Jack Sidneii

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Graduate Department of Anthropology

University of Toronto

O Copyright by la& Sidnd 1998

National Library Bibliothèque nationale du Canada

Acquisitions and Acquisitions et Bibiiographic Services services bibliographiques

395 Wellington Street 395, nie Wellington OttawaON K1AON4 OüawaON K 1 A W Canada Canada

The author has granted a non- L'auteur a accorde une licence non exclusive licence aliowing the exclusive permettant à la National Library of Canada to Bibliothèque nationale du Canada de reproduce, Ioan, distriiute or sel1 reproduire, prêter, distribuer ou copies of this thesis in rnicroform, vendre des copies de cette thèse sous paper or electronic formats. la forme de microfiche/nlm, de

reproduction sur papier ou sur format électronique.

The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts fiom it Ni la thèse ni des extraits substantiels may be printed or otherwise de celle-ci ne doivent être imprimés . reproduced without the author's ou autrement reproduits sans son permission. autorisation.

Gender, Space and Linpistic Practice in an Indo-Gnyanese Vülage

ABSTRACT

Doctor of Philosophy 1998

Jack Sidnell Depariment of Anthropology

University of Toronto

This is a study of language use and ianguage socialization based on ovelve months of field

research in an IndoGuyanese village between 1994 and 19%. The dissertation includes an

historical account of the village, a description of the language varieties currently in use and

discussion of fieldwork and methodology. A number of chapters are devoted to a

consideration of gender and linguistic practice viewed fiom various theoretical paradigms.

This includes a consideration of pronom vmiability as well as analyses of several disputes.

Finally, 1 consider the uçe of affective strategies basic to the process of language sociaîization.

[ examine two moments in the affective shaping of children's communicative bodies. In the

first case, long before the child uses language productively, caregivers begin to inculcate a

certain way of understanding the body. The values that 1 have labelled autonomy and age-

graded/gendered solidarity are embedded and passed on through the very practices of holding

and caring for children. With emerging linguistic skills cornes exposure to a range of caregiver

verbal strategies which more precisely direct children in community based moral and social

education. In the second moment which I discuss, children move out of the position of

dependents. This happens early for aii children but has Merem consequences for boys and

girls. Ail children develop peer groups but for girls the development of nich relations tends to

be restricted by the fact that early on they are expected to contibute to the household

economy. For some this means a very rapid transition tiom dependent (up tili age four) to

caregiver (by age seven they may be caring for their siblings). For others it meam contributing

to household labour in other ways - for instance cooking and cleankg. Boys on the other hand

are not expected to work in the same way. They are enwurageci to engage in various semi-

productive activities like fishing and birding. Between the ages of 7 to as old as 18 they

experience a kind of personal autonomy and fieedom that is UtViVaiied in the village. The

dissertation concludes with a discussion of the pewasive genderiag of language aiad action in

the community.

TABLE OF CONTENTS Acknowledgements A note on local terms and transcriptions

Chapter One Introduction: Approaches to Lanpge and social iife in the Anglophone Canbbean 1 .O Introduction 1 1.1 Ethnographie and sociohguistic perspectives on the

speech community 1 1 .1 .1 Traditional approaches to speech commhty 1 1.1.2 Reassessing speech community h m the mole

perspective 4 1.1.3 Communities of practice 5 L. 1.4 Ideologies of commUXLity - egalitarianism 7 1.2 Gender and egalitarianism 1 1 1.3 Sotialization and egalitarianisrn 12 1.4 Between practice and ideology - the mediating role

of spatial organization 13 1.5 The speech commuxüty reconsidered - more

methodological precautiow 13 1 .6 Variation, discourse and self-cornportment 17 1.7 The stmcture of the present work 19

Chapter Two The village: Rhythms of work and the culture of üved spaces 2 -0 lnuoduction 22 2.1 Historical background 24 2.1.1 The histoly of nce famiing in Guyana and the

East Indian family as corporate group 26 2.1.2 Whynce? 34 2.2 Contemporary econornies 35 2.3 The public face of kinship: Portraits of three

patrilocal groups 40 2.3.1 Bo10 43 2.3.2 Teeli 46 2.3.3 Pakar 48 2.4 Mamiage. power. gender 50 2.5 House and yard 54 2.6 The road 6 1 2.7 Themshop 64 2.8 Conclusion 65

Chapter Three Methodological pos i tionality 3 .O Broken in fieldwork 66 3.1 Rac4gendeidass: soaal locations and fieldwork 74

3.2 Informants as CO-authors 3.3 The field period and the sample

Chapter Four

Chapter Five

Language variety, language use and language ideology Introduction Variation in the creole continuum 'Creolese~/I3asiiecta1 sample Engiish sample Contrasts between basilectal and acrolectal varieties Native perspectives on the language Local tems for linguistic varieties The speech economy Native awareness of variation Regional variation Social variation Stvlistic variation A&cu of the interaction between language and context in CC discourse Dimensions of context Spatial organization of talk Temporal organization of taJk Participants and person deixis Stance Me talinguistic predicates and linguistic ideology The Iexico-grammatical structure of metapragmatic descrip tors Conclusions

Taking nativespeaker awareness seriously: Morphological variation and gender

Preamblc 1 ntroduction Sex, gender and linguistic variation in Anglo-Caribbean speech communities The variables used in this study Differences between men and women Variation within and between categories Metalinguis tic awareness Subjects - pragrnatic values of 1st penon subjects Third penon object variation AnirniV and gender effects Interpretations of complex gender-based patterns of variabüity 179 Conclusions 186

Plates 1. Typical working dass wooden house 2. The yard. the road and two roadside houses 3. The yard 4. Bottomhouse 5. Children by the road 6. Brothers of a single patrilocal goup retuming 7. Grandmother and grandson

Chapter Six Cussing up and taking hard: Anective dïsplays of 'passion' as challenges to a gendered territoriality 6.0 Preamble 6.1 Introduction 6.1.1 Women and disputing in the Cuyanese context 6.2 Ambiguities in the gendehg of self-cornportment 6.3 Background to the transaipt 6.3.1 The rranscript 6.4 Affective displays of passion 6.5 Conclusions

Chapter Seven Developing a habitus for everyday iife: The contexts of childhood and adolescence

Introduction The place of chiidren in the world of adults Chiidren's bodies - an education before the onset of language Affect and social control Directives - indexing age graded and gendered solidarity Threats. fear and physicai violence Teasing and self assertion Boys and girls: the social construction of gender Expectations of gendered behaviour in later childhood and adolescence Shame and shaming Adolescence and the heterosexual market Space. socio-econornic relations and the transition from childhood Socialization and local theones of language and rneaning Conclusions

Chapter Eight Learning social and spatial location through dicitations 8.0 Introduction: elicitations, spatial orientation and

social relations

8.1 Body schema, space and habitus 2 72 8.2 Leaming to kaal aaoss spatial boundaries 275 8.3 Leaming to km and ciaiming rights to inhabit interactive

space 280 8.4 Conclusions 286

Chapter Nine Conclusions 9.0 Power, stereoSrpe and the talk 9.1 Modes of resistance 9.1.1 The 'traditional" pattern 9.1 -2 Other fonns of resistance 9.2 Sociaiization, gender and power 9.3 Variation, discourse and ethnography - the

reciproaty of perspectives

Appendix A brief sketch of Guyanese Creole 0 .O 1 ntroduction 1 -0 Phonology 2.0 Lexicon 3.0 Morphosyntax 3.1 Pronouns 3.2 Deixis 3.3 Articles 4.0 The simple verb phrase 4.1 Copular and atuibutive predication 4.2 Tense. mood and aspect categories 5.0 Passives 6 .O Corn plementation 6.1 Complements with se 6.2 Complements withfi 7.0 Verb serialization 7.1 SVCs and orher multipredicate constructions 8.0 Front focussing and copying structures

B ibliography

Tables and Figures Table 1.1

Figure 2.1

Figure 2.2

Table 2.1 Table 2.2

Table 2.3 TabIe 3.1 Table 3.2 Table 4.1 Table 4.2 TabIe 4.3 Figure 4.1

Figure 4.2 Figure 4.3 Table 4.4 Table 4.5 Table 4.6 Table 4.7 Table 4.8 Table 4.9 Table 5.1

Table 5.2

Table 5.3

Table 5.4

Table 5.5

Table 5.6

Table 5.7 Table 5.8 Table 5.9

Speech nroms and speech use in Cane W& Guvana (from Rickford I986a) ~ k d drawn map of the village showing houses, public road and population estimates Extraa from Nineteenth Century Map showing the region of East Coast Demerara Summary of lncome Producing Strategies Viliage exogamy and pauüocal residence for 20 domestic units Categorization of some f%h commonly eaten Schedule of fieldwork Children in the sample Contrasts between texts l&2 Metaiinguistic vocabiilaiy for code varieties The speech economy Associations between soaal categories, personal characteristics and structural variation Men's and women's evaluations Working and Nonworking subjects' evaluations 1 m plicational arrav of addressees by ~ e ~ r e p o r t i n g CC rhree-term -stem for locative adverbs The system of mot and pst-posed deictics in CC Indusiveness of pronominal referential categones Hierarchy of characterizhg features of pronouns Lexicqram.matical coding of addressee Morphological basilectal. mesolectal and acrolectal pronoun contrasts Morphological basilectal, meso led and aaolectal pronoun conuasts adjusted to include only relevant ca tegones Distribution of basilectal sub jects ( I st) . objects (3rd). and genitives. by sex-of-speaker Distribution of basilectal I st person subjects by sex-of-speaker Distxibution of basilectal3rd person objects by sex-of-speaker Distribution of 3rd person objects. for animate referents ody, by sex-of-speaker Feature marrix for indicators, markers and stereotypes Effects of animacv on the variabiiity of 3rd person objects 17 1 Effects of referen; gender on the variability of 3rd person ob jects 172

Figure 5.1

Table 5.10

Table 6.1 Table 7.1 Figure 8.1 Figure 8.2 Figure 8.3 Figure 8 -4

Table 9.1

Appendix Table I Table 2 Table 3 Table 4 Table 5 Table 6

Effects of the gender of the referent for male and female speakers (3rd person objects) Categorization of speakers according to frequency of nonbasilectal variant in two subcategones Partiapants' movements through the spaces of road, yard and house Conflicting oppositional formats Child centered and child decentering caregiver svategies Saussure's exchange circuit Spatial organization of the k t kaal-elicitation routine Spatial organization of the second kaal-elicitation routine Spatial arrangements and participants (transcript 8.4. h e 1) Some formal features of disputes ovrr four instances

Vowel inventoxy for GC Hindi kin terms in common use in the village Words considered prototypidy ''creolese" Interrogative pronouns "Copuiar" structures in CC Basic meanines for TMA markers

This work could not have been completed without the help, assistance. encouragement

and support of a great many people. I t is a great pleasure to have a chance to thank them.

For hanaal assistance 1 thank the Werner-Gren Foundation for Anthropologicai

Research, The Soad Saences and Humanities Research Cound, The University of Toronto

Aiumni Foundation and the Ontario Graduate Scholanhip program.

Mv understanding of a sociaily conceived linguistics and a linguistically conceived social

science is derived in large part from the excellent instruction 1 received from Hy Van Luong at

the University of Toronto. At both undergraduate and graduate leveis his teachïng has inspired

me to critique power by Iooking at the details of Iinguistic and social phenornena. His teaching

technique is one that encourages people to explore new ideas for themselves rather than adopt

the positions of othen. 1 would also üke to thank Hy for his careful and aitical readings of

many draft proposals. 1 have been lucky enough to find another excellent teacher in my time as

a graduate student. Bonnie McEihhny has introduced me to a world of ideas about the

workings of language, power. authority and gender. Her instruction has been patientiy critical

and personal. Her comments have always been detaiied and illuminating. 1 thank both these

people for their encouragement and attentiveness to my shifting needs. Both of these people

read an earlier draft of this work and provided very helpful suggestions for improvement 1 want

to thank Bonnie particularlv for her very detailed comments on issues of both syle and

substance. w . .

John Rickford and Donald Winford provided invaluable help in planning the fieldwork

portion of this research. Both have also given encouraging comments on my, at times ha&

baked, analyses of hguistic phenornena. Thanks also go to lohn Rickford for his support in

other areas of academic life.

For her detaiied reading of two chapters and vexy helpful eciitorial advice, 1 thank Chene

Robertson. For comments on versions of these chapters presented at conferences, I thank

Jennifer SchIegeI. Lanita Jacobs-Huey, Margaret Field. Jennifer Reynolds. Cyndi Dunn, Candy

Goodwin, Chu& Goodwin. Elinor Ochs, Pete Haney, Paul Garrett, Robin Oakley and Jack

Chambers. For detailed comments on individuai chapters, 1 thank Jack Chambers and Margaret

Field.

The ideas presented here have developed, in part, through conversations with my peers.

I would specifically Iike to acknowledge the important roie played by Paul Garrett and Warren

Olivo. I'm looking forward to many more years of discussion with these scholars. Thanki to

both for su unsefishiv sharing their expertise and providing so many novel thoughrs on topics

of mutual interest. Thanks also to Warren for his unstinting fnendship and companionship.

Fieldwork is a very persona and uansfonnative experience - some rnight Say traumatic.

I t is hard to find anybody who can understand i t I was vely lu* to have Rdph Premdas as a

fnend in thir regard. Ralph and I have spent many hours "shooting the b d n and despite the

casualness implied bv the expression these conversations have been absoluteiy formative for

me. 1 am impressed not only by the brilliance of his analyses and the breadth of his scholarship

but also by his generositv of spirit and his wonderhl sense of humour.

A huge debt is also owed to my hosts and fiends in the field. Although we corne form

very different worlds and although we frequently do not agree, my 'family" in the village has

ix

never given up on me. M y 'buddiesn P.S., G.S.. and N.S. deserve speaal thanks for taking it

upon themselves to teach me not only the v a r but &O the habitus of Creolcse. S.S. and

S.S. I thank for their care and gossip. for their humour and their witlinigness to accept

ciifference. A#, S.K , F.K., S. K I thank for their incredible hospitality, their concems for m y

welfare and their constant championhg of my cause. Avery s p e d thanks goes to ffivita for

whom I cannot think of a suitable pseudonyn Words faü me. Despite our differences we

became the dosest of friends. To ail my other &ends, acquaintances and opponents in the

village. 1 thank p u for dlowing me to stay with you.

Thanks to Mom, John, Ann, Dee. Chris. Jessie, Randy. No one picks their f d y but I

sure am happy with the one 1 got - thanks for ev-g (ged 'er in 'a 'ya bud!).

Fùially 1 want to thank my life-long friends. Michael Grant, thanks for putting up with

mv seif-absorbed, thesis-writing state for the past three years. 1 promise to get over myself in

the near hture. AUison Greene has taught me so much and also forced me to teach mvself. I

want to sav sorry for all the nighu spent in the study with the clatter of the keyboard. Thanki

for vour patience, vour support and for taking me down a notch when I needed it. Last. but

certainiy not least. thanks to Sula for allowing me to live and learn in Cuvana while she was

leaming to live. You are the joy of my Life - I love you iittle buddy.

CHAPIER ONE

INTRODUCTION: APPROACHES TO hNGVAGE AND SOCIAL

L m IN THE ANGLOPHONE CARXBBEAN

1.0 INTRODUCTION

The following is a consideration of the soaal uses and functions of language in srnail.

rural. Indo-Gyanese community. The work is derived fiom my fieldwork expenence in Guyana

between 1 994 and 1 996. In this introduction I situate the present work in relation to the

existing literature on language and social life in the Anglophone Caribbean.

1.1 ETHNOGRAPHIC AND SOCIOLINGUISTXC PERSPECTIVES ON THE SPEECH COMMUNfIY

'Speech cornmunity' is, arguably, the centrai concept in both quantitative

soaolinguistics and anthropological approaches to discourse. Data coming from creole speaking

communities in the Caribbean has, howwer, raised a number of problems for existïng

theoretical models of speech community. The study that follows contributes to this ongoing

rechinking of the speech community concept from a creoiist perspective. In this introduction 1

discuss some previous work on the speech commUN~ which focuses on CEC (Caribbean

English Creole. cf. Winford 1993) data before introducing the major uajectories of the present

work

1.1.1 TRADITIONAL APPROACHES TO SPEECH CO-

Perhaps the most weli known definition of speech community is that advanced by

Gumpexz ( l962:3 1):

[A linguistic or speech community is][ ...] a social group which may be either monolingual or multilinguai, held together by fiequency of social interaction patterns set off h m the sumounding area by weaknesses in the lines of communication.

Whde Cumperz emphasizes frequency of social interaction. Hvmes ( 1967: 18) suggested that

the speech community be defined in terms of what was shared. He wxites ( 1967: 18):

A speech communiry is defined then, tautologidy but radical@ as a community sharing knowledge of d e s for the conduct and interpretation of speech. Such sharing comprises knowiedge of at l e s t one fom of speech. and knowledge ais0 of its pattems of use. Both conditions are necwary. '

Labov's understanding of speech cornmunity is more complex and more pnxîse. For Labov. the

speech community is manifest in shared noms of evduation.

That New York City is a speech community, and not a collection of speakers üving side by side. borrowing fiom each other's diialects. may be demonstrated by many kinds of evidence. Native New Yorkers cliffer in their usage in terms of absolute values of the variables, but the shifts between conuasting styles foiiow the same pattern in almost evely case ( 1966:7).

Discussing the production and evaluation of p s t - v o d c ( r) in New York City, Labov

This result is typical of many other empirical findings which confimi the view of New York City as a single speech community. uniteci by a unifonn evaluation of linguistic features, ye t divenified by inaeasing stratification in objective performance.

Labov contrasts shared noms of evaluation, the defining feature of the speech community.

with the heterogeneity of production.

The term community assumes something held in comrnon. The heterogeneity that the speech community shows in speech production is matched by a relative homogeneiv in noms of interpretation: both rest on shared linguistic knowledge ( 1980:369).

1 Hymes definition of the speech community, Iike his notion of communicative cornpetence, was, in part, a reaction to prevailing Chomskian notions. Chomsky ( 19652) had wrïtten "Linguistic theory is concemed primarily with an ided speaker-listener, in a completely homogeneous speech community."

3

I t is this relation between shared evaluation and different noms of production which dows the

soaoiinguist to constnict a community grammar in the form of variable d e s for groups rather

The constmction of complete grammars for idiolects, wen one's own, is a fniitless and unrewarding task; we now know enough about language in its soaal context to realize that the grammar of the speech community is more regular and systematic than the behaviour of anv one individual. Unless the speech pattern is studied within the overail system of the community. it will appear as a mosaic of unaccountable and sporadic variation (Labov 1 9 6 9: 75 9).

Labov's conception of the speech community is thus intimately tied to his fomuktion of

variable rules. Bickerton's ( 1973a) rejection of the variable rule paradigm for GC can thus be

seen as a rejection of the Labovian notion of speech cornmunity. For Bickerton, different

productions are the result of the operation of differmt rules - these rules being uneveniy

disvibuted across the commriNty thus giving rise to a continuum. In his work on AAVE (BEV).

Labov recognized the possibiiity for the uneven distribution of rules within a ucommunîty".

While Labov suggested an overlv dose relation between white and black copuiar production

d e s (Baugh 1983). he posited absolute differences between the two groups in other areas.

For some variables, New York ci* Negroes [sic] participate in the same structure of social and s~l is t ic variation as white New Yorkers. For other variables, there is an absolute diff&ntiation of white and Negroes which reflects the process of social segregation characteristic of the &y (Labov 1 968: 248).

Labov's early work on the relation between white and black groups seems to prefigure his later

arguments concerning de facto segregation and divergence (Labov and Harris 1986). Suikingly.

Labov ( L 98Ob), while suggesting that Philadelphia consists of two distinct speech cornmunities

(white and black). argued that it was possible to posit a single creole speech communitv for m . .

Belize.'

1.1.2 &USESSING SPEECH COMMUNlïY FROM THE CREOLE PERSPECTIVE

Two major critiques of the notion of speech community have emerged from work on

CEC comrnunities. W d o r d ( 1988). taking issue with Labov's notion of speech communitv,

argues that the existence of a continuous dimension of Linguistic variation (i.e. a continuum)

and shared noms for the evaluation of linguistic variants does not impv a single community

grammar. Winford's ( 1990, 1993) most recent work, in fa- has dernonsuated significant

discontinuities in basilectai and upper mesolectal grammars (arguing against Bickerton's 1973a.

1 9 73 b anaiyses which assume underiingS identical grammaticai categories for aeole and

English 'copularn variants - see Appendix 1).

Rickford ( 1979, 1985. l986a. 1986b) has developed this notion of discontuiuities in an

extended ethnographic case study of Cane W a k For Rickford the speech cornmunity of Cane

Walk is characterized by both concord and cof ict . Memben of CaneWalk's two major social

classes (Estate Class and Non-estate Class) agree in some areas and disagree in othea.

Shared u s e h terpretation of raPing (public verbal disagreement)

Concord Agreement on assotiation of speech sqdes with certain occupations (statu measure)

Talkin' Nansi (telling African-derived €olk?aies - done by estate dass only)

Connict Disagreement on which speech styles are most 1 ikely by those they would label " friends" (solidarity measure)

Table 1.1 SPEECH NORMS AND SPEECH USE IN CANE W U , GUYANA

2 Labov's argument is based on the analysis of wo speakers. One is a Belizean accommodating to acrolrctal targets while the other is white speaker of AAVE.

One important aspect of Ridcford's andysis of the speech cornmuniqr is the way it combines

research on variable patterns of use and evaiuation with research on discowse level patteming.

In later work ( 1986a. 199 1) Rickford has shown the way in which EC and NEC Cane Wallcers

exhibit 'dramatic ciifferences' in language use. specifically in their use of nine singuiar pronoun

categories (cf Riddord 1 986b:2 1 7). The work of Winford and Rickford forces a

reconceptualization of speech community dong two dimensions. Winford's work emphasizes

the problematic notion of both a 'communily grammar" (Labov 1969) and a 'seadess

contuiuum" (Bickerton 1973a. 1975). Riddord critiques the notion that a speech communi~f

must necessarily be characterized in terms of shared norms of evaluation.

