of a range of structural differences, e.g. absence of certain functional elements, absence of allomorphs, etc. (Leglise and Migge 2006). Keliki taki is the language traditionally used in church books and during mass. It is a stylised, non-native variety of Sranan Tango used by the early missionaries. Foisten nenge is the language used by the early slaves and runaways. Today, it is heard in spirit possession ceremonies when the spirits of the early slaves and runaways communicate with present-day Maroons. It is also a stylised variety that resembles Sranan Tango in several respects. Maroons also talk about aftkaan tongo. The speech thus designated is in fact a group of three languages referred to as Popo, Kromanti and Loango in Suriname. They are based on African languages that were spoken by the early slave population, namely varieties of Gbe, varieties of Akan and varieties of Kikongo, respectively. These languages are no longer widely spoken among Maroons and appear to be mainly used on occasions involving ritual ceremonies. Further research is needed to clarify the use of these languages among Eastern Maroons. 3 3· Richard Price has discussed some issues in relation to the use of these languages in his works on the Saamaka Maroons.
On the emergence of new language varieties
4· The EMC migrates to the coast Since roughly the 1950s and particularly due to the civil war in Suriname in the late 1980s, members of all Maroon societies have increasingly been migrating to the regional urban centers in Suriname and particularly French Guiana. This has brought about the relative de-population of the traditional villages and has given rise to changes in the social and linguistic practices of the Maroon populations. 4.1
Changes in the linguistic repertoire and in-group linguistic practices
Traditionally, Maroons employed their variety of the EMC for all their communicative needs. Only a small number of men who regularly engaged with members from other communities for purposes of trade and/or who had spent some time doing cash labor on the coast acquired active competence in one of the main regional lingua francas, Sranan Tango or Creole Guyanais. Competence in the official languages of the region, Dutch (Suriname) and French (Guiana), was even less common among the rural Maroon population since formal schooling did not become widely available until the 1970s and was severely disrupted during the 1980s due to the civil war in Suriname. This situation changed, however, when Maroons started migrating permanently to the coast. In both rural and urban coastal contexts, they entered into more or less regular interactions with other members of the Surinamese and/or French Guianese multi -ethnic societies. For instance, they came to live in mixed neighborhoods, entered cash labor and subsistence work networks with members from different local communities, engaged in trading with members from other local ethnic groups, and attended educational and training institutions and social venues frequented by members from all local social groups. Linguistically, this shift in interactional patterns meant that Maroons had to expand their linguistic repertoires. They had to acquire at least a passive competence in the related urban Creole and regional lingua francas (Sranan Tonga or Creole Guyanais) to interact with the members of the other local populations. For those attending educational institutions and seeking work in more skilled jobs, it also became a necessity to acquire competence in either or both of the official languages (Leglise and Migge 2005). The expansion of the linguistic repertoire has also led to changes in individual and community linguistic practices. On the one hand, we observe a greater incidence of borrowing. There is a noticeable increase in the use of mainly lexical items from either Sranan Tango and/or Dutch and French to denote things that are relevant to people's life in the new urban context, e.g. terms related to administration, food, new habits, locations, etc.
217
218
Bettina Migge & Isabelle Leglise
On the other hand, we find a greater use ofSranan Tongo in in-group encounters. EMs increasingly employ code-mixing and code-switching to construct newly emerging social distinctions and meanings. Elders, for instance, employ codeswitching with Sranan Tongo to downplay the traditionally strong hierarchical nature of their social relationships with each other and with younger members of the community; by code-switching, they project a peer group or friendship-type relationship with their interlocutor. In Example 4, Kabiten Anton, one of the village/sublineage heads, wants to talk to one of the authors, Bettina, who was a bystander at a semi-formal meeting between the paramount chief (Gaaman) and another village/sublineage head, Kabiten Obi. Gaaman and Kabiten Anton are in their seventies, while Kabiten Obi is in his late fifties. To talk to Bettina, Kabiten Anton has to seek permission from Kabiten Obi and Gaaman. Kabiten Anton switches to Sranan Tongo, or a Pamaka saliently influenced by Sranan Tongo, when addressing Ka.biten Obi. By switching, he is invoking their personal rather than their positional identities (kabiten-hood) and emphasizing the lack of social distance between him and Kabiten Obi- Sranan Tongo is widely used in the multiethnic work contexts in which most of them would have worked during their 'youth'. In foregrounding his peer group or friendship-type relationship with Kabiten Obi, he conveys to Ka.biten Obi that his desire to talk to Bettina is non-threatening. He is asserting that he is not trying to compete over Bettina's attention. (4) Pikin kuutu 1 Anton: 2
3
Bettina: Anton:
Ma, daaa, fa de e kai a flou nen? But, so what's the woman's name? Bettina!
Soo Betna. (.) Okay, Bettina.
4
(to 0)
So, da (.) ka.biten, mi kan (ST) taki ptyin tori (ST) Okay, well kabiten, can (ST) I chat a little bit (ST)
anga a uman pikinso yere (ST)? A no (ST) wan mulikimuliki toli.
5
6 7
Obi: Obi:
with the woman, right (ST)? It's nothing (ST) bothersome. (nods approvingly and laughs)
Young people generallyemploycode-switching to construct themselves as sophisticated and urbanized Maroons and code-mixing to assert membership in the social group of young men whose salient properties are modern urban sophistication
On the emergence of new language varieties
(Migge 2007). Example 5 is an example of code-mixing. It comes from a meeting between Pamaka men in their mid-30s who head a local Maroon cultural group. They are discussing the activities of their group. Note that in this turn, the speaker is frequently alternating between Pamaka, the matrix language, and Sranan Tongo, Dutch and French forms. 4 (5) Yunkuman kuutu
B: lni a pisi toli san (ST) u taki fu a d.n sedusu (ST) u d.n libisama. In this story that we are talking about if we are not honest we are not human. A ini a pL~ ape mi o taki, mi o piki oo! Te yu nanga (ST) u an man e wooko I will say something about this part, I will respond! If you cannot work with us ma a de ini i konde oo, da i mu luku a wooko fini. but it [the event] takes place in your village, then you have to carefully consider the job.
Efu i lobi a waka dati, u an, da i o gwe go namo na. a wooko. If you like that kind of travel/event, we don't, then you'll definitely go and take up the job.
Di i sabi di i no sabi, i o gwe go na a wmko omdat (STID) i wa.ni teki a. wooko, Whether you know [the job] or not, you'll take up the job because you want to take up the job, a. de a. ini i sikin. Da.ti u, a. ini wa.n dey di i basi no (ST) de, i mu man du wan
sani, it is your desire. That we, every day that your boss is not there, you have to do something o ehee leki {a (ST) u e taki a toli fu den skoro a yari disi (ST). A yari san psa, (ST) Ahm yes, like we were talking about the schools this year. Last year, u luku a yari disi (ST). U akisi kon fu go ini wan skoro ma omdat (ST) a pamplia fu organisasi (D) no (ST) be herken door (D) lanti we were contemplating this year. We were asked to go to a school but because the organization had not been officially recognized by the government pe den ma.n fu education national (F) musu (ST) stm·t (D) a. moni gi u.
4·
Only one turn is provided since turns are very long with these kinds of meetings.
219
220
Bettina Migge & Isabelle Leglise
where the people of the [French] ministry of education have to send the money for us, den no (ST) man sabi pe den mu stort (D) en gi i pea sama meki they don't lrnow where they should send it for you, when the person a. poking (D) fu Awibenkiiben mi a be abi en. Den naki a.na gi en klopklop (ST/D), attempted to transfer to Awibenkiiben, I did not have it They clapped for him.
a. ma.n, ne en nen a e meki kaba. Dus (D) na so a dansi de, the guy, he's already promoting himself. Thus that's what professional dancing is like, na so wan~ (D) de, na so wan libi mu de. Kwolon! Mi na e taki moo. that's what a [dancing] group is like, that's the way it should be. Finish! I am not saying anything else.
42
Toward koineization
In the coastal context, the members of the difterent Maroon groups are in much more regular contact with each other and with members from other local social/ ethnic groups (e.g. Haitians, Amerindians, Creole Guyanais, metropolitan French) who are culturally and linguistically relatively different from them. Here, Maroons develop a different sense of ethnic belonging. Unlike previous generations of Maroons who generally emphasized inter-Maroon differences, in the current context, especially young Maroons foreground and emphasize the similarities that exist between the different Maroon groups. They construct themselves as members of the social or ethnic group of 'Businenge or Maroon: a pan -Maroon identity that transcends the traditional ethnic divisions. One piece of evidence in favor of the existence or emergence of such a panMaroon identity comes from the results of a school survey (Leglise 2004) where primary and secondary school children were asked by a metropolitan French researcher to discuss their linguistic practices. When discussing their linguistic repertoire, the children generally only referred to their native language using terms such as Takitaki and Businenge Tongo, which are strongly associated with a panMaroon identity. Language names associated with specific ethnic groups and identities (Aluku, Pamaka, etc.) were either only supplied on repeated questioning by the researcher or not at all (Leglise and Migge 2006).
On the emergence of new language varieties (6) During the interview Res: Quelle la.ngue tu pa.rlais ava.nt d'alla· l~cole? Which language did you speak before starting school? Child: Takitaki. Res: Lequel? Which one? Businenge. Child: Res: Oui mais lequel? Yes, but which one? Child: Ben Businenge, Takitaki cest pareil. Well, Businenge, Takitaki, it's the same.
a
Linguistically, this process of identity formation seems to be giving rise to processes that are also associated with koineization. According to Siegel (1985), koineization involves mixing of features from different related regional dialects, levelling of such features, formal reduction, and finally focussing of a new 'mixed' variety. Examples 4 and 5 suggest that dialect mixing involving Sranan Tango and the Maroon varieties has become rather widespread even in in-group encounters. We also find that processes of levelling are in progress. Essentially; ethnically or rurally marked linguistic features are increasingly levelled towards more 'neutral' forms. For instance, Maroons frequently remark on the fact that the down-river Ndyuka do not realize word-finallong vowels (ct~ Example 1) in the same way as up-river Ndyuka; the vowels in down-river Ndyuk:a are noticeably shorter though still different from those associated with Aluku and Pamaka varieties. The realization of very long vowels has become associated with a rural and traditional Ndyuka identity. Moreover, strongly ethnically marked morphosyntactic features such as the verbal negation marker (Ndyuka: d, Pamaka/ Aluku: an) and the negative potential marker (Ndyuka: poi, Pamaka, Aluku: man) are variably replaced with more ethnically neutral equivalents that usually come from Sranan Ton go, namely no and kan, respectively. Levelling does not only affect ethnically marked features but also those that are closely associated with the rural context. Greeting procedures are a case in point. The EMC has a number of different greeting procedures. Some require careful interplay between interlocutors. are rather formulaic, and are specific to particular times of the day (see Example 7). (7) Some traditional Maroon greeting procedures We have awoken, paramount chief. a. A: U weki oo, gaaman. B: Iya, u weki yee/baa. Yes, we have awoken! A: EeyalIya. Yes b. A: U miti oo, mma. We meet, [female] elder.
221
222
Bettina Migge & Isabelle Leglise
B: Iya, u miti (baka) yee!baa, papa. Yes, we meet (again), elder! A: Eeya/Iya.. Yes c. A: Dda, a tapu u (baka) oo. [Male] elder, it [the night] has fallen (again). (lit. 'it covers us again') B: Iya, a. tapu u yee. Yes, it has fallen! A: Eeya/Iya.. Yes d. A 1: Mma, da u de (mooi)? Mrs. G., then, are you well? (lit. 'then we/you (pl.) exist welf) B1: Iya, u de (mooi) yee/baa! Yes, I am well! (lit. 'we/you (pl.) exist nicely!) U seefi de (mooi)? You yourself, are you well? (lit. 'we/you (pl.) self exist nicely?') A2: Iya, u de (mooi) yee/baa. Yes, I am well! (lit. 'we/you (pl.) exist nicely for sure') B: Iya/Eeya. Yes (from Migge 2005) In actual practice, greetings 7a-c are generally combined with greeting 7d, though each of them can also be used by itself. While these greetings are regularly used in the rural context, they convey a relatively great social distance between the interlocutors, pay each other respect, and construct each other as respectable persons. There are also two main, shorter greetings that can be applied throughout the whole day. (8) Additional Maroon greeting procedures Mrs Moiboto, how are you? a. A: (Sa Moiboto),fa i tan? B: Saajlio/saafsaafi/Mi de. WelVI am well. b. A: (Baa Aseengi.), fa a ego? How is it going? B: Mi de/a e go/saafsaafi!saajlio I am well/as usual/well. These greetings, particularly 8b, is associated with the urban context and Sranan Tonga. These greetings convey social closeness and relative social equality between the interlocutors, essentially a friendship-type relationship. In the rural context, these greetings are only employed among young people who are all part of the lowest social category and tend to maintain relaxed kinds of relationships. In the urban setting, however, EMs of all generations and social groups increasingly use the greetings in Example 8 for regular everyday interactions. The longer greetings are increasingly being reserved for special occasions associated with situations that require heightened attention to negative face such as demands for help. complaints, formal events, strongly hierarchical relationships (e.g. with in-laws) (Migge 2005)
On the emergence of new language varieties 223
Finally; we also observe the emergence of new, mixed varieties. In recent years, in both French Guiana and Suriname, new radio programs are emerging due to grassroots efforts. These programs are broadcast in local languages and are targeting specific local ethnic population groups. A case in point is the radio program lowema.n paansu. It is broadcasted in and around St Laurent du Maroni (French Guiana) for two hours in the late afternoon Mondays to Fridays. It is produced on a voluntary basis by young, urbanized Maroon men with a Ndyuka ethnic background. The program primarily targets the extensive (Eastern) Maroon population of the area, but members of other ethnic groups (e.g. Amerindians that are familiar with the Surinamese Creoles) also listen to it It covers a range of topics that are of interest to all sections of the EM population such as presentation and discussion of local popular music, discussions of health issues, political matters, local news. birthdays, obituaries, etc. In particular, news programs and discussions oflocal socio-politicalissues are carried out in the EM respect variety (Migge 2004). However, there are a couple of salient ditrerences between the traditional respect variety and the one used on the air (Migge in press). First. the radio variety does not employ discontinuous speech and does not have a dialogic nature (i.e. it does not involve a ritual responder). Second, speakers make greater use of rurallymarked forms rather than the special respect vocabulary and figures of speech to encode negative politeness. Third, there is a greater use of foreign lexical items in the radio variety than in the traditional variety. In the latter, use of anything other than the local EM variety is considered problematic in that it is easily interpreted as a lack of alignment with the local community (Migge 2005). 4·3
Vehicularization
Especially since the civil war in Suriname in the late 1980s, Maroons in general and EMs in particular have emerged as one of the largest 'ethnic groups' in western French Guiana. Members of other local ethnic groups, including metropolitan French, have either felt obliged to or have desired to learn the EMC to integrate in some areas, such as the west of French Guiana, where EMs are numerically very dominant (Leglise 2004; Thurmes 2007). To a small extent, these L2 practices are also used for communication between non-Maroons, e.g. members of ditrerent Amerindian communities, Hmongs, etc. Non-native or L2 practices are quite diverse, ranging from relatively reduced learner varieties to near-native-like practices (Leglise and Migge 2006). Generally, L2 varieties differ in two main respects from native EM practices. First, they show varying degrees of structural reduction and/or variation that is mostly not found in native practices. For example, relatively acoustically imperceptible functional elements whose functions can also be inferred from the context are either
224 Bettina Migge & Isabelle Leglise
completely or variably absent (e.g. the imperfective marker e and the future marker o in Example 9), or are replaced by elements from another language (e.g. the conditional element efu which is replaced by French si in (9)). Inherently variable forms tend to be regularized to a perceptually salient form. For instance, while the 2nd person singular pronoun in the EMC is realized either as i, y or yu depending on the phonological environment, it is generally realized as yu in L2 varieties (Leglise and Migge 2006). (9) Si no 0 teki dresi, yu dede mama. (L2) na e teki/diingi den deesi, i o dede mama. (EM) efu i if you NEG IMP take/drink DET tablets you FUT die female elder 'Grandma, if you don't take your medicine, you may die: Second, L2 speakers tend to mostly select Sranan Tongo lexical items rather than EM ones in those cases in which the two differ, despite the fact that most L2 speakers would mainly interact with Maroons. The relative preponderance of Sranan Tongo-derived lexical items is most likely due to the fact that EMs tend to shift to Sranan Tongo in all kinds of public out-group contexts. Moreover, Sranan Tongo figures prominently in urbanized speech by young (male) Maroons (see above). Socially. the most immediate effect of its increasing acquisition and use by non-Maroons is the maintenance of the language. From the point of view of the native speakers, its acquisition by non-natives contributes to the social valorization of the language as a valuable means of communication. This, in turn, contributes to its maintenance within the native speaker communities. Linguistically, depending on the overall frequency and importance of exchanges with non-native speakers, its use as an L2 variety may further reinforce patterns of variation and change that are already in progress in native practices. For instance, the increased use of lexical elements from Sranan Tongo and simple as opposed to complex prepositional phrases in interactions as a way of accommodation to non-Maroons may lead to a further establishment of their use in intra-Maroon practices.5
5·
The role of social and linguistic processes in language diversification
The developments discussed suggest that linguistic diversification in the case of the EM linguistic space is due both to processes of social and linguistic convergence and divergence. Convergence involves processes that lead to the reduction or 5· Complex prepositional phrases: na NP locational N (11a a tafa tapu 'on the table'); simple prepositional phrases: P NP (tapu a tafa; na tapu a tafa)
On the emergence of new language varieties 225
elimination of social and linguistic differences, while divergence refers to processes that bring about the creation of social and linguistic differences or contribute to their emphasis. Social processes such as urbanization (i.e. displacement and reorientation in a new environment), new identity formation (i.e. pan-Maroon identity), and emergence of new interactional patterns (i.e. increased out-group interaction with L2 speakers and native speakers of related varieties) lead to convergence. Socially, they bring about new contexts of interaction, practices, and inter-group relationships that transcend or cross-cut traditional ethnic boundaries. Linguistically, they give rise to contact and mixing between existing varieties, the leveling of differences between them, and possibly to the eventual emergence of a new 'compromise' variety such as a koine that would be added to people's existing repertoire. Two kinds ofleveling processes seem to take place:
In interactions between speakers of different native varieties (Aluku, Ndyuka, etc.), speakers tend to level marked differences between their varieties (e.g. relative shortening oflong vowels in Ndyuka, adoption of Sranan Tongo forms to replace ethnically-marked morphosyntactic differences) towards a common ethnically 'neutral' norm. 2. With respect to interactions between native and non-native speakers, L2learners only acquire the most widely/frequently available practices while native speakers tend to select only those practices that they deem to be widely/easily understood (e.g. the use of Sranan Tongo vocabulary items). 1.