1.1.3 COMMWNITIES OF PRACTICE

The notion of community that I employ in the current work ciiffers somewhat from

those reviewed above. AU of the models reviewed attempt to define speech cornmunitv as

something evidenced in speech behaviour. Rickford's important revision is to point out that

whether such norms and usages are shared or not is an empirical issue. However. that

'communities" exist in advance of the linguistic and social practices in which they are made

manifest is assumed in ail cases (even if the problems of defining boundaries are readilv

acknowledged). The title of Labov's ( 1980b) article is somewhat misleadhg given that the

question that he asks is not "is there a creole speech communi~?" but instead "is there a single

creole speech community?". Thus far, then. speech community has sexved as a useful heuristic

device in sociolinguistic and anttuopologicai Linguistic work - that is, it has operated as a terrn

meaningful to the researdier. In recent social theory (and some socioiinguistic work), a general

acknowledgement of the way in which community is also a meanin@ category for s o a d

actoa has emerged. Anderson ( 1983) has provided a biilliant account of the way in which

'imagined communities* selve the ends of political powers. In the course of his discussion he

...ail communities larger than the primordial vülages of face-to-face contact (and perhaps even these) are imagined. Communities are tu be distinguished, not bv their falsiv/genuineness, but by the sqle in which they are imagined.

Edcert and McComell-Ginet have dweloped a simüar notion of 'community of praaice" in

several papers ( L 992, 1995). Thev note that Gumpea's earlier definition of the speech

community 'does not directiy address social ceLations and differentiation among members of a

single comrn unity ( though implicitly ueating differentiation as rwealing 'sub-" cornmunities) .

Nor does it make f d y expiicit the role of practïce in mediating the relation between language

and sociev ( 1992:464)." In contrast, they suggest ( 1992:464):

A community of practice is an aggregate of people who corne together around rnutual engagement in an endeavour. Ways of doing things, ways of talking, beliefs, values. power relations- in short, practices- emerge in the course of this mutual endeavour. As a social construct. a community of practice is different from the traclitionai community. primarily because it is defined simultaneously by its membenhip and bv the practice in which that membenhip engages. (This does not mean that communities of pracuce are necessarily egalitarian or consensual- simply that their membenhip and practices grow out of mutual engagement.) In addition. relations between and among communities of practice, and relations beoveen communities of practice and institutions are important: Individuals typicdy negotiate multiple memberships (in families, on teams . in workplaces. etc), [ . . . ]

Combining the suggestions of Anderson and Ecken and McConnell-Ginet, in this work, 1

attempt to show the way in which ideological notions of community ernerge in and through

Lingustic practice. Sudi an approach has a number of implications for the study of v ~ a û o n ,

discourse and soad process.

One of the central ideas in the presmt work then. is that ideologies of community are

multiple. Rather than posit a single, overarching community within which certain aspects of

verbal behaviour and gramxnar are shared while others are not, 1 suggest that participants

invoke particdar ideologicd notions of the community in Cramtig interactions. This is possible

because no single notion of social relations and community is overwhelmingly dominant to the

exclusion of all others (as both Rickford 1985 and WiIliams 199 1 have demonstrated for

Cuvana). In the first case. villagers draw on corn peting notions of epaiitarianism and hierarchy.

Cockaiorums cowistentiy express two views of the nature of social Me. one positive. the other negative. ünked respectively to the precepts of egaiitarianism and to the precepu of hierarchy[ ...] Therefore, with reference to these competing sets of precepts, an individuai's actions are aiways subject to at least two contrasting ùiterpretations. Anv action raises the question whether the person is tryïng to adhere to egalitarian nomi of equaliq. solidaritv, and generaiized reaproaty, or to hierarchical noms of inequality. individualism . and cornpe tition.

Wilson ( 1973). noting the gendered aspect of this dynamic relation, capturcd the tension in

tems of an opposition between reputation and respectabiüty. For Wikon ( 1973:98) notions of

respectabilitv are both the justification and the ra tionale for differential distribution of

When some people in a society become and remain wealthier than others, or when they occupy positions of power and prestige over others. they must establish a right to sudi advantages. Why should they, rather than others, be so well-off? Innumerable doctrines in various societies of the world develop rationales to answer this question: the rich and powexfd may be descended from first settlen or mythical ancestors; they mav sUnply be more devout in their worship of Cod. so that their prospenty is no rn&e than just a reward: thev may be supenor human beings because they are white, or Chinese, or avilized. On ~rovedencia a large part of the rationale, which is itself a form of social value. is bound up with the idea of 'being respectable."

Respectabiüty, for Wilson. is part of the syiboiism of colonial order - its chudes , its

pigmentocracy. its manners of speaking etc.. In dynamic tension with such social values is the

idea of reputation. Reputation is focwsed on male viri1.i~ manifested in offipring. I t is

symbolized in both the sweet t d k that a man uses to entice women and the bawdy stones that

he reserves for all-male Company. To his aedit. then, Wikon makes evely effort to show the

wav in which this ideological universe is not simply floating above and somehow informïng

everyday life. but is înstead deepIy embedded in both the mudane and exuaordinq practices

of social actors. This said. there are a number of outstanding problems with the wwork As

Besson ( 1 993) has demonstrated, the phenornena which corne together in Wilson's sphere of

"respect" are only contingently and circumstantially linked. Thus in Provedenaa. Wilson fin&

tha t the church is strongly assoaated with women and the elitist values of respectabdis. and

has Little revolutionary potentid. Besson. howwer, discussing her research in Jarnaica. shows

that the Baptist Churches of Trelawny have. since slaveiy, been involved in left-wing,

anticolonial and anti-elite struggles and that women have been instrumental in such

movements. Whatever the fit in Provedencia, Wilson's attempt to extend his thesis to the

rntire Caribbean must be treated w i t h some trepidation.

In this study 1 look dosely at the penrasive notion of egalitarianism in the village. In

fashioning community through hguistic practice. villagers consistently draw on this ideological

characterization of cornrnunitv. However, the way they do this points to the instabiiity of any

one formulation of egalitarianism and its potential for multiple interpretati~ns.~ E~gaütarianisrn.

9 . . '~i l l iams ( 199 1 : 170) writes

Egalitarianism is net.[...], a set of precepts aimed at speafying the conditions or mode of conduct for absoiute equality. Instead,[ ...], it is (in proverbiai language) a code requiring

in the G u y s e context. is in fact a cluster of interrelated ideas and values about the place of

the individual within a community of othen. Thus, on the one hand. egalitarianism indudes

the expectation that individu& should be free to go about their daüy actMties without undue

hindrance from others. As such, egalitarianism ernphasizes the value of individual autonomy.

On the other hand, egalitarianism indudes a strong ethic of soliddty whereby individuais are

expected to play an active role in maintaining enduring social relatiomhips based on age-grades,

kinship. residential proximity and economic cooperation (note that both of these aspects are

incorporated in Wilson's ( 1973) 'reputation" notion). This is, as Wiiliams ( 199 1 ) notes (cf.

footnote 3). a highly qiialified egalitarianism - one built partly around the asymmetrical

obligations of kinship. The values of solidarity are also suongly rooted in the social and

economic relations of traditional Indo-Guvanese village Me. Up und the mid part of rhis

centuw, many rurai Indo-Guyanese people were stül farming rice in a traditionai manner. This

meant sowing and weeding rice by hand and ploughing fields with oxen. Labour, for this labour-

intensive economy, was organized around the patriiocal domestic unit, as was most ntual and

much of evervday soaal activity in general. The traditional pattern gave rise to a parricular kind

of local village solidarity. one that is suongly gendered and age-graded. I t is this cluster of

individuais to recognize that "wan hand kyan wash or clap." That is to Say. everyone should understand that al1 accomplishments and their en joyment are made possible by assistance attained from others: therefore. everyone should be willing to assist others. As they stniggle to become somebody, they should stmggle equally hard to demonstrate. through their rveryday actions. ( L ) their respect for the hurnan dignity of others without reference to dass position. (2) their willingness to engage in generalized reaprociq. (3) their desire to avoid actions that damage their own reputations and those of others and (4) their wiilingness to submit to and participate in processes of mediation that treat formal litigation as a 1st resort. The characteristics of soaai action are necessary because ail humans, as humans deserve one another's respect. In short,[...], we may note that egaiitarianism, when applied to evaiuations of individual differences and of interpersonal conduct, provides cri teria for a moral stratification of the social order that Cockalorums attempt, with limited success, to align with the economic criteria on which class stratification is otherwise based.

10

values which indudes an emphasis on personal autonomy and soiidarity (or family corporatism)

that 1 refer to as a qualüied egaütarianism.

Egalitarianism. then. includes strong expectations that an individual will activelv

maintain relational ties based on friendship and kinship through mutual engagement in

activities ranging from religious and domestic work to drinking and gossiping. For example. at

least once a vear most Hindu f d e s in the village do a Jandhi or Puja to make offerings to the

deodau and ask for blessing. Even a poor famüy wiiI invite large numbers of people and is

obiiged to feed all of them. The familv hosting the religious event will. at these times. d on

fiiends and kinspeople to assist in the cooking and in a number of other tasks. Many people are

thus expected to involve themselves in some way in the work - either as guests or as helpers.

Those that do not attend are often seen as unwüluig and s e f ih and may be subject to suong

social sanction from their network of close associates. Similady, each member of the household

is expected to fi1U3.U certain obligations based on their gendered and age-graded position. Those

that do not comply are considered mannirh 'disagreeable, antisocial" and may be forced to

Ieave the household permanently.

This is not to say that indnnduals must comply to every whim of public opinion. There

is an operating counter-force to this strong ethic of structmd cooperation in the value placed

on penonal autonomy (particularly in relationships between peen). Indo-Guyanese adults in

the village do not take kindly to those who beüeve themselves authoritative enough to order

and direct a supposed equal's action. They may remark explicitly that such behaviour cames

the implication of an hierarchid relationship. Of course, relationships based in hierarchy do

manifest themselves, espeaally between older and younger siblings, husbands and wives, a

I I

husband and his wife's fimdy etc.. However. even in cases where the subordinate is cornpliant.

they frequently criticize the other for treathg thun iïke a small M d . orgetting them likc wan

as. Cenerally, adulu vigorouslv defend theK nght to unfettered action and persona1 autonomy.

Consequently an attempt to coerce another wiü often be interpreted as an act of ve-pas , a

denid of one's personal integrity and dignity. Egalitarianism is thus a highly complex and. in

some ways, contradictory or tenuous duster of ideas and values. I t implies both the repudiation

of hiemchy through the defence of personal autonomy and the active maintenance of reiatively

solidarv relationships through mutual engagement in task activity.

1.2 GENDER AND EGAUTARIANISM

Ideologies of the egalitarian comrnunity are firmiy embedded in the soaal and economic

(i-e. domestic) organization of the community. Given that women and men are differently

positioned vis-a-vis these social and domestic structures, their relation to the multiple ideologies

of the egalitarian communitv also differ. In interaction, men and women invoke notions of

community strategically and within the iimiü set by their relative social positions. When

communitv is envisioned as smctured in tems of an age-graded and gendered solidaritv,

women are iikely to find themsehres in a disempowered position relative to senior men. In such

cases women often respond with daims to autonomy. that is, an altemate way of

understanding the egalitarian communi~ (for an example see Pria's remarks in chapter 6).

There are two important points to make on the relations between gender and egalitananism. In

the first case, it shouid be noted that a woman's position within the community envisioned in

terms of an age-graded and gendered solidarity is not solely determined by her gender. Some

12

women are the heads of their own householb and tend to frame interadon in tems of age-

grading and asymmetrical obligation as fiequene as do men in a simüar position. Notions of

autonomy, on the other hand, are more oofte ned bv those in a relatively disempowered

positions. Although this group often indudes women, particuiarly the wives of men who iive in

their pauilocal yard. it also indudes junior men and children. Gender is thus not a category

whidi can be undentood in isolation from other kinds of sociai and economic categones. This

said, @en that there is a dominant system of marriage and kinship which structures the

domestic organization of most households, certain generaüzations about women in the village

can be made. In fact. if we want to explain patterns of variable pronoun usage for men and

women it is necessaxy to look at both the predominant position of women and the exceptions

to the d e (that is, we need to Iook at both the differences beniveen men and women and the

variation within each group, see chapter 5).

Similar issues arise in the discussion of socialization. Here again I consider the wav in

which multiple notions of an egalitarïan communiry are invoked in interaaion. In an extended

discussion, 1 discuss the possibilities for emphasizing autonomy on the one hand. and age

graded and gendered solidarity on the other in interactions with children. In socializing

practices, such as elicitations, caregivers mode1 projected scenes in which the chüd is expected

to occupy a particular social role. These routines and foxms of interaction provide the M d with

an opportunity to play out social relations based in both conceptions of the egalitarian

community. This kind of social leaming is thus important to the reproduction of the dynamic

tension benveen ideas about autonomy. on the one hand, and age-graded and gendered

solidariq, on the other.

I -4 BEILWEEN PRACllCE AND IDEOLOGY - THE MEDUITING R0L.E O F SPATIAL ORGANIZATION

Partiapants to an interaction rarely. if ever, Say '1 am autonomous" or some local

equivalent. Nor do they Say 'you are obliged to do this for me because 1 am your big-brother."

That is to Say, they rarely uivoke notions of the epaiitarian community using expliut means.

The reasons for this are not entirely dear. However, it suffices to say that speakers use other

devices for achieving these interactional goals. One means to this end that 1 have found to be

very common is the linguistic consmction of iived space. The way people are related to

particular spaces is an index of more general sociai relations and participants exploit this

potentid in framing interaction. We thus find speakers, at various times, attempting to locate

themselves, duough a variety of linguistic means includuig defis, in spaces which thev daim to

own or to which thev daim some privileged access. Related to this linguistic characterization of

space is bodilv cornportment. Movemenu of the bodv involve, minimdy. access to the

ùnmediate space which it is inhabiting (and inhabiting in a certain manner) . Gestures, postures,

and movements can also index multiple notions of egalitarianism in addition to perfonning a

number of other conversational functions. These issues corne together with the emphasis on

g e n d e ~ g practice in my discussion socialization and caregiver suategies (chapter 7 and 8).

1.5 THE SPEECH COMlMUNlTY RECONSIDERED - MORE METHODOLOGICAL PRECAVnONS C . '

In Leaming How to Ark, Briggs ( 1986) provided a groundbreaking anaiysis of the

kteMew as a c d t d y specific kînd of speech event The central argument of that work as 1

understand it, was that cornpetence in the interactional d e s and metalinguistic cues assocïated

with the i n t e ~ e w was uneqdy dûuibuted across different speech communities. The use of

the i n t e ~ e w in communities where it is not recopized as a distinct speech event thus malizes

a set of power relations and, at the same time. reproduces a kind of scientific hegemony. In

later work. Briggs has developed a notion of speech communî~ and language ideology which

emphasizes not only the differences between, but also the differences within, particular groups.

1 WU argue against viewing ideologies of language simply as pan of the linguistic background shared by members of a speech community. 1 will suggest that such a perspective is not only empir idy unsound and unenüghtening but ais0 enters into the process through which scholars naturalize their own interpretive authority.

Briggs develops his position as a critique of Rosaldo's ( 1982) paper. He suggests ( 1992:388)

like a nurnber of students of ünguistic ideology, she presenu Ilongot thinking about language in essentialist terms. positing a single linguistic ideology that wodd seem to be distributed homogeneously through out the cornrnunity.

In opposition to such dthem ideas about linguistic ideology. Briggs ( 1992:388) argues that:

ideologies of language are sociaily distributed. I would argue. however. that the relationship between conuastive ideologies and social differentiation is vastly more complex than a mere correlation would suggest I t is not simply the case that men conceive of language in X terms and women in Y or that shamans assert Z about language.

In the preceding pages, I have argued that such a view might be extended to the notion of

speech commmiy itseif and that social actors are not in complete agreement about the ways in

which the speech community and its noms are constituted. With Briggs, I want to argue that

ideologies of community are sociaiiy distxibuted and that these conuastive ideologies cannot be

15

easily assigned. one-to-one. with some identifiable soaal group within a @en cornmunity.

Rather, people partiapate in each other's iinwtic ideologics. They tatitiy tatify them in

conversation to suit their own ends, that is. strategically. Besides this, members of the group I

am discussing share a great deal in ternis of knowledge and beliefs about speech and Ianguage

(Le. linguistic ideology). Howwer, this body of shared knowledge, this linguistic ideology, is

neither whollv consistent nor neutral. By Uiconsistent, I mean to Say that it contains within in

it contradictions - like the one between solidarity and autonomy - which participants can

manipulate in interaction, first stressing one aspect, then another. When 1 Say it is not neuual.

I mean that it expresses and realizes relations of power and privilege in the comrnunity. This

raises a number of methodological problems to which 1 now tum.

If linguistic ideologies (and power) are s o a d y distributed with differences being

manifest both between anthropologist and subject group (Briggs 1986) and within the subjea

group iueif, we must, as researchen, be careful to recognize just whose metacommunicative

devices ( Briggs 1 986:53.83) we are cmploying in our research methodologies (and whose

interests they serve, i.e., who they frame as authoritative. irrationd, etc.) Given the existence of

multiple notions of how the speech community is constituted, researchers should be aware of

the potential for multiple norms and multiple metacommunicative suategies, each inhatek

aed to relations of power and authority.

This danger is compounded by the fact that. when researchers enter a field site. they are

Wely to encounter the dominant speech community ideologies h t since its proponents are

usualiv in the most powerfui and most public social positions. The temptation at this point is to

focus on acquisition, and to avoid distraction by seemingly deviant norms and usages. HOW

16

many authoritative anthropological voices have been built on the borrowed authority of an old

and respected shaman, a chief or a local poiitician? Sudi politicai figures are often at pains to

impose their own reading of the speech comrnunity. And of course, in so far as this is the

dominant (or hegemonic) ideology, badced by sikificant resources ( poiitical, militaxy.

econornic) . other members of the group are iikely to ratify the anthropologist's choice to study

the 'authenticated" representation of their 'culture ".

Ideologies of language have the capaaty to mturalize relations of power. When they are

naively absorbed into anthropologicai methodology. anthropologists participate in the

reproduction of diose power relations. Consider, for ucample. the emphasis on named speech

events in the ethnography of speaking (Hyrnes I968[ 19621)- Such an approach does not

acknowledge that lexicalization patterns are sensitive to the s o d distribution power as weli as

to objective differences between the referents. In Guyana, for instance. women's taik about

others who are not present is often referred to as tut name. Talking name is considered a

ferninine activity des pite the fact that men engage in very similas verbal activ*. When a group

of men are asked what they are doing in such cases. the normal answer is abi a gyaf 'We're just

chatcingn. I once made a senous f a n pas when. in the Mdst of a friendy conversation where 1

was asked "what do you think of person Xn (a person my intedocutors were at the time

denigrating), I responded mi na waan taak nem '1 don't want to talk name." The person who

had asked me was a major poiitical player and a senior man in the viuage. He looked shocked

and replied mi na taak >rem bai - mi oonli akr w a y c tink 1 don't talk narne - I only asked what

vou think." The unintended implication of my remark was that the men were engaging in

womanly/feminhe behaviour. Of course. such ideologies not oniy legitimize male 'gossip", they

17

also vivialize waiuative talk by women (an imporrant mode of social control in the viUage).

Ideologies of this kind can be invoked to make women seem essentiaiiy petty, divisive and

backbiting. W e 1 am not denying that some practices that women engage in f a within the

semantic field covered by a term like tnlk name. 1 argue that such metalinguistic terminology is

necessady ideological and embedded in relations of power. Terms such as talk name gender

practice and the participants engaged.

Methodological sophistication (cf. Briggs 1986) then. depends on the researdier

attending not only to possible communicative hegemonies between groups but also w i t h

them. The naive acceptance of named speech events and dominant linguistic ideologies can

lead to the reproduction of existing power relations and the presenration of existing stereotypes

in the scholarly iiterature (an issue to which 1 return in the conclusion).

1.6 VARIATION, DISCOURSE AND SELF-COMPORIMENT

Ir seems fair to assume that disciplinaxy boundaries have been erected dong the h e s

defined by a dominant linguistic ideology in the western inteilectual culture (Rumsey 1990).

Thus Lve divide up the labour of studying referentid meaning and structure (the linguists), from

that of studying usage (the variationists. the socioiinguis ts) , from that of studying practice and

behaviour (the Linguistic anthropo10gists and conversational analysts). While there is obviously

some leakage between these disaphes. it is not hard to find people that will defend both the

'realig* of the classification (in the facts) and its heuristic value. WhiIe there are a few

exceptions (Ricidord 1979. Edcea L 989. Eckert and McConneU-Ginet 1995). most studies

focus on one or another aspect of the total linguisti~communicative f a n Very few studies have

18

attempted to integrate these h e s of inquiry (cf- Eckert 1997 for a recent example, also Ochs

1 986). In this w o k 1 suggest that a complete undentanding of variation cannot be undustood

without reference to the wav in which variants are part of larger stretches of discourse. Without

an understanding of what social and interactional work variants do in t e ~ s we are not able to

explain theu distribution. But the issue goes deeper than t h . To paraphrase lohn Haviland

( 1986). anthropologists, like the speakers they study, meet words. not abstracted from the

evervdav a + lives of soaal actors, but in the course of doing work, achieving sociai ends and

punuing interactionai goals. To separate out on or another aspect as the most legitimate is

necessa* subuactive. While such methods may fit weii with the dominant linguistic ideology

of North Amencan academic culture, they often clash with the expectations and beliefs of the

people studied. In this work (beginning in Chapter 4) 1 have attempted to show the wav in

which villagers' undentanding and awareness of Ianguage and language variation is inùmatelv

tied to their understanding of both soaal acts and soaal actors. ViIiagen do not readily accept

the separation of "linguistic varïety" from the evervday uses to which it is put. Discussions

about people's attitudes towards "creole". 'English". 'AAVE" or some other variety. abstracted

from the uses to which thev are put. are thus somewhat misguided and &ciai in many

çontexts. For the villagers 1 discuss in this work, Cnoksc is not an abstraa code but a way of

being, acting, seeing and feeling - it is, in Bourdieu's ( 1977) words, a habitus. People thus

display cornpetence in Credlce not onlv duough rnastery of a grammar but also in ways of

holding the body. in postures, gestures, glances, and even in precisely timed rhythmic

inhalations. Speaking, being part of this complete body hexis, is corporeal and is properly m . =. ..

studied ui the contes of the bodv and its movements. As Briggs (1992:399) rerninds us. our

own scholarly terrninology ofien characterizes communicative practice in a way that is

completely a t odds with native realities:

I t would be highly inaccurate [ .. .] to suggest that such 'linguistic ideologies" refer exclusively to dimensions of linguistic structure and use. These ideologies focus just as squarely on soaal relations, particularly soaal inequality, concepts of the person. epistmolo@d categones, and the lïke. It may thus be misleading to speak in a comparative vein either 'linguistic ideologies" or "ideologies of language" in view of the fact that these ideologies wiil often mcompass cultural foQ that appear quite extraneou. If analysts simply pull out those elements that confonn to their definitions of language and linguistia, the character of these ideologies - and particularly theK funaionai relations with the discourse practices in which they emerge - will be displaced by an empiricaily and theoretically irnpoverished caricature that offen the advantage of ready incorporation into existing anaiytic hameworks. We have certainly become sufficiently sophisticated to know that we cannot look for direct counterparts to analvtic concepts in situated discourse practices, whether in Western or non- western societies. I fear that referrïng to ideologies that regulate and ernpower discursive suategies as 'linguistic ideologies" may excuse researdiers from exploring dimensions that extend beyond even a broader definition of the term 'linguistic."