In French Guiana (and Suriname), these processes operating in native-native encounters and in native-non-native encounters happen to be linked through local language ideologies to similar kinds of linguistic practices. For EMs, relative social neutrality in interethnic encounters and simplicity or accommodation in interaction are encoded by drawing on Sranan Tongo -associated practices. This then suggests that these difterent social and linguistic processes lead to similar linguistic outcomes: they reinforce the use of Sranan Tongo practices. In contrast, the assertion of various individual (e.g. respectable), traditional (e.g. Aluku, Pamaka, etc.), and new group (e.g. sophisticated urban EM) identities gives rise to patterns of social and linguistic divergence, differentiation, and fragmentation. Essentially, in accordance with their goals, speakers draw in selective ways on the locally available linguistic and social resources, including ethnically neutral forms, to construct unique individual and group identities. Linguistically, this leads to a kind of reorganization and linguistic instability that over time develops into stable variation and code-switching and gives rise to the emergence of new styles.
226 Bettina Migge & Isabelle Leglise
5·
Conclusion and implication
The discussion in this paper suggests the following things about the development of the Surinamese Creole linguistic space. First, on the microlinguistic level, we showed that speakers creatively and strategically draw on different varieties and emblematic markers in accordance with a range of social factors such as the communicative situation, the interlocutors, their self-positioning, their goals, etc. All of these constitute good examples of individual agency. While these are the activities of individuals, they may lead to the expansion or diversification of the community's linguistic repertoire through social and linguistic focusing (Le Page and Tabouret-Keller 1985). Cases in point are the emergence of the new styles that we discussed. The individuals' mixing of native EM linguistic practices with external ones (e.g. from Sranan Tango) has led to the emergence of the EM style wakaman taki, and the blending of two different native styles ( wakaman taki + lesipeki taki) is leading to the formation a new formal style (radio talk). These new styles are added to the existing styles and classic language varieties/styles (cf. the social and ethnic varieties discussed in Section 3). This suggests that the EM linguistic repertoires are open ended and non-static; new styles or varieties are continuously added to existing ones. Second, Creoles, like other languages, involve different types of variation. Taking a dia- model of variation (Coseriu 1956; Oesterreicher 1988; Gadet 2003), our case shows diatopic (geographic) and dialectal variation, and diachronic and diaphasic (stylistic and register) variation. As for diastratic variation (due to social factors and social stratifications), our case shows the same results as for diaphasic variation. We assume that this is due to the social structure of the EM societies: their social structure is less based on social groups. in which a variety could be isolated, than on social events (such as kuutu, Migge 2004) and ways of addressing the elderly, etc. Third, on a macrolinguistic level, we observed linguistic variation and change and the emergence of new varieties: 12 varieties and new dialect varieties through what could be a process ofkoineization. The findings from our investigation challenge two of the main tenets of the Creole continuum theory. First, it examines the assumption that changes in the linguistic repertoire of a Creole community and in the Creole's structure are largely due to contact with an external variety or erstwhile lexifier, e.g. English. Our case shows that changes may also occur due to contact between different varieties of the same language induced by a range of social forces and by contact with a related language, Sranan Tonga. The influence of the official languages of the region (French and Dutch) is relatively minor despite their overall symbolic and economic importance. Second, this study questions the assumption that linguistic
On the emergence of new language varieties 227
change is unidirectional. Our study showed that di:trerent social forces give rise to two kinds of (contradictory) linguistic developments. On the one hand, we found a reduction in diversity (e.g. through levelling and the reduction of the differences between ethnic varieties in favor of an emergent pan-Maroon variety) and on the other hand, we observed an increase in diversity (e.g. with the emergence of new EM styles, new non-EM styles, and new linguistic structures). Moreover, the present case seems to be an instance of a mix of two classic cases of dialect contact: contact involving L1 and 12 speakers on one hand and contact among L1 speakers of di:trerent dialects on the other (Kerswill and Williams 2000). The investigation of the Surinamese Creole linguistic space makes it quite clear that current models of the Creole continuum are not applicable in situations where "the" Creole is no more in contact with its erstwhile lexifier. Would some sort of a continuum of practices or a continuum of varieties be a more relevant model? Or would maybe a complex (or a galaxy) of practices and varieties, organized by social forces and always in action, be a more suitable model to represent the dynamics of variation?
References Arends, J. 1999. The origin of the Portuguese element ln Surinam Creoles. In Spreading the Word: The Issue of Diffusion among the Atlantic Creoles, M. Huber & M. Parkvall (eds.), 195-208. London: University of Westminster Press. Bickerton, D. 1975. Dynamics of a Creole System. Cambridge: CUP. Britain, D. 2002 Space and spatial diffusion. In The Handbook ofLanguage Variation and Change, J.K. Chambers, P. Trudglll &N. Schilling-Estes (eds.), 603-637. Oxford: Blackwell. Brown, P. & S.C. Levinson. 1987. Politeness: Some Unl1'ersals In Language Usage. Cambridge: CUP. Calvet, L-J. 1993. Veb.icularite, veb.icularisation InLe Fran~ais dans ll;space Francophone, Tome 1, D. de Robillard, M. Beruamino & C. Bavoux (eds.), 431-449. Paris: Champion Carlin, E.B. 2001. Of riches and rhetoric: Language in Suriname. In 20th-Century Suriname: ContinuiNes and DiscontinuiNes in a New World Society, R. Hoefte & P. Meel (eds. ), 220-243. Kingston & Leiden: Ian Randle Publishers & KITLV Press. Coseriu, E. 1956. La Geografia Lingillstlca. Montevideo: n.p. DeCamp, D. 1971. Toward a generative analysis of a post-creole speech continuwn. In PidginlzaNon and Creollzatlon ofLanguages, D. Hymes (ed.), 349-370. Cambridge: CUP. Dieu, M. & Renaud, P. 1979. A propos d'une etude statistique du multilinguisme au Cameroun: Quelques problemes methodologiques. In Plurilinguisme, Normes, Situations, Strategies, P. Wald & G. Manessy, (eds.), 61-99. Paris: I.:Harmattan. Fattier, D. 1999. Contribution a l'Etude d'un Creole: llltlas LingulsNque d"Haiti, Cartes et Commentaire.s. Paris: Presses universitaires du Septentrion.
228 Bettina Migge & Isabelle Leglise
Gadet, F. 2003. Theoretical discussions in progress: Is there a French theory of variation? InterJ~ational Joun1al of the Sociology ofLanguage 160: 17-40. Garrett, P. 2000. 'High' KweyOl: The emergence of a formal Creole register in St. Lucia. In La11guage Cha11ge and Language Co11tact in Pidgins and Creoles, J. McWhorter (ed), 63-102. Amsterdam: John Benjamins. Goury, L. 2002. Pluralite linguistique en Guyane: Un apen;:u. Ameri11dia 26/27: 1-15. Goury, L. & Migge, B. 2003. Grammaire du Nengee: Introduction aux La11gues Aluku, Ndyuka et Pamaka. Paris: Editions IRD. Kerswill, P. & Williams, A. 2000. Creating a New Town koine. Language i11 Society 29: 65-115. Kortmann, B. & Schneider, E. (eds.). 2004. A Ha11dbook of Varieties of English: A Multimedia Rejere11ce Tool. Berlin: Mouton de Gruyter. Labov, W. 1963. The social motivation of a sound change. Word 19: 273-309. Leglise, I. 2004. Langues frontalieres et langues d'immigration en Guyane Fr~aise. Glottopol 4: 108-124. Leglise, I. 2005. Contacts de creoles aMana (Guyane frans:aJ.se): Repertoires, pratiques, attitudes et gestion du plurllinguisme. Etudes Creoles 28: 23-57. Leglise, I. 2007. Des langues, des domaines, des regions: Pratiques, variations, attitudes linguistiques en Guyane. In Pratiques et Attitude.s Linguistlques en Guyane: Regards Croists, I. Leglise & B. Migge (eds.), 29-48. Paris: IRD Editions. Leglise, I. & Migge, B. 2005. Contacts de langues issus de mobilites dans un espace plurilingue: Approches croisees aStLaurent du Maroni (Guyane). In Mobilites et Contacts de Langues, C. Van denAvenne (ed), 75-94. Paris: L'Harmattan. Leglise, I. & Migge, B. 2006. Towards a comprehensive description oflanguage varieties: A consideration of naming practices, ideologies, and linguistic practices. Language In Society 35: 313-339. Leglise, I. & Puren, L. 2005. Usages et representations linguistiques en milieu scolaire guyanais. In Ecole et Education, Univers Creoles, Vol 5, F. Tupin (ed), 67-90. Paris: Anthropos. LePage, R. & Tabouret-Keller, A. 1985. Acts of Identity: Creole-based Approaches to Language and Ethniclty. Cambridge: CUP. Mesthrie, R. 1991. Language in Indenture: A Sociolinguistic History of Bhojpuri-Hindiin South Africa. London: Routledge. Migge, B. 2003. Creole Formation as Language Contact: The Case of the Suriname Creoles. Amsterdam: John Benjamins. Migge, B. 2004. The speech event kuutu in the Eastern Maroon community. In Creoles, Contact, and Language Change: Linguistic and Social Implications, G. Escure & A. Schwegler (eds.), 285-306. Amsterdam: John Benjamins. Migge, B. 2005. Greeting and social change. In Politeness and Face in Caribbean Creoles, S. Miihleisen & B. Migge (eds.), 121-144. Amsterdam: John Benjamins. Migge, B. 2006. Tracing the origin of modality in the creoles of Suriname. In Structure and Variation in Contact Languages, A. Deumert & S. Durrleman (eds.), 29-59. Amsterdam: John Benjamins. Migge, B. 2007. Codeswitching and social identities in the Eastern Maroon community of Suriname and French Guiana. Journal of Sociolinguistics 11(1): 53-72. Migge, B. in press. Negotiating social identities on an Eastern Maroon radio show. Joun1al of Pragmatics (Special issue on postcolonial pragmatics). doi:l 0.1 016/j-pragma201 0.06.021
On the emergence of new language varieties 229
Oesterreicher, W. 1988. Sprechtiitigk.eit, Einzelsprache, Diskurs und vier Dimensionen der Sprachvarietiit In Energeia u11d Frgo11: Sprachllche Variation, SprachgEschichte, H. Thun (ed.), 355-386. 'llibingen: Narr. Price, R 2002. Maroons in Suriname and Guyane: How many and where. New West b1dia11 Guide76: 81-88. Queixal6s, F. 2000. Les langues de Guyane. In As Linguas Amazo11icas Hoje, F. Que!xal6s & 0. Renault-Lescure (eds.), 299-306. Sao Paulo: IRD-ISA-MPEG. Reisman, K. 1970. Cultural and linguistic ambiguity in a West Indian village. In Afro-America11 A11thropology, N. Whitten &J. Szwed (eds.),129-144. New York NY: The Free Press. Siegel, J. 1985. Koines and koineization. LanguagE in Society 14: 357-378. Siegel, J. 1988. The development of Hindustani. In LanguagE Tra,ISJllanted: The Developme11t of Overseas Hindi, RK. Barz & J. Siegel (eds.), Wiesbaden: Harrassowitz. Siegel, J. 1990. Pidgin Hindustani in FijL In Pacific Islands Languages: Essays in Honour of G. B. Milner, J. Davidson (ed.), 173-195. London: School of 0 riental and African Studies. Siegel, J. 1997. Mixing, levelling, and pidgin/creole development. In The Structure and Status of Pidgins a11d Creoles, A. Spears & D. Winford (eds.), 111-149. Amsterdam: John Benjamins. Smith, N. 1987. The Genesis of the Creole Languages of Surinam. PhD dissertation, University of Amsterdam. Smith, N. 1999. Pernambuco to Surinam 1654-1665: The Jewish slave controversy. In Spreading the Word: The Issue of Diffusion among the Atlantic Creoles, M. Huber & M. Parkvall (eds.), 251-298. London: University of Westminster Press. Thomason. S. & Kaufman, T. 1988. LanguagE Contact, Creolization, and Genetic Linguistics. Berkeley CA: University of California Press. Thurmes, Marion. 2007. Pratiquer une langue locale pour s'integrer. Pratiquer deslangues locales et representations de fA.utre chez les metropolitains de Guyane. In Pratiques et representations lingulstiques en Guyane: Regards croises, I. Leglise & B. Migge (eds.), 171-192. Paris: Editions IRD. Winford, D. 1997. Re-examining Caribbean English Creole continua. World Engllmes 16: 233--279. Winford, D. 2003. An Introduction to Contact IJnguistics. Oxford: IDackwell.
'Creole' and youth language in a British inner-city community Susan Dray & Mark Sebba
This chapter draws on ethnographic research among adolescents in an inner-city area of Manchester, England, which has traditionally been seen as the centre of a 'Caribbean community: We describe how young people in this area from a variety of ethnic backgrounds are using Creole language forms, and show how this poses challenges for existing models which aim to relate ethnicity, geographical origin and language. We suggest that the language behaviour of individuals is more usefully explained by considering their involvement in specific language-based practices valued by the peer group than by reference to their membership of an ethnically defined 'community~ In their discourse, participants construct themselves (and their preferred practices -linguistic and non-linguistic) as 'blac.k: where being 'black' is a position in the social order, defined by their practices, not by their ethnic background We conclude that 'black' now more than ever is a social construct and not a skin colour.
1.
Introduction
This chapter draws on ethnographic research undertaken in 2004-2005 among adolescents in an inner-city area of Manchester, a large city in the northwest of England. Although no linguistic research has previously been undertaken in this area, the fact that it has traditionally been seen as the centre of a 'Caribbean Community: together with knowledge of how Creole has been part of the linguistic repertoire of other inner-city areas in Britain, has led us to expect that we would find Creole in some form being used by Caribbean-heritage adolescents there. What we actually found was surprising and in some ways more interesting. Though there were still plenty of'Caribbeans' in the area of Manchester where our research took place, it was hardly possible to find any social networks which were exclusively Caribbean, or even 'black.' Among the adolescents in our study, none were actively using Creole for everyday communication. However, Creole language forms were being widely used, not just by black adolescents, but by their peers in
232 Susan Dray & Mark Sebba
multi-ethnic social networks. These findings pose challenges for existing sociolinguistic models such as the variationist one which attempt to show correlations between language and ethnicity and/or geographical origin.
2.
Background
Caribbeans have been part of the British linguistic landscape for some centuries, since the slavery period, but the main wave of migration from the Caribbean to Britain took place in the 20 or so years from 1949 onward. During this period, mainly adult migrants came to Britain to take up employment in services and industries where there were labour shortages; once labour was no longer in demand, immigration controls started to be imposed and the stream of migration diminished. Linguistically, the first generation of settlers from the Caribbean reflected the diversity of their region of origin: a majority would have been speakers of a Creole variety, usually a mesolectal one in the case of those from places like Jamaica and Guyana where a continuum exists. In the period following their arrival, these migrants accommodated linguistically to their new environment to varying extents. Wells (1973) shows how Jamaicans variably adapted their pronunciation to match that of London English; we can assume that similar variable adaptation took place in this generation in other places and with respect to other aspects of language such as grammar and discourse. By the late 1970s, there was a second generation who had been born and/or spent most of their early life in Britain. The evidence suggests that this generation was very difteren t linguistically from their parents, growing up with first -language competence in local varieties of British English. At the same time, this generation had a Caribbean 'heritage language' in which they had complete or almost complete passive competence, but active competence which varied greatly from individual to individual. This heritage language was not necessarily the vernacular of their parents, as it was based on Jamaican Creole even for those speakers who did not have Jamaican ancestry. This generation were thus competent and prolific code-switchers between local English and a local variety ofJamaican Creole. Following early reports of a 'London Jamaican' variety in the late 1970s, several studies showed similar patterns in different parts of the country: for example Edwards (1986) for the West Midlands, Sebba (1993) for 'London Jamaican: and Tate (1984) for Bradford in the North of England. Tate's research showed clearly that Jamaican ancestry was not a prerequisite for using a variety ofJamaican Creole, as her subjects were a in reggae band and, like most of the Carib beans in Bradford, were of Dominican descent and thus the children of French Creole speakers. The members of the band used Jamaican Creole and Bradford English among themselves.