While we should. then. be conscious of the fact that formal vaxiation and the exploitation of

heteroglossia provides 'me talinguistic means through which members demons trate their

cornpetence. as weU as validate othea', in a speech community" (Jacobs-Huey 1997). we

should also not lose sight of the way in which such devices are integrated into a more complete

set of semiotic strategies. Furchexmore, as 1 argue in Chapter 5, it is important that we rethink

language variation itself in tems of the soaalized and gendered body and local expectations for

self-cornportment.

1.7 STRUClURE OF THE PRESENT WORK

In Chapter 2, I introduce the village, focwing on historical developments which have

led to the overwhelming importance of nce farming and partïcular forms of social and econornic

organization. Specifically, 1 trace the development of patrilocal groups in the village and point

to the ways in whkh they now corne into confîict with the system of wage labour. In this

chapter, 1 introduce the analytidly important notion of a qualified egalitarianism. In the

second chapter. 1 also inuoduce the spaces of the house. the yard and the road and discuss their

ideological importance.

Chapter 3 discusses my persona1 and professional relation to the village, its spaces and

its inhabitants. 1 have nied to give an account of the way in which the fieldwork was itself co-

suuctured and co-authored by particular village members. 1 &O discuss the importance of race

and gender in the village with specific reference to my own negotiated and emergent social

location.

Chapters 4 and 5 are concemed with the language varieties spoken in the villages. In

the first. 1 @e a bnef structural description of the most focussed varieties before moving to

discuss native perspectives on the language and the variation contained therein. In Chapter 5, 1

tum to a quantitative analvsis of pronominal variation focussing on the effects of gender. Here.

1 give a pragrnaticallv and ethnographicallv based solution to a quantitative problem of

inconsis ten.. In these chapters, I also lay the ground for a more contextual account of

language use in the village.

Chapter 6 develops the contextuai approach to language with an in-depth analysis of a

parricular text. Here 1 give an account of women's smggles against restrictions on their

movement. 1 am concemed to show the way in which two women collaboratively produce

oppositionai frameworks in the course of challenging the dominant ideology regarding women's

rights to movement and the gendering of space. Thus, while this chapter further develops the

account of language-use in relation to gender and spatial organization which 1 inuoduce in

Chapter 4 and 5. it also senres as an introduction to the Iart two chapters whidi focus on

language socialization.

Chapters 7 and 8 are concemed with the reproduction of spatial organization and

ideologies of community through socializing Linguistic and cultural practices. In the fmt of

these two chaptem. 1 look at the manipulation of dllldren's bodies as a precursor to later

regdations on the movement of bodies through space. I argue that men before thqr become

cornpetent users of language, dllldren receive an education in the ideologies of autonorny and

age-graded soiidarity. 1 then tum to linguistic strategies in the socialkation process induchg

teasing and sharning. This then leads to a discussion of the social construction of gender in

adolescence, l o o h g at the way in whidi the contexts of everyday Me change dramaticaliy for

boys, on the one hand, and girls. on the other, as they make the transition to adulthood. This

chapter condudes with discussion of local ideologies of communication and the role they piay

in the soaakation process. Finaily, in the last substantive chapter. I look at one kind of

language sociaiization practice: elicitations. In such routines carrgivers mode1 projected scenes

for novice language-users. These routines provide detailed characterizations of "typified" am of

speakuig. Through such interactions children learn about the meanings of kinship, spatial

organization and language use and the way the three phenornena are interrelated. In the

conclusion 1 return to the overarching themes of the work in a consideration of ideological

hegemony. linguis tic practice and modes of resistance.

CHAPTERTwo

THE VILLAGE: REIYTHMS OF WORK AND THE CULTURE OF LIVED SPACES

2.0 INTRODUC~ON

Just bevond the Mahaiconv River, on the road which leads through East Coast

Demerara to Berbice. is a string of four unus* hght 'lively" villages - Huntley, Airy Ha&

Dundee and Novar. Some three miles from this road, along sometimes treacherously mudcS,

backdams. lies the village of Champagne. This duster of settlements form the focus of the

present study of linguistic practices and language sociaiization in rurai Guyana.

Despite a rather particuiar history, the villages in the area. at least superfiaaliy,

resemble others up the East Coast (cf. map 2.2). Along the road are the biggest houses. A few

are grand even bv North Amencan standards, with high fences locking in cars, trucks, tractors

and trucks along with the house. More comrnon are the two-story painted wooden or cernent

block houses which belong to the more or less well-to-do people of the road. Often these people

have a shop. or minor clerking jobs or, perhaps. wealthy relatives Living outsidc ('living in

Amenca or Canadan). Behind these houses are about a hundred one and two bedroom houses

that belong to working people. These wwden houses are usually raised on stiits to keep from

being flooded. The bottom house is a hard-packed mud floor that is daubed every week It is

here that a family will spend most of its day with the man, if he is home from work, in the

hammodc. the wife cooking at a fireside or upstairs on a green stove and the children and

friends, visitors and kin corning and going. When the man is out, the women will often gather

to sit and gyaf here in the cool shade of the house after the moming work of washing out and

cooking breakfast (the mid-day) meal is done. Two can codortably redine in a harnmock

together to talk.

The village itself is dÎvided into l o d y recognized spaces quite dearly indicated on map

Figure 2.1 HAND DRAWN MAP OF THE VïLïAGE SHOWING HOUSES, PUBLIC ROAD AND POPtnAnON ESTIMATES

24

People generaliy accord the road a prominent position in their descriptions of the village and

M., the author of this map, began by drawing in the road through the centre of the page.

Inland, the village is bounded by the old railway iine. now simply referred to as "the iine" which

is actually a canal cariying water from the Lamaha Conservancy to the fields of nce and

gardens kept by the villagers. Seaside. the village is bounded by the canai. Eastern and Western

boundaries are marked by the dams which are both ~ d c s into the a g r i d t d areas and

waterways used for drainage and eventually leading to the sea kokus. Within this one village

we calculated that there were about 600 people resident at the tirne.

Somewhat of an anomalv. the area remained relatively unpopulated and untouched bv

the ravages of a colonial regime hungry for land, productnnty. people and profit Colonisation

of the area was so insigruficant that Schomburgk, whose A Desrripiion ofBriti~h Guiana was

published in 1 840, devotes onlv a short paragraph to it:

The hamlet of MAHAICONX, on the small river of the same name, consists of about thirty houses, with tradesmen and shops of different desmptions. Some settlements are s p ~ g i n g up near the srnaii river Abari; and as the high road from Georgetown to New Amsterdam leads through these villages, they are chiefly reîommendable for industrious e m i p t s ( 1840: 76) .

Of course. by the time Schomburgk wrote his description, one colonial venture. the production

of cotton. had already been attempted and, by ail accounts. failed due to falling international

prices. Rodway ( 18 9 7: 76). the most prolific contemporary historian of colonial life in Guyana.

wro te:

The downfall of cotton was most disastrous to those concerned in its planting. fhree- fourths of the coast estates were abandoned aitogether and have remained undtivated

ever since. Al1 that long stretch between the Berbice River and the Corentyne. as well as the lblahaiconv Coast, once glowed with the flowers of that handsorne s h b . and some cotton estates tumed into cattie farms on the East Coast have remained uncultivated since its downfall.

Plantations were abandoned and the area was left in the state desaibed by Schomburgk for

some t h e . Meanwhile, in other parts of British Guiana. planters continued to invest in sugar

plantations and the labourers needed to keep such an industry 'profitable." East Indian

indentured labourers became the planters' worker of choice in the colony, and a course was set

in which a strong historical association berneen 'the coolie" and sugar cane wodd be

established in the min& of manv the colonv's residents. With the end of slavery in 1833 came

the establishment of free Afncan villages such as Buxton and the emergence of a particularly

African cultural adaptation to "free" life in colonial Guyana (see Bartels 1977. Moore 1995.

Rodnev 1 98 1, Williams 1 984). Thar the contemporary socio-cultural and s~cio-iinguistic

situation owes much to these developments has been weli documented by other researchers (cf.

Javawardena 1963. 1968, R.T. Smith 1995, Wiiiiams 199 1. Despres 1967. 1970. 1975a.

1975b. Rickford 1987b) and is not in question here. However, the Mahaicony region (and

possiblv parts of West Coast Berbice. cf. Siiveman 1976. 1978. 1979, 1980) remains

anomalous with respect to this general histoncal path. In 1 8 83. a number of East Indians

acquired the Estates of Dundee and Novar and began to plant rice. setting the region on its

own particular histoncal trajectow. This was not a world of b e n and plantarion overseeren.

Instead, it became the world of East Indian corporate family groups tied together bv relations of

hiendship, kuiship and cornpetition.

O . -.

The colonial histoqr of the Mahaicon~ area would have perhaps remained unremarkable

had it not been for the inevitable cvdes of international economic crisis that plague capitaiist

modes of accumulation (Hawey 1989). I t was largely the international market that caused the

failure of the monoaop plantations of the area and the fortunes of their owners, although

environmental conditions. labor problems and conflict between colonial powers also assisted in

this. As it was, however, and as Rodway ( 1897) notes, by the mid-nineteenth century, the

estates of the Mahaica-Abary region had fden h t o ruin. The planters had left, the labouren

had been removed, and the once rich, husbanded agridtural land had tumed to bush. Given

the rather precarious position of sugar in world markets, planters m u t have been reticent, if

not completely incapable, of turning these abandoned estates to large-scde cane production. In

fact. in other areas at the time. most planters were content to ride out the duonic labour

shonages and falling prices by maintaining a suipped d o m production process ( Rodney 198 1 ) .

It was in this contes that a group of East Indians. some from Windsor Forest, othea

from Haarlem and Cane Grove. A having completed the texms of their indenture. bought the

land of Novar Estare for 64 an acre in 1883 dong with MO Africans (Potter 1975). The land

had been sweyed and divided into plots in 1880, and it is ükely that rice was already under

cultivation there by that tirne. Although it is impossible to fix the date at which rice-growing

aaually began in the region. it is apparent that. within a few years of buying the land, East

Indians. who had now bought out the original Ahican partners, were the cultivaton of

flourishing rice fields which were often pointed to as a great success by white coloniais.

From : T h e Journal of the Royal Geographicai Soaety, Volume 2. 1832 to accompany "Notices of the Indians settkd in the interior of British Guiana By William Hilhouse, Esq. Available from the Perry Casteneda Map Collection. University of T w s

William Russeil ( 1886) gave a lengthv description of the rice production process in these eady

East Indian villages. He concludes his discussion by drawing attention to Novar and Dundee:

Harvest arrived, with need of additional hands, to reap the aop while it was crisp and

* * - dry: and here came the first clash between manager and rice grower. the one wishing to keep his mill supplied. the other wanting to Save his rice. 1 deaded in favour of thé rice grower. as the reaping was not likely to employ too long a tirne. The worken. with a srnail toothed reaping hook, smaller in size but much the same as the now obsolete tool formeriy used for reaping in the old countly, cut the head of grain off with about

a foot of st&. which being made into s m d sheaves afker remaining in the sun for a tirne. were f i n a h conveyed to the barn. a rough structure thatched with cane bands. A stake was drive* into the ground a t one end of the bam; by fieely mmming the smunding earth, a threshing flwr was secured say about 12 feet in diameter. The buliocks were yoked close together and made to wak round this stake, while shed after shed was thrown under the5 feet and shaken up so as to bring evexy part under the tread of their feet I t was simply astonishi. how quiddy the grain was by this means separated fiom the straw, the paddy being h m time to time swept into heaps and put into bags. for the winnowing operations. This was done in a dear space. exposed to the wind. by the well known antient system of letting fall from a sieve. The paddy once ready for market, a ready sale was f o n d for i t on the estate.

Harvest over, water was let on, and a fine ratoon aop came up as by magic, little inferior to the first After reaping this m p , the land was again treated in exactiy the same way by puddling as at kt, of couse the work being much easier.

For want of labour, in 1872 the rice culthration ended. The whole of the above description can be appiied to the venture in nce dtivation which has been carried on for several seasons on Novar and Dundee, in the Abarv district, by the coolie propnetoa of those estates ( 1886: 106- 107).

Despite Russell's ailusions to 'as by magic" and 'much easief (and his persistent omission of

an agentive sub ject) . it is obvious that such procedures were highly labour-intensive.

Traditional nce growing in Guyana required the labour of a large group of adults willing and

able to work according to the needs of a seasonal growing cyde. Russeil was one of a group of

highly innuential sugar planters who argued for the continued support of rice dtivation in

Guvana by the colonial government '

W e Russeli and othen tended to portrav themselves as &ends of the 'coolies." in

fact, their prirnary motive was once again to secure a labour force for sugar plantations. By the

late nineteenth and early twentieth centuries, planters again faced a labour shortage (Potter

1982, Rodnev 198 1). The first and second waves of East Indian indentured labourers had now

finished their service and most were looking for opportunities off the sugar estates (Porter

'Some did not agree. A correspondent for the 'Daily Argosy' wrote in 1883: "[ ...] after aossing the Mahaicony the land is mainly devoted to rice-growing and cattle-rearing by small Indian proprietors. al[ presenting a dreary sight" (àted in Potter 1975: 156).

29

1975, 1982). To import more Ïndentured labourers. if possible, would be costly, so plantes

made every effort to maintain some hold on the labour force composed of ex-indentured

laboures. One suategy was to allow these persons a plot of land on the estate which could be

used for the cultivation of nce as a cash aop. Another was to encourage the establishment of

East Indian nce-growing villages which were within a shon distance of a plantation. Dundee

and Novar became such villages. In cornparison with the kinds of bitter struggles waged against

emancipated Amicans intent on securing land for the establishment of k e viuages. East Indians

were encouraged by the local plantoaacy, although this support did have its ümits. The

colonial govemment and the local plantocracy made every effort both to see that rice was

estabiished as a major crop in the colony and to insure that they maintained control over the

market on whidi it was to be sold. To this end, the Rice Growing Company was established in

the year of 1 8 95 (Potter 1 9 75). 1 ts objectives were:

1 .To acquire and empolder land (Le. prepare it) for nce growing,

2.To grow the rïce a d o r let the land:

3 .To erect a miil in Georgetown;

4.To purchase paddy, mili it and sell it: and

5. To assist, under certain conditions, individual efforts to grow rice.

T h e Rice Company was the fint official body to argue for extensive 'rationalization" of the

rice-growing indusw. Such processes had widespread effects on the politid-economy and

social organization of East Indian, rice-growing communities. and it is not surprishg that the

first steps were taken by eager coloniais who believed they had discovered the East Indian's

After investigating several districts where rice was being cultivated by East Indians. and especiallv reporthg enthusiastically on the progress being made at Novar, Mahaicony. the comkttee recommended that a block of land be selected on the Abary Creek. where a rice plantation wodd be set up, complete with ail necessary reaping and threshing rnadiinexy and a purnping engine for drainage and irrigation purposes.

The planters reported that although the East Indians had done well to get the industrv started. their methods were very primitive and they lacked the necessq hple&nts for efficient production. The company's plantation could be envïsaged as a kind of demonstration model, and would provide an "object lessonn to the s m d growers (Potter lWS:239).

Well before the planter eiite expressed anv interest in rice and rice-growing, East Indian

families were dtivating two OC three aops a year on old estates and newiy established village

plots iike those surrounding Novar and Dundee. But mechanization was to introduce new

demands on nce growers in Guyana. As noted above, traditional rice-growing requires high

invesunents of labour but relatively little in the way of cash outlay. I t has been noted by several

scholars (sec particularlv Despres 1967) that there was a good fit between the kinds of labour

demands of traditional rice-growing and the East Indian farnily or household group. Despres

( 1967: 89) writes:

The pattern of these econornic activities is functionally related to (the) kinship structures [ ... 1. Rice cultivation is phar i ly a f e proposition. This is partidarly the case among sugar workers who combine wage employment with cash crop farming .

Families combined into larger patrilocal corporate groups of the son desaibed by Despres

( 1967). With a group of this size, rice cultivation on a large-sale could be managed efficientiy

and cost effectivelv. Oral reports confirm that such corporate groups did hinaion as both

econornic and as religious units. In the early nineteenth century, residents of Dundee s t i l l

iC performed the Gau Puja (lgau puuja/ "village ritual") in large corpora te groups. . *.

OZd people say that whm t h 9 w m mal1 it was usual for the viüàge to do a yeariy wmk caUrd the Gau Puju. Gau, thty Say, meuns viikge and this war the wu& thnt wouU mlirt the help of one m mon deodats ( ie . divinities) in ProMing the villagefint m n g and mrrfortune. Thc site o~fthe work w u some borind-p. Peoph would &il1 one or two goutrfor the %orrndagmtcuta: AI2 the mm w d attmd - black, MosIm and Hindu- Tuday the buunda'pmaster no longer pmtects the village ar a whuk Imtead,

he hangr on mes in a partimlaryard and the houe owner must do h t wont abne. I t hm become a guarded and semtive rihuil. At night, the man will go out and offrr a pmyet for the home and fami&. m n g s are made of white mm, b& cheere and cigarettes. Mnny peopie conri& thic work d a n p u s - if one forgets or Lrpses,for w h a ~ e r reason, the boundarptastm is likely b 'get va" and will hold ontu a home rnember. Zn such cases, i t ir necessug? to bring in a medium who ïs abk to contact the spirit and become its voie. rite vilhge work has b m e individual wu& and the botindaria have been redrawn ritually amund the house w h m thtg ured tu be amund the viihage. A work that was ona done for the safecv of al1 members of the wntntunity kas becorne a dangerou individual venture. Whm I ask why this hm happencd the oldcst man in the villap telk me l u t nowadays people a n dfford to do the wunt themelves and when one pmon does the work thfg min? mon? blnsing - the blerring ir mon

concentrated ( Fieldnotes, IL23 ).

The dissolution of such corporate groupings was never completed and they are today still

important for the organization of ntual, work and authority as 1 show later in this chapter.

Hotvever, colonial obsenrers considered such traditional rice dtivation methods and labour

organization both inefficient and "primitive." The first moves made toward mechanization.

however. vielded disappointing results (Ha* 1975, 1979). Between 1909 and 19 15, the

colonial governmen t. working with American capital. granted extensive concessions along the

Abarv creek and inland from coasral villages. After some problems with the equipment. rice was

cultivated along a ma of 7920 acres. With the coming of World War 1, the Americans left

and East Indian tenants took over the land. retuming to traditional rnethods of cultivation:

But the example was not lost. for in the same year an East Indian with a 99-year lease of 602 acres on the Mahaicony Creek, imported a tractor and ploughs. He had laid out what was described as a 'weil-organised estate with pumping station. imgation uenches and a complete rice-milling plant ... with several large ranges for the accommodation of labourers'. Aithough others did not readily foilow his example. and

it was to be some vears before 'mechanical appliances' replaced animal power. the general level of cuitivation among the East Indian famiers improved considerably by 19 15 (Potter 1975297).

In a report which argued for continued mechanization of the rice produhg sector in Guyana.

Douglas. nevertheless. described the American venture as *dl-fated and unsuccessful

( 1930:3 7). " However. he did note that one enterprising cdtivator had adopted many of the

mechanical innovations successfully:

A most interesting Wit was paid to the property owned by Jugdio Bros., on the Mahaicony River, on both sides of which considerable are& are sown to rice. This venture occupied two years; but by means of this canal, coupled with pumping plant instalied on the banks of the river. it is possible to irrigate a considerable area of land in times of defiaent rainfd. Here &O the maximum development was observed in the application of d t i v a ting machinexy consisting of tractor-operated ploughs and hairows. mechanical threshen, and reapen and binders-all. be i t noted, of Amencan origin.

nie Abary River nce scheme was, inadentally. promoted by Americans, and it is rather more than probable that the Mahaiconv idea of mechanical mage, harvesting and threshing, together even with some of the e&lier tractors themselves, was derived from that ill-starred venture ( 1 93O:3 7-38).

The experirnents in mechanization at Abary and in West Coast Berbice, although not

themselves whoilv successful, had the effect of usheruig in a new way of rice cultivation in

Guyana. After the experience at Abary, a number of farmers began to modemize the dtivation

process. In this way. people like Jugdio not O* irnproved the quality of the rice grown. but

also greatiy expanded the sale of production. The tum to mechanization had wide-ranging *

economic and sotial consequences. First of all. it tended to edge out the srnail producer or

patriiocal extended familv in favour of larger landholden who could use mechanized technology

more efficiently. Those s m d producers who did embrace mechanization were &IV to get

adequate return on their investments (Hanley 1 975, 1979). But more often than not. s m d

producers (those famiing between 3- 10 acres) wodd not have been able to put up the cash

outla- hvolved in medianized production Tractors. pesticides, insecticides. spetialized

paddies. e tc - this was a very different business h m the one described by Russell in the late

nineteenth century. As small producers were edged out, and smd-scale production became

inaeasingly less cost-efficient. a few corporate groups began to buy up rice land and continued

to rnodemize the production process and increase the scale of production. Big rice farmers

turned more and more to wage labour. The overaii eEect was the establishment of two

relatively stable. intenelated, labour systems - one based on wage labour and dass relations, the

other based on domestic labour. kinship and famiiy based corporate groups.