'Creole' and youth language in a British inner-ctty community 233
Research carried out during the 1980s indicated that there was a strong link between language use and ethnic identity for this generation of Caribbeans-bydescent. While virtually indistinguishable from other locally-born British people when speaking the local English, they could mark themselves out and show solidarity with other Caribbeans by using the local form of]amaican Creole. The extent of their use of it could vary, however, from fluent and confident to almost purely symbolic and consisting of common phrases and discourse markers. However, as Hewitt noted, it was a powerful identity marker: The use of creole by black adolescents is arguably the strongest single marker of black youth culture and it is consistently present throughout the different forms assumed by that culture. (Hewitt 1982: 226)
As mentioned above, the available reports suggest that among the second generation of Caribbeans in Britain, the patterns oflanguage use were similar in different parts of the country. However, the picture we have from research is only a patchy and partial one. The only areas where fairly extensive research projects were carried out were London (Hewitt 1986; Sebba 1993), the West Midlands (Edwards 1986), and Bedford (Sutcliffe 1982)- all in the early 1980s. Some cities with longestablished Caribbean communities, like Bristol had no linguistic research carried out at all. Likewise there was no research done in cities with smaller Caribbean minorities, like Ipswich and Preston, where different patterns of settlement might have led to different outcomes. 1 And strikingly, two large cities in the North of England, Leeds and Manchester, each with an identifiable Caribbean community settled in an identifiable area, received no attention from researchers. At the same time as African-Caribbeans were found to be using Creole and English in urban areas like London, researchers observed that Creole or at least some elements of Creole were being taken up by non-Caribbeans. Hewitt documented the use of Creole by white adolescents who were involved in peer groups with Caribbeans (1982, 1986) but also observed that in London a local 'multiethnic' vernacular had emerged: 'the language of white as well as minority youth' (2003: 193). This included identifiably Creole lexical items and some discourse markers. In an area with a relatively high proportion of Caribbeans, he recorded 30 such lexical items in use (1986: 129), noting a consistency in the phonological influence of Creole on just a few words (1986: 133). However, as Hewitt's study did not comprehensively address the extent to which the syntax of the 'multi -ethnic' vernacular was influenced by Creole, we have to assume that this vernacular was essentially recognisable through the occasional use of Creole tokens and formulaic expressions within the syntactic frame of a local British vernacular, to which
1.
Recent work in Ipswich has redressed this particular omission (Braila-Straw, this volume).
234 Susan Dray & Mark Sebba
all speakers had been exposed as small children. Unlike Creole (London Jamaican), this multi-ethnic vernacular did not carry symbolic meaning in terms of race or ethnicity and could be freely used by white or black adolescents without fear of overstepping ethnic boundaries. In a subsequent study, Rampton (1995) described the phenomenon of crossing, in which young urban speakers in peer groups used limited amounts of languages or varieties which were spoken locally but which they did not 'own.' For example, a white child of English-speaking background might use Creole or Panjabi, possibly in a highly restricted or formulaic way, in interaction with peers in an ethnically mixed group. To summarise, research in the period approximately 1980-1990 indicated that local varieties of Caribbean (specifically, Jamaican) Creole had developed in some urban areas of England and were in regular use by young members of the Caribbean community alongside local varieties of English. At the same time, Creole had been taken up in a limited way by non-Caribbean youth, with some Creole vocabulary and discourse markers being used within the ethnically undifferentiated youth language, as well as in 'crossing' styles among peer groups. London was probably in the vanguard of both of these trends.
3·
Manchester
Manchester is a city (population just under 400,000 according to the 2001 census) within a much larger conurbation (Greater Manchester, population almost 2.5 million in 2001) in the North-West of England, some 200 miles (320 km) from London. Manchester followed the familiar pattern of rapid industrial growth (centred around textiles) in the 19th century followed by a slow decline of traditional industries in the 20th century. The areas surrounding the central business district remain (in spite of urban renewal projects) generally areas of low status, with low-quality housing occupied largely by low-income families and with social indicators such as health, education, and crime all considerably worse than the national average.2 Despite its importance as a Northern city with a distinctive character, Manchester has attracted surprisingly little interest from linguistic researchers. Manchester English is poorly described. We were unable to find any authoritative description of the phonology or grammar of the English associated with Central Manchester (i.e. the residential districts immediately south of the current central 2. See, e.g., Central Manchester Strategic Regeneration Framework (Manchester City Council n.d).
'Creole' and youth language in a British inner-ctty community 235
business district). Wells (1982) remarks that Greater Manchester is in an area (the 'middle north') characterised by 'typical' northern accents (1982: 350) but says nothing about Central Manchester specifically. The nearest variety which has been well-described is that of the outlying town of Bolton (Shorrocks 1998), which is 15 km or more from our 'target area' and which historically has been a separate administrative entity and is not geographically contiguous with the area where we were based. Though the accents, not surprisingly, are similar, Bolton speakers would be heard as coming from outside the area of our research site. We are therefore not able to say much about the traditional forms of Manchester English, other than that it is a typically northern variety of English, with relatively few highly localised features. However, it is easily distinguishable from Southern British English varieties, from the variety of its large neighbour to the west, Liverpool, and from all the Creoles of the Caribbean.3
4-
Our research site
Manchester was a destination for many new arrivals from the Caribbean who settled in Britain after the Second World War. By the 1980s and 1990s, there were several areas ofcentral Manchester that were strongly associated with the Caribbean community. However, at the time of our study in 2005, these areas were no longer predominantly 'Caribbean? In the areas covered by this research, there was still a strong 'Caribbean' presence, but there was also a large, well-established Asian population, and in one area in particular, a very recent and still continuing influx of new arrivals from Somalia. According to reports from local inhabitants who had witnessed the changes in the local population over the last two decades, the demography of the area began to change significantly following the demolition of several housing blocks during the 1990s. As new migrants arrived from Somalia and elsewhere, some of the established residents moved out to other areas and
3· Several vowel features distinguish Manchester English from RP and Southern varieties of British English, most saliently: (1) the lack of a distinction between the vowels of put (RP [u]) and putt (RP [A]), both [u] in Manchester; (2) the lack of a distinction between the vowels of trap (RP [a]) and bath (RP [a:]), both [a] in Manchester; (3) the use in Manchester ofalowback rounded vowel [n] in unstressed final syllables where RP would have[~] (lettER); (4) the use in Manchester of a lax vowel [1] or [e] in the final syllable of wordsllke happY, where southern varieties have a tense vowel (Wells 1982: 350; Beal2008: 136). In all of these respects, Caribbean creoles (and Jamaican in particular) pattern with RP/Southern British English, but most of the other features which differentiate Caribbean accents from those of Southern British English also differentiate Caribbean accents from Manchester English (see Wells 1982 and Sebba 1993 for details of these features).
236 Susan Dray & Mark Sebba
were dispersed. Thus by the time of our study, the one-time 'Caribbean' districts of Manchester had become a more culturally hybrid, ethnically mixed community in which it was no longer possible to identify a distinct Caribbean community. As no thorough linguistic study of Manchester has been carried out, there has been no systematic research on its Caribbean inhabitants and their languages. However, we have indirect and anecdotal evidence from Manchester that Jamaican Creole was used there up to the mid-1990s, and furthermore was used in a similar way to how it was used in other British urban communities. We have reports from informants, now adults, who grew up there, that they used Jamaican Creole with their peers when they were teenagers. We also have evidence from a number of'Yardie' novels (crime novels with Caribbean protagonists), set in Manchester and written by black Manchester authors, which appeared in the early- to mid-1990s. For example, the following passage gives a literary depiction of the language used by the young Manchester-Caribbean character Teeko and his peers, making it explicit that he has command of two varieties: "Whe' me deh? I can't concentrate on this with them girls watching;' he muttered, shifting his beefy frame uneasily in his seat. Like most of his friends, he spoke in a mixture of Jamaican dialect and English, with a strong Manchester accent. (Smith 1994)
If this is an accurate depiction of the speech of a Manchester-born adolescent ca. 1990- and it is certainly credible given the author's local connections- it suggests for Manchester in the late eighties/early nineties exactly what research had shown for London about a decade previously: that young African-Caribbeans were using Creole and English, in a code-mixing mode in everyday conversation among themselves.
5·
Our study and its participants
The original intention of our research project funded by the Economic and Social Research Council, 4 reflected in its title 'Language and literacies ofyoung Caribbeans in Manchester: was to look at the language ofchildren and adolescents of Caribbean heritage. But in the area of Central Manchester which was the focus of our interest, immigrant communities from a variety of religions, cultures, and races had been living alongside a white British working class population for up to four generations. In 2005, mixed-race families had become commonplace and the corollary of 4· Reference number RES-000-22-0681. We are grateful to the ESRC for providing funding for this exploratory project. We are also grateful to the editors and the anonymous reviewer of this chapter for their many useful comments.
'Creole' and youth language in a British inner-ctty community 237
this was that some members of the younger generation were exposed to a variety of cultural influences both at home and in their peer groups. Many had black Caribbean and white British relatives, some of these also had a further influence from an Asian or African parent or grandparent. One teenage participant in the study described himself as "quarter-casf' Not only did we observe a large proportion of mixed-race young people, we also discovered that some young people (who were not necessarily mixed-race) lived with families of a different ethnic heritage as foster children. The social networks of the participants in the study were therefore noticeably diverse with regard to their contact with people of other ethnicities and ethnic heritages and we were not able to identify any ethnicallyexclusive Caribbean social networks of young people. Given that there were very few young people of exclusively Caribbean descent and even fewer that were not British-born, using ethnicity as a category by which to select participants became problematic. Nevertheless, in order to present the research findings and to support our claim that ethnicity is an inappropriate category to explain language use in this case, we give an indication of the ethnic composition of the sample in the Table 1 below. For ease, the ethnicities of the participants have been described according to their parents' ethnicity. The ethnicity of their grandparents or of other family members has not been taken into account. The total number of participants with Caribbean heritage (Categories 1 and 2) is 22 (55% of the total). Of the 'mixed race' and 'black' participants (Categories 1-3, total= 28), the mixed heritage group (total= 12) accounts for almost 43%. This figure is very close to the proportion of the 'mixed' population in the overall 'black' ('mixed' and 'black/black British') group in the 2001 census data for Manchester (41.7%), suggesting that this sample is a fair representation of the local demography. Table I. Ethnic composition of the sample Ethnic description 1. Black British (all of Caribbean heritage) 2. Mixed heritage with Caribbean parentage 3. Mixed heritage not with Caribbean parentage 4. White British 5. British Asian 6. Unknown
Total
Male
Female
Total
11
5 2
16
4 6 5 3 2 31
2
6 6 7 3 2
9
40
238 Susan Dray & Mark Sebba
Of these 40 participants, we worked closely with a core sample (total= 22) comprised of 15 members of a young men's group and seven young people coming from four different families. The ethnic composition of this core sample was as follows. Young men's group (13-15 year olds): 1. 6 black British (Caribbean heritage) 2. 5 mixed heritage (4 black Caribbean/white British and 1 Asian/white British) 3. 3 white British 4. 1 Asian Four families with children and teenagers of primary- and secondary-school age (6girlsages8-15; 1 boyage 12): 1. 6 black British (Caribbean heritage) 2. 1 mixed heritage (black Caribbean/white British) We took an ethnographic approach to data collection. We observed and participated in the activities of the youth club as much as possible, and this was supplemented byvideo recordings made by a youth worker in our absence. We enlisted the help of a youth worker and a school governor, both of Caribbean heritage, living in the local community and already known to the participants in order to collect naturally occurring spoken language data in authentic and typical settings. The kinds of data we collected included: group interviews with some of the participants, which were facilitated and audio-recorded by the youth worker with whom the participants were familiar; direct observation of spoken and written language; video recordings of peer-group interactions both in the youth club and while undertaking organised activities such as paintballing, football, cycling, trampolining, etc; photographs of written language including lyrics and graffiti; and audio recordings of spontaneous spoken interaction in the family setting and in the youth club. The latter were either collected by the youth worker or the school governor, or recorded by the participants themselves. The conversational language data amounted to approximately 19 recorded hours. In addition, in order to contextualise this language data, we conducted 24 interviews with teachers, community workers, youth workers, musicians, and ex-gang members.
6. Linguistic behaviour As we said above, there is no systematic documentation of the extent to which Creole was used among young people of Caribbean heritage in previous generations in Manchester. As far as we are aware, our research is the first to explore the
'Creole' and youth language in a British inner-ctty community 239
role of Creole in the language of the younger generation, some of whom have links to the Caribbean through their grandparents or parents, while others with whom the former are closely networked have no direct links whatsoever. To our surprise - as we had come looking for and expecting to find Creole continuing to be used as a heritage language at least to some extent -we found no evidence of a systematic or extended use of Creole in any of the peer-group interactions between the young people in our study. However, there was evidence of the use of Creole fm·ms by all the participants. Importantly, some of these forms were used by all peer-group members regardless of their heritage, and these forms did not appear to be seen as 'Creole' by the speakers, but rather as a 'slang' form of English. Other Creole forms were observed less frequently and appeared to signal a shift in style, indexing more overtly a Black British/Caribbean identity. Such style shifts were not necessarily confined to those with Caribbean heritage. An example of a Creole form that had become re-appropriated into the everyday linguistic repertoire of the participants was the discourse marker man, which was used for emphasis (e.g. "Naa, I don't play football man, I ain't playin' football man~ - Keith, 14, white British). The realisation of the vowel in man varied according to speaker and context. The vowel shifted between the British English realisation [a] and the Black American Vernacular English realisation [ei] and sometimes fell somewhere between the two (e.g. [ax]). Further examples of Creole-origin forms, occurring several times in the sample data and used byatleast three participants, were 'vex (to be angry) and 'mash up' (destroy, damage, injure, etc). Again, the application of these forms was variable. Sometimes they were adapted to English rules and inflected (e.g. "The community's gonna get mashed up"- Harvey, 14, Caribbean heritage), and sometimes, following Jamaican Creole grammar, they were not inflected in the past tense (e.g. "I just get vex" -Nick. 12, Caribbean heritage). Furthermore, many of the participants had/dl as the most common realisation of RP /l!J/ in word initial position (e.g. the, there, though). The frequency and ubiquity of this phenomenon in the local area was observed not only in speech but was consciously reified by some users themselves in their written language, particularly text messages and grafitti, by using the spelling for Standard
. However, there was little evidence for /t/, the Creole realisation ofRP /9/, which tended to be realised either as [9] or [f] 5 (e.g. "Don't you think that girl finks she's bum (i.e. great)"
5· [f] is increasingly a realisation of /8/ among younger people in most nonstandard varieties of British English in England. See, e.g. Kerswill (2003).
240
Susan Dray & Mark Sebba
- Daveena, 15, Caribbean heritage).6 On only one occasion did we observe the use of Creole [t] in the realisation of the word 'thick.' and this was by a white adolescent with his friends (of Caribbean heritage). In the example that is given below, it is clear that all those involved in the exchange understand the utterance and are familiar enough with it to contribute to the joke that he makes in implying that he has become thickset ('tick') from working out in the prison ('pen') gym. (1) Lee:
[Firing 'gunshots' into the air enthusiastically with his fingers] what's up with 'im? Youthworker: he's excited Ben: Youthworker: told you man, you shoulda stayed in Spain for an extra week Ben: naa, he's been in Spain for - he's been inside for about four weeks Lee: I've just come outa pen [autapen] (.)Tick! [t:rk] Andre: your head's tick [tik] Lee: so's your bottom lip! [Lee, 14, white British; Ben, 14, mixed race (black Jamaican/white British); Andre, 13, mixed race (black Jamaican/white British)]
There is no evidence in the rest of our data to suggest that this is a form the young people would draw on regularly as part of their everyday repertoire, but the fact that it was accepted and understood without comment indicates that it is recognised and understood as a variant of It/. Several participants demonstrated a more active use of a broader set of Creole forms than the rest, but these were nevertheless limited to short phrases (e.g. "(Let's) catch some joke,- Harvey, 14, Caribbean heritage) and to curse words (e.g. "Ai! You bomba!" - Chris, 13, Caribbean heritage; "Rass" - Lee, 14, white British). These young people tended to come from homes where family members used Creole and/or had contact with relatives in the Caribbean, although this was not always the case. From the peer-group interactions we observed, none of the young people could be described as speakers of Creole, nor did they describe themselves as such, and there was a general consensus among them that people from the Caribbean were difficult to understand. 6. Irvine (2004) found a similar difference in the use of /d! and It/ among speakers in Jamaica, concluding that speakers of Standard Jamaican English would 'carefully avoid voiceless TH-stopping'but 'freely vary [d-
'Creole' and youth language in a British inner-ctty community 241
7·
The legacy of Creole
Finding that there was no regular use of Creole as a language among our participants was a genuine surprise. Much of our research was carried out in a young men's group with a high proportion of Caribbean-heritage members. In London in the 1980s, for example, in such a con text it would have been easy to find fluent users of Creole alongside others with a less active knowledge. However, as we have said, though we found evidence of Creole forms, we found no evidence of Creole in use as a language in its own right. What we did find was a number of practices in which language is central, which involved Creole as a frequent (though optional) element, and where the practices themselves appear to have roots in the Caribbean. There is a large body of linguistic literature relating to verbal rituals, and in particular verbal duelling, among the African diaspora - see for example Labov (1972), Smitherman (1999), and Mendoza-Denton (2008). In Sections 7.1 and 7.2, we discuss two of the practices we observed in Manchester, MC-ing and murking, in more detail. In both of these practices, the primary goal is to verbally beat your opponent. MC-ing (i.e. rapping) is a strongly language-based practice, involving spoken and (potentially) written language. It is a part of hip-hop culture, which is international, but has developed local forms reflecting local cultural practices (see Mitchell 2001; Pennycook 2006). Access through the media and new communications technology to the 'international Hood' (i.e. traditionally Black (American/British) youth life experience and cultural forms such as hip-hop) means that the young people in Manchester have alternative linguistic resources to draw on which may encourage linguistic 'experiments' -lexically and phonologically. The linguistic styles and strategies drawn on and valued in the MC-ing 'performance' are frequently also found in peer-group speech, blurring the boundaries between spontaneous communication and 'composed' performance. The practice known locally as murking ('mocking'? 'murdering'?) is variously defined as 'stealing,' 'beating someone to death: and 'to beat someone at rapping by having better lyrics:7 Neither the word, nor the practice, is confined to Manchester. We observed murking in its third sense above, as a verbal duelling competition. The young people's exchanges are highly intertextual and often explicitly so. They frequently draw on different voices, and this makes a distinction between 'natural' speech and performance difficult to sustain. Performance frequently plays a part in everyday peer-group interaction and thus contributes significantly to the negotiation of individual identities and relationships between group members. Intertextuality is thus a key element of the group's value system, as it is one linguistic mechanism by which group members are able to signal belonging and achieve status. 7·
See, e.g. entry for 'murk' in Urban Dictionary.
242 Susan Dray & Mark Sebba
7.1
Language practice 1: MC-ing
MC-ing is dominated by males. Perhaps because of this, it is a gendered practice in that it is associated with and imbued with the male value-system (sexual prowess, survival, wealth). That said, if a girl does it (and she must be good), she is very highly respected by the boys. The problem for the girls is that they don't appear to have the same 'apprenticeship' opportunities as the boys. If they are not good, they won't practice because of the fear of ridicule. The values that are attached to lyrical ingenuity/innovation in the MC-ing practice spill into everyday communication and result in a view of language that is not that of the 'schooled' standard, but rather one in which non-standard, vernacular forms are equally, if not more, prestigious. Embodied in this 'lyrical prowess which is found in performance and normal conversation is a value-system which reflects a lifestyle in which masculinity is constructed as 'survival of the fittest' (as embodied in the title of US rapper 50 Cent's 2003 album Get Rich Or Die Tiyin'). The recurring theme of killing is metaphorically represented in the murking practices (see below) of younger males, who draw on the practice of verbal duelling (a central element of the MC-ing performance) in their peergroup interactions. The extract below, performed by Harvey (14, black British of Caribbean heritage), highlights some of these points. Perhaps most importantly is that, although it is a performance, it highlights the styles, language forms, and combinations found in 'normal' conversations. However, very few Creole forms are actually used -probably only neck back (line 3), run back (line 5), and bus (line 10). (2) MC-ing by Harvey 1 ...boy, you backstabber 2 all you better get back 3 before you get capped in your neck back 4 (inaudible) ifl see gunshot 5 come against me and you will get run back 6 if I take it to violence you will get smacked 7 [ ... ] 8 you should give me an encore 9 you will get dropped to the floor 10 bus your head on the dance floor 11 your spinal cord will get cut 12 (inaudible) better fuck off 13 or you will get hacked off
'Creole' and youth language in a British inner-ctty community 243 This extract also demonstrates the focus on 'survivaf!killing/dissing. There is potential for another MC to come back at Harvey with a verbal response to his 'death threat' (i.e. murking, see Section 7.2). 7.2
Language practice 2: Murking (verbal duelling)
This is also more of a boys' activity, but not exclusively so. However, there is a gender di:ffurence. The girls reported that boys do it with their friends - almost as a way of expressing friendship/intimacy. When girls do it (which is not frequent), they are serious about it. It can get nasty. In the extract below, Harvey (14, black British of Caribbean heritage) had begun murking Keith (14, white British) while he was attempting to video the other boys playing football. Keith moved the camera onto Harvey who apparently scratched Keith with his 'sharp nails' as he pushed Keith away. (3) Murking between Harvey and Keith 1 Harvey: sharp nails!? [laughs] your ass smells ofBO man 2 Keith: no man Harvey: you're like a skinny penguin 3 4 Keith: no man 5 Harvey: pigeon ches' 6 Keith: peppergrain NO. [sa.id with emphasis a.ndfinality] 7 Harvey: 8 Keith: yeah yeah 9 Harvey: your skin - tongue looks like off milk 10 Keith: why; why are you murking me? You want me to get Martin to murk you again? This is recognisably a version of the African-American practice extensively described by Labov (1972: 305-353) and variously named sounding, signifying, and playing the dozens. In this extract, there are references to physical attributes -body odour, size, hair, and colour, which are based on their in-group knowledge of what counts as offensive, and also what counts as acceptable. Harvey appears to be more adept at murking than Keith, whose one attempt at name-calling- 'peppergrain' - refers to Harveys hair and oversteps the line of 'friendly banter. 'Peppergrain' is a term used in Jamaica to describe short hair of a bobbly texture and was considered a strong insult by the participants in this study; as indicated in the extract by Harveys definitive response. Keith is unable to find any further come back at Harvey and ends the event by threatening to involve the murking expertise of Martin.
244 Susan Dray & Mark Sebba
8. Global, local, and translocal
In their language practices, young people in Central Manchester can be seen as drawing on both local and global linguistic repertoires. The majority of their everyday vernacular is 'local' in the sense that it is the Manchester variety of English, clearly marked as such by features of accent and (to a lesser extent) grammar. The global aspect of their repertoire is evident in the ways they draw on lexis from the Caribbean and African-American vernacular, as well as other varieties and accents of English (such as RP, for example) for stylistic effucts. More importantly, we saw above how the practices of MC-ing and murking, which originated in overseas 'black' (African-American and/or Caribbean) culture, are part of the linguistic culture of many of our participants. A third aspect to the linguistic repertoires of the people in our study is what we might call the translocal. 8 By this we mean language forms or practices which are not 'global' (in the sense ofbeing widespread around the world) but which are localised in two or more places, and transmitted between them through social networks. For example, a number of distinctively Creole words were in use in this area of Manchester at the time of the study, but with apparently localised meanings different from the ones they would have in the Caribbean or elsewhere. For example, the Jamaican Creole adjective 'dutty' ('dirty: JC pronunciation [d:Jti]) was being used by young people in Manchester to mean 'good' or 'cool' (i.e. 'that cars dutty'). In Manchester, it was pronounced [duti], with the Northern /u/ vowel, while in London, where it appeared to retain only its original meaning, it was being pronounced with the Southern British English !AI vowel and also reflected the trend for intervocalic t-glottaling among young British people, thus [dAti] or [dA'li]. This kind of localised meaning of a word is exemplified in the next extract, in which three people talk explicitly about a culture-specific lexical item. A family consisting of a mother, her son Nick (12, Caribbean heritage), and her daughter Keisha (10, Caribbean heritage) are talking about a homework project (involving choosing between three courses of action, see line 1) when Nick unexpectedly starts rapping and introduces the address term blood (line 3). The mother is confused, not because she has never heard the word used in this way (as she explains later, she has) but because of the way her son pronounces it. (4)
1 Mother: I'm sorry Keisha. What were the three choices again? 2 Keisha: pick up the bag and run to the nearest= ah come on blood [blAd], come on, come on blood [blAd]! 3 Nick:
8. See Grlllo (2007) for a discussion of meanings of this term within the anthropology of transmigration.
'Creole' and youth language in a British inner-ctty community 245
=no-no-no-no-no. Just wait man! come on what? come on blad [blAd] blad [blAd] what's 'blad'? [gasps] blood [blud] blood [blud] blood [blud] oh! not the one that comes out= yeah yeah, as in 'me blood' [mi blud] as 'me blood [blud] brother' as in blood [blAd] as in family as in respect 17 Nick: 18 Keisha: mi blood! 19 Mother: get offKeisha! erm (.)that's how you greet(**) now? yeah. 'Come on blood [blAd]' 20 Nick: 21 Mother: Ohgod! The mother's difficulty is caused by the fact that the word 'blood' (the substance in the body), which would be pronounced [bbd] in Jamaica, is always pronounced [blud] in Manchester English, as elsewhere in the North of England. In fact, the phoneme !AI does not occur in 'normal' Manchester English, as the cup/put distinction is neutralised in Manchester as in other northern dialects. 9 The pronunciation [blAd] (a term of address) uses the vowel characteristic of RP, of London and the South. It would easily be understood (and recognised as 'Southern') by a Manchester native. Nick and Keisha never use IAI in other words. It is therefore quite remarkable that a vowel 'alien' to the Manchester system is being used here to make a distinction which has cultural significance. 10 At this point, talk turns temporarily to other topics: the children talk about a style of dancing which is popular in London and begin to compare Manchester unfavourably to London. Nick talks about the fact that he was born there, 11 after 9·
For descriptions of the phonology of Northern (including Manchester) English, see Wells (1982), Beal (2008).
Older boys ( 14+ years) at the youth dub reported using this pronund.ation 'for a laugh' contextuallsing it as jocular by stylising it Used in this way, it appears to serve a distancing function between the Northern and Southern 'hoods:
10.
n. The family have London connections, and the mother was living there when her son was born. However, all three have Mandlester accents and Nick has dearly grown up in Mandlester.
246
Susan Dray & Mark Sebba
which his mother returns, 22 turns after the last mention, explicitly to the topic of the address term blood. (5) 43
Mother:
44 45
46
Keisha: Nick: Mother:
47 48 49
Keisha: Mother: Nick:
50 51
Keisha: Nick:
52
Keisha:
the last time I heard that word- erm blood [blAd]- is when I was down in London and I heard AB saying that= ey bloody blood [blAdiblAd] yeah what's wrong with that? no no I just remember them saying that. I didn't even know youa heard it ey bloody blood [blAdiblAd] C stop being stupid 'ey bloody blood [blAdiblAd]'? What the heck are you talking about? Hey bloody blood [blAdiblAd] How do you- Manch- the proper English (in) Manchester would say= = 'hello blood [blud]'
Here, the speakers actively acknowledge that the word in this specialised sense, with its particular pronunciation, is an 'import' from London. Other than in the above example, we did not observe the use of this London form of 'blood' among the participants, but there were several occasions when the discourse marker man was, as we have already mentioned, pronounced with a stylised American accent (e.g. [s:]). We found no evidence of a Caribbean pronunciation of this word. 1his 'imported' form of man was unique in that, unlike other imported pronunciations, its occurrence was not limited to performance genres. During conversational exchanges, switches between genres such as rap or song or documentary-style commentary were common. For example, Lee (14, white British) was asked to call everyone to the table at the youth club to eat and informed them by shouting out "The food's ready n-o-o-o-w Ina:!; singing the last syllable with falling intonation. The influence of performance styles in everyday interaction also manifested itself in the structures of utterances. In Example 6, Chris (13, Caribbean heritage) is commenting on a girl's hair. (6) Look at that knotty [nn?r] pick head. That knotty [nn?r] pick head. Look at that knotty [nn?r] picky head. U-ugly [s:rgh] (.) Blap-blap! His utterance is structured as a series of repetitions which creates a rhythm, and he reinforces his 'performed' point by drawing on a pronunciation of 'ugly' [s:rgh]) which resembles an American accent. He ends the turn with a Creole discourse marker (blap-blap) - the onomatopoeic sound of a gun -salute. 1his expression is
'Creole' and youth language in a British inner-ctty community 247
commonly recognised and used by dancehall goers in Jamaica and Britain alike to indicate their appreciation of a particular tune. Furthermore, he draws on Creole, combining two contrasting Jamaican Creole adjectives, 'knotty,' which tends to have positive connotations (and usually refers to dreadlocked hair, although not in this case), and 'picky' (messy), which has negative connotations. The first of these is uttered with a local British accent ([nn?I]). With 'picky' he drops the final vowel in two of the three occurrences. Voices representing 'others' can be made identifiable by a shift in pronunciation which distances the form from the speaker's own accent (cf. Goffman's notion of'say-foring: 1981: 150). When this intertextualityoccurs and it is not marked by a switch in accent, such forms appear to have become appropriated into the everyday linguistic repertoire of the participants and no longer carry 'import' status. Examples 6 and 7 demonstrate the convergence of linguistic forms that are drawn from the global and/or translocal repertoires available and manipulated within the local context until some of them are no longer explicitly marked as 'imported.' The comment in Example 7 below was made by Harvey (14, Caribbean heritage) during a trip to an ice skating arena. (7) We're thuggin' [fugm] it out in Blackburn, 12 runnin' tings [tiiJz] The phrases 'to be thugged out' (i.e. to identify with the urban/ gangsta lifestyle, cf. US rapper Tupac's 1994 Album entitled 1hug Life) and 'run ting(s)' (i.e. to be in charge) index American hip-hop and Jamaican cultures respectively but have been brought into line with the local vernacular pronunciation (i.e. [fugm], [rumn]) and grammar. Such 'reassembled' forms are not 'ad hoc: 'mixed: or 'impure' in the sense described by Hewitt a generation ago ( 2003: 193) but are localised styles which result from the convergence of local and non-local repertoires and which index membership in the peer-group and affiliation to the 'international hood: Importantly, these styles are not necessarily comprehensible to non-local contemporaries undertaking similar practices elsewhere in Britain or the world. Nor are they necessarily comprehensible to those living locally but outside of the peer-group. In Example 7, Harvey's comment is a form of conversational 'performance; a verbal claim to superiority in which he distances himself from a 'Blackburn' identity, but which is not necessarily intended as a realistic appraisal of the situation. This 'distance' reflects the geographical limitation of 'localness' for most of the participants in this study whose movements tended to remain within central Manchester. A few of the participants claimed not to have travelled much beyond the boundaries of Greater Manchester. Several had not travelled extensively in the 12.
IDackburn is the town in Lancashire where the skating arena is located.
248 Susan Dray & Mark Sebba
northwest region of England where they lived, but may have been abroad, either on holiday or to visit relatives. Physically then, 'local' networks for our participants tended to be concentrated within Greater Manchester and 'trans-local' networks could be, for example, as near as Liverpool (the neighbouring city) or as far as Jamaica. What is striking is that these globally and translocally informed styles, whether performed or re-appropriated, are legitimate language forms in the contexts of peer group interaction. This suggests that in these social spaces, acceptable linguistic behaviour cannot be described by reference to local regional varieties alone or to ethnicity (cf. Meyerhoff and Niedzielski 2003). The styles the participants use with each other are more easily associated with young people who have an affiliation with the 'international hood: and for our participants, this affiliation was not a function of an individual's ethnicity or race but of their shared lived experience in the inner-city or the 'life on the streets' as some of the participants called it.
9· Conclusion
The notion of'ethnicity as a fixed and formative inheritance' (Harris and Rampton 2003: 5) which informed early variationist studies has - along with other social categories like social class, gender, and age - been substantially problematised within linguistics and neighbouring disciplines in recent years. The study of Creole languages has itself been part of this process (e.g. LePage and Tabouret-Keller 1985; Hewitt 2003). Furthermore, even inherited ethnicity becomes problematic in a situation where social networks are relatively unconstrained and intermarriage is common. To complicate things further, through globalisation, some styles which were once local and ethnically marked, in particular those associated with Creole and African-American Vernacular speakers, are now subject to reappropriation and recontextualisation in many different contexts around the world through hip-hop culture (Pennycook 2006). Example 7, with its incorporation of hip-hop and Jamaican Creole phrases into a sentence characterised by strictly local vernacular pronunciations (and a local place reference) suggest that here and elsewhere we may be dealing with what Meyerhoff and Niedzielski (2003) call 'a broadening of the vernacular base' in which speakers expand the range of forms to which they have access in their least self-monitored and most local contexts. Our research in Manchester shows that the pattern of English/Creole codeswitching reported in England in the 1980s is not part of the behaviour of young people today in Manchester. Creole forms remain in the vocabulary and in traditionally 'black' practices like murking and MC-ing (rapping) but they do not constitute 'switches' between codes. Rather, they are convergences of globally and
'Creole' and youth language in a British inner-ctty community 249
translocally available forms which may be re-appropriated into the local vernacular as styles. Our naturally occurring language data demonstrates that Creole forms and associated practices are not distributed on the basis of ethnicity. Indeed in the research site, where young British people of a variety of ethnicities and mixed heritages live, it would be impractical and misleading to attempt to account for language behaviour using ethnicity as a category. Speakers are legitimate users of these vernacular styles as a result of their lived experience in the inner city and their interest in and exposure to cultural forms such as the global hip-hop culture and its music, from which linguistic forms may be drawn and locally interpreted, reappropriated, and valued. This process is enabled through advances in information and communication technologies, as well as through a speaker's local and translocal social networks.