Depending on their cornmitments. scholars tend to see the East Indian cultivators in

Guyana as very conservative or as péutidarly modem. Thus, Potter ( 1 9 75:îW) suggests that:

The diffusion of new ideas and techniques into the traditional industry of nce-growing was bound to be a slow process with the conservative East Indians. The rniiiers with their strictly commercial approach encouraged the move away h-om subsistence attitudes.

On the other hand. Hanlev ( 1979: 175- 179) argues:

ïhere can be no doubt that the traditional system was extremely arduous and made heavy demands on the memben of the f d y and neighbours, so that most people were quite happy to move to a more convenient fonn of production ... In addition there were status elements involved in the move to more modem cultivation methods. The tractor came to be seen bv fanners as the symbol of modemity and progess, with the result that to have such a machine parked conspicuously underneath the house became just as much a status symbol as the motor car parked in front of the suburban house. and in neither case was very much attention paid to how effectively the expenditure could be iustified ... Although farmers still speak nostaigidy about the old system. the women are almost uniMrsaUy glad that it has passed. since they had to do so much of the planting and reaping in addition to their mual household duties and many state that they would never go back to the fields again ... anyone chwsing to revert to traditional methods when mechanized ones are available would risk ridicule and dension from fellow farmers.

It would seem that Potter and Hanley are taking about w o quite different situations and. in

faa, their remarki are directed to periods separated by fifsr yean. Howwer, it is more

important, I think. to recognize the validity of both descriptions and the possibility that two

value systems were operating within a Iarger economy. The consenrativeness of East Indian

faxmers can be at least partially atvibuted to the facts of structural integration. whereby family

structure, community organization and rice production are aU fünctionaiIy interrelated. On the

other hand. the keenness wîth which modemity, in the form of the tractor and other tools for

mechanization, was embraced is more obviously linked to the emerging cash economy, wage

labour system and associated systems of hieraxchicai status distinction and dass aügnment.

2.1.2 WHYRICE?

With the exception of the large-scak production schemes. nce-growing is no longer a

highiv profitable business in Guyana ( Hanley 19 79). The tum to mechanization and the move

away from traditional fonns of labour organization meant that input costs began to outweigh

returns for small cultivators. In a senes of artides, Hadey ( 1975, 1979) has attempted to

explain the reasons for continued govemment and international aid support for nce-growers in

the face of a growing recognition that such agridtural projects are no longer economicallv

viable. Also in question is the devotion of agriculturalists to nce despite the fact that three

quarten of them had total incomes below the US target for aid assistance of Gâ600 (US$270) . in 1980. Hanley ( 198 7) suggests four interrelated reasons for the continuing importance of

rice and small-scaie cdtivation in Guyana. Of particular importance are the vast infrastructural

invesunents that have already been made by Guyanese and colonial governments as weli as by

small and large-scaie producers. These indude the extensive systems of drainage and irrigation

in the form of canals. kokers, sluices and water conservancies as weil as machinery such as

pumps. tractors, combines and even s m d airplanes. Another aspect of the preexisting

infrastructure is the bureaucratie and informal organization of productive activity in the fom of

fanners' cornmittees. miilers' associations and rice-growers' lobbies. Added to this base are the

historical developments which have taken place in reaction to the plantation society. Plantation

agriculture had an enormous and long-Iasting effect on the Guyanese economy, such that most

productive activity is focwed on export value rather than on local, national use- this despite

the efforts of PNC govemments to create a selfsuffiCient peasantry in the 70's and 80's (the

'feed. dorhe and house" campaign cf. Premdas 1995). Hanley further notes the strong relations

betsveen ethnic ideology and certain kinds of productive activiw. 'Indianness* has become

stronglv associated with rice-fanning such that to grow nce is in fact a performance of

"Indianness" (Hanlev 198 7).

For fanners, justification for nce growing, even on a smaii scaie, is found in the faa that

ir allows one to Live and work without the constant presence of a bossmnn (espeaallv since the

cmergence of mechanized cultivation). To be able to "work to please oneseif" is a goal many

Guvanese, farmers or othenvise. hope to achieve. Such an emphasis on autonomy has far-

reaching consequences for soaal process, as 1 detail in the foliowing chapters.

The situation in Mahaiconv is perhaps rather different from that in other rural areas.

~Mahaicony was one of the vesy fint areas to see mechanized production and the

commercialization of the rice industry seems to have advanced W e s t here as a result. Today w . .

very few people are small independent cultivators (unüke say Black Bush Polder). Instead. most

36

are wage labourers. workùig for the few people (or companies) that have large holdings in the

area. One such producer is Mankad. He lives in the village. owns the rnill. hundreds of acres of

rice land and rents out as mu& more from others. A few others grow on a large. yet not quite

so impressive, scale. ïhere are aiso the large government schemes. Of coune, there are still a

few smd-sale producers who rely on &ends and f a x d y members to assist a t shai (/shai/

" sowing rice") and nrt (/kat/ "hawesting rice") times. The retums here are minunal and ail the

small producers that 1 know supplement their nce-growing income with a number of other

activities.

For most people. the rhytiun of work is verv regular despite the variesr of productive

activities in which viilagers involve themselves. The character of the village and the mmment

of bodies tkough it changes throughout the course of the day. In most cases, it is the woman

(cldest daughter or mother) who wakes tùst to put on nce and begin sweeping out the dead and

dying mosquitoes and c o d e s that litter the floor. Soon after. the men wake and go to dean

their teeth on the road or from the back porch. As the trucks. minibuses, cars and b i d e s come

out ont0 the road, the hum of activity grows more intense. People appear at landings.

toothbrush in hand. caiiing out mamin "momingn and wa hapm =what's up." Croups of men

gather bv the road to smoke cigarettes and read the paper while women start to ready the

ühildren for school. A few weli-dressed individuals stand up to stop transportation into

Georgetown. inevitably. some gyaf is initiated conceming the day's activities, work to be found

andior who made noise (was taUEing hard) the night before. As the trucks come out. labouring

men humiedhr drink tea (usually one or two roti. cuny and a hot drink) and catch a ride to their

work site in the rice fields. After the men leave. the women get a diance to take tea and gyaf

sornething with a neighbour. Children have to be ciressed and escorted to school before the

work of washing out the house. daubuig the floor. attending to gardens and fetdiing eggs can

stah Wîth the labouring men gone, by about seven or eight o'dodc the pace begins to slow

somewhat Some women work as domestics in other people's houses and they WU huny to

dean out their own house so as to kee up the rest of the day for wage labour. The carpenters

and shop owners, the men that work in Georgetown. iine the road waiting for minivans to c a r y

them up to town or down to Berbice. Women often make one or two stops in the rnorning by a

hiend's h o w to ta& name or stoty (i.e. gossip) and get the latest news. Throughout the day,

c m pass carrving greens or fkh or dothing, and perhaps some religious pictures. The more

confident wiii c d for the ciut to enter the yard while others rush to the road canying bags.

Sometime in the aftemoon, the rice workers retum from work Many, if not most. stop

by one of the rumshops strung along the road. Now is a time for serious talk as weii as the

joking banter of male peer groups. If their wives don't send someone to fetch them earlier. men

usualiv make their wav home for dinner around five or six o'clock. by whidi time the curry and

rice are waiting. The women have spent the aftemoon houn in gyu$ deaning and cooking. etc..

Table 2.1 gives an account of the varieg of income producing strategies which members of a

single household mav engage in.

GreendiCitchen Garden 1 for hounhold

Crcole Fowl

Ncady cvery house has somt kind of kitchen gardcn. Ochro. boa. 1

Crcok fowl arc a hardy bncd which tend to reproduce th& numben with lide hwnan intervention. They can bc caun or kcpt for the tggs. Thcy arc r a d y sold for mcat. 20 fowi can be expeacd to give about 10 eggs eveqr two tiays. I

Whiv Fowf

Shctp or Goat 1

Box (a local btwry) I

4-5 pound bud sek for $500

7 pound bird seils for % 1260-

prie variCs according to season

O t h e m known as "nuat bird" thêy arc kept in d penr and tlkt about 6 wttks to -ch maturity. htkd cas is about $70 for the c h i d and fced c m total about $200-300 per bird Nua to dudc this is the n u i s t f i v o d p o u i t r y i n d ~

Du& rrqilliics the same paddy and spa?l feeds as do white fod They maturc at about 6 months- The art much l e s rcsilicnt to distast chan both the white fowi and the atok fowi. iMost people consider raising durkaNkybusinesj.Itis~Itto~eaptdeaofmonryh loQlnuiduds(with50-100d&) butifa'rick"hitsitisquitc possrble to lose aiï one's duck overnight

These animais arc minded in soue households. Most of the goat and shctp consumcd cdnv h m hoIders of largt tladu.

Sime nwt people in the v&gc are W u . beef is usuayI not e n . ahhough therc is a srnaJi Moslem market for i t Richer follcs may keep a cow for milk. A cow usu- costs about % 15-25.000.

The canai. the l i . and the vzrious hxdy-awned ponds are d for local fishing. Whcn water is low. 6shing tan be very productive. yidduig catches of up to 60-100 hasar in an aftemoon. in addition to mixrkneous srnaii fish (patois. hourii, nhterbait. cuffom and spring kurass). Howevcr. whcn the water is so lm and 6sh arc pkntïfd. such catches arc to diffidt to sel1 and msî is uscd for houschold consunpion with a good deal of the catch king rtdisviburtd (gtnenlPcd rcciprwty). Whcn the water is high, hasar xii for about $80 a p i e (relative to sïze). An initial outhy of $3000-5000 î. neccssary to buy a cast n e t

Stands x U a varieq of items besides the usual ginio or chicken foot (which are crut~chy. &ep 6ied sticks made h m Oour. water and d t

and caokcd in ooconut oiI) induding cigarettes. pira ( l d M g e ) and biscuits. Initiai outlay is minunal although vendors arc officiaiiy ucpccred to get a hanse. S o m people d and WU with similîr goob.

This indudts a range of EpMalities including fixing bicycles. umbreUas. and other items invohring rom technical skill and a ftw tools. ,Mcrnbtn of household may acquirr a xtputation for fixing a panicular item over a numbcr of gcnerations.

Box money is usuaiiy paid weclcly. but this varies according to the circumnances of the conuibuton. Unial dcposits arc $500 or O 1000. Otpcnding on the number of people using the box. it rnay go for 10- 12 wceks rcdting in pyoffs of $10,000 or S 12.000 for cadi person. People must 'put in" CO rcceive theu box when they join up - çome want it earfy for houehoid rcpairs e t c

Daywork

Shop

Rice Land

The gandtning business can't make you rich! There are the unial problems of suppiy and &mand (on national. regionai. and pcrsod lcvcls - if p u çcii LOOLb &dot evcryixdy wants them. if you ECU S00ib you have to cany them home with FU), A sin& gaxdencr is abIe to plant about LOOIb shaü~t cvcry 8 wetks ( 14 be6) whLch wiR yreld about $ 1 0 0 . 0 if they sen at top pria ($ Iûû/lb). but the= arc a numkr of complicating fima besidcs those of the market, induding neig)ibours. animais eating them out. droukfit. probiuns with transportation to marice~ A number of txpensts m m atso be amuntcd for, induding fercSlizer. disease spny. land transportarion. Iabour for wtcding. Ni in an. the average fdi-tirru gudener is probzbiy 1- if hJ&e rnakcs the (fower) avcxagc &y-pay of $500-1000.

M a hqucntly, thiE den to bagping and ioading paddy the payanent for whicfi m d y uccttds SIS a bag rrsultuig in a da@ pay of $500-8ûû. This usuayl laves a substantial portion of the &y frte for other actnnties which makts it partrarkrfy auxactive for youngcr men who m t to spcnd at lean çoxxtc of the aftemoon Liming by the mad. jobwork indudes wccdùig, for which peopk m d y ~cc ivc between $500- 1500 for a half &y's woiic Contract jobs arc as6 considcd "pb work". This indudes arpcntly and mapr pubtic worlcs (weeding and deaning trenches, bui ihg kokers and bridges. etc-). Paymcnt usuaüy worlLs out to about $ I200/&y for a woriunan and $2000 for a bossman.

Private and govcrnmcnt ricc producers unially pay by the day. Bosses and workmcn usluyI meive the sarm $600/day (bosses arc rcquind to do less manual work) at p-v bm. The work is rcgJar and there is rhe opportunity to make ovenime pay ( 1 Yi). Gowmmcnt p d u c e n pay les. ($450-SSO/day). but the work is lighttr and the= is more job securiry and benefits (induding paid hoiidays and sïck days). Womcn aiso do daywork. Thû i . uniaUy domestic iabour in one of the big houses- Payment for wo& gcnded ferninine is considcrabiy lower. rarciy exaeding $20-300/day.

This indudes salaricd employmrnt with p e n u n c n t (offitt2. schools. mspectors) and pxivav businesses (innrrance king the mon common. and Eomc ricc exmutive jobs). The local gwcrnmrot employs a t I ~ z t two peopie to nin the village office as weii as nurses, media and, of coune. teachers. Paymcnt is minimai with junior ttachcn earrwig about $SOO/day and hcadmasc.cn about $2000/&y. The more pccnigious lobs (such as woriung for the regional diainnan) usually pay much betrtr (bcwecn 2500-HXM/day) and often inclt.uk 'side moneys*.

Owning and opcnting one's otvn &op is the aspiration of many villagers. I t does not invohm the kind of acquiewncc to cxua-local authority as do the jobs citd as "sahicd emptoynunt". but at the same tinu. canics with it a good dcai of sooal and matcriai rcwards. Shops can be quite profitable if they have local badung in the form of capital. patronage and the willingness of the fd authorities to tum a blind eyc to smd infractions (opelaring wirhout a Liquor licence). Weil-to-do EImilits wili oftcn set up a shop, If the woman is to nu\ it. the shop will cany d kinds of provisions. perhaps a fcw artides of dothing and oftrn druiks and biscuits. If the man is to manage the shop. it wiii most kcly be a ruxnshop with his fnends and mak bmdy being the primary PaVoN-

The problems and possibilitics of ricc produars have aktady k e n

2.3 THE PUBLIC FACE OF KïNSHIP - P0RTRAiTS OF THRE]E PATRILOCAL GROUPS

Despite changes in the organization of labour. the traditional communal and patrilocal

y t e m is still very much a part of the social character of the village. Thus, while the main areas

of productive labour are not organized around the corporate group of a pavilocally extended

family. many other kinds of work and activity are2 House-painting, various construction

projects, car mechanic work, and tomb-building wül often involve men who belong to an old

corporate group (or two allied ones). More pmnounced. perhaps, is the assembly of corporate

groups at various kinds of reIigious work. induding death rites. jandhi and the various pujas.

Despite the weil known and often ated expression an1 a h a wanfamalii 'AU of us are farniiv."

people have a fairlv firm sense of who they can, and. who they can't. cal1 family. Thus although

the notion of family is often fiqively extended. the importance of 'bloodn is made quite explicit

in evervdav . - tak This exclusive and corporate nature of family is expressed in the proverbfumlii

stik a ben bo ii na bmok "the family-stick can bend but it can't break" As 1 discw, with reference

to a particular case, in the next chapter, such family solidarity is motivated, at least in part. by

the recognition that the pauilocal group has a very public face. People in the village are not

often considered in the singdar - more often their actions are considered to be expressions of

the famiiv character. This public face of the pavilocal group has serious implications for its

membea. Political positions, favours. and respect all hinge on the name and reputation of the

famiiy. Besides its importance in terms of reputation. the patrilocal group figures prominently

in the organization and disvibution of both propeq and domestic authority. In the k t case.

' It should be noted that wage labour (day work or job work) is often donc in "crews." Ricr producers do not hire inilividuals but groups of men who are used to working together. The composition of thrse crews iisuaily shows the influence of patrilocal group allegiances.

41

although some groups are landles, pauilocai group property is still gener- inherited by the

sons. As 1 have discussed earlier, patrilocal group land holdings pxwîously functioned as the

main productive resources in a system of labour-intensive peasant agriculture. Today, few

people dl f a m their patrilocal group land holdings (although there are exceptions - see

below). With the move to wage labour. landholdings have thus become l es important as

economic resources. Often they have been sold or are rented out to large rice producen. Stiu,

the idea of a pauilocal group landholding, embodied most obviously in the sentiments attached

to the yard. is of central cuiturai importance. The rights to houses. house plots. and yards are

reckoned in tems of patxiiocai group membership. The death of a younger, propertkd male is

often followed by a number of disputes in which relatives make daim and counter-daim to

various properties. In the case of older household heads, the structure of authority between

vounger and older sibhgs has usualiy had t h e to work itseif out and disputes are avoided. But

land in the village is more than an econornic resounie. I t is central to an individuaïs sense of

persona1 identitv. Kinship and other kinds of soaai relations are cont indy expressed and

realized in the dailv movements of people through the spaces of the nllage (Chapters 6-8).

The patrilocal group is also, of corne, vital to the organization of domestic author*.

Because they hold propeq in common, members of the group often have a reason to meet and

make decisions as a collective. During these meetings. one senses a tension between the official

distribution of authority and the practical realization of social power. Offiaally. the oldest

resident male is the household head (the patriarch). Practicdy, decisions are made not a

lone individual but through the alliance of several siblings. Members of a family will hrquendy.

and quite openlv, state that they have a favourite sibling. Often this is the one to whom they

42

are closest in age or. altemativelv, the one to whom thq feel materndpatemal. These speciai

affective bonds are important in politicking, and a favourite sibiing can uswdy be c d e d upon

to champion one's cause (in fact. it is expected and others will note that "those two stick

togethern). Despite the possibilitv for such suategic ailiances. the ideology of age-graded and

gendered solidaritv is a powerfid force structuring authority. I t is expressed in a multitude of

everyday activities, some of which I discuss in the following chapters.

Relations between patrilocal groups are also of great importance. I t is here. in fact, that

the public face of kinship is most obvious. Some groups are thought to be aiiied or otherwise

iinked while others are opposed. Relations of alliance between groups may corne about rhrough

marriage. econornic cooperation, chüdhood Mendships, and are often reinforced through

r i tud3 In such cases. the ideology of age-graded and gendered solidarity is extended to another

familv and chikiren will be insuucted to cail this person 'unde" or "auntn4 In generai the

authority that senior adults and household heads draw from their domestic group is extended to

non-kin relations. Thus, senior people often send out unrelated chüdren on errands. Refusal to

complv is inevitablv interpreted as y e - p a s "an act of flagrant disrespect." This said. it is not

hard to sense how the com peting structures of hierarchv wïthin groups and those between

groups result in a tense field of social interaction. Thus. young members of the most powerful

patrilocal groups wiU sometimes daim authority over much older memben of lower-ranking

3 Members of ciosely dlied groups take a very active role in one another's yearly religious work (sometimes strving as stand-in hosts, helpers. cooks etc.). During dead-work they often spend up to 13 days with the family. A specid ritual day called rd i i involves the making of speaal fictive kin bonds between women and the men they take as brothers by tying bracelets on their w r i s t

4 Hindi kin terms are generally reserved for use with "actual" kin. The exception is bmji i *brother's sisters's wife" which is often used creatively to express a relationship not so much to the woman but to her husband ( i x . brother).

patrilocal groups.

In contlict situations. members of a pavilocal group wiii attempt to raUy support fiom

their group. Although disputes often take place within the patrilocal group, when they occur

between two groups thev tend to be more public. Inwitably, at some point. the two families

meet on the road to have it out. Such conflicts. though relatively infkquent. are remembered

for a very long t h e and the relations of 'not talking" that result sometimes extend over more

than one generation.

The overall character of pavilocal groups varies dong a number of dimensions including

their size. their s tatu, their level of intemal integration and the degree to whidi they act as a

corporate entiry. Some patriiocal groups have even dissolved and their members, for all intenu

and purposes, are individualS. To give a sense of this varying importance of the patrilocal group,

I now give three bnef portraits.

2.3.1 BOLO

This is one of two families revîewed here that trace their pauiline back to the original

purchase of the estate in 1883. Aithough Bolo offiaally encompasses a number of other

hmiües (Bolo was father to four sons and three daughters). it is the family that now occupies

his house that is most assoaated with his name and reputation. These are the children of

Michael. Bolo's voungest son. People report that Bo10 was rich by local standards. The famiiv

had sigdican t plots of nce land. a large yard and a fairiy grand house. Bolo's authority is also

legendarv in the village. During Midiael's time as household head. peasant nce f a m g was

generailv abandoned. and the first important steps towarb wage labor were made. The familv's

44

fortunes gradually diminished. Today, the 1 1 k g children of Michael occupy a range of social

and economic positions. While Nancy and Shanka üve in two of the srnailest houses in the

village and face penodic shortage, Kota and Joe both own relatively success€d bakeries. Papso is

a prominent local politician, construction boss, govemment employee and sweet-man. Dada

has moved to America and occasionaily sen& remittances. The rest of the siblings ocmpy

stable working class positions depending on the wages of at Ieast one regdar worker and a

number of other supplementq forms of income. Although ail members occasionaüy visit the

family propertv and their mother who h e s there, some are more invoived than othen. Gobin

and Papso play an important role in the maintenance of the patdocal group (which now

extends over three generations) and its property (Midiael's house and the yard). This is

particuiarly tme for Gobin because he h e s on this propeq in his own house. Papso iives across

the road. His involvement stems fkom his prominent role in the vülage and his fairiv high

position in the network of adult men. Many people consider him responsible for the familv and

the actions of its memben. He d e s on this role frequently and often acts as the final judge

and arbiter in familv disputes. Complainants seek him out to resolve any Merence involving

individual members of the patriiocai group. Gobin and Papso also play an important role in

policing the actions of the inhabitants of the pauilocal group house. This indudes their mother.

Kay. and several of her adolescent granddddren. Papso is often cded into the yard to d k to

his mother and his nephews. When there is a problem with noise or undeanliness, it is Papso

that usuailv receives cornplaint. In more extreme cases. or in cases where one or more members

see Papso as biased, Kota, the oldest brother s t ü l lMng in Guyana may be called in to settle

thuigs. Each of the three sisters expücitly allies herself with one or more of the bmthen. Such

45

alliances are reinforced by close relations between the brother and the sister's husband. This

affine, known as bonai. is almost always more effectively intepted than are in-manying &es.

In the case of the Bolo. the in-marrying wives make alliances primaniy amongst themselves

rather than with the sisten (Le. the daughters of Midiael). In the one case where this was not

m e (Le. where the in-marrying wife was doser to the sisters than to the other brother's wives),

the other in-maiiying wives pursued a fairly relentless strategy of scandal against her.