References Beal, J. 2008. English dialects in the North of England: Phonology. In Varieties of English 1: The British Isles, B. Kortmann & C. Upton (eds.), 122-144. Berlin: Mouton de Gruyter. Edwards, V. 1986. Language in a Black Community. Clevedon: Multilingual Matters. Goffman, E. 1981. Forms of Talk. Philadelphia PA: University of Pennsylvania Press. Grillo, R 2007. Betwixt and between: Trajectories and projects of transmigration. Journal of Ethnic and Migration Studies 33(2): 199-217. Harris, R & Rampton, B. (eds.) 2003. The Language, Ethnicity and Race Reader. London: Routledge. Hewitt, R 1982. White adolescent Creole users and the politks of friendship. Journal ofMultilingual and Multicultural Development 3(3): 217-232. Hewitt, R 1986. White Talk Black Talk. Cambridge: CUP. Hewitt, R 2003. Language, youth. and the destabllisation of ethnictty. In R Harris & B. Rampton (eds.), 188-198. London: Routledge (originally published 1992). Irvine, A 2004. A good command of the English language: Phonologkal variation in the Jamaican acrolect. Journal of Pidgin and Creole Languages 19(1): 41-76. Kerswlll, P. 2003. Dialect levelling and geographkal dtlfusion in British English. In Social Dialectology: In Honour ofPeter Trudgill, D. Britain & J. Cheshire (eds.), 223-243. Amsterdam: John Befl:lamins. Labov, W. 1972. Language in the Inner City: Studies in the Black English Vernacular. Philadelphia PA: University of Pennsylvania Press. Le Page, RB. & Tabouret-Keller, A 1985. Acts ofIdentity: Creole-Based Approaches to Language and Ethnicity. Cambridge: CUP. Manchester City Council. N.d. Central Manchester Strategic Regeneration Framework. . Mendoza-Denton, N. 2008. Homeglrls: Language and Cultural PracNce Among LaNna Youth Gangs. Oxford: Blackwell.
250
Susan Dray & Mark Sebba Meyerhoff, M. & Niedzielski, N. 2003. The globalisation of vernacular variation. Journal of Sociolinguistics 7(4): 534-555. Mitchell, T. (ed.). 2001. Global Noise: Rap and Hlp Hop Outside the USA MlddletownCT: Wesleyan University Press. Pennycook, A 2006. Global Englishes and Transcultural Flows. London: Routledge. Rampton, B. 1995. Crossing: Language and Ethnlclty among Adolescents. London: Longman. Sebba, M. 1993. London Jamaican: Language Systems In Interaction. London: Longman. Shorrocks, G. 1998. A Grammar of the Dialect of the Bolton Area. Bern: Peter Lang. Smith, K. 1994. Moss Side Massive. London: The X Press. Smitherman, G. 1999. Talkin That Talk: African American Language and Culture. London: Routledge. Sutcliffe, D. 1982. British Black English. Oxford: IDackwell. Tate, S. 1984. Jamaican Creole Approximation by Second-Generation Dominicans? The Use of Agreement Tokens. MA thesis, University of York. Urban Dictionary. Entry for 'Murk: <.www. urbandictionary.com/define. php?term=Murk>. Wells, J.C. 1973. jamaican Pronu11ciation i11 Lo11don. Oxford: Blackwell Wells, J.C. 1982. Accents of English, VoL 2: The British Isles. Cambridge: CUP.
Le Page's theoretical and applied legacy in sociolinguistics and creole studies John R. Rickford
This paper reviews Robert B. LePage's contributions to creole linguistics, sociolinguistics, and applied linguistics, combining a critical and academic viewpoint with the personal perspective of a sociolinguist whose career and work have been deeply influenced by Le Page. The conclusion reflects on how linguists and educators can ideally build on Le Page's legacy. Following his example, we can make our most useful contributions by continuing to make sociolinguistic research bear on curricula in schools and universities. Meanwhile, the field of Caribbean sociolinguistics continues to offer vast opportunity for original empirical research and theoretical thinking that should develop and supersede Le Page's most lasting contribution in this area, the Acts of Identity model.
1.
Introduction
When Robert Broderick Le Page passed away in 2008, he left behind a rich legacy of contributions- descriptive, theoretical and applied to sociolinguistics and creole studies. 1 His work is especially familiar and valuable to Caribbeanists, but no contributions also reverberate beyond the confines of this geographical area. The Dictionary ofJamaican English, which he coauthored with Fred Cassidy in 1967, is an invaluable reference work, not only for Jamaica, but for the several other Anglophone Caribbean varieties (Guyana, Barbados and Trinidad, for instance) with which it shares lexical cognates and phonological and grammatical parallels. His (1960) account of the development of Jamaican Creole was one of the earliest - if not the earliest- in a series of detailed socio-historical studies that came to define creole studies and (should) make it the envy of sociolinguistics. Le Page was one of the founding fathers of sociolinguistics, along with pioneers like Charles Ferguson and William Labov, and his Acts ofldentity theory - honed on data from Belize 1.
This paper is a significantly revised version of a paper presented at the Le Page conference held
at Stanford University in July 201J7, within the Linguistic Institute of the Linguistic Society ofAmerica.
I am grateful to the editors of this volwne for pushing me to develop and defend my ideas.
252
John R. Rickford
and St. Lucia - should attract increasing attention as interest in identity as a sociolinguistic construct continues to grow (cf Bucholtz and Hall2004). Although Le Page's applied work is not as well known as his descriptive and theoretical research, it was significant too, being among the earliest to advocate bilingual education, contrastive analysis and other strategies to help creole speaking pupils bridge the gap between their vernacular and standard or mainstream varieties. It was also personally meaningful to me, his (1968) paper exciting my imagination as an undergraduate, and leading me in 1969 to switch from literature to a self-designed major in sociolinguistics. Finally, Le Page played a major facilitative role in Caribbean creolistics and sociolinguistics, serving as co-editor of the two-volume Creole Studies books that were among the leading readers in creole studies until Hymes' (1971) Pidginization and Creolization of Languages came along. He also helped to train - both at the University (College) of the West Indies, Jamaica, and at the University of York, England - several of the Caribbean scholars who went on to become the leaders in research, teaching and university leadership in the Caribbean. In this paper, I will provide a more detailed account ofLe Page's contributions in each of these areas, along with a critical assessment, in the hope that it will enable us to extend and enrich his rich legacy.
2.
2.1
Le Page's theoretical legacy
Sociohistory
First of all, and less generally known or emulated than it should be, Le Page demonstrated the value of careful socio-historical research in elucidating current sociolinguistic variation and change in his long (125-page) article, "An historical" introduction to Jamaican Creole in Creole Language Studies I, ed. by LePage in 1960. Although Le Page's work on this topic has now been extended and to some extent superseded by Kouwenberg (2009), it remains an important pioneering study. Creolistslike Philip Baker and Chris Corne (1982), John Singler (1996, 2007), and Sarah Roberts (2004, 2005) have followed Le Page in conducting detailed socio-historical research, but most sociolinguists working outside of Caribbean and creole-speaking communities have not. As a result, I believe that many sociolinguistic studies present a relatively "flat" picture of synchronic variation, not sufficiently attuned to diachronic forces and events that might have led community members to display the social attitudes and linguistic behaviors that they do. For the creolist, for whom sociohistorical issues are woven into the central "life cycle"
Le Page's theoretical and applied legacy in sociolinguistics and creole studies 253
conception of their subject matter,2 soda-history is ahnost unavoidable. But for non-creolist sociolinguists, sociohistory can and often is neglected - one reason, perhaps, that "Socio-Historical Linguistics" in the sense of Romaine (1982) has never realized its full potential. Let me develop this point a bit further. Even in Romaine's excellent (1982) text on Sociohistorical Linguistics, most of the attention is devoted to the linguistic analysis of variation in relative clause marking and its constraints, as attested in texts from Middle Scots and earlier, and discussion of the social and extra-linguistic context is minimal. By contrast, Romaine's also excellent but "creolist" book on urban and rural Tok Pisin in Papua New Guinea (1992) has far more soda-historical data and analysis. In a similar vein, the admirable studies of historical morpho-syntax conducted by Anthony Krach and his students at the University of Pennsylvania are rich in linguistic detail, but much less so on sociohistorical context. Indeed, Krach and Taylor (1997) simultaneously acknowledge the importance of such context while acknowledging that it is not a primary concern: Although we are not primarily concerned with the historical and sociolinguistic dynamic that established the Middle English dialects, the sociolinguistic history of population contact and diffusion which underlie them is a matter of considerable interest, and it sheds light on why the dialect difference we have uncovered should exist.
They go on to give a one or two-paragraph account of the sociolinguistic history. Perhaps a longer account might not have shed any additional light on the linguistic developments. But a creolist - in the mold of Le Page at least - would be hungry for more. 3
2. Who were the speakers of the different languages that came into contact to form an initial pidgin. and when, where and how - via trade, on colonial plantations - did the contact occur? Under what conditions, did the pidgin "creolize;' perhaps by becoming the primary language in its community, and/or acquiring native speakers? And how did the resultant creole go on to extend in social role and expand in "inner" and "outer" form (Hymes 1971) and/or "decreolize" in subsequent decades/centuries?
3· One area of sodo-historicallinguistics outside of creole studies that does appear to pay more attention to socio-historical context is work on "historical sociolinguistics" and the history of English being conducted by various individuals and groups in Europe, for instance, the work ofTerttu Nevalainen (e.g. Nevalainen and Raumolin-Brunberg 2003) and her VARIENG group in Helsinki (http://www.helsinki.fi/varieng/index.html), the historical sociopragmatics research of Andreas Jucker (e.g. 2008, and see http://es-jucker.uzh.ch/), and the social history/network research of Susan Fitzmaurice (e.g. 2000, and see http://www.shefac.uk/linguistics/members/ researchprofiles.html). I wish to thank Elizabeth Closs Traugott for drawing my attention to this work, while disassociating her from the larger claims being made here.
254 John R. Rickford
2.2
The Acts ofldentity model
Indisputably, however, Le Page's most valuable theoretical legacy to sociolinguistics and creole studies is the Acts of Identity model he pioneered and developed with students and colleagues in his (1974) article, and with Andree TabouretKeller in their (1985) book. The central thesis of principle of this model is that "the individual creates for himself [/herself] the patterns of his [/her] linguistic behavior so as to resemble those of the groups with which from time to time he [I she] wishes to be identified or so as to be unlike those from whom he [/she] wishes to be distinguished~ (LePage and Tabouret-Keller 1985: 181). This central principle is subject to four riders or qualifications, described by Le Page and Tabouret-Keller (1985: 182): in these terms: "We can only behave according to the behavioral patterns of groups we find it desirable to identify with to the extent that: i. we can identify the groups ii. we have both adequate access to the groups and ability to analyze their behavioral patterns iii. the motivation to join the groups is sufficiently powerful, and is either reinforced or reversed by feedback from the groups iv. we have the ability to modify our behavior:' Without going into the details of the evidence (chiefly from sociolinguistic surveys of Belize and St. Lucia) on which this model is based, I would like to discuss what I consider its principal plusses and minuses. Plusses of the Acts ofIdentity model. The first plus of the Acts ofldentity model is that it focuses attention on the social forces and socio-psychological factors that motivate sociolinguistic variation, more so than aey other variationist framework, e.g. Labovian quantitative sociolinguistics (with the exception ofLabovs (1963) study of Martha's Vineyard). It is thus a truly socio-linguistic model and in privileging the social dimension, could make fruitful contact with social theory in the social sciences. and bring about the socio-cultural linguistics that a number of modem sociolinguists (e.g. Mary Bucholtz at UC Santa Barbara) envision. Deborah Cameron (pers. comm., July 1993) criticizes it for seeing language use as reflecting (predetermined) social identity rather than constituting it (who you are and are taken to be depends on the language acts you engage in), but I think the latter perspective is wholly within the spirit and reach of the model The very fact that the model emphasizes the role of the individual as a creative agent ("the individual creates for himself the patterns of his linguistic behavior so as to resemble those of the groups with which from time to time he [/she] wishes to be identified"), projecting various identities through his or her speech acts. places this model in the category of more recent
Le Page's theoretical and applied legacy in sociolinguistics and creole studies 255
approaches in which the constitutive, agentive role of language is emphasized. In the words of Coupland (2001: 208-209), one of the advocates of this agentive approach: It is often the case that we can only identify a "contextual type" by virtue ofthe stylistic attributes of FH's speech. He is the orchestrator of contexts [...] Variation in his speech and in particular his dialect should therefore be said to be not only styled but stylized[ ...] His styles are not situational reflexes. They are ways of subtly activating multiple simultaneous dimensions of meaning potential. [Emphasis in original.]
A second plus of the Acts of Identity model is that in favoring work with small often self-selected groups over correlations with broad social categories (class, race, gender and so on- see Edwards 1983: 309, fn. 6), and being open to the distinctive choices of individuals, the model is in tune with the ethnographic 'communities of practice' (CofP) approach that a number of sociolinguists (e.g. Eckert 2000, Davies 2005) now consider a fruitful additional or alternative way of accounting for sociolinguistic variation. Mallinson and Child's (2008) description of two small communities of practice in Texana, Texas (the "church ladies" and "church sitters') is, to my mind, one of the best exemplars to date of the potential value of the CofP approach, and I think it is squarely in line with the Acts ofldentity model although CofP advocates may not recognize or cite it as such. Minuses of the Acts of Identity model. One weakness of the model, however, was its statistical reliance on cluster analysis, and the strategy of working entirely from the linguistic data to the social groupings rather than the other way around, or through an intermediate strategy. The idea was "to cluster the children according to similarities in their [linguistic] behavior," and then look for significant correlations between the linguistic clusters and "various cultural and socioeconomic indices" of the children and their families (McEntegart and LePage 1982: 106). The authors' decision to approach the analysis in this way; rather than beginning with the social categories was explicitly justified as follows: The decision to avoid the procedure followed by Labov (1966, 1972a), Trudgill (1974b) and G. Sankoff (1974) of dividing the population according to pre-established socio-economic categories and then sampling each category, was taken for several reasons. In the first place, such a procedure precludes discovering anything about the emergent social structure other than in terms of these 'imported' and pre-set categories; in the second place, we did not feel we knew enough about the cultural and economic and ethnic stratification of the two societies to arrange our sample in this way; in the third place, it was part of our aim to discover what social mechanisms were at work, what groups were emerging according to the linguistic symptoms, rather than vice versa. (1982: 107)
Creolists - especially Caribbeanists - might hear in these objections echoes of a similar complaint voiced by David DeCamp in his pioneering (1971: 355) implicational analysis of the Jamaican continuum:
256 John R. Rickford
Why [ ...] have sociolinguists so often correlated their linguistic data to preconceived categories of age, income, education, etc., instead of correlating these non-linguistic variables to the linguistic data? Both the varieties and the defining features of a linear linguistic continuum can be ordered without recourse to the sociolinguistic data, so that these data may then be used to interpret the continuum without circularity of reasoning.
However, while LePage and Tabouret-Keller (1985: 112-153), do go on to present clusters of linguistically similar individuals from their fieldwork and data analysis in Belize (five primary clusters) and St. Lucia (8 clusters, half in interview style and half in informal style), and while they do go on to report and discuss interesting relations between the linguistic clusters and their extra-linguistic correlates,4 they also refer to "weaknesses in our method revealed in McEntegart (1980)" that prevent them from treating their findings and hypotheses as "firm results" (147). And while LePage and Tabouret-Keller said they continue to hold the view that cluster-analysis might provide "a reasonable analogue" for the kinds of multidimensional acts of identity that people might make in everyday life, they end with the observation that: "What we cannot dispute is that owing to faults both in the design and execution of the survey, the statistical results have been meager and inconclusive" (153). The earlier paper by McEntegart and LePage (1982), to which LePage and Tabouret-Keller (1985) also refer, is even more negative. Indeed, it is the most brutally honest and admirable admission of the limitations of the analytical techniques of one's work I have ever seen, concluding that "the survey was far too ambitious in wishing to take account of all social and psychological factors relating to the linguistic behavior of the children at once" and that "there are severe limitations on the usefulness of cluster analysis in sociolinguistics" (123-124). Not having access to the more detailed report by McEntegart (1980), it is not clear whether the weaknesses admitted are really inherent in the method of cluster analysis itself, or the ways in which it was implemented in the gathering and analysis of data from LePage's two research sites. 5 This is a problem that is worth further study, since it may affect what subsequent researchers attempt. 4· For instance, the primary correlate ofllnguistic groupings in Belize is where each individual lives, followed by father's occupation, and more weakly, the family's socio-economic profile (LePage and Tabouret-Keller 1985: 131 ). ForSt Lucia, associations or correlations are related in more complex ways to sex/gender, school level, crowdedness and location of the home, claims to own domestic goods, degree of political activism claimed in the home and language use claimed for the home (ibid: 145). 5· For instance, the difficulties may have had to do with the scale of the project; there are suggestions in McEntegart and LePage (1982) that they needed a bigger sample, but the 280-student sample in Be1lze is already much bigger than that used in most sodolinguistic surveys, even though only 164 students were analyzed in detail The difficulties may also have had to do with
Le Page's theoretical and applied legacy in sociolinguistics and creole studies 257
A second limitation of the Acts of Identity model is that although it provides
in principle for the existence of limits on what speakers can do, in terms of analyzing and adopting the language use of groups with which they'd like to be identified (see riders i-iv above), in practice sociolinguists working within this model assume a much greater competence than their speakers' performance warrants. 6 Contrary to this assumption, note the gaps in pronominal competence at both the basilectal (Creole) and acrolectal (Standard) poles revealed by Correction Tests in Cane Walk, Guyana (Rickford 1987a: 163)7 Or the case of Steve Kin New York City, whose attempts to show that he could produce 100% r-ful pronunciations ended in frustration (Labov 1972: 104-105). Or the Hemnesberget (Norway) speakers who heard themselves switching between Ranamal and Bokmal on tape despite not wanting to (Blom and Gumperz 1972: 430). Although many of us tend to conceptualize sociolinguistic behavior as entirely a matter of choice and strategy, there are real limits on speaker's performance in practice based on limits to their ability
the limited amount of speech analyzed from each child (only 400 words in the case of Belize), or the way in which the envelope of variation was drawn. LePage and Tabouret-Keller 1985: 147-148 discuss the alternative results for St. Lucia obtained by formal versus semantic criteria, and McEntegart's failure to establish "any truly meaningful duster" using the semantic criteria that he favored. Finally, the weakness of the results obtained may, as LePage and TabouretKeller suggest (1985: 132) be due primarily to "a general diffuseness of 'the language of Cayo district' [Belize] and of[... ] a society in a high and rapid state of flux." 6.
Responding to this comment, the editors of this volume have noted (pers. comm.) that: On the point of over-estlmated competence, bear in mind that the act of identity is a psychological act, hence actors might perceive themselves to be competent in whatever variety they are targeting. There are some markers which indicate to speakers that they have 'got it: It is likely that the linguistic markers that linguists (or even the target group) use as a defining feature of the target variety may not be as important (unconsciously) to speakers for them to perceive themselves as competent in the target variety.
This is an interesting possibility, but it may then be impossible to distinguish between an individual who is trying to imitate or adopt a variety with which he or she identifies, and one who is not I think we do have to recognize the limits on individuals' competence to move around in multi-dimensional linguistic space, while conceding that there may be valuable theoretical payoff to distinguishing between the psychological and physical or actual. 7· The Correction Tests were a part of the Controlled Interviews, conducted after all my Spontaneous Sociolinguistic Interviews in Cane Walk, Guyana were over. While the goal of the latter had been to record samples of informal or spontaneous speech, the goal of the former was to elicit speakers' intuitions about language and the socio-cultural milieu. The Correction Tests included a "Creole to Standard English" component in which respondents were asked to translate creole sentences into the "most" standard English they could produce, and a "Standard to Creole English" component in which they were asked to translate Standard English sentences into the deepest creole they could produce. For examples and further discussion, see Rickford ( 1987a).
258
John R. Rickford
and competence - precisely the kinds oflimits LePage and Tabouret-Keller identified in theory. A third and final weakness of the model - contrasted with, for instance, the quantitative variationist sociolinguistics ofLabov and others (see Bayley 2002), or the implicationaVwave model of C.}. Bailey. Bickerton, and DeCamp (see Rickford 2002) - is that it does not pay enough attention to internal linguistic constraints. The key constraints in the four riders above are social and psychological (identification, access, motivation), and although internal constraints might conceivably be included under the fourth rider, involving ability. in practice they are not. There is little if any discussion of internal constraints or conditioning on the linguistic variables in Le Page and Tabouret-Keller (1985) - seep. 135, for example. And Edwards (1983), an excellent introduction to the Acts of Identity model that is required reading in my sociolinguistics graduate seminars, consistently rejects internallinguistic in favor of social or socio-psychological explanations, as in this quotation (300): Rather than looking solely towards a linguistic explanation for the pattern in Table 4 and the other Tables, I take the position that these salient linguistic items form part of the sociocultural milieu of rural and urban Guyana and that individuals and groups of individuals choose to use forms on the basis of their perception of the social value of these forms in the community. In the present case, (i) habitual marker [as against habitual doz or a] enjoys widespread social approval in both rural and urban communities.
This particular case is especially relevant since the linguistic constraint against which Edwards was arguing - a restriction against the use of habitual markers in conditional and temporal clauses first formulated by Bickerton (1975: 30-33) turned out to be supported by all the textual data on Guyanese Creole examined by Rickford (1987b: 126-127). To my mind it takes nothing away from external social and socio-psychological constraints to acknowledge that language is simultaneously influenced by internal, linguistic constraints; and identifying both internal and external constraints is a crucial part of accounting for and predicting sociolinguistic variation, as the Acts ofldentity model seeks to do.
3· Le Page's applied work Robert B. LePage's (1968) paper, "Problems to be faced in the use of English as the medium of education in four West Indian territories," is rarely cited. But it had a profound influence on my career, inspiring me to switch from Literature to a selfdesigned major in Sociolinguistics at the University of California, Santa Cruz, in
Le Page's theoretical and applied legacy in sociolinguistics and creole studies 259
1969.8 Because this paper is relatively unknown, while LePage's theoretical work on the Acts of Identity framework is much more widely cited, I'll begin this assessment ofLe Page's applied legacy by recapitulating its main points in some detail. As he explains in an introductory note, Le Page's (1968) paper was inspired by a seven-week tour he had taken of Jamaica, Trinidad, Guyana (then British Guiana), and Belize (then British Honduras), between March and April1966. Jamaica and Trinidad became independent of Britain in 1962, Guyana did so in 1966, and Belize in 1981, so issues of national development and redefinition were central at the time. The article was also based on Le Page's ten years of teaching and research at the University (College) of the West Indies, especially as Director of the Linguistic Survey of the British West Indies. The list of scholars with whom he had collaborated over this period reads as a veritable "who's who" in early pidgin -creole studies: Beryl Loftman Bailey, Jack Berry, Frederic G. Cassidy, David DeCamp, Morris Goodman, Robert A. Hall, Jr., Douglas Taylor, R. Wallace Thompson, Jan Voorhoeve, and "former students" Miss Jean Creary (later D'Costa) and Mervyn Alleyne.9 Le Page begins his paper in much the same way that other recent commentators discussing the educational challenges facing students who speak creole and vernacular varieties of English have done (see D. Craig 1999, H. Craig et al. 2009, Labov 2006, Rickford and Rickford 2007) - by pointing to the low rates at which such students often succeed in exams in English and other subjects. In the case of the Anglophone Caribbean territories LePage was considering, 10 the data came from students'
8. & noted in Rickford (1997: 162): "[... ] what really helped me to abandon English literature and design my own major in sociolinguistics was a paper by Le Page 1968 which dealt with the high failure rate (70% to 90%) of Caribbean high school students on the English language GCE (General Certificate of Education) "0" level exam set by London and Cambridge Universities. Having worked as a high school teacher in Guyana for one year before setting off for college, I was aware of the problem, and I was convinced by Le Page's arguments that it resulted partly from the fact that teachers could not recognize the differences between local creole and Standard English, nor help students to shift smoothly between the two varieties [... ] Armed with Le Page's guidelines [for training English language specialists to help teachers improve their methods] and assisted by a liberal and innovative college environment [at the University of California, Santa Cruz], I combined courses in linguistics, anthropology and other fields, and graduated with a self-designed major in sociolinguistics in 1971." 9· From Romaine (2006), we learn that LePage also had as students the Nobel prize-winning Caribbean poet and playwright Derek Walcott, and theoretical linguists David Baker and Geoffrey Pullum. Of Walcott, LePage had this to say in his 1998 biography: "I certainly dont feel that I taught him anything, although he seemed to enjoy reading Anglo-Saxon and Middle English poetry with me" (54).
to. LePage also introduced comparable data from Cyprus, and for other subjects (Geography, History, French, Biology, Physics and Chemistry) besides the ones listed in Table l.
260 John R. Rickford
Table 1. Pass rates on London G.C.E. "0" level exams in various subjects, 1962 (adapted from Le Page's 1968: 433 Table 1) Subject
English Language English Literature Pure Mathematics
Guyana
Belize
Jamaica
Trinidad
Barbados
19.6% (2483)
21.0% (24)
19.4% (661)
23.1% (1521)
10.7% (150)
21.5% (1245)
0% (1)
13.0% (79)
12.3% (349)
24.2% (91)
24.6% (898)
60.0% (5)
7.4% (135)
17.3% (659)
8.9% (45)
performance on the 1962 London University General Certificate of Education "Ordinary» level exams, set and graded in England at the time (see Table 1). Le Page does not dwell on the statistics, but these uniformly low pass rates- or conversely, these uniformly high failure rates of 80% or more- clearly represented a problem of enormous proportions, especially since the GCE exams constituted "the passport to a Civil Service job, entry to a training college or to the University [... ] like a sluice gate controlling the flow of manpower into the educated roles that should provide the dynamism for the economic and cultural growth of the countries concerned" (432). Le Page provided a shrewd analysis of the problems to be faced in teaching Standard English to native Creole speakers in the classroom, and in expecting them to perform we 11 on exams that require a high degree of competence in Standard English without preparing them to bridge the gap between these varieties. Neither the children, nor their school teachers, nor their training college and university educators, he felt, were equipped to recognize the systematic phonological, grammatical lexical and semantic differences between Creole speech and "the spoken dialects that underlie the standard usage of the textbooks and of the examiners" (435). And "Instead of being able to keep the two systems separate,[ ... ] the children try to make one composite system out of the vernacular they know in their homes and the model language they are supposedly taught in schoo~ the result naturally satisfies nobody» (ibid.). Furthermore, notes Le Page, teaching of language skills is a time-consuming, prescriptive and unproductive task, made worse by the fact that many of the teachers were untrained, and their own competence in the standard was often precarious. Creole speech was often dismissed as "bad talk," and children were often "inhibited from any kind of creative expression [... ]and the prizes go to the best mimics rather than to the most talented" (438). LePage's proposed solutions to these problems included establishing specialist posts in English language teacher training at teacher training colleges in the West
Le Page's theoretical and applied legacy in sociolinguistics and creole studies
Indies, with terms comparable to that of a university lecturer. Such specialists should have a thorough basic training in linguistics, psychological and sociological aspects of linguistic behavior, the psychology oflanguage learning, the processes of creolization, the principles of contrastive analysis, and the structure of the languages involved in their situation (e.g. Creole English, Creole French, English, Spanish, Maya). They must also be trained in the general principles of education, in the preparation of teaching materials, and in the use of audiovisual aids, radio and television. (440)
And their goals should be to develop trainee-teachers' competence in educated standard English, to prepare culturally relevant teaching materials that provide contrastive analysis between the creole and the standard, to provide retraining for existing teachers, and to develop radio and television programs to disseminate new approaches to teaching English and the language arts more widely. Although now over forty years old, LePage's diagnosis of the language-related educational challenges facing creole speakers in the Anglophone Caribbean is still relevant, even though attitudes to the creole are not quite as vituperative these days, and many (not all) teachers are more informed and progressive than the ones he encountered and quoted. 11 Like other contemporary scholars in sociolinguistics and creole studies (cf. Cassidy 1970, DeCamp 1972), LePage did not go significantly beyond the diagnostic and programma tic level in dealing with the educational issues. But he would undoubtedly have welcomed the studies of educational policy and practice, the instructional texts, edited collections, teacher guides and other works to help creole and vernacular speakers in the classroom that were produced by subsequent researchers, including Lawrence Carrington (1976), Pauline Christie (2003), Dennis Craig (1980, 1999, 2006), HubertDevonish (e.g. Devonish and Carpenter 2007), Shondel Nero (2001, 2006), Velma Pollard (1999), Ian Robertson (1996), Hazel Simmons-McDonald (1996, 2006), McDonald et al. 1997), and Jeff Siegel (1992, 1999, 2007), among others. 12 Before leaving the applied area, where fve focused on the 1968 paper that was most personally meaningful to me (see footnote viii), I should mention for completeness some ofLe Page's other publications in the area of applied (socio)linguistics, broadly conceived. Perhaps the one most related to the 1968 paper in its focus was "Caribbean Connections in the Classroom: A pamphlet of guidance for teachers concerned n. But see Craig (1999) for evidence that negative attitudes and poor pass rates persist. 12. See also Rickford and Rickford (2007, 20 10), and The Readi11g Road instructional program for AAVE and Latino-English speaking students in the USA developed by William Labovand the Penn Reading Initiative, which can be downloaded free of charge at http://wwwling.upenn.edu!pri/.
261
262
John R. Rickford
with the language problems of children of Afro-Caribbean descent~ Produced in 1981 by the University of York, where Le Page worked from 1964 until he retired in 1988, it was available for 75 pence from the Institute of Linguists Educational Trust [ILEA] in London. It was intended for teachers trying to help Caribbean immigrants and the their British-born children overcome linguistic and other barriers to success in British schools, and included a sketch of the pronunciation, grammar and vocabulary of Jamaican (the dominant Anglophone Caribbean variety among immigrants to Britain), with texts, and notes on language variants in other territories. LePage felt (16-18) that West Indian language was indeed part of the "Language Problem in Britain;' in part because of direct interference or transfer from the West Indian vernaculars to Standard English, but even more so because of the indirect effect of language stereotyping and prejudice on the basis of the children's dialects, 13 and because of children's proclivity to use "London Jamaican" as an Act of Identity. 14 His advice to teachers (20 ff.) was to try to understand as much as possible about their pupils, their parents and their home language, and he encouraged them to visit and talk with them in their homes. He also expressed the hope (22) that sociolinguistics might be introduced as a high school and exam subject, because There is probably no discipline which deals more effectively with, and more effectively diffuses, racial or class tensions. [... ) To have an effective "language studies and linguistics" programme will involve enlisting the cooperation of all the teachers in the school, because their demand for [...] "correct English" may well be called into question. But [...] the understanding which can come from a "language studies and linguistics" programme may itself be instrumental in providing some of the motivation for children to improve their command of Standard English, and it can help teachers understand that the worst way of achieving good results is by concentrating on "mistakes':
Note also the pedagogical value of the Dictionary of]amaican English (coauthored with Cassidy in 1967, second edition 1980), which was an invaluable predecessor to more general works like Richard Allsopp's Dictionary of Caribbea.n English Usage, and Jeannette Allsopp's The Ca1·ibbean Multilingual Dictiona.ry, which were even more deliberately geared towards classroom teachers. 13. In this sense, Le Page's assessment was similar to that ofJ ustice Joiner in the "Black English" trial In Ann Arbor, Michigan, who in his 1979 ruling against the Ann Arbor, Michigan school district for failing to take African American Vernacular English into account, cited potentially negative attitudes of school teachers and district officials towards this variety rather than its structural features as posing the greatest barrier to equal educational opportunity for its speakers.
14 Indeed, Le Page drew substantially on his Acts of Identity theory In this pamphlet - four years before the publication of the (1985) book with Tabouret-Keller that would make it more widely known to the field of Linguistics.
Le Page's theoretical and applied legacy in sociolinguistics and creole studies 263
In the category of Applied Linguistics must also be placed Le Page's short (81 pages) but widely read (1964) book: The National Language Question: Linguistic Problems of Newly Independent Sta.tes. In the spirit of other macro-sociolinguistic works of the 1960s that focused on sociolinguistic surveys of newly independent nations and the role oflanguage in national development (cf. Ferguson 1962, 1966, and of course Fishman et al. 1968), LePage's book deftly discussed general considerations in choosing between indigenous, local and international options in the selection of an official or national language, and illustrated it with case-studies of Malaysia (where he had taught for four years), and India. LePage mentions in his (1998) biography that this book. reprinted in 1967 and 1971, "was to affect my future in quite unforeseen ways, leading as it did to invitations to conferences in the United States and Canada and East Africa and London at which I met my future collaborator from Strasbourg [Andre Tabouret-Keller] whose interests and resources dovetailed very closely and in a complementary way with my own" (147). Which indeed brings us to the final work in this category, his (1997) Vernacular Literacy book. co-edited with Andre Tabouret-Keller (as lead editor), Penelope Gardner-Chloros, and Gabrielle Varro. This book, which remains an invaluable reference work, represents a re-evaluation of the famous UNESCO (1953) monograph and its central axioms that "every child of school age should attend school and that every illiterate should be made literate" (6), and that "the best medium for teaching is the mother tongue of the pupie' The first part of the book is made up of four essays dealing with general issues, including political, economic, social, orthographic and pedagogical considerations. The second part includes four more essays that are all case-studies, drawn from Europe, various parts of Africa and Asia, and pidgin and creole-spealdng communities. Tabouret-Keller wrote the Conclusion, while LePage wrote the Introduction, and authored/edited the longest of the essays in the first part, a 59-page essay entitled "Political and economic aspects of vernacular literacf. 15 The essay is a complex work that surveys vernacular literacy in Britain (especially Scotland), Europe (especially among the Basques), India, Africa (including pidgin varieties in Cameroon and Nigeria), and Latin America. A couple of its "tentative conclusions" -less sanguine than the axioms of the UNESCO (1953) report- are worth quoting here: in many of the countries which we have considered, and in some which we have not, political and ethnic and economic and religious turmoil is at present so great
15. Because Ven1acular Literacy is the outcome of a series of workshops from 1986 to 1992 and its essays draw on the contributions of various participants, each essay is described as "edited" by one or two researchers, but each "editor" must have served as the primary "author" of their essay as well In the case of"Political and Economic Aspects ofVernacular Literacy; twenty-one other "contributors" are listed, besides Le Page.
264
John R. Rickford that academic prognoses about literacy are likely to remain academic; and the academic community, with its Unesco and other conferences, resolutions and recommendations, seems frequently to operate in a sealed world. [...] it is economic motivation which appears the most powerful in affecting vernacular literacy. [...] Literacy is to some extent a by-product of prosperity and this makes it appear to many the only escape route from poverty at least for their children; whether vernacular literacy will provide as effective an escape as literacy in some other language is often a puzzle, so complex is the interplay of conditions. [... ) What is certainly true is that a literate vernacular, replacing a previous language of dominance, is often, as with the Basques, the leading banner around which those seeking a power base focus nascent ethnic identities and nationhood. Yet, in spite of the very considerable efforts made among the Basques, the Welsh, the Irish and the Scots to encode national identity in their languages, labour migration and the economic pull of other languages make their successes seem very limited and temporary while their nationalism and hunger for power becomes more threatening. One other aspect of Le Page's work that might be considered "applied," but also "theoretical," is the significant role he played in providing fellowship support and training for Caribbean and Third World sociolinguists at York University (see Christie 2008: 138, Romaine 2006: 669), and in making it possible for them in turn to train others. The list of linguists who got their doctorates in Linguistics under Le Page includes the Caribbean creolists Pauline Christie, Hubert Devonish, Walter Edwards, Kean Gibson, Donald Winford and Colville Young. Christie and Edwards in tum became major contributors to the Acts ofldentity model that was Le Page's theoretical tour de force - Christie as a central field worker and data analyst in the Caribbean communities in which the model was developed (she is coauthor ofLe Page et al. 1974, and editor of his 2001 festschrift), Edwards as a major expositor of and advocate for the model itself~ drawing on data both from Guyanese Creole (Edwards 1978) and African American Vernacular English (Edwards 1992).