Although th*; ~ o u p has a very public face and. people believe. a svongly ucpressed

coilective 'character," memben frcquently corne into conAict with one another. Confiïcts most

often arise over property, perceived ge-pms, and talk which leads to scandal. The brothers often

feel that one or another of the patrilocal group memben has not livcd up to their obligation as

kin. Such a failure to iive up to an obligation based in kinship (such as attending religious work)

is taken as disrespectfd and as an indication that the offending Party thinks he is too important

to bother with the others.

Despite this. it is hard not to notice the way in which the group coheres. The patrilocal

group is a source of strength and soaal power for its members. This is learnt v e y early. When a

voung man threatened to beat her, a nine year old girl, a member of the Bolo group, tumed to

hUn and said quite p r o u d v p buun biit Bolo - nonbadii &am biit Bo10 'You can't beat the Bolo.

Nobodv can beat a Bolo." Conflicts invohg senior members of the group often escalate into

battles of a corporate nature. The Bo10 tell several stories that explah present relations with

other famiiies. In one case. a problem arose because a neighbour to Gobin and Papso beiieved

that Papso had infomed his supervisor at work when he took extra long breaks and days off. 0 .

Every night this neighbour would drink rurn and, when he got home hom the rumshop. wouid

46

begin to cuss Papso. One day, Gobin asked Papso if he had never heard the man cussing him?

Papso was funous. That night he took the man out drinking himself. M e n they paned. Papso

seaetly hid himself in Gobh's bottomhouse. But he got impatient waiting for the man to start

cwing him and before he heard anything he jwnped out, kn-d down the fiont gate. and

threatened the man with a cutlass. When the men began to cuss each other, they were joined

by other members of the5 respective patrilocal goups. Eventually, the police were c d e d into

break up the fight.

Much smaiier than Bo10 is the group known as Teelü. The core members of the group

are the chiidren of Baker John. five brothers and one sister. Of the five brothers, two have gone

'outside." one to Amencan and one to England. The latter, the eldest and a doctor, has been

very successful. When he visits, he is treated with respect and some curiosity. After twentv

vears outside the country he seems, in his speech and demeanour generaiiy. more Engi.ish/white

then C uvanesdIndian to many people. As one brother died early. and the sister lives with her

husband, the real core of th is group consists of two brothers who occupy the house of the now

deceased Baker John. While OLd Baker John buiit a reputation on his knowiedge of several

knguages (Hindi. Arabic) and a profound understanding of Hindu theology, his sons are known

more for their jokey and good natured sense of humour. The two brothen are both slight but

assertive and charismatic. The youngest. Tus, operates the shop which is located at the front of

the yard. This is one of the most popular shops in the villages and it is where both the more

powemil bosses and the tightly knit aews of workers corne to drink One group of senior men

47

bges 30-40). who f o m a highly integrated p u p . cornes here everyday either to arrange work

crews. to organize other kinds of coiieaive labor. or just to sit. gyaf and drink nun in the

aftemoon. Although the core patrilocal group is small, the brothers have a very reüable and

tight &de of long-standing hiendships. The older brother. Pa& is one of the few remaining

smaii-scale rice producen in the area and such fnendships are crucial to his economic sumival.

The core group of friends regUi..iy assist Pank in the field. In return, those who help out are

accorded special privileges in the shop. They are given creck, allowed behind the counter. given

free rum and food. These men are accorded the statu of fictive kin in the household- Both

brothers place great importance on their friendships. Unlike many other men in the village the

Teeli do not involve themseives in public politics and other iwtïtutions of legibated authority

like the Mandir. Rather, they draw support form their central role in the male soiidarily

network.

Both brothers married women from the immediate area. This has had rather different

implications for each. The older brother's wife belongs to a large local family. The mamage

aiiowed Pank to caii upon many people at times when labor is needed for rice dtivation and

harvesting. In general, relations between the Teeii and Pank's wife are very good and there is a

high degree of cooperation between the two families. The younger brother. Tus. married a

woman kom a once time quite weaithy farnily. The group has. however. recently seen some

financial difficulties and have called on Teeli for assistance. When Tus eventually refused. his

wife took monev from Teeli funds to give to her sister. Police were d e d in and for some cime

things were quite tense within the pauilocal group. However the problem was eventually

resolved without surfacing as a public dispute between the two qoups.

2.3.3 PAKAR

In contrast to the families reviewed above. the patrilocal group known as Pakar is

essentidy a female enterprise. Currently the p a t r i i d group land is occupied by Zizaan and

her daughters Ashan. Shana, Fasia. and o c c a s i o ~ Shamaroon. Bat, the youngest of three

bmthers. also lives here. The cenaal economic interest is a weil stodced shop and lurnber yard.

Shana and Bat also keep gardens in which they farm s h d o t Fasia does some seamsuess work

One of the two non-resident brothers is a water carrier. Using a donkqr and cart, he coliects

water from a pump near the family shop and c-es it up the road to a village where there is no

public water faeüty. The other brother is weil known 1ocaUy. He is the owner of a very large

shop in New Amsterdam and has strong connections, locally with the police and with customs

officiais. and, intemationally, (in Surinam) with traders and exporters. Pakar himself, now

deceased. is remembered by his famiiy as a pious and learned man who was poor but fluent in

Arabic. Hindi as weil as Sanskrit. Others remember him as a 'nak man" who pretended to

know ritual curing in order to get close to various young women in the village. Pakar's children

are less involved in the local mosque and viiiage ritual life than theix father was. They draw

social power mainly from their partidarly stable economic position. Ashan recounts that they

were not always well-off. In 1964. during the racial nots of the Burnham years. flour was

bamed as an import. Taily, however, had a contact in Surinam who could provide him with

flour. He brought it aaoss and Ashan sold it. According to Ashan, people came from ail over

Guyana to buy flou hom her shop.

Pakar keep a veqr tightly knit corporate group and. although they have a great manv

acquaintances and business contacts. they eschew the pursuit of 'friends" (i-e. popularitv).

Those constantly in pursuit of friendships, they note, usually end up "brooksn (without

resources). Pakar advocate keeping company (and their private 'business") within their own

PUP-

In their relations with other patrilocal groups, Pakar attempt to appear generous and

benevolent. Loyal and reliable employees are rewarded with reasonable pay, consistent work

and assistance in times of need. Pakar are quick to assert that they cannot be txifled with,

however, and when somebody crosses them, they are relentless in their pursuit of 'satisfaction."

When a boy who had worked for them (whom they had taken pity on) stole jewellery, they not

only tracked him down through several villages but also found out his accomplices and pushed

the issue to public attention. Later, they arranged to interrogate the boy so as to get all the

details on how their neighbours had been involved in the crime. After a year or so, they

employed the neighbour as a casual laborer secure in the knowledge that she would not uy to

steal from them again. In disputes, Ashan calls on the authority inhering in the ownership of

pavilocal land and in the leadership of a powerful patrilocal group. When a worker for one of

the large-scale rice producers came to buv cement, a dispute arose over how many bags he had

loaded into his cart. When Ashan went to count the bags, he held her hand. Ashan was furious

and sent him out of her yard without his cement. Later. the man's superior had to come and

apologize. Ashan complained that. by holding her hand, the worker had taken his eye and

passed her. Ashan resented the implication, which she drew from this action, that, in relations

of status, gender was more important than age and social position (as the head of a powerful

patrilocal group Ashan was c e d y the young man's superior). His hand-touching, she felt,

implied the possibiliw of a sexual/romantic relationship. In this, he was very much over his mark

(i.e. 'he had exceeded hunself 'he was mistaken"), Ashan remarked.

Ashan. Lke other household heads. was thus qui& to draw on the authorin, and social

position that cornes from the owneahip of property. In one case. a man was building a new

toilet for his house which was on land that belonged to Ashan. Ashan told the workmen to put

the toiiet to the side rather in front of the house where she would have to look at it. When a

report got back to Ashan that the man's wife had been burin h i . w e r this (asking if Ashan was

not his d e ) . she went straight down to their home and told them that. if it happened again.

she would throw them out,

I t shodd be noted that although Pakar ülustrates the range of patrilocal, groups within

the communitv, this situation is quite rare. In most cases, patrilocal group authoritv is. at least

officia-. smctured around a male voice.

2-4 MARNAGE, POWER, GENOER

A central theme of the present work is the way in which sociai relations between men

and women are negotiated in interaction. [t is thus necessq to give some details regarding the

impact of kinship and marriage on these social relations. Rather iittle attention has been paid

to gender and domestidty in East Indian Caribbean comrnunities. This is particularly striking

in çoniparison with the enormous literature on mauifocaiity in Afro-Caribbean comrnunities

( for an ove~ew. see Smith 1988). As is the case with matrifocal Afro-Guyanese comrnunities,

in East Indian vilkges one h d s a strong association of women with the domestic. In a

pauilocal setting, this is less a source of sociai power than it is a hegemonic device for , 9 . .

maintaining gender inequalit-. Many researchers of Afro-Caribbean households have argued

51

that a pubiidprivate dichotomy is incapable of adequately describing the social o'ganization of

evervday We. In their day-to-day iives, women bridge the public and private, domestic and

political (see Berleant-Schüler and Maurer 1 993).

In the patrilocal context of the Indo-Cuyanese MUage. however, a nurnber of powerful

forces attempt to keep domestic and pubiidpolitical separate. At the local level. women are. in

most cases, exduded from overt participation in public poiitics. As 1 have desaibed above,

women are &O generaily exduded from oven participation in the workings of the p a a i l d

group as a politid unir

As I show in later chapters. gender inequality seems most extreme for married women

who live with their husband's patrilocal group. Female affines are not s t m c t d y integrated in

the same way as male ones. Whiie the wives of senior patrilocal group men draw authon. from

their husband's position (and must be accorded a certain respect as boujii 'big brother's wife or

father's brother's wife), the wives of junior members often find themselves in a di f f id t position

vis-a-vis the patriiocal goup with whom they live. Slight. or even perceived, breaches of what is

considered proper for a mailied woman in terms of movement, self-cornportment and social

interaction can tum into senous charges of misconduct as they &date . Women who join a

patrilocal group as an in-marrying wife ofren fxnd themsehes at odds with both the senior men

and the other women in the patrilocal group. The relatively disempowered position of the

mamed women is exaggerated in the cases where village exogarny is practised. M e n a woman

cornes from far awav to live with her husband's patriiocal group, she can't easily cal1 on her

own group for support when she finds herself at odds with her husband. Not surprisingiy,

perhaps. men often express a preference for this kind of marriage arrangement One young

52

man told me explicitly that, mi na p mariid wan n dcnr gpzl wa de omnggu - mi gu maniid wan p l

fan far "1 won't many a girl from around here - 1 a m going to marrv a girl fiom far wayn 1 asked

him why he said this. He repiied. w m mi biit shi shi rnos buan gu hoom ' M e n 1 beat her she

must not be able to go home." When the wife's patrilocal group is both relativeiv integrated

and iiving close-by, they can provide support in conflict situations. In one case, a girl mmied a

boy from a n e i g h b o h g village. As is customary, they started their married life a t the home of

the husband's father. They brought all the gifts that had been received a t the wedding, dong

with other propeq that they owned individuaiiy. However. after a few months. relations

began to break down between the new wife and her mother-in-law. The husband's mother

claimed that the now quite-pregnant wife was too lazy. argumentative, and disrespectfui. On a

number of occasions, this new wife showed up a t her father's house in tears. Fuiaily. the

conflict became so heated that, while the husband was at work. the mother toid the wife to

leave the house. At this point. the husband and wife went to the wife's father and asked if he

would accept them into his house. He did. That night. a group of t h iq men from the

patrilocal group (and myseif) were rounded up. The wife's father borrowed a tractor and trailer

from his boss and ail of us made the uip to the next viUage. I t was a noisy aowd, and when

the crew reached the husband's fémdy's house, a row quickiy ensued between the two

household heads. In the meantime, the propeq was gathered up and everybody returned to

the wife's father's house feeling vindicated. In a minority of cases, then. a husband and wife

will move into the home of the wife's famihr. But this usually happens only after some

altercation with the husband's group.

- pp -

HOUSE OF ... (&EREN= NUMBER FOR HOUSE)

Brarnmie and Kavita (2)

Gobin and Seeta (2) -- - -. - - -

Gobin and Seeta (3)

Shana (4) Shana (1)

Bagi and Dari (5) Bagi and Dari (4)

Pank and Baby ( 6 )

- -- -- -

Pank and Baby (2)

Tus and Cita (7) Tus and Gita (2)

Dredi and wife ( 8 ) - - - -- . -

Dredi and wife (O)

Marna (2 )

Rohan and Channo ( IO) -

Rohan and Channo (2)

Ben and GooIin ( 1 1 ) Ben and Goolin (3)

Ben and Goolin ( 1 1 ) - - pp

Naga and Siima ( 1 )

Nancy and Shanti (12) Nanq and Shanti (2)

Baby and Husband ( 14) Baby and Husband ( 1 )

Baby and Husband ( 14) Deo and Shalini (O) - - - - - -

Papso and Isha ( 15) Papso and Isha (3)

Dinna and Ma ( 16) Dinna and M a (2)

Sun (17) ves

Table 2.2 VUGE EXOGAMY AND PATRILOCAL RESLDENCE FOR 20 D O M E S ~ C W T S .

Looking at table 2.2, i t is apparent that, despite the preference of some men mentioned

above. there is no dear pattern of village exogamy. However, there does seem to be a dear

preference for patxiiocal residence. Of the twenty nudear units in the sarnple, 1 1 (55%) were

fonned through village exogamy. Of the same twenty nudear units, fourteen (70%) were

54

praaising pauilocal residence dehed as. in the case of marriage. residence with the husband's

father's farniiy, and. in the case of unmarried household heads, residence on their father's

property.

2.5 HOUSE AND YARD

With houses spread almg both sides of the road and stretdiing badc. both towards the

bad<dam (the üne side of the road) and into the Pasture (the sea-side of the road). Dundee-

Novar is a densely populated village. Houses are generally of wood or brick construction.

consist of one or two floors. and may or may not be recentiy renovated depending IargeIy on

the recent fortunes of the f d y (cf. Plates 1-2). Along a single plot (a parcel of land usually

sweyed during the colonial penod) , a corporate group's history can be roughiy estimated. At

the road corner is the biggest of the family's houses and o h it incorporates a shop, a repair

service for cars, a carpenter's work place. or some other business.. Most of the road houses have

two floors. are fairly recentiy painted and have large yards with a tractor or car strategicaliy

parked for public viewing. At the centre of the Dundee road stands the local 'bossman's"

house. impressive in size and dwarfing all that surround it. Mankad's house was built in the

booming years of rice dtivation in the 1980's. Not far away is his family's rice mill whidi

stretches out into the backdam area to the iine. Tractors. combines. jeeps and traders are

scattered across the area between two large processing plants. Other f d e s have smaiier

enterprises (ten tractors, two jeeps etc.) and as such. these people make up the local village

elite. They live somewhat removed fiom the small dramas of eveyiay Life dong the road. They

are more mobile and their n e ~ o r k extends to others in a simiiar dass position outside the

55

village. niqr clah that they spend more thne minding the books rhan they do other people's

affairs. In faa, rhey ofien attribute their success to this ability to iive separate from the

'common" people in the vülage who "drag others dom." While their presence is somehow l e s

physical, it is nevertheless powerful in its mystique. They are seen dming out in expensive

vehicles imported from the U.S., they entertain important people behind the thidc walls and

iron gates of the5 compound. and when they talk to you, you stop, Listen and comply. There is

a distinctly feudal character to the social relations between bossman and village. The bossman

has a phone and enough economic resources to support local iniuativcs and smaii projects. A

tight &de of loyal worken are crucial to the production process. These are the managers who

run cultivation in the field, production in the miü and marketing in town. The managers üve in

modest houses away kom the road. Amving earlv and leaving late. the manager is distinguished

b his loyalty and respowibility.

As 1 have discussed. the domestic unit in Indo-Cuvanese villages is . p i cdy a pauilocal

one and uiheritance is patrilineal. This means that. aftcr marriage. a woman often leaves her

parents' home to go üve with her husband's family - or more specifically to go Live on her

husband's father's land (cf. Despres 1967). At first, the new couple may live in the husband's

father's house. If this is the case they usudy take over the bottomhouse (/batmhous/) which is

the space between the concrete or mud ground fioor and the h t raised fIoor of the bungalow

which is about twelve feet off the ground on four corner posts.

It is more usud to construct a srnall dwelling at some distance from the main house

though still within the yard. This new houe often - in fact, in most cases - offers veiy meagre * . .

amenities. It is usualiv some wav from the mad. If it is not on risers the usualiy hard packed

56

mud and daub floor can become pit of mud during the rainy season. Most often it is a simple

one-room cowuuction with a s m d overhang under which is kept a green (kerosene) stove.

Warer rnay or mav not be available dose by. Fowi. dogs. and cau rnay cohabitate the general

iiving area even if they are kept out of the house.

The yard. endosed by a six foot fence. rnay contain several houses belong to memben of

the same patritocai extended unit. I t is thus a fairly direct mapping of a saiient kiwhip grouping

ont0 physical space and has become strongly associated with the notion of an extended

patrilocal corporate group. Members of this p u p (both men and women) rnay enter freely

without being diailenged by others. This is m e for just about everybody who can trace a direct

link through their father or mother to the patriüneage. Non-resident *es of men who belong

to this pavilineal unit enjoy cowiderably iess freedom. certainly less than male affines. They

rnay c d out at the gate, they rnay only corne infrequentiy (and then for a speaal purpose) or

thev rnav r e h e to enter altogether. But. as noted above. in many cases wives iive in the vard.

These women are in a particulaxly tenuous position. They are ever under the watchfd eyes of

their husband's extended f d y , they are often economically dependent on this family. and

thev are alwavs in a subordinate position vis-a-vis this unit The vard, even more than the

house, is not their own and they do not enjoy the sense of security that memben of the

e-xtended pauilocal unit draw fiom it. For members. the yard symboiizes shared obligation

based on kinship, as weU as generalized reciprocity and corporate identity. With respect to

these meanings. the wife is always a suanger.

The house and the w d are, in this way, very much delimited spaces. Viiiagers are quite

aware of who goes in and out of both their own and their neighbour's yards. at what tirnes and

for what reasons. People ohen keep dogs and this ad& to the feeling of impermeabiliw @en bv

the yard fencing (cf. Plate 3). Depending on the relationship, a visitor will often cal1 out to the

house resident before entering the yard. Once thev see that somebody is at home and they are

invited to come through the gate they will then go into the yard space. Often, however. thev

will wait at the road for the resident to come out and talk at the corner. Such behaviour is

subject to a range of interpretations:

G. w u settling d o m for the ewrning in the hammoat In fmnt of him wnr a parter botth of white rum.

a f i puces 0ff;ch in a p h t i c bowl, watm and n littLe g k s ofwca-rola. S. rat on a bench stitching a

srnull square of chth. n i e childmr wem dming on t h wncretefiwr of the botfumhouse. Suddrnly, we

hewd yelling nt the fmnt gate. One of the young mm who lived i n e t came mnning back calhg for G. to come to the @nt. " W a kedii, mi en noo shiiw teplied G "shi e n kmn ya? Shi tuu big fi< kom ya". He stmted gnimbling. He was a littlè boy that should jump when th$ ladg callr/or him? No he won't go to thefiont. He workr hard and now is h b time to ~ l a r nint S. m u t urne h y o u n g man says and he

goes in fmrrt. Pria g r ~ c up labouriousiy fmm what she ii doing she muttm smething about people not

having rnnnnen - but it is unclear who she ic talking about. G. gets up and/ollows hm. 7hm b some yelling so Iget rïp and go to thefront. I t is a dûpute about rhr visiter's hurband and G.'s cousin. Much to the visitoh displeasure the two have becorne involved while living outside (New York) and have

decided to live together. nip vbiior wantr G. to tu& to his cousin. She won't corne in the yard till she gets

some scitisf~fctiun~ A nowd hm now gathmd in front and G., wa'y of anuthm scanda1 quickiy gathm eveybody up and hun+cs inside theyard to the houe (Fieldnotes II:38).

The house is the place of respectabilitv. good relations with family and neighbours. and

the site of religious work While at times it is üable to rock on its stilts with the hoilering of a

passionate husband and wife. it is nevertheless populariy conridered a haven h m the nickus

goings-on of the road. As Jayawardena noted (1963). when conflicts between IWO viUagers

reach a cenain level thev are taken to the house of one of the offending parties. But this is not

a venture to be undertaken lightly. To take a quarrel to a person's doorstep is to push it to the

next stage. Evervbody has a cutlass at home and, when pushed. they will wield it. although it is

rarely the case that anybody is senously injured.'

As 1 discuss in later chapters. the delimited spaces of house and yard interact in

interesthg ways with the emergent spaces in whkh interactional encounters take place. For

instance. calling out to passers-by on the road is an important means by which the boundaries

imposed by the yard kncing become penneable. Caregivers spend a great deal of Ume

instnicting chiidren to use 1anguage in this way (Chapter 8). Often. conversations wilI be

camied on aaoss the yard fencing. In ihis way. linguistic practices selve as bridges across

physical. economic and legal boundaries and. although the bounds of house and yard serve as

weil-understood demarcation lines establishing public from private, the ambiguous nature of

these h e s is strking1y illustrated by the practice of nrrsing and talking hard (discussed in

Chapter 6). Despite the dominant belief that men (are supposed to) t a k louder than women.

women are well-known for their abiüty to nccs and pumL Women sometimes use Loud, shrill

voices to engage an audience whidi is outside the immediate interactional field of the house. If

a man has been d d for days. or won't help around the house. he is in danger of receiving a

nüsing u p from his wife. This leu surrounding neighbours know not onlv of his behaviour. but

ako of his inabilitv to control his wife.

These spaces. then, are not simply functionai. Rather. they are imbued with a peat deal

of meaning and sentiment A married or otherwise gown man who is homeless. unless the

house was destroyed by some disaster. is likely to endure much tomenting from others. Often

people h e in houses which are not. suictly speaking, their own. They may be looking after it

5 Of the 20+ cutîass wielâing inadents that 1 witnessed during the course of the fieldperiod, only two ended with somebody being seriously injured (Le. cut). There is proverb relevant to the discussion insofar as arguments often involve family members: fmily mrlacs na cut deep.

59

for relatives who have gvne outride (Le. America &Canada) or they may be living on the charity

of a patron (sometimes ernployea wiU lend houses for particularly loyal workers). In such cases,

people frequently remark that the man plays big ("acts important, successful") but that it is not

his own house. In contrast to the world outside, in whidi the male worker is obliged to follow

the directives of his superïors both at work and within his own kin-group, in his house the adult

male is the self-professed 'boss". This is expressed in an authoritative way of inhabiting the

house - usually redined in the hammo& standing on the veranda. taking breakfast or drkiks.

The adult male is a picture of relaxation and comfor~ Others swirl around a m e r h g questions.

fetching refreshments, doing chores. Status and authority is thus naturalized in the indexid

meanings of movement and imrnobility.

Women seem to be less concemed with statu meanings attached to houses chan are

men. Women frequently note the aesthetic quaüties of different houses, the spaces of privacy it

affords. the modernity of the kitchen etc. Women also tend to occupy the house rather

differendy than do men. For a d e , the home is a place of work, not one of relaxation and

remove from the world of obligation to others. As a man relaxes in the hammock the woman of

the house usuailv keeps busy deaning, cooking and directing chiidren. She rnay sit, but it would

be highly unusual for her to relax whüe her husband was engaged in some task As Smith and

rayawardena ( l959:33 7) remark, the 'offiaal* structure of familial authority is expressed in

eating arrangemenu?

6 I t is important not tu overemphasize the completeness of maie control in the house and family. In some families men are absent from the house a geat deal of the time and this obviously changes things. Also one should note that women often play a pivotai role in constructing the power relations within which they may be subordinate. This is more than just compliaty. The active and constructive role that subordinates play in constmcting relations of power has frequently been over looked (cf. Goodwin and Goodwin 1990 on the collaborative construction of hierarchy in children's

Eating arrangements emphasize the authoritarian position of the males, and partidarky of the husband-father. As head of the household he is usualh/ served alone at the table whikt women and ChiIdren eat their food sitting on the floor, probably in the kitchen. Other adult males mav eat at the table. either with the head or after he has finished [...] it is a general rulé for the wornenfok to see that the men and children are senred before they eat themsehns.

Most importantly the house is not her own. A wife sïmply cannot make d a i m s to patnüneally-

inherited property except through her husband (and of course except in ber own pa*e cf.

Ashan). Even a sixyfive year old woman who has iived in the same house for dose to fifiy

years camot claim to own the propeq. As 1 have discussed above. major decisions conceming

the house. such as those conceming its upkeep, who should be aiiowed to üve there, the mord

standards the inhabitants must maintain while they lnre there etc.. are almost alwaw made, at

least "officidy. " bv the grown sons. Smith and Jayawardena ( 1 959~339) remark

When a male head dies.his wife should manage the household and propeay until such tirne as the eldest son can take over the headship of the househoid. obtainuig assistance in the meantirne hom her husband's male kinsmen ... Most of the imponant detisions are made bv the adult son who also acts as the f d y ' s male representative.

One of the responsibilities of the male-head is to regulate and police the movernent of people in

and through the house and yard. Male authority, localized in patriiocal group's landholding, is

thus diredv tied to the way in which women's everyday movements are restnaed. I t is thus

wvithin the house that m h e d women spend most of their time. Rickford's ( 19 79: 140)

description of the situation seems to be accurate for most rural communities:

One important difference between men and women is that the men enjoy more freedom to move beyond the house and the village. They can often be found outside the home or lirning the streets- whether as young boys off to the seawall to catch bu&. as teenagers on their way to a movie, or as adults congregating by a corner for a

0 . . gaff. By conuast. the female's place is considered to be in the home, and it is there rhat

argument and Krating 1994 on the role of one woman's active participation in the expression of power - from which shr is excludcd - in spatial arrangements.)

the women of the community can most nlpcally be found: deaning, caring for young children, preparing meals, washing dothes. and so on. They are seen l e s often on the sueets than men - usually when they are out shopping, or on their way to the home of a fnend or neighbour for a visit (Women never socialize in nun shops.)

I t shouid be noted that women are kept in the home not only by the pressure of household duties, but also by the protenive attitudes of their fathers and husbands. These attitudes are fairly well-enmnched in Guyana as a whole, but particuiarly so in East Indian communities. A suiking m p l e of this protectiveness of chawinism is the high proportion of women who had been takcn out of school at an early age because their parents feared that their education might lead to invoivement with members of the opposite s e x

The social relationships of kinship and domestic authority are thus embedded in the

organization and occupation of parti& spaces. As I show in later chapters. space is an

important resource which participants to an interaction draw on to express a range of social

relations and interactional roles.

2.6 % R o m

The road is popuiated both by villagers and passen-by. People wak up and down the

road from Berbice to Georgetown selhg everyrhrng h m vegetables to the latest fashions.

Women gather in smail groups dong the road caiiying bowls in which to collect the fish sold bv

men on bicvcles. Buying fish, as mundane as it may seem, is an important arena for status

cornpetition among neighbours. There is a locdy understood prestige s a l e according to which

fish are ranked. While individu& may have penonal Ues and dislüces, everybody is aware of

the prestige values associated with certain types of fish:'

7 Thanks to Ralph Premdas for helpful insight on the organization of prestige with regards to local fish markets.

Unscaie-Fish

Scaie-Fish

Table 2.3 CATECORIZA~ON OF SOME FISH COMMONLY EATEN

Unscale fish. such as kurm and catf;Fh. are beiieved to be bottom-feeders which scavenge dead

Salt Water

Snapper Banga Basha

anirnals and fish from the bottom of bays and harbors. Scale-f%h are usually found in deeper

water and in smaiier numbers. Similady, urcept at certain times of the year, haar is much less

plentifid than houri. The regular diaracter of market availability is superimposed with cultural

meaning. n i e cheap and low-status f i h are interpreted as and, in some ways. unfit for

human consumption. When women go to the road to buy fish, they regularly engage in status

cornpetition with each other and at the same time haggle with the seller. Such t r a i e n t arenas

give the road its flavour as a place where people are engaging in social acfvity where

reputations are won and lost. Groups of men, young ad& and children often gather by the

roadside to uuf'or discuss some business. Often a family or a couple of young women wiü be

seen waiking down the road, sheltered from the sun with an umbreila. dressed in their best

dothing. Wuau -yu a wu& out "What? - vou're taking a stroll!" somebody will cail a t them.

The road is very much a communal place where evexybody who one knows and is on good

terms with must be greeted. I t is a place to lime 'hang out" and enjoy the company of friends,

but i t is also a place which can be dangerous especially to young women:

Fresh Water

Hasas Talapia Houri

I Both R and G had kcpt a sntull stand by thc madfor smnr time. Being bmdim, dun was m m thon a lit& rompetition to the mdeavoux W h o WOU^^ get the saurïi needed tu nt& nhc chutney? Wlio wouL1 attruct m a of theymug m m who like to stop and o a f bcfm thcg go in for suppm? Dunng hanmting

tirne, the m m w m f o m d to wod Ltte and thcy tlrc stalLr to drc womar. G's wq% S and RS daughter M. began seiling by the nad and both attracted unurualb la'ge numbm ofnrttomen who mouid sir on Iogs or benches fading the mad by th& r e spec th stands. Lopalties w m stnmg to one m the other linring

standr. As tim pawed, peopk b e ~ n to tdk about smnc apecialllp mgular patmnr nt both stalls. Som thm wen rumoun that S. was umy ing on with ayoung man f r m n nei@buunng village. She suggestcd

that hm rival M. had staNd the rumour in order to tarnish hm nputatimi and stain h a nnmr. Soon

people w m s u p t i n g duct M. hmeqhad bem cuught in the h u e ofnpmrng man who lived a m s s the

mad fmm hm stand. The mrnours aploded one day in a public busing bg the rotid, the two w m e n

grabbing and purhing one another until M. i m o k intentened and wus treated to a smack in the face m hm daugfrtrr. nint night G., impind bg afuU bottk of white mm, went e the stand and bmke it

apnrt. People watrhedfrorn thnr verandm as G. yekd about Air wijèi i n t q and srnashed the

=Queen Bee" irrtofi~~wood. @wune, th& was tnkm as confinnation ofthe m m u r that S. had in fuct i- been canying on wàth that boy fint the neighbouratg vil+. n e n n t moming G. announccd that he

would not build back rhc stand - i t war too dangerou and the mad was not a phcefor womm. A f i

months later, M. i father and m o t b leunted ufhet poor rfanding in seconda y school and they stopped selling by the mad. Fumily mentbm complained about the irnpnrdmce of lrtting n young girl stand up

[(and sel1 thinps bv the rond - whnt did the wrmtr emect? (Fieldnotes 156)

The movement and activity of the road is considered dangerous in relation to the stabiiity and

structure of the house. Men ofken Say that the road makes women get wild and uncontroilable

and that the house is their proper place. And while women may not share this attitude

completely. they usually deride a girl who spends too much tïme on the road instead of in the

house doing her work. One girl was given the false name 'roadrunner" (after the song rather

than the cartoon) and this became a well-known index to her reputation as an unconrrollable.

wild and dtimately dangerous individual (see chapter 7). The road. dong with the nimshop, is

considered public space. and, while there are certain exceptions to this general rule. public

space in w a l Guyana dtimately belongs to men. This division of spaces is sometimes hotiy

contested but, in general. it is weU-established and &es with it the force of tradition.

Rum and the rumshop are pivotal in the culturai life of the villagem. As one enten. the

male voices immediately d r o w out the sounds of the road. Around a table. on which is sirting

one or wo large bottles of white mm or vodka, a jug of water, a phte of ice. some Coke and

perhaps a plate of mttm (/katas/ =snacks eaten whiie drinicing"), sit t h e to five men of various

ages. They look up and greet you - if they know p u or know something of you, then a s t o ~ is

likeiy to ensue - if they don't, then you will be treated to a friendiy greeting, and questions as to

where you corne from and who is your fq (unless. of course, you are not Indian). Everybody

experiences the effects of rum and the pull of the rumshop at one time or another, be it duectiy

or indirectly.

To sues its cultural importance is not to underestimate the degree to which alcoholism.

a Iegacy of a colonial regime in whkh ail productive activity was focussed on the manufacture

of cane. is a reeal and very destructive force in the community. The shop is. of course, the proto-

typical male domain - a place where male gynf is given a place of great importance. where

<allQng out some storv can Iead to violent encounten when a man reaches badc home. where

the talk man and the joke man reign supreme above ail othen. People gather here to hear their

own stories again and again.

Women are expected not to enter a rumshop. On occasion. however, a wife will fetch

her husband. Even at these Urnes she does not enter. Standing up on the road outside the shop

she may c d in to the man who is sitting dosest to the entrante. The husband will usually go

home quietly at this point (if a woman is ready to go and fetch her husband it is u s d y

because he has been there for the better part of a day), trodding dong the road behind his wife.

65

Rurn and rum drinkîng are deeply embedded in daily We. Rurn is used. in a more pure

form caiied highvine, as a remedy for colds. I t is &O used in reiigious work for the spint known

as the Bounday-master. Most importantly, perhaps. it is used in the rituals of male peer p u p s

performed dady in the shops dong the public road and in bottomhouses of private homes.

2.8 CONCLUSION

In this chapter, I have introduced the major fonns of social organization and have

illusuated the way in whidi these social structures are embedded and realized in locally

recognized spatial divisions. 1 have also uied to give an account of the historiai developments

leading up to the present situation. In the foUowing chaptem. 1 look dosely at the way in which

spatialized linguistic practices both reproduce and contest the dominant models of kinship.

gender and domestic authority.

Because anthropologists and iinguists oftcn studv communities other than the ones in

which they h e , enterhg and assuming a role in the host community becomes an important

part of anthropologicai and sotiolinguistic methodology. A great deal has been wrïtten about

this process from both methodological and theoretical perspectives (Labov 1972, Rickford

1979.1997, Clifford 1983, Scheper-Hughes 1992: 1-30). In the fol3owing 1 want to give an

account of the way in which I was integrated into the cornrnunity 1 studied and how the nature

of that integration affected the present work as weii as its author.

One never knows all the members of a community - be it a village. a town or a aty - in

the same way or to the same extent. People ueate webs of relations in which some individuais

are central and othen are relatively more peripheral. My fieldwork experience has led me to

believe that this web has something of a concentric character, whereby the relations at the

center tend to suucture the relations which are more peripheral. In the foilowing, 1 di these

centrai people 'broken." Brokers are people who open up possibilïties for the expansion of a

web or network, but in douig so. exert a good deal of idluence on the way in which these new

relations are manifested. As rehtions become more locaüled, and centered around a pdcular

group of people, the influence of the broker becomes more specific. However, when one is first

entering a countxy or nation-state, the influence of the broker is very general and global.

In my case, entry into rural Guyana was faàlitated by a woman whose name had been

given to me by another linguist ShaIini had worked with various linguists studying the

67

Ianguage of the area that I was interested in - East Coast Demerara. Because she was raised in

the area. and was also university-educated, she made an ideal field assistant for linguîsts

working on grammatical structures. She \vas able to provide them with naturalistic recordings

made in her home community. which is a s m d setdement about 3-4 miles in h m the public

road. It is s m d and isolated and the aeole variesr is rather more homogenous than that found

in more denselv populated areas. The members of the cornmunity are semi-peasant farmers

who own small tracts of land and divide their tirne between working their own nce land and

working that of the big fanners'. 1 soon found that although S h M was both knowledgeable

and helpful, her ability to assist me in sociohguistic work was in part Limited by her

membenhip in the small backdam community. To begin with, she was unable to make

recordings and do ethnographie research in the main village community by the road. Because

she was alreadv known in this community, entering for the purpose of research was next to

impossible. Known as teacher-&val. Shalini. like just about eveybody else from neighboring

vUages. was the object of some suspicion and derision to the people on the road who believed

thar she \vas biggitty (/bigitiV ''full of oneself). This reputation was in part due to the fact that

she had received a coilege education and lived in town. and in part due to the fan that she

belonged to an e-xtended family who had left the roadside vülage, gone to Arnerica and had

subsequently severed ail ties with their feliow "counuymen." I t was arranged that I should.

with mv partner and daughter. iive in the fairly grand house that this famiiy had left behind.

After five davs in the village and a number of walks up and down the public road. 1 was called

off bv a group of men taking a break from work They called me off saying ee b a i , p u a wair bai

urpotugii? "Hev - are vou white or Portuguese?" The exchange was short and, at the cime, 1 had

68

trouble understanding the importance of the question. Later, 1 came to understand that such

distinctions were vital to the way in whidi people in the village undeatood others with whom

thev were not famiiiar. That same day. I was called off by another group of men liming (/lairn/

' to hang out") in front of their familv vard. Thev called me over and quickiy found out my . .

name, mv reason for coming to the s m d village of about 600 - which I gave as wanting to leam

aeolese - and where 1 was staying. I was made an honored guest in the eypicaf Guyanese

manner - beer and mm was bought on aedit, a dinner was prepared and taken. After dinner. I

was invited to accompany my hast Raja to a wake which was taking place in neighboring

village. 1 went and met a number of other men and women in the context of a ntual known as

dead- work or th i~een day .

Raja. his immediate nuclear famiiy and memben of his extended patrilocal unit. became

rnv most important contacts. AU became brokers - opening up possibüîties for research with

other people with whom they were associated. Raja. and his brothers Nancv and Papso. also

took it upon themselves to teach me Creolere and the varieiw of the language with which 1 am

familiar is the one that they felt was most appropriate to informal. c a s d relations between

men of roughlv equal status (cf chapten 4 and 5).

When I mentioned that 1 wanted to find four children to record, friends and relatives

were contacted. I was introduced and times for recording were a m g e d . The initial seledon of

people who would be appropriate was thus not mv decision. When 1 made recordings in other

people's houses, Pria, Raja's wife. would almost always accompany me. This greatiy facilitated

the ease with which 1 was able to incorporate myseif into the ongoing domestic Me of several

different households. Pria and I also spent many hours talking about children, caregiving and

69

language learning. When 1 mentioned 1 was interested in some rituai or event that tmk place

in the vülage, the appropriate 'older headw was located. A steady stream of teachers and Local

experts on language, rice-growing, house-building, and village history were invited by my hosts

to corne and ùistruct me. For this guidance. 1 am extremely grateful. My point is. however, that

the assistance of Pria (Raja's wife) , Raja. Dada and Nancy was integral to the completion of

this work and their perspectives on Me and language in the viuage are M y embedded (with

littie chance of being dislodged) in the fouowing chapters. Their selection of informanu, their

influence on my way of thinking, and their consistent guidance in the field work process. has

not onlv imparted knowledge but also a partidar bîas. AU are strongly committed to the

village in which they iive. Uniike others, with the possible exception of Dada. they do not strive

(or. in some cases, believe i t possible) to leave the village. They do, of course. look to

opportunities beyond the village, but for the most part are concerned to improve their lot bv

staying put. Other villagers are different in this way. Some show much less cornmitment to the

focussed networks of the village in which Dada and Nancy are centrai. Others are less involved

than Raja and Dada in the local svuggle for power which operates through the transmission

and displav of esotenc religious knowledge and the making of appointments in local vülage

councils. Unlike Pria, many women are demure, soft-spoken and (at times) subse~ent - they

suive to be "proper" at al1 times. Many wives, also unlü<e Pria, have comfortable relations with

their husband's family. SUU other women, such as Shanka, cultivate j o b and 1ocaUy

charismatic public selves engaging in behavior more typical of "reputationn-seeking men (cf.

Wilson 1973). The peculiarities of each person's life have thus shaped the present work in

innumerable ways.

70

My integrauon into the patrilocal group known as Bo10 (see chapter 2) and the nudear

family headed by Raja and Pria was extremely rapid. An arrangement was made for me to eat

tvith the fw and for tape recorduigs to be made in their house and another in the same vard.

1 began to spend a great deai of time in their Company. 1 was @en a book nnme ("name

considered to be one's right name. usuaUy derived from Hindin) and a f i name ("name uxd

for informal. everyday contexts- usuailv indexid or iconic" cf. chapter 7) by Dada as a sign of

community and family membership. Membership in the patrilocal p u p was n tudy confexred

on me through ruki (/raaku/ "a xitual in which bracelets are exchanged*), a day when women tie

bracelets on men they take for brothers. 1 was accorded the status of a younger brother in the

group. 1 was younger rhan ail the 9 brothers and 3 sisters but was accorded a special position on

account of my education and economic standing. I t was agreed that the house in which 1 was

living was both too expensive and too far away so an arrangement was made for me to move

into a new house close by.

The house 1 moved into was owned by a group of sistea and their mother coilectivelv

known as 'Pakar dem" (see chapter 2). These people had a shifüng set of alliances with the

Bo10 who the? generaliy acknowledged as *goodn people despite a tendencv towards theft and

p l ~ i n g big. Ashan and Shana. the printipal Pakar, provided exceptionally important

commentary on village events and persons which often confiicted and differed from that which

1 received from the Bolo. They &O introduced me to a wider range of people than 1 had been

in contact with before. Most important about the move fiom one house to another was the fact

that it effective. cut me off from one group of contacts (with the village elite who managed the

&t house) and made my assoaation and integration in the other group much more complete.

71

When 1 moved a heated conflict ensued with Indi, the woman who was in charge of the house.

She believed that 1 was breaking a contract with her and the altercation furthered the

impression that 1 did not belong in the elite group. (This group is of course &O divided into

factions and the disagreement with Indi did effectively open up alliances to sorne other opposed

members of the eiite group).

The expulsion from the elite &de in the village forced me to rethink the project design

in which 1 had proposed to investigate dass merences in language socializaiion strategies. 1

had already realized that the midcile and elite dass groups in the vülage would probably not be

willing to participate in the study. 1 found that they believed that the study was a usefd and

interesting one but would not acknowledge that they, dong with the working class people, were

possible objects of study. Thev infomed me that t h did not speak the Creolese and that they - .

behaved just like 'ciMüzedn people anwhere when taiking to childxen! Also, the non-laboring

population in the village have a very different attitude towards their houses. The house is not

open to visiting in the same wav and people are not usually called in off the road as they are bv

the working people. In the end 1 deuded to focus comparison on the difference between the

peasant farmen of the backdam communitv and the wage-laborers of the road.

AE 1 became more local and more a part of a pauilocal group, 1 became a more common

target for suspiaon, derision, and gossip. Like local people generaily, my actions and behavion

were subject both to public scrutinv and evaluation. niere were those who believed 1 had

exceeded myself, that my intentions were Iess than honorable. or that 1 was g d t y of ye-pars.

One group was particularly put out by my integration into the Bolo. This was the group of 0 . .

young men of roughiy my own age (the oldest was about two years younger than me but

72

generally they were much doser in age than were those who belonged to the g o u p into whidi 1

was adopted). 1 had been adopted into the age-grade diredy above them and they were

expected to pav respect to me (which entailed not only various forms of deference but also

being ready to run errands at my request). Part of my instruction in vülage Life was in giving

directives to lower status others. At first, 1 tried to maintain an uneasy balance by which 1

accepted membership in both age-graded sets - 1 drank. joked and c w e d with both sets of men.

However, eventually confiict arose when members of the younger group were told to Leave by

those in the older group, or to behave themselves in my presence.

Eventuallv. the younger group found a way of indicating their displeasure with the

situation. While 1 was awav on a trip to visit members of another f ' y in Berbice. thcy broke

into mv house and stole a number of items. When 1 returned and discovered what had

happened and who had done it. 1 was left in a quandary. Should I pursue the matten with the

authorities and N k furrher dienation from the younger set as well as hom the patrilocal unit as

a whole? Or. should 1 leave it and risk loss of face that could be damaging to my ability to

conduct research, perhaps even jeopardUing my own safeqr? In the end 1 made more of an

emotional response to the situation than a calculating one. Feeling betrayed by my own group

and Mends. 1 charged the young men and even came to a public standoff and minor scuffle

with one of them. The incident changed the course of my fieldwork

From that point, my relationship with the unit known as Bolo became strained. Though

some memben of the poup attempted to distance themselves from the event, downplav its

importance or even, to some degree. side with me over their own family, the theft introduced a

kind of contradiction that couid not be resolved. There is a proverb. famlii stik a ben bo ii na brook

73

'The familv stick can bend but it can't break" which mv friends who were not Bo10 thought

was panicdarly relevant in this case. Any concession to me by members of the Bolo logidy

entailed an admission of their own 'guüt" or, more properlv, shamefulness (on guilt and shame

see chapter 7). At the same t h e . untii some local version of justice was handed down (social

sanctions for those responsible) 1 was left without, in local tenns, satisfctïun. More to the point,

1 was. according to a number of hiends, made to h k likr wan as. But the Bo10 codd not

publiclv acknowtedge the incident without public humiliation. The charge that "they" (that is

members of the patrilocal unit - members under their authority) robbed someone who liwed nice

with thm. who iived like a bmther with them is quite serious. The incident was vezy embarrassing

to the senior members of the group and they did theY best to c q on üke it never happened.