4· Conclusion: Problems to be faced in building on Le Page's legacy The 1968 publication of Le Page's that first attracted me to his work and drew me into linguistics had an alluring but antiquated (verbose) character: "Problems to be faced in the use of English as the Medium of Education in Four West Indian Territories~ The notion of"problems to be faced" provides a good conceptual perspective from which to ask how we might build on Le Page's theoretical and applied legacy in sociolinguistics and creole studies. It should first be noted that we are at a period in the history of both subfields in which some of our pioneers are either retiring or passing on - besides Le Page,
Le Page's theoretical and applied legacy in sociolinguistics and creole studies 2.65
Richard Allsopp. Frederic Cassidy, Dennis Craig and Dell Hymes have died in recent years. Each of these stalwarts has left a scholarly legacy, but the extent to which these legacies are being developed varies, partly depending on the extent to which these pioneers were survived by colleagues and aficionados interested in continuing that line of work, partly depending on how students are being trained and what they are being exposed to in universities in the Caribbean, North America, and elsewhere, and partly depending on the whims of scholarly fashion in academia. For instance, the lexicographical work of Richard Allsopp is being actively carried on by his collaborator and widow, Jeannette Allsopp, and her students at the University of the West Indies, Cave Hill, Barbados, despite problems to be faced in terms of funding. Dennis Craig's attempts to deal with the challenges of vernacular literacy in the Caribbean are being extended, and in some respects surpassed by the continuing work of Hazel Simmons-McDonald, Ian Robertson, Hubert Devonish and other faculty members and students at the University of the West Indies in Barbados, Trinidad and Jamaica. To the extent that each of these areas is developed, those parts of Le Page's legacy that overlap with them are also assured of growth. Indeed, the work of Devonish and his colleagues and students at the Jamaica Language Unit at UWI, Mona, Jamaica is a significant development of an idea that LePage had only passingly proposed in the 1950s that Jamaican school children be taught in creole rather than standard English for the first year or two of schooling. Cassidy (1970), citing LePage's proposal, noted that Kingston Star (Jamaica) columnist Vere John condemned it as a "pernicious and insulting idea~ Today the idea of initial bilingual instruction in Jamaican Creole as well as English still attracts some criticism in the local media, but it also attracts some praise, and with government support it is being implemented experimentally in several schools (Devonish and Carpenter 2007). Meanwhile, bible translation and experimental TV programs in Jamaican Creole are underway. On the descriptive and theoretical front, it is more difficult to be sanguine about the development of Le Page's legacy. 16 On the one hand, many sociolinguistics scholars outside of the Caribbean are not familiar enough with Le Page's Acts of Identity model or they have not tried hard enough to integrate it with newer approaches to language and identity or other variationist models. On the other hand, despite the fact that the Le Page/Tabouret-Keller model was developed with Caribbean data and that competing quantitative and implicational/dynamic models were also extensively applied to Caribbean situations in the 1970s and
16. Except with respect to detailed socio-historical research, where, as noted above, the tradi-
tion that Le Page helped to start continues to thrive in creole studies, if not sociolinguistics.
266 John R. Rickford
1980s, 17 variationist approaches are rarely taught or practiced on university campuses in the Caribbean today. Although there has been quite an efflorescence of variationist research on English-based Caribbean varieties over the past decade, 18 with one or two exceptions (e.g. Irvine 2004, 2008), this has not come from graduates or faculty of Linguistics programs in Caribbean universities, and the work has been broadly quantitative or corpus-based, rather than drawing on the Acts of Identity or Implicational models. To some extent the neglect of Le Page's Acts of Identity model may be due to the fact that the last major work in this framework (Le Page and Tabouret Keller 1985) was published twenty-five years ago. In this respect, as with work in the implicationalist and continuum frameworks from the 1980s, it may seem to be passe. But the theoretical and methodological plusses that each of these variationist approaches offered are real and significant, and neither their age nor the fact that each has minuses of some sort should deter us from attempting to develop them. In the case of the Acts of Identity model one problem to be faced in this respect is developing our understanding of anthropology and social and cognitive psychology so that we can capitalize on the cultural and sociocultural aspects of sociolinguistic behavior and the concern with individuals as well as groups that Le Page's approach came closest to capturing. Another is understanding whether the difficulties with using cluster analysis and working from the linguistic to the social that Le Page and his colleagues so honestly admitted were a function of the specific ways in which they attempted to implement these approaches, or whether they are more general perhaps inhering in the approaches themselves. Neither of these problems can be solved without perusing the publications Le Page and his 17. The marked linguistic variability which the region displays makes it an ideal laboratory for studies of sociolinguistic variability. For instance, the linguistic span of English-based varleties in Jamaica or Guyana is broader than that in Norwich or New York or most metropolitan cities where some of the classic sociolinguistic studies were done. And this is not even including places like St. Lucta and Trinidad and Belize where substantial segments of the population speak another language (e.g. French Creole, Spanish, Garifuna) besldes (creole or standard) English.
18. See for instance Aceto and Williams (2003), Blake (forthcoming), Deuber (2009), Fenigsen (2002), Gooden (2008), Hackert (2006, 2008), Hinrichs and White-Sustafta (forthcoming), Irvine (2004, 2008), Lacoste (2008), Leglise and M!gge (2007), Mair (2009), Meyerhoff and Walker (2007), Patrick (2009), Poplack and Tagliamonte (2001), Rickford (2010), Seymour (2009), Sidnell (1999, 2002), Straw and Patrick (2007), Walker and Meyerhoff (2006), Wassink (2001), and Wassink and Dyer (2004). A number of these projects draw on the ICE-Jamaica corpus that was assembled in collaboration between the University of the West Indies Mona and Frelburg University (on the corpus see Mair 2002 and 2009). This list of recent variationist or sociolinguistic works focuses on Caribbean English-lexicon creoles and vernaculars, but even so, is representative rather than exhaustive. For additional references, see individual articles within edited collections in this list, e.g. Aceto and Williams (2003), and references within each of these works.
Le Page's theoretical and applied legacy in sociolinguistics and creole studies 267
colleagues bequeathed to us- including his (1998) autobiography and the more inaccessible report of McEntegart (1980) -and by conducting new research that pushes the model to its limits and beyond. It may also be useful to discuss in detail these and other problems with colleagues like Pauline Christie, Andree Tabouret Keller, and Damian McEntegart, who worked closely with LePage on the Belize and St. Lucia projects, and who may have inside insights to share with us. Finally, it is worth remembering, in closing, that among the front-runners in sociolinguistics and creole studies, only a few scholars - for instance, Hubert Devonish, William Labov, Suzanne Romaine, Jeff Siegel and Walt Wolfram - resemble Robert B. Le Page in combining theoretical, descriptive and applied research at a sustained level over several decades. We should continue to develop the pioneering initiatives and ideas he developed on all three fronts.
References Aceto, M. & Williams, J.P. (eds.). 2003. Contact Englishes of the Eastern Caribbean. Amsterdam: John Benjamins. Allsopp, J. 2003. The Caribbean Multilingual Dictionary of Flora, Fauna and Foods, In English, Fre11ch, French Creole and Spanish. Kingston. Jamaica: Arawak. Allsopp, R (ed..). 1996. Dictionary of Caribbean E11glish Usage. Oxford: OUP. Baker, P. & Corne, C. 1982. Isle de France Creole: Ajfinitle.s and Origins. Ann Arbor Ml: Karoma Bayley, R 2002. The quantitative paradigm. In The Ha11dbook ofLanguage Variation a11d Change, J.K. Chambers, P. Trudgill & N. Schilling-Estes (eds.), 117-141. Malden MA: Blackwell Bickerton, D. 197 5. Dy11amics of a Creole System. Cambridge: CUP. Blake, R Forthcoming. Not as clear as Black and White: Race, class and language in a Barbados community. In Black Language - The United States and the English Speaking Caribbean: Education, History, Structure and Use, A. Spears (ed..). Malden MA: IDackwell. Blom, J.P. & Gumperz, J. J. 1972. Social meaning in linguistic structure: Code-switching in Norway. InDirections in Sociolinguistics: The Ethnography ojCommunicatlo11, J.J. Gumperz & D. Hymes (eds.), 407-434. New York: Holt, Rinehart & Winston. Bucholtz, M. & Hall, K. 2004. Language and identity. In A Compa11ion to Linguistic A11thropology, A. Duranti (ed..), 369-394. Malden MA: Blackwell Carrington, Lawrence. 1976. Determining language education policy in Caribbean sociolinguistic complexes. International Joun1al of the Sociology ofLanguage 8: 127-43. Cassidy, F. G. 1970. Thaching Standard English to speakers of Creole in Jamaica, West Indies. In Report of the TWentieth Annual Round Table on Languages and Linguistics: Linguistics a11d the Teaching of Standard E11glish to Speakers of Other Languages and Dialects, J.E. Alatis (ed..), 203-214. Washington DC: Georgetown University Press. Cassidy, F.G. & LePage, RB. 1980[1967]. Dictionary ofJamaican English. Cambridge: CUP. Christie, P. (ed..). 2001. Due Respect: Papers on E11glish a11d E,~glish- Related Creoles in the Caribbean in Honour ofProfessor Robert LePage. Kingston. Jamaica: University of the West Indies Press. Christie, P. 2003. Language in Jamaica. Kingston, Jamaica: Arawak.
268
John R. Rickford Christie, P. 2008. Obituary: Robert Brock Le Page, December 8, 1920 - January 12, 2006. Journal of Pidgin and Creole Languages 23( 1): 136-40. Coupland, N. 2001. Language, situation, and the relational self: Theorizing dialect-style in sociolinguistics. Style and Sociolinguistic Variation, P. Eckert & J. R. Rickford (eds.), 185-210. Cambrldge: CUP. Craig, D.R. 1980. Models for educational policy in creole-speaking communities. In Theoretical Orientations in Creole Studies, A. Valdman & A. Highfield (eds.), 245-95. New York NY: Academic Press. Craig, D.R 1999. Teaching Language and Literacy: Policies and Procedures for Vernacular Situations. Georgetown, Guyana: Education and Development Services. Craig, D. R. 2006. The use of the vernacular in West Indian education. In Exploring the Boundaries of Caribbean Creole Languages, H. Simmons-McDonald & I. Robertson (eds.), 99-117. Kingston. Jamaica: University of the West Indies Press. Craig, H. K., Zhang, L, Hensel, S. & Quinn, E. 2009. African American English-speaking students: An examination of the relationship between dialect shifting and reading outcomes. Journal ofSpeech, La11guage and Hearing Research 52: 839-55. Davies, B. 2005. Communities of practice: Legitimacy not choice. Journal ofSociolingulstics 9( 4): 557-581.
DeCamp, D. 1971. Toward a generative analysis of a post-creole speech continuum. In Pidginizatlo1! and Creolizatlo11 ofLanguage.s, Dell Hymes (ed), 349-370. Cambridge: CUP. DeCamp, D. 1972. Standard English books and creole speaking children: A linguist's appraisal. Paper presented at the UWI/UNESCO Conference on Creole Languages and Educational Development, University of the West Indies, St. Augustine, Trinidad, July 24-28, 1972. Deuber, D. 2009. 'The English we speaking': Morphological and syntactic variation in educated Jamaican speech. Journal ofPidgl11 and Creole Languages 24(1): 1-52. Devonish, H. & Carpenter, K. 2007. Towards full bilingualism in education: The Jamaican bilingual primary education project. Social a11d Economic Studie.s 56: 277-303. Eckert, P. 2000. Linguistic Variation as Social Practice: The Linguistic Co11struct1on of Identity 111 Belten High. Malden MA: IDackwell. Edwards, W.E. 1978. Tantalisin and Busin in Guyana. Anthropological Li11guistics 20(5): 194-213. Edwards, W.F. 1983. Code selection and shifting in Guyana. Language in Society 12(3): 295-311. Edwards, W.F. 1992. Sociolinguistic behavior in a Detroit inner-city IDack neighborhood La11guage i11 Society 21(1): 93-115. Fenigsen, J. 2002. Language ideologies in Barbados: Processes and paradigms. Pragmatics 13(4): 457-481.
Ferguson, CA. 1962. The language factor in national development. Anthropological Linguistics 4(1): 23-27.
Ferguson, C.A. 1966. National sociolinguistic profile formulas. In Sociolinguistics, W. Bright (ed), 309-314. The Hague: Mouton. Fitzmaurice, S. 2000. Coalitions and the investigation of social influence in sociolinguistic history. European Journal ofEnglish Studies 4(3): 265-276. Gooden. S. 2008. Discourse aspects of tense marking in Belizean Creole. English World- Wide 29(3): 306-347. Hackert, S. 2006. Oral narrative and tense in urban Bahamian Creole English. In Structure and VariaNon In Language Contact, A. Deumert & S. Durrlem.an-Tame (eds.), 225-242.
Amsterdam: John Benjamins.
Le Page's theoretical and applied legacy in sociolinguistics and creole studies 269 Hackert, S. 2008. Circumscribing the variable context in quantitative studies of past inflection Language Variation and Change 20: 127-153. Hinrichs, L. & White-Sustafta, J. Forthcoming. Global Englishes and the sociolinguistics of spell~ A study of]amaican blog and email writing. English World-Wide 32(1). Hymes, D. (ed). 1971. PidglnizaNon and Creolization of Languages: Proceedings of a Conference Held at the University of the West Indies, Mona, Jamaica, Aprll1968. Cambridge: CUP. Irvine, A. 2004. A good command of the English language: Phonological variation in the Jamaican acrolect. Journal of Pidgin and Creole Languages 19: 41-76. Irvine, A 2008. Contrast and convergence in Standard Jamaican English: The Phonological architecture of the standard in an ideologically bidialectal community. World Engllshes 27: 9-25. Jucker, A 2008. Historical pragmatics. Language and IJnguistlcs Compass 2: 894-906. Kouwenberg, S. 2009. The demographic context of creollzation in early English Jamaica, 16551700. Gradual CreolizaNon: Studies Celebrating Jacques Arends, M. van den Berg, H. C. Cardoso & R Selbach (eds.), 327-348. Amsterdam: John Benjamins. Kroch, A & Taylor, A 1997. Verb movement in Old and Middle English: Dialect variation and language contact Parameters ofMorphosyntactic Change, A. van Kemenade & N. Vincent (eds.), 297-325. Cambridge: CUP. Labov, W. 1963. The social motivation of a sound change. Word 19: 273-309. Labov, W. 1966. The Social Stratification ofE11glish in New York City. Washington, D.C.: Center for Applied Linguistics. Labov, W. 1972. Socioli11guistic Patter11s. Philadelphia PA: University of Pennsylvania Press. Labov, W. 2006. Spotlight on reading. Presentation to various groups, including the West Philadelphia Thtoring Project and America Reads, and to the Voices of African American Students group of educators in Los Angeles. . Lacoste, V. 2008. Learning the Sounds of Standard Jamaican English: Variationist, Pedagogical and Phonological Perspectives on 7- year-Old Childrerls Classroom Speech. PhD dissertation, University of Essex. Le Page, RB. 1960. An historical introduction to Jamaican Creole. In Jamaican Creole, R B. Le Page (ed), 1-124. Creole Language Studies 1. London: Macmillan Le Page, RB. 1964. The Natlo11al Language Question: Linguistic Problems of Newly I11dependent States. Oxford: OUP. Le Page, RB. 1968. Problems to be fuced in the use of English as the medium of education in four West Indian territories. In Language Problems of Developing Nations, C. Ferguson, J. Fishman & J. Das Gupta (eds.), 431-442. New York NY: Wiley. LePage, RB. 1981. Caribbea11 Com1ections in the Classroom. London: The Mary Glasgow Language Trust (Distributed through the Institute of Linguists Educational Trust). Le Page, RB. 1998. Ivory Towers: Memoirs of a Pidgin Fancier. Kingston, Jamaica: Society for Caribbean Linguistics. LePage, RB. & Tabouret-Keller, A 1985. Acts of Identity: Creole-Based Approaches to Language and Ethnlcity. Cambridge: CUP. LePage, RB., Christie, P., Jurdant, B., Weekes, AJ. & Tabouret-Keller, A 1974. Further report on the Sociolinguistic Survey of Multilingual Communities: Survey of Cayo District, British Honduras. Language in Society 3: 1-32. Leglise, I. & Migge, B. (eds.). 2007. PraNques et representations lingulsNques en Guyane: Regards Croises. Paris: Institut de Recherche pour le Developpement. Mai.r, C. 2002. Creolisms in an emerging standard: Written English in Jamaica. English WorldWlde23: 31-58.
270
John R. Rickford Mair, C. 2009. Corpus linguistics meets sociolinguistics. In World Englishes- Problems, Properties and Prospects: Selected Papers from the 13th IAWE Conference, T. Hoffmann & L. Siebers (eds.), 39-60. Amsterdam: John Benjamins. McEntegart, D. 1980. Final report and appraisal ofthe collection and analysis of the data in the sociolinguistic survey of multilingual communities, stages I (Belize) and II (St. Lucia). London: SSRC. McEntegart, D, & LePage, R.B. 1982. Appraisal of the statistical techniques used in the Sociolinguistic Survey of Multilingual Communitles. In Sociolinguistic Variation in Speech Communities, S. Romaine (ed), 105-124. London: Edward Arnold Meyerhoff, M. & Walker, J. 2007. The persistence of variation in individual grammars: Copula absence in 'urban sojourners' and their stay at home peers, Bequia (St. Vincent and the Grenadines). Journal of Sociolinguistics 11: 346-66. Nero, S. 2001. Englishes in Contact: Anglophone Caribbean Students in an Urban College. Cresskill NJ: Hampton. Nero, S. (ed). 2006. Dialects, Englishes, Creoles, a11d Education. Mahwah NJ: Lawrence Erlbaum Associates. Nevalainen, T. & Raumolin-Brunberg, H. 2003. Historical Socioli11gulstics: Language Change 111 Tudor and Stuart England. London: Pearson. Patrick, P. 2009. Pidgins, creoles and linguistic variation. In The Ha11dbook of Pidgins and Creoles, S. Kouwenberg & J.V. Singler (eds.), 461-487. Malden MA: Blackwell Pollard, V. 1999. Beyond grammar: Teaching English in an Anglophone Creole environment. In Creole Ge11esis, Attitudes and Discourse, J.R. Rickford & S. Romaine (eds.), 323-333. Amsterdam: John Benjamins. Poplack, S. & Tagliamonte, S. 2001. African America11 E11glish in the Diaspora: Tense a11d Aspect. Malden MA: Blackwell Rickford, A. & Rickford, J. R 2007. Variation, versatility, and contrastive analysis in the classroom. In Sociolinguistic Variatio11: 1heorie.s, Methods, and Applications, R Bayley & C. Lucas (eds.), 276-296. Cambridge: CUP. Rickford, A. & Rickford, J.R. 2010. From outside agitators to inside implementers: Improving the literacy education of vernacular and creole speakers. In Eth,wlinguistic Diversity and Education: Language, Literacy and Culture, M. Farr, L. Seloni & J. Song (eds.), 241-259. London: Routledge. Rickford, J.R. 1987a. The Haves and Have Nots: sociolinguistic surveys and the assessment of speaker competence." Language in Society 16: 149-177. Rickford, J.R. 1987b. Dimensions of a Creole Co11tinuum: History, Texts and Li11guistic A11arysis of Guyanese Creole. Stanford CA: Stanford University Press. Rickford, J.R. 1997. Unequal partnership: Sociolinguistics and the African American community. Language in Society26: 161-197. Rickford, J.R. 2002. Implicattonal scales. In The Handbook of Language Variation and Change, Chambers, J. K., Trudgill, P. & N. Schilling-Estes (eds. ), 142-167. Oxford: BlackwelL Roberts, S. 2004. The role of style and identity in the development of Hawaiian Creole. In Creoles, Contact, and Language Change: Linguistics and Social Implications, G. Escure & A. Schwegler (eds.), 331-350. Amsterdam: John Benjamins. Roberts, S. 2005. The Emergence of Hawai'i Creole English in the Early 20th Century: The Sociohistorical Context of Creole Genesis. PhD dissertation, Stanford University. Robertson, I. 1996. Language education policy (1): Towards a rational approach for Caribbean states. In Caribbean Language Issues Old and New: Papers In Honor of Professor Mervyn
Le Page's theoretical and applied legacy in sociolinguistics and creole studies
Alleyne on the Occasion of his Sixtieth Birthday, P. Christie (ed), 112-119. Kingston, Jamaica: University of the West Indies Press. Romaine, S. 1982. Socio-hlstorical Linguistics: Its Status and Methodology. Cambridge: CUP. Romaine, S. 1992. Language, Education and Development: Urban and Rural Tok Plsin in Papua New Guinea. Oxford: ClarendoiL Romaine, S. 2006. Obituary. In Memoriam, Robert BrockLe Page, 8 December 1920- 12 January 2006./ournal ofSociolingulstics 10: 667-673. Seymour, K. C. 2009. "Dis How it Does Go": The Organisation oflmperfective Aspect in Urban Bahamian Creole Engllsh. PhD dlssertation, New York University. Sidnell, J. 1999. Gender and pronominal variation in an Indo-Guyanese creole-speaking community. Language in Society 28(3): 367-399. Sidnell, J. 2002. Habitual and imperfective in Guyanese Creole. Journal of Pidgin and Creole Languages 17(2): 151-182. Siegel, J. (ed). 1992. Pidgins, Creoles, and Nonsta11dard Dialects 111 Education [Occasional Paper 12]. Canberrra: Applied Linguistics Association of Australia. Siegel, J. 1999. Creoles and minority dialects in education: An overview. Journal ofMultilingual and Multicultural Developme11t 20: 508-531. Siegel, J. 2007. Creoles and minority dialects in education: An update. Language in Education 21: 66-86.
Simmons-McDonald, H. 1996. Language Education Policy (2): The case for creole in formal education in St Lucia. In Caribbea11 Language Issues Old and New: Papers i11lfmor ofProfessor Mervyn Alley11e on the Occasion of his Sixtieth Birthday, P. Christie (ed), 120-142. Kingston, Jamaica: University of the West Indies Press. Simmons-McDonald, H. 2006. Vernacular instruction and hi-literacy development: French Creole speakers. Rtplorb1g the Bou11daries ofCaribbean Creole Languages, H. Simmons-McDonald & I. Robertson (eds.), 118-146. Kingston, Jamaica: University of the West Indies Press. Simmons-McDonald, H., Fields, L. & Roberts, P. 1997. Writing 111 English: A Course Book for Caribbean Stude11ts. Kingston, Jamaica: Ian Randle. Singler, J. V. 1996. Theories of creole genesis, sociohlstorical considerations, and the evaluation of evidence: The case of Haitian Creole and the Relexification Hypothesis. Journal ofPidgin and Creole Language.s 11, 185-230. Singler, J. 2007. Samana and Sinoe, part II: Provenance. Journal of Pidgi11 a11d Creole Language.s 22: 309-346.
Straw, M. & Patrick. P. 2007. Dialect acquisition of glottal variation in /t/: Barbadians in Ipswich. Language Sciences 29: 385-407. Tabouret-Keller, A., LePage, R. B., Gardner-Chloros, P. & Varro, G. (eds.). 1997. Vernacular Literacy: A Re-Evaluation. Oxford: OUP. UNESCO (United Nations Educational, Scientific, and Cultural 0 rganization). 1953. The Use of Vernacular Languages In Education. Monograph on Fundamental Education VIII. Paris. Walker, J. & Meyerhoff, M. 2006. Zero copula in the Eastern Caribbean: Evidence from Bequia American Speech 81: 146-163. Wassink, A. B. 2001. Theme and variation in Jamaican vowels. Language Variation and Change 13: 135-159.
Wassink, A. B. & Dyer, J. 2004. Language ideology and transmission of phonological change. Journal of English Linguistics 32(1): 3-30.
271
Name index A Alleyne, M.C. 259 Allsopp, J. 262. Allsopp, R. 78,142, 1.45, 15S, 172., 262,265
G Grillo, R.D. 244 fn. Gwnperz, J,J, 149, 257-25S
H
Halliday, M.A.K. 1oS B
Hemnann, T. 83 Hewitt. R. 233,247 Hinrichs, L. 135-136, 141-142, 14S, 154 fn., 158-159. 257 fn. Hoffmann, T. S5, 99 Hymes, D. 1, 7, 252, 253 fn., 265
C. J, 25S Bak:htin, M. 11o-113, 118, 121 Ball,C. 84 Bauman, R. us Bell, A. 149 Biber, D. S5,165 Bickerton, D. 133, 163, 179,191, 196, 209, 25S Blake, R. 66 Boberg, C. 61-62 Bolton, W. 43 Brafia-Stra~ M. see Straw. M. Briggs, C. 11S Buchol~ M. 254
K
c
Kouwenberg, S. 252 Kroch, A. 253
Baile~
Callender, T. 115 Cameron, D. 254 Carrington, L. 261 Cassi~ F. G. 111, 251, 259, 261, 265 Cheshire, J, 42 Christie, R S1, S6-87, 99, 261, 264> 267 Clarke, A. 114, 116, 124 Collymore, F.A. 116
D DeCamp. D. 5-6, 255, 25S-261 Deuber, D. 141 Devonish, H. 261, 264-265, 267 Donicie, A. 14, 16-18
Irvine, A. S0,145 fn., 240 fn., 266
Johnson, D. 2
L Labov, W. 41, 196, 207, 241, 243, 257-259. 267 Levey, 82 Lobanov, W. 62
s.
F Fox. B. S2
s Sand, A. 136 fn., 137-1.41 Sankoit G. 6o,62,76, 255 Schilling-Estes, N. 154. 160 Schneider, E. So, 133 fn., 208 Schumann, C.L. 15 Sebba, M. 108, 163 Seuren, R 14, 16, 18--lo, 23, 28,32 Shields, K. So, 81, 87, 96, 97,101 Simons, R.D. 14, 16 Singler, J, 252 Smith, K. 2J6 Strang, B. 41 Straw 62 fn., 64. 74,76 Szmrecsanyi. B. 47
T Tate, S. 232
v M
van den Berg, M. 34.35
Mail; C. So, S7, 97, 100, 133, 134 fn ., 140, 142, 145. 163, 167-168, 174 Mallinson, C. 255 McEntegart. D. 255-256
Van Herk. G. 65,75 Voorhoeve, J. 14.16--19.23, 27-2S,30,J2,J7,259
McPhee, H. 42
p E Edwards, V. 232--233 Evans, B. 6o, 73 Farquharson, J.T. 257 fn.
R Rampton, B. 234 Rickford, J.R. 1-2, 6, 110-112, 118, 132, 137-139· 25]'-259, 261 Roberts, S. 252 Romaine, S. 253, 259, 267 Robertson, I. 265 Robinson, F. 41
Patrick, R 1 fn ., 79--80, 86-S7, 13J-140 Payne,A. 6o Perkins, M. 30 Price, N. 44 Price, R. 211, 216 fn.
w Walcott. D. 259 fn. Weldon, T. 43, 50 Wells, J.C. 6o, 63, 64, 65, 74, 232,235 Wendelaar, W. 14, 20, 21, 23, 24.2S Winfurd, D. 1S, 34, 36,39, 42, 43, 134. 208-209, 264 Wolfram, W. so, 119, 267
Subject index A
Creole English see English creole
French 165, 166,210, 217, 219,
D
French creole Dominica 232 St. Lucia 266 fn. Trinidad 262 fn.
English American 83,84-5,93-4 Barbadian 65,111--::1 British English 61, 65, 73, 81, 83, 84--6, 8!}, 91>-5· 97· 98--9,101 Early Modem English 41 Ipswich 57, 58 Jamaican 4, 5o 58, 59> ;79-So, 88,135 Middle English 41 New Englishes 134 New Zealand English 81, 8J,
c codeswitching 5, 149,218, 148 competence communicative 196 linguistic 112, 117, 134. 198, 217,237.257.258,260,261 stylistic 159,192, varilingu.al 191, 198, 204
continuurnn see creole continuum continuurnn of meaning 154 fn. corpus British National corpus 83, 84.91
International Corpus of English (ICE) 4. 83, 8s, 88, 89, 92, 95. 96, 97--9. 101, 134. 136-160, 164.165, 17-\, 173, 174 fn ., 266 fn. Santa Barbara Corpus 83, 84 creole continuurnn 2, 3-4, 5, 6,
North American English 113 Suffolk English 65 English creole AAVE 41,43 Barbadian 65, 108, 111--3 Guyanese 34, 43 in Bequia 3,39-53 Jamaican 3-4. 79, 81, 135,140 Trinidadian 43 F
In the series Creole Language Library the following titles have been published thus far or are scheduled for publication:
36 35 34 33 32 31 30 29 28
26 25 24 23 22
20 19 18 17 16 15
HINRICHS,Lan and Joseph T. FARQUHARSON (eds.): Variation in the Caribbean From creole continua to individual agency. 2011. vi, 276 pp. MIGGE, Bettina, Isabelle UiGUSE and Angela BARTENS (eds.): Creoles in Education. An appraisal of current programs and projects. 201 o. vii. 356 pp. ABOH, EnodJ. 0. and Norval SMITH (eds.): Complex Processes in New Languages. 2009. vii, 409 pp. SELBACH, Rachel, Hugo C. CARDOSO a ad Mugotvan den BERG (eds.): Gradual Creolization. Studies celebrating Jacques Arends. 2009. x. 392 pp. MICHAEUS, Susanne (ed.): Roots of Creole Structures. Weighing the contribution of substrates and superstrates. 2008. xvii, 425 pp. HUBER, Magnus and Viveb VELUPILLAI (eds.): SyndJ.ronic and Diachronic Perspectives on Contact Languages. 2007. xii, 370 pp. BAPTISTA, Madyse and Jacqueline GuERON (eds.): Noun Phrases in Creole Languages. A multi-faceted approach. 2007. x. 494 pp. SIEGEL, Je1f, John LYNCH and Diana EADES (eds.): Language Description, History and Development Linguistic indulgence in memory of Terry Crowley. 2007. xv. 514 pp. DEUMERT, Ana and Stephanie DURRLEMAN-TAME (eds.): Structure and Variation in Language Contact. 2006. viii, 376 pp. CLEMENTS, J. Oancy, Thomas A. KIJNGLER, Debonh PISTON -HATLEN and Kevin J. ROTTET (eds.): History, Society and Variation. In honor of Albert Valdman. 2006. vi, 304 pp. ESCURE, Gen«Wilme and Armin SCHWEGLER (eds.): Creoles, Contact, and Language Change. Linguistic and social implications. 2004 x. 355 pp. MOUS, Maarflen: The Making of a Mixed Language. The case of Mia/Mbugu. 2003. xx. 322 pp. MIGGE,Betlina: Creole Formation as Language Contact. The case of the Surinan1e Creoles. 2003. xii, 151 pp. MVHLEISEN, Susanne: Creole Discourse. Exploring prestige fonnation and change across Caribbean English-lexicon Creoles. 2002. xiv, 332 pp. SMITH,Norval and ToajesVEENSTRA (eds.): Creolization and Contact. 2001. vi. 323 pp. NEUMANN-HOLZSCHUH, Ingrid a ad Edgar W. SCHNEIDER (eds.): Degrees of Restructuring in Creole Languages. 2001. iv, 492 pp. McWHORTER, John (ed.): Language Change and Language Contact in Pidgins and Creoles. 2ooo. viii, 5 03 pp. RICKFORD, John R. and Suzanne ROMAINE (eds.): Creole Genesis, Attitudes and Discourse. Studies celebrating Charlene J. Sato. 1999· viii, 418 pp. SPEARS, Arthur K. and Donald WINFORD (eds.): The Structure and Status of Pidgins and Creoles. Including selected papers from meetings of the Society for Pidgin and Creole linguistics. 1997. viii, 461 pp. ESCURE, Gen~ve: Creole and Dialect Continua Standard acquisition processes in Belize and China (PRC). 1997. X. 307 pp. THOMASON, Sarah G. (ed.): Contact Languages. A wider perspective. 1997. xiii. 506 pp. CLEMENTS, J. aancy: The Genesis of a Language. The formation and development of Korlai Portuguese. 1996. xviii, 282 pp. ARENDS, Jacques, Pieter MUYSKEN and Norval SMITH (eds.): Pidgins and Creoles. An introduction.
1994 XV, 412 PP· 14 KIHM, Alain: Kriyol Syntax. The Portuguese-based Creole language of Guinea-Bissau. 1994 xii, 310 pp. 13 ARENDS, Jacques (ed.): The Early Stages of Creolization. 1996. xvi, 297 pp. 12 BYRNE, Francis and Donald WINFORD (eds.): Focus and Grantmatical Relations in Creole Languages.
Papers from the University of Chicago Conference on Focus and Grantmatical Relations in Creole Languages. 199 3. xvi, 329 pp. 11 BYRNE, Francis and John HOLM (eds.): Atlantic Meets Pacific. A global view of pidginization and creolization. 1992. ix. 465 pp. 10 WINFORD, Donald: Predication in Caribbean English Creoles. 1993. viii, 419 pp. 9 BYRNE, Francis and Thom HUEBNER (eds.): Development and Structures of Creole Languages. Essays in honor of Derek Bickerton. 1991. x. 222 pp.
8
6 5 4 3
BAILE~
Guy, Natalie MAYNOR aad Pttrida CUKOR-AVILA (eds.): The Emergence of Black English. Text and commentary. 1991. x. 352 pp. FABIAN, Johannes (ed.): History from Below. The "Vocabulary of Elisabethvi.lle"by Andre Yav: Text Translations and Interpretive Essay. With the assistanc.e of Kalundi Mango. With linguistic notes by W. Schicho. 1990. vii, 236 pp. SINGLER, John Victor (ed.): Pidgin and Creole Tense/Mood/Aspect Systems. 1990. xvi, 240 pp. JACKSON, Kenneth David: Sing Without Shame. Oral traditions in Indo-Portuguese Creole verse. 1990. xxiv, 2 57 pp. LIPSKI, John M.: The Speech of the Negros Congos in Panama. 1989. vii, 159 pp. BYRNE, Francis: Grammatical Relations in a Radical Creole. Verb Complementation in Saramaccan. With a foreword by Derek Bickerton. 1987. xiv. 293 pp. SEBBA, Mark: The Syntax of Serial Verbs. An investigation into serialisation in Sranan and other languages. 1987.XV, 218 pp. MUYSKEN, Pieter and Norval SMITH (eds.): Substrata versus Universals in Creole Genesis. Papers from the Amsterdam Creole Workshop, April1985. 1986. vii, 311 pp.