The most immediate resdt of the incident was that 1 f o n d mvself in a social institution

known as not-tnlking with several village members induding the young men who had broken

into mv house and a few others who had pubiiciy sympathized with them. Not-taiking involves

more than a conscious refusa1 to engage in verbal interaction. i t is more properly characterized

as a total avoidance suategy. Two people who are not-talking tatitlv agree upon certain divisions

and borden within the village. Each person stays within die space aiiocated to them. AUiances

are quicklv made. and netwarks contracted. so as to d o w Lfe to carry on as nomal with the

one exception being that each person completely exdudes the other from their daily activities.

As one can imagine. such an arrangement îs dificult to maintain in a village of six hundred. It is

even harder to maintain within an extended family. Inevitably, conflicts arise.

While the conflict iimited the scope of my research somewhat, I think it made it much

more reaiistic for just this reason. Vülagers are not free to go anywhere they iike and they

74

themselves 1iM lives iimited in scope by vinue of social relations of the sort 1 have described. I t

was the conflict with the young men that reaily drwe home the importance of space and its

occupation for me (cf. chapter 6-8). I t was a rather ciifficuit Iesson about the way social power

was inscribed in local geographies. I found myself sticking much doser to my ailies. their

homes, my lit* temtones. and in this way 1 began to live a He much more like everyone else

than I had previously. My group of friends. outside my adopted patrilocal group. thus became

much more important as companions, instnictors, infonnants and as broken.

Let me make a few finai comments about the d e of brokers in shaping anthropo10gicaI

fieldwork. The first point is that, unless they are employed by the researcher. brokers genedy

have the upper hand in the relationship. I found it was very difficult to persuade my brokers to

do anything that thev did not alreadv want to do. This meant that recordings were not

scheduled despite my intention to conduct a longitudinal study. Brokers and other informants

essentiallv set the schedules and the field researcher must attempt to foIlow it. Furthemore.

because alliances between people are never stable, my abiiiv to visit and record the people with

whom 1 had fkst made arrangements for recording were constrained by the m e n t srate of

their relations with the relevant broker.

3.1 RACE / GENDER / CLASS - SOCIAL LOCATIONS AND FIELDWORK

One's social location is also a process. Constructed in real-time discourse. it nevertheless

draws upon both individuai history and ideologidy normative ideas about who people are or

who the? can legitimately claim to be. Social location is thus at one and the same time both

fixing and shifting. Guyana is a post-colonial nation at the bottom of the economic order both

intemationailv and regionally. Rural Indo-Guyanese villagers. at the bottom of the national

hierarchy (with the possible exception of Amerindian peoples) . are well aware of their

international position even if they don't have the tools necessary for an analysis of commodity

feùshism and globaüzation that would give some explanation of their predicament Every day,

villagers sit down to watch n i e Young and the Resths tiuough which they are exposed to

"American culture" and 'American concerns." even if it is a fairly reified representation of such

phenornena. The important point is that viiiagers recognize 77ze Young and the Ratlas and

similar American exports as cultural products rather than as representations of daily Me. nie

Young and the Resths thus gives viiiagen an idea of the irnaginary possibilities of American Me

(partidarly for Leisure, recreation and privacy) but does not serve as their main reference point

for "what üfe is iike in the U.S." (the often exaggerated reports by retumed immigrants seem to

figure mudi more centrally here). The representations on American television (and other forms

of media) fit weii with the weil established and long-standing cultural tradition of racial

stereotvping. Racial stereowes have been mu& discussed in the literature on Guyana (see

Bartels 1977, Drummond 1980, Williams 199 1 ). As Williams argues, an individuai's actions

and behaviors are often interpreted on the basis of a set of enduring raaal stereowes which

introduce a ranking of the various contributions made by particular ratiaiized groups to the

Guyanese national identiq. These stereotpes are sites of contestation between rival goups.

Such contestation manifests itself both at the local and interpersonal level, in arguments and

narratives about race as well as at the national level in public and policy and development

agendas. One scheme of stereo-pes places whites a t the top of this hierarchy. assoaating them

with the elite institutions of education, governent and international administration. In the

same scheme. Afro-Guyanese are assigned a relatively high role on the basis of their role as

middle dass civil senrants and bureaucrats, etc. Indians are renowned for theiir business sense

and their hard work As Williams and Bartels have documented, the cultural struggle between

those who place themselves in the Indian group and those who piace themsehres in Afio-

GuyanesdBlack group takes place on a number of different levels. My fieldwork in Guyana

was. of couse, suuctured in part by such racial stereotypes. In the community where I worked.

whites were generally renowned for the erudition. their appreàation of high culture and their

abiiities in the international marketplace. Many people expressed, at various times. the belief

that Indians were much 'closer" to whites than they were to the hGuyanese (a number of

aiteria were use for making such judgements - induding hair? attitudes towards sex, rnarriage.

ethics, etc.). At the same tirne. whites are often considered physicdy inferior to Indians and

lacking in various foms of commonsense. When I first d e d in Guyana. 1 was bombarded

with questions about how whites rtayed (i.e. 'what they are Like." 'what are their essentiai

qualities"). Despite my many long winded expianations for why culture and behavior could not

be so easilv correlated with race or skin color, many people persisted in the belief that ail my

actions were the necessarv expression of mv whiteness. During my s t a y in the vülage. a srnail

bov to whom I had become personally attached died as the result of severe burns on his Legs

and lower torso. His death was very troubling to me. as were the reactions of the other villagen.

Radier than engage with the situation at an emotional level, 1 became very involved with the

work that follows a death in Guyana. M e r spending the day digging a grave and canying Sand,

stone and cernent for the tomb, a discussion took place b e ~ e e n a number of villagers. They .

believed that I was some kind of racial anomaly, that this behavior - to work for somebody else

77

- was raciallv . anrpicai. - At some point, in order to preserve their s te reove of what white people

are me. many people felt it necessasr to recategorize me as ndman ('a penon of mixed African

and European ancesty") or as Portugziese (/potogii/ Portuguese were brought ta Guyana as

indentured laboren and people consider them distinct fiom the European elite). Many people

to whom 1 was introduced as a whiteman remarked that they had never heard a whitebuy ta&

raw. For the most part. as long as I spoke Creolese, 1 went unnoticed with people assuming I

belonged to one of the local heages of Portuguese people. Whiie this absorption into a p u p

of people who are considered 'Guyanesen in a way 'whites" never couid be made life somewhat

easier (and was somewhat satisfymg on a personal level). the reaüzation that, despite

anomalies. racial categorïzations are firmiy embedded in people's everyday understandings of

the world was distressing. I t seems to me that very few people actuaiiy rethought racial

categorization because of my presence. They generally believed that whites were unwilling to

iisten. uninterested in anvthing but themselves and highly nomative in the values they place

on speaking and behaving properly. My anomaious behavior seems to have had Little impact

even on the people with whom I was most dosely associated. I do not want to suggest that

people in the village are incapable of rethinking racial stereotvpes nor do 1 want to re* their

rxpenence of the world of racial differentiation and oppression. My point is a simple one - it is

easier to recategorize an individual than it is to rethink an entire set of categories especiallv

when there may be social institutions whkh hold them in chedc.'

' h d certainly 1 do not want to deny the very powerfui social and historicd forces which have gone into shaping this racial ideology.

78

As 1 have indicated in the forgoing discussion. my control over the fieldwork process was

far from complete. My relations with people in the MUage were stmctured in part through a

coionial discourse about the nature of race and raaal Merence. These relations were of course

also suuctured in pa.rt by assumptions (my own induded) about the relevance of my gender. 1

was qui* integrated into a maie group despite my interest in chiidren and the women who

care for them. The assumption that 1 would be most cornfortable taking with reiatively high

s ta tu men at first made conduaing research on language soaaüzation rather difficult. 1 soon

impressed upon people that mv work involved obsenhg drüdren. Some people deaded that

this was in fact a ruse. and that my real reason for spending time in the houses was to get close

to mamied women. Aithough rumon quiddy spread about me being rhere with (/de wid/

"sleeping with") vanous women. 1 managed to maintain my contacts with the mothers with

whom I worked. This was in large part due to Pria's abiiity to persuade hiends and her

wilh,gness to champion mv cause with various women in the village.

In the present study 1 have uied to take native speaker intuitions and explanations

about linguistic practice seriouslv. My training in linguistics and anthropology has taught me CO

alwavs look for the deeper meanings in what people Say - the way what they Say reveals

something unsaid. Yet, for a variety of reasons, I have tried to relax this hemeneutics of

suspicion (Dreyfus and Rabinow 1982) and take the people in the village who were willing to

spend niany hours talklig about their language. history and culture more as collaborators and

CO-authon than as informanu. At various points in the following, 1 finci traditional scholarly

explanations lacking and instead turn to explanations oRered by native-speakers and villagers. I

am guided to do this by the realization that rationalizations for behavior very often affect the

way in which that same behavior is smctured (Shrerstein 1979). At the vely least, 1 wouid

argw that a complete understanding of practice is unlikely to be attained if the researcher is

unable or unwilhg to attend to the phenomenal and subjective world within which actors h e .

In so far as many villagers were willing to share their insights with me they have become co-

authors. 1 suspect that even those few people who r e b e d to ta& to me have, in some way.

influenced the writing of this work

3.3 THE FIELD PERIOD AND THE SAMPLE

The field period for this research was spread out over three caiendar years. There were a

number of reasons for organizing the field researdi in this way. Initidy. 1 was concemed that 1

should be able to return to Canada a t some point to Wit with my daughter (then just w o .an

old). After some time spent in the village. 1 reaiized there were other good reasons for breaking

up the field component. First, it was nearly impossible to write while 1 was in the field. During

daylight houn and much of the evening, various soaal engagements daim one's attention.

Maintainhg social relations becomes a full-time occupation. Bv the time 1 retired to mv house.

it was inevitablv dark and infested with mosquitoes. Nthough I kept a journal, there was Little

opportunitv for sustained descriptive writing. Making transcripts in the field also proved

dificult. primady because of the many demands made on rny tirne bv neighbors. fiends and

adopted kin. 1 t was therefore necessary to leave the village in order to assess the progress of the

research projea. Breakhg up the field penod ako aliowed the study to incorporate a more

extensive time frame. This. in fact. turned out to be quite important since a iiumber of social

relations went through extensive change during the field period. Furthemore. the longer

temporal frame dowed for a Wer account of the sochihtion and dwelopmental processes

documented in chapters 7 and 8. After an initial field p e n d in 1994, during which time I

decided upon a research area and g h e d primarv exposure to the language variety, 1 organized

the research into two penods over 1995- 1996 (cf. Table 3.1)

MA.Y Field period begins

NOVEMBER Mando tum 3 years Nankaran dies

MAY Jm Return to field live in Raja's house

JANUARY Preliminaiy analysis of field materiais (Toronto)

AUGUST M' Kumar cornes to iive in the village

SEPTEMBER 'Theft" stoxy cthis chapterz

S E ~ M B E R retuni to Toronto.

OC~OBER Pria's Story cchapter 6 z

Table 3.1 SCHEDULE OF FIELDWORK

Most of the naturalistic recordings were made in the fint period while the second was spent

engaged in i n t e ~ e w s and elititation sessions with adult villagers. In addition. I conducted rnost

of the ethnographie work on the organization of space in the second penod. As 1 have

rnentioned. the composition of the sarnple of diildren (whose interactive contacts form the

focus of chapters 7&8) was not random. Rather i t was structured by the nature of my

integration into the cornmunity. Generally the children corne from two patrilocal groups of

roughiy equal social standing.

-- - - - -

Patrilocal Group I age

iii Kumar Mando

Alicia Süon Lex Taiga

Babv ~ w & r Kavita

8-24 mos 30-46 mos

0- 15 mos 4-5 years 7-8

10-1 1 years

4 5 yean 7-8 years

9-10 vears

- -

Patrilocal Gmup 2

age

Leezi 7-8vears Deevii 4-5Gears Linda 0- 1 O mos

Shamir Shabir

Table 3.2 CHILDREN TKE SAMPLE

During the fieldperiod, 1 recorded over eighty hours of n a t d y occuning interaction and just

under ten hours of interviews and pedormances of stories. The vast majority of the naturaiistic

recordings were made surreptitiously. 1 had. at the beginning of the fieldperiod asked a nwnber

of people if I could make recordings with their children and various caregivers. In most cases

this involved securing pexmission fiom both the mother and the father and anv other addts

who were resident. Some people refused but most people agreed. 1 explained that thev would

not know when I was recording and when 1 was not because the tape recorder would not be

openiv displayed. As such. 1 think it is safe to sav that in most cases participants were rither

unaware that they were being taped or did not care. In the few cases where participants did

become aware. for one reason or another, a marked shift in style is quite obvious. The

naturalistic recordings include examples of a number of different genres oniy a s m d section of

which are represented in this work Included are toasts on a wedding annivenary, folk songs,

0 . . arguments between men. between women and beween children, oral histories, pedagogical

discourse, caregiver-child interaction, heated debate, casual conversation (i.e. PUB, recitation of

82

religious doctrine, the proceedings of a puja and words spoken in the course of possession by a

Dutchrnan i spirit.

Most of the i n t e ~ e w s were conducted fairly late in the course of fieldwork and 1 was

able to use my knowledge of metacornmunication derived from partitipation in some of the

above events to direct and structure the interview in ways that were locaily acceptable. Such

metacommunicative n o m s indude the pacing of question and answer. the system for overlap

and tum-taking, and the use of parricular lexical items which can be used to cue culturai

knowledge. In this regard, I found that it was particuiarly important to frame the interview

situation as pleasant gyaf rather than as question-and-answer. I n t e ~ e w s are, for some peopie in

the village. svongly assoaated with the interrogations carried out on a rather darmingly

fkequent basis bv the local police/militia. For others, the thought of being asked questions about

language and Me in the village conjured up unhappy mernories of the Guyanese coioniai and

post-colonial school wstem. I t was thus important for me to stress that 1 only wanted to gyf

and that the questions 1 was asking na ga rang an mit ansa 'don't have a wong and a right

answer." I used different metcommunicative means to achieve these ends depending, in part,

on the gender of the interviewee. If a man was being interviewed, I would usudy ask thern to

corne to Raja's house and we would have a few drùiki, eat Lunch and then graduaily move into

the i n t e ~ e w . With women, the situation \vas somewhat different. In some cases, 1 was invited

to the woman's house to conduct the interview. This happened more often in cases where the

woman \vas not mamed. On severai occasions, 1 arrived to meet not one but several local

women busiiy engaged in t a k In each case, these turned out to be exceptionaily good

interviews. With no men present (other than mvself), the women would cook, dance. sing, joke

83

and drink M e r some thne ac this. 1 would be dowed to ask my questions and received long

and detailed answers. Not aU the interviews 1 conducted went so smoothiy. however. In one

case, the i n t e ~ e w e e spent the entire time looking out toward the road to see if her husband

was coming. In most of the i n t e ~ e w s with women. a fairiy relaxed atmosphere was jointly

created through the exchange of greetings and a few jokes. Ofien, one or ~o hiends were aiso

present and this helped to aeate a more relaxed atmosphere. Of course, it would be wrong to

Say that participants undentood these situations as somethuig other than interviews. However.

interviews are complex genres and have. embedded within them. conversations. asides, and

casud gyaf.

1 also spent good deal of my t h e 'Interviewingn children, most often, a t times when

thev were left in mv care by rheir parents (1 too became a caregiver in the village). These

communicative situations were markedlv different in stnicture than those conducted with

adults. Usuaiiv thev range over a great many genres and topics, the partiapation is fluid and

the overail stvle of discourse îs verv Lively. Most of these children had little experience with

forma1 i n t e ~ e w s and thus tended to treat the communicative situation simply as playful

interaction with a caregiver. ï h e ease with which these "interviewsn were conducted was also

facilitated by the la& of ambiguity in terms of role structure and partiapation. As Briggs

( 1986) notes, the i n t e ~ e w presupposes the interactional roles of an interviewer who asks

questions and an i n t e ~ e w e e who answers them (even if we d o w for some fiexibility) .

Problems arises when such interactional roles conflict with local role expectations which divide

up interactional work Meren*. Thus. in the Mexicano community he studied, Briggs found

rhat younger people (himself induded) were expected not to ask repeated and direct questions

84

of their elders (except in particular circumstances). ln the interviews I conducted with children.

however, expectations regarding questioning and those concerning local status relations

coincided neatlv with the interactional d e s of interviewer and inte~ewee. Because 1 was of

higher status in the system of age-grading, 1 was aiiowed and expected to ask the questions

necessarv for conducting the interview.

The social relations within whidi I was engaged by members of the village and that 1

have sketched out in this chapter are quite typid and the reader will h d that. in the foilowing

chapters. the same social roles reappear in slightly different guises and with different occupants.

The marner in which I was inugrated was thus not only important for achieving ethnographic

research goals, it in fact became part of the research itself. My participation in joint activity was

thus not just a means to an end but a major part of the process through whidi ethnographic.

interactional and linguistic data was "coliectedn.

CHAPTER FOUR

LANGUAGE VARIETY, LANGUAGE USE AND LANGUAGE IDEOLOGY

4.0 INTRODUC~ON

In this chapter. I give an account of GC as a pragmatic system. As 1 have suggested in

the introduaory chapter. this means integrating analyses of forma1 variation with other aspects

of tak which are more comrnonly understood as pragmatic (deixis, affect and stance etc.). I

attempt to achieve this integrarion through an extended discussion of native speaker awareness.

In the first section. then, in order to relate the present work to previous research on linguistic

variation in Guyana, 1 look at the nature of sotiolinguistic variation in the village and.

specincaliy. at the way native speakers understand this variation as reflecting (and,

occasiondv. as constituting) social structure. My partidar interest in this discussion is the

wav in which formal variation is related to the way in which vülagers understand the2

comrnunity - both its boundaries and the way it is intemaily differentiated. Following the

description and aneses of variation. 1 consider in more detail the interactional negotiation of

context in talk. As 1 have discwed in the inuoduction, one way in whidi panitipants frame

interaction as occurring in an egalitarian community of a certain type is through the linguistic

consmal of space and the participants' relations to it. I t is therefore worth spending some t h e

describing the structural devices which GC uses for relating speech event participants to the

context of speaking. With regard to this last aspect of contextuaiization. in the later part of this

chapter 1 discuss the use of formal variation in GC for marking speaker's orientation to the

message in terms of affect or stance and thus revisit the initial discussion of variation and code-

86

choice from a rather different perspective. Firtally. to dose the discussion of native spe&r

attitudes and awareness, 1 consider the way in which metalinpuistic items (verbs of speaking)

categorize and characterize speech-interaction dong a number of dimensions.

4.1 VARIATXON IN THE CREOLE CONTKNUUM

To give a general sense of both the ciifferences bemen basilea and a a d e a and the

abiüties of speakers to move between them, consider the foiiowing two te= which were

produced by the same speaker as samples for a matched guise test'

Jack: vu noo Çu kuk dok ko*? & p u know how to cook du& curry?

W: ves bai mü noo fu kuk a dok korii ves, I know how to cook d u k curry.

Jack mi wan fu noo tel mü nou I want to h o w - tell me willyou?

W: vu \van mi fii tel vu hou h kuk a dok korii? Yuri tuant me to tell you how to cmk duck curry?

lack: vea -ves

W: arait p W - vu gu get yu dok an yu kot a m op an yu wash am Wii::n

' M ~ first attempt to elicït two earnples of speech for a matchcd guise test with a different speaker had failed because the woman was unable to approximate rither the basilect or acrolect consistcntlv. The speaker who produced these te- was orîginally from a countxy area but had moved to

0 . . Georgetown whrn she was eighteen. A hairdresser. she enjoyed returning to the c o u n t ~ ~ for a day to gyaf with friends. Matched guise and subjective reaction tests almost invariably use men if they use only one voice. 1 wanted to use a wornan because of my interest in beliefs about women and language and socialization.

Jack:

w:

aln'ght child - you will get a duck andyuu eut it up nnyou wash it chan

an den yu goo lait op a faia an yu put an a kahaarii fu hat- le ii hat and thot you ligiht up a j h andyou put on a kahari to heat - kt it get hot

an wen a kahaarii hat VU chroo in a ail an bai taim a a3 a hat and when the kaharii i$ ~ O I ~ O U - J O U t h m in the oil and by die tinte the oil ir gefting hot

pan a faia vu pu grain op yu masala - yu noo wa masala? on afireyou will grind upyour mnrala - p u know whnt masula is?

mi noo masala a wa yes I know what masala is yrr

wel vu gu grain op a masaia wid a gyarlik an a onyan an a pepaa - yu well~orc will @d up the masala with rhc gttrlic and the onion and tlic peppet -you

noo abc-diiz piiopl laik plentü hat pepa? soo wen yu don grain op a mas& h o w we likr a lot o f hot pepper? So w h y u hawjnirhed grinding up the masalu

yu go0 bonjee am in a kahaarii. an wen don bonjee a masah gud den you will bunjay it in the kahan. And whmyor< havefinished bunjaying the masais then

vu gu teek a dok miit an diroo am in an yu gu bonjee a dok miit a n yu gu you will takr the duck meat und thmw it in andyou will bunjay the duck meat andyou will

lef am Li l'am a bom out lil bit biika a-a rank waata de dee so p u gu le am bom out Li lmve it n littlc and Ict it bum o f because there ir tank watm in it so you will kt it bum out a little

bit an wen am don bom out den yu gu teek hat waata an chroo am op tap a da le a müt bit and when its finished then you will take hot water and thmw it on top of that so that the meat

bai::I an wen a miit don bail an a dok saf iinof da p u kyan chuu a m boik and when the ment is finirhcd boiling and the duck is soft enou@ so that you can chew it

den vu teek am dong pan a stoov an yu üt am wid a rais an yu enjai am then vozi take it o f f the stove and you eat it with rice and eniov it.

Jack ... kukdokkorii? ... cook duck curry?

W: wel p u wud laik tuu noo hou tuu kuk dok korii? wel yu get di miit - yu get a dok an welï you would like to know how to cook du& curry? Wellgu get the meut - you gtt a du&

w u klün it an y u kot it op Wiin an yu wOsh it - okee? wen p u finish wûshing ciii you clean it ~ndyor i eut it up c k n andyou wush it - O.K? whm youfinirh warhing the

miit an soo forth yu get dii ingriidüents tuu kuk di korü wich iz di masala d i garlüc di meat and so forth you get the ingndimtr tu wok the mrry which is the masala the gariic the

onyon di pepor along wid di korii poodar - kee? den yu lait op di stoov an yu put on di onion the pepper along with the n<ny powder - O.K? then pou Ziglit the stove andyou put on the

di p o t weer yu gona kuk di korii in to hot an yuu chroo a lit1 ûil an put the pot in which you a n going to wok thr currp in to heat up arïdgou tkmw in a fit& oïl and put

it fo hat kee? den yu grain op di masala aloon g wid di ada inpnidüents az ai tOld yuu it to get hot - O.K.? thmyou @d up the masah along with the o t k ingredin ts as I toldyou

the garlik an di onyon an di pepaa an soo fort wen yu finish grainin di masala an di Oil the garlic and the onion and the pepper and so forth wm you jniih gngnnding the masala and the oil

iz hat vu chroo it in tu di 0i1 an p u kuk it a Lid yu ondastaand? p u bonjee it soo wen ir hot& thmw it in to the oil andyou cook it a little -you understund?pou bunjay it so when

vuu finish bonjeeing it vuu chroo in di dok miit in tu di p o t an den yu kuk it wel wid vozijînish bzrnjaying ityou throw in the duck m a t in to the pot and then you cook it well with

dat vu wil get a lit1 w0t.a soo vu iiiv di wOta to sima dong - okee? soo wen its finisht that&rr will get a lit& wnter s o h lave rhc watm to rinimer d o m - O. K. ? so whm its finished

sirnmaring den vu ad hO:t wOtaa fo di miit tu bOil an wen di mïit iz tinisht bOiling *yu s irnntoing then you add hot water for the meat to boil and when the meat b finirhed boilingyou

gona chek i t an den yu-yu ad di di styuu on top ihof stvuu an den wen y u finish will check it and they&-you add the rtew on top - mough r& and then whenyou finish

bOüing y u teek it af an dat iz di wee tuu kuk di dok korii boilngyou tnke it off and that i s the wuy to cook the du& cuny

PHONOLOGY

or>a

O>a

C#>O

CC#>C#

tl>l

dh>d th>t

n>ng

Oi>ai

LUCICAL

kaharii-pot

predicate serialization

MORPHOSYTACnC

pronouns ( 1 st singdar)

pronoms (possesive)

pronouns (3rd neuter sing.)

cornpiemen tizer

copula

compktive

irrealidfuture

articles

DISCOURSE

address

discourse market

a eg. faia

a eg-hat

O eg. biika

C# eg. saf, grain

1 eg.1il

d eg. "d

ng eg. dong

ai eg.ail,enjai

kaharii

wash-kiiin

mi

yu

am

Fu

O cg. wen a kaharii hat

don eg-wen amiitdon bail

gu

a (generalized)

piknii

-

a-or

a-O

C

CC#

tl

d t-th

*g

Oi

Pot

ai

it

fuhu

iz

eg. sima, pepaa-pepor

eg. hat-hot, d s h

eg. dat

eg, tOld, onciastaand

eg. lid

eg. dat cg. forth-fort

eg. dong

eg. Oïl

eg. di 0 i 1 iz hat

is finished eg. iz finisht bOiling

w d g o n a

di

ookee, yu ondastand

90

As shown in table 4.1. the texts above ihstrate many of the salient differences between more

basilectal and more a a o l e d varieties in Guyana. Phonological Merences, even in this s m d

sarnple, tend to be more often expressed as statistical tendenaes than as categorial differences.

Thus, although we find non-aeole. aaolectal segments in text 2 (eg. or, O. Oi, th. etc.), in

most cases. these altemate with the aeole variant (e.g. or-a, O-a. th-t). In some cases, the

speaker did not approximate the aaoleaal phonology for text 2 (e-g. she realized dh as d in

dut). In terms of Lexical selection, the speaker distinguished a basilectal variant kahani 'pot used

for making curry" from what she considered the more English and metropolitan pot. In f a n

kaharii wi used in the mesolectal and aaolectal registers of East Indians in Georgetown. The

speaker here seems to be equating the basliîect with the expression of an East Indian racial

identitv (and. bv default, the acroiect thus becomes e t h n i d y 'unmarked"). ln terms of lexical

differentiation, there is the hint of a very productive basilectai process of predicate serialization

in the expression wash am kliin. The two texts are most dearly differentiated at the level of

morphosvntax wirh the speaker employing a range of devices to distinguish basilect and

acrolect. Included here are variation in the pronouns (mi-ai. am-it), complementizers

( hi- tu). and copula (O-iz) . Completives are marked with the basilectal tense-aspect marker

don in te.= 1 but with adverbialftnsht in text 2. Further. in tems of tense-modaliq+aspect

marking, we find that, in text 1. ali unrealked clauses are marked with gr or goo. whereas. in

text 2. the speaker distinguishes between conditionai clauses marked with wud (eg. p u wud

laik tu noo ...) from future wealized marked with wil (eg. yu wii get a lit1 wOta) orgona (eg. yu

gona chek it). FinaUy, the speaker marks ail noun phrases with a generalized a in text 1 but

uses di in text 2. This pattern seems to indicate that the speaker has not fully acquired the

basilectal ?stem for articles (see the appendot). Finallv, in terms of discourse features. the

speaker uses the address form piknii in text 1. Like the equation of basilectal forms with East

Indians (evidenced in the use of Miauri ody for text 1). this usage seems iduenced by the

ideology whkh associates aeole varieties with the oldest generation in the community (see

below). This said. drawing attention to the way in which social relations between participants

(speaker and hearer) are determined by the system of age-graded and gendered solidarity with

the use of kin t a s as weii as age and sex sensitive tems (püonü. bai) is very much a part of

the discoune-pragmatic system of basilectd creole. The absence of obvious basilectal discourse

tags in te.= 1 (checkhg the receipt of information and its comprehension. such as.yu onastan?.

yu noo?.yu noo hou?) is rather striking when compared with the numerous uses of ooke 'OK" and

-mi ondastand in text 2. Whether this reflects. on the one hand. a s te reoge of aeole discoune

as more monologic and creole speakers as less responsive to the audience. or. on the other. a

lack of pragmatic competence on the part of the speaker, is unclear.

The nvo texts are quite usefd in ülustrating the contrasts between basilect and acrolect

at a number of levels. In fact, when test subjects in the matched guise listened to the two

samples thev were invariablv impressed a t the speakers' (sic) abilities. None of the respondenrs

doubted that the two texts were in fact produced by different speakers. The texts also illustrate

the wavs in which speaker competence is limited and the way in which the speaker

compensates for such limits. In the activity of making a recording of "real" or "raw" (see below)

Creolese. the speaker drew on certain stereosrpes both in places where her competence was not

complete (eg. the articles) and in places where contextualization was required (eg. in her use of

lexical selection to ethnically mark one of the texts, and in address to indicate the social

92

relations between participants). The texts thus. while ülustrating salient contrasts. go further to

reveal local ideologies about the social characteristics of the speakers who typicaily use these

varieties and the contexts in which thev occur.

4.2 NA^ PERSPECTIVES ON THE LANGUAGE

Structural differences of the kind illustrated above do not go completely unnoticed by

native speakers. 1 say completely because native speaker awareness, Lüce that of the analyst, is

alwavs parriai although certain individuais do show an acute understanding of the nature of

variation in the continuum. The question of what native speakers take into account when

assessing language variation and what they find sigdicant or noteworthy in this respect is a

fascinahg area of study and one which touches on important recent developments in Linguistic

anthropology (the generai concem for reflexivity, metapragmatics and language ideology). Of

course. native speaker evaiuations have been apart of sociolinguistic study at least since Labov

( 1966) and Lambert ( 1967). Matched guise tests measured people's evaluations of different

speech varieties in a community showing that memben cypidy and routinely correlated

linguistic variation with different social and persona1 characteristics of speakers. In the

foliowing I report on research 1 conducted into the question of native speaker attitudes towards

the varieties Guyanese Creole.

4.2.1 LOCAL TERMS FOR LINGUISTïC VARIETES

As 1 argue in the following chapter. language. in the rurai Indo-Guyanese villages with

which 1 am familiar. has a particular importance as both a symbolic field and as an instrument

through which battles for symbolic capital are waged. After more than thirry years of empirid

sociolinguistic research in a range of speech communities, there can be linle doubt that

of whatever Ianguage ("so-called" monolinguals and multilinguaS alilce). So without wishing to

dispute the universal nature of sotiai and stylistic variation in language. 1 think it is necessq

to stress the particularly salient role that language plays in so-caiied 'post-colonial" situations,

panicularly those in whidi creole Ianguages are spoken by the majority of the population. The

combination of soào-political and linguistic factors in Guyana. for ewmple, has @en N e to a

situation in which language is vexy much a primary index of soaal position. Dale Bisnauth, a

member of the late Jagan's PPP govemment, wrote in the PPP (People's Progressive Party)

paper. the Sunday Mimx

1 went to a smaii dinner party not so long ago, by invitation. A properly sari-clad woman ranted and raved about those who had recently come into the h e light so to speak. of Guyanese life by the acadent of politics. The recency of their arriva1 was apparendv an offense as was the perception that 'thef' were not as educated as she \vas with her IWO degrees and ail. or as possessed of the social graces as she was, la de da and so on[ ...] She berntrd some of ur for our incapacïty to speak the Quem i Engliih properiv and for our penchant to mis-match noum and vmbs; and othm ofzisfir being so obviousiy Philistine adtrcralh[. .JI know that few ofw, given our slave and indentund back~ound, are redk o f the rnannerlmanor born, when you come d o m tu it, whatewr we ntay pntmd and ivlratever may be our affectation ... But the new snobbq revolver amund education in Engfish. ability to speak that langziuge properlg, some capacity to appreciate culture and to support its development, conspicuous consumption, the ability to acquire things and possession of relatives abroad. The abiiity to daim some connection with persons in perceived high places, a h counts ( 1 995:35, emphasis added).

The gender of the participants in the reconstmcted scene is probably not a coincidence. Suffice

for now to say that language, specificaliy the clifference between English and Creolese. is on just

about evervbodv's mïnd at some time or another.

I t is not surprising, then. to 6nd that native speakers have a fairly elaborate set of tems

94

for talking about language and Linguistic differentiation. At fimt sight the native metaiinguistic

vocabulary for structural linguistic vanation seems to be somewhat "overlexified" (H&day

1976). However, on closer examination, it appears that most of these terms do not refer to

stnictural characteristics alone. If we were to make a first approximation of the metaünguistic

vocabulary for code varieties it would incorporate the foiiowing terms arranged from Creole

(left) to English (right) @es.

Creolese raw talk broad talk bra wlin ' talk patwa Coolie talk

Table 4.2 METALINGUIST~C VOCABULARY FOR CODE VARIETTES

On the basis of the correlations 1 have made bennreen native tenns and those of the creolistics

literature (e-g. basileml-mesolect-acrolect) one might be led to a number of erroneous

conclusions regarding the way native speakers dassify different kinb of linguistic production.

While the set of ternis in the top row (Creolere, b r u h down language. deep Engfish) do seem to

take as theu primaxy aiteria for ciassification structural features, the others generally do not.

The other terms, in characterizhg and dassifying different h d s of linguistic production.

associate structural features of language varieties with various functions of speaking and salient

characteristics of the speaker, situation and context (thus, it is rather difficult for native

speakers to anmrer a question like "when is it appropnate to use raw talkn because the speech

m i e - is itseif part of the context). Most obvious is the relation of structural characteristics

95

and language hinctions. For example. as 1 show later in this chapter. the acrolect is assoaated

with politeness while the basilect is iinked to with cussing and "unrulyn (often anti-

authoritarian) behaviour despite the fact that the situation is much more complicated than

such a one-to-one mapping impiies.

4.2.2 THE SPEECH ECONOMY

English and aeole varieties do not exhaut the resources available within this speech

community and despite the fact that this is not the focus of the present work some mention

should be made of other varieties in use in the speech community and the way the total

repertoire (Gurnperz 1968) is natively understood as constituting a speech economy (Hyrnes

Deep Englirh MU-up Talk CreoIRre/Ra w Talk

English-Based Coolie Talk ESOTERIC VAR~ET~ES Gibberish

Non-EnglishBased Hindi Arabic

Table 4.3 THE SPEECH ECONOMY

I have divided the esoteric l anpges into EngIish and non-English based varieties. However

such a classification is in some ways hadequate. So-cailed Coolie-talk is really Crealese (ie. CC)

with a greater contribution from Hindi and/or Arabic in the lexicon and some Hindi

phonological featiues such retroflex alveolar stops. Ody a few people in the village speak a

variety which othen would cal1 Coolie talk. In generd it is a remembered variety but not one

that speakers command. Gibbmsh (/giberii/) is, according to local accounts, the Afncan version

96

of Coolie talk. According to the few speakers that I know. gibberish was a secret language used by

the reai 'congo niggersw in Guyana. The most profitient speaker I was able to h d was a

woman in her eighties of rnixed African. Indonesian and Dutch origin who had strong ties to

Surinam. Her father, she told me. was an independent landowner of mixed Indonesian and

Dutch heritage who had found himself in Guyana and once there took an African mistress (her

mother). I t was the mother who spoke the variety Cousin Dina cailed gzbbhh. In fact. the

present-day variety of gibberish which 1 recorded is CC with syllabic andior consonantal

insertions. Although it sounds suange to speakers of GC it is not a gcammatically or

phonologicailv distinct system. iohn Rickford (P.C.) tells me that it is quite commonly used as a

chïldren's play language in Guyana. What is interesting about both these varieties is the way

thev take an essentidv edinically unmarked variety (Le. CC) and make it ethnically charged

through the addition of features that are interpreted as essentidy "Indian" or "Afikann.

Hindi ( r edy Bhojpuri - cf Gambhir 198 1. 1983) and Arabic are languages restricted

prïmarilv to religious expenence in Guvana. While most Hindus. including a number of local

pandits. have only supeficial knowledge of Hindi, educated Moslems often have fairiy

developed levels of competence in Arabic and some are able to converse quite freeiy in i t This

difference no doubt has to do with the religious traditions as they are manifested in Guvana.

Devoted Hinduism does not depend on a sophisticated knowledge of Hindi. Rather devotion in

Hinduism is expressed through dailv prayer. annual ritual, fairlv regular church attendance and

an involvement in the institution of the church as weU as knowledge of Hindu myrhology.

. . Knowledge of Hinduism is regularly transmitted in English or Creolese. Hindi generauy being

used onlv for prayers and bajan (religious music). Moslem religion. on the other hand. is

97

stronglv embedded in the learning of Arabic and ckildren attend language dasses as part of

their reiigious instruction. AU prayer, and some conversational dialogue. in the mosque is in

Arabie. Manv ma@. in contrast to the pandits. show a f&fV deep knowledge of Arabic.

Despite the presence of a number of esoteric languages, the most salient dimension of

linguistic differentiation, the one that local people take into account when evaluating a speaker

on the grounds of their speech. is the Creole-to-English one. I discuss the degree to whidi

people are aware of su& variation and in what capaâty in the following sections.

4.3 NATIVE AWAREmSS OF VARIATION

4.3.1 REGIONAL VARIATION

ï h e variation that native speakers rem on most often, and seem most sure of, is

that berneen inhabitants of different regions in Guyana. The viUage studied is located at the

border between East Coast Demerara (ECD) and Berbice (it is the last major settlement before

Berbice). There are a number of commonalities between the communicy studied and the

neighbouring Indo-Guyanese villages in West Berbice üke Bushlot. Rice fanning has been

particularlv important in both areas. Inhabitants of the village, when questioned about

language use. however, tend to exaggerate differences between themselves and their Berbician

neighbours. When 1 initially mentioned to people that 1 was interested in the creole language 1

was more often than not told that 1 was in the wrong place and that I needed to go to Berbice

because people around "here" don't speak like that anymore. The Berbecian' reputation for

speaking the "real" creole is so widely accepted it is interesting to speculate as to its origin.

Even Bickeaon ( 1973a) seems to accept this myth. In fact, of course, people in Berbice show

98

just the same kinds of variation as is found in ECD and in other d areas (some speakers are

primariiy acrolectai while othen are primariiy basilectd). The idea that the basilectal speakers

are somehow more basilectal than people from other areas does not seem to be m e based on

mv own observations. But mmy inhabitants of the village where 1 conducted fieldwork argue

that Berbitian aeole preserves a number of features which aeoiists wodd c d archaic. Induded

here are 'sing-song" intonation patterns. the use o f ' rather thanfu as complementker. and the

expression bring out mii iit 'give me my food." This Last supposed characteristic of Berbician

aeole is so often Qted as evidence for the more raw nature of the variety that 1 am led to

wonder if there was some othenvise forgotten story or joke behind it. Despite native speaker

intuitions of this kind, the growth of major cenues like Rose Hd, New Ams terdarn and

Rosignol (not to mention smaller. but nonetheles significant, villages and Settlements like

Black Bush Polder and Bushlot, Blairmont) has no doubt led to inaeasing linguistic

diversification in the area and the adoption of more acrolectal varieties in a sigruficant portion

of the population. StiIl. vîilagers' association of raw talk with Berbice is worth investigating

m e r . I t is interesting to note, for instance. that language in t h case seems to index more

general ideas about social relations. Villages believe that Berbicians are not only more creole-

speaking but also that thev are more traditional. backward and superstitious. In this regard it is

instructive to note that the stock 'berbicianism", bnng out mi iit. plavs out stereotyped and

traditional gender roles (a man demanding his food from a wife who has spent the aftemoon

cooking). This, thrn, is not ody a statement about language variety but also about an

interactive conte* and about social relations.

People are also aware of. and occasionaiiy comment on, something loosely labelied tong-

taak " town talkn. This variety is probably not indigenous to Georgetown (although cf. Walter

Edwards 1 9 84, 1 994). Rathet, it is the fairiy upper meso l ed vaxiety spoken by the large

nuxnbers of people who live ourside Georgetown but relv on the capitai to make a living. This

heterogenous group includes aii kinds of seliers and hucksters and, m a t prominently, the

minibus drivers and conductors. A reguiar column in Guyana Chronide, appropriately titled

Right h m in G/T, is written in a rather stylized version. The foliowing sample is representative:

. . . Whatever the cause though. I could feel it coming on. ..a stxong attack of nostalgia. The his' signs does show up when you gyaffing old-tune story with aidren what you ain't reason foh a long tune, an yu find that ndf a you sentences starting with: 'You could remember when..?" The other dav ('Remember ... The Radian B." Guyana Chronicle, November 14- 1995).

Or,

And you remember. wha' you looking at as a fast conquest today. might tum-out foh bu p u death-warrant in tights. After all. half the fun 'a them action is looking back with satisfaction, right? So wha' sense it mek if you ain't gon deh-bout foh reminisce? niink, me-brother. We gon catch-up ('Death-warrant wearing tightc." Guyana Clrroniclti, Oct 2 1, 1995).

The varietv, which features near categorical use of non-basilectal markers such as fint peson

subject ai 'In. habitua1 marker doz (lirnited to non-past contutrs) , progressive -hg, quotative and

adnominal laik "iiken as weli as a great many coiioquial expressions (ai de man "I'm hanging in

there") is similar in many ways to what has been labelled AAVE both in its contexts of use and

a nwnber of its saiient structural features (for instance, copula patterns cf. Bickerton 1 9 73 a,

Wïnford 1992). As a recognizable style of speaking or register, it holds a good deal of nvay

among the punger male set in the village who are concemed to present a more experienced

and less d self image (cf. Eckert and McConnell-Ginet 1995 on the parallel behaviour of

" burnoutn boys). The variety is incorporated into a larger set of non-verbal signs and overaii

body-hexis (cf Bourdieu 1 99 1 ). When combined with a certain mode of dress (loose firting

monocolor pant-SM combination, expensive brand name tradr shoes or 'boots". and the

ubiquitous kangol furgora hat) and a certain way of holding one's self (in certain places

parricularly the road but also in chinese food restaurants. rumshops etc) it indexes a local

notion of bad boy/man.2 This person is someone who shouid not be troubled. someone who is

dangerous and likely to do anything if pushed. The body-hais seerns derived from what is

typically understood as a maie Afro-Guyanese mode of comportmentr3 the musculature is

relaxed but ready, the han& and arms always exposed (never in the pockets or under each

other). the head and the hat are often tilted down hiding the eyes. legs, if sittingD are

outsuetched, if standing, may be toosely crossed. The cigarette is held loosely in the mouth

between hard and somewhat dramatic inhalations. The overall effea is a self-projected

immobility or calm which is quite at odds with the dominant stereotypes of East Indian

rnascuünity4 Such modes of cornportment, induding assoaated ways of speakingD are reacüly

' Williams ( 199 1: 107) discusses a similar kind of person whom she labels "virago":

In many respects. the term virago coinades with the US Alro-Amencan designation "bad-ass nigger." In this sense, a virago is a person who is quick to anger and equally qui& to settle disagreements violently. Unlike bullies, viragoes do not usudly instigate fights, but like them, when confronted or "crossed" by others, they are unlikely to seek a compromise resdution to the disagreement.

I do not want to reib racial difference here. It is more proper to say that sorne young Indo- Cuyanese men mode1 themselves pady after a particdar interpretation of urban masculinity which is very much influenced by Afro-Guyanese culture. Aithough cd turd practices are not distinctiy African or Indian, people's interpretations often are. See D m o n d ( 1980) on the nondiscreteness of cultural practice in Guyana and Bartels ( 1977) and Williams ( 199 1 ) on the enduring importance of racialty polarized stereotypes. It should also be noted that the young East Indian men in t o m (those who work with Afro-Guyanese and mixed people) take a deadedly anti-racist position. Unlike the young men in the rural villages. urban East Indians who belong to the minibus aews are quite vocal about the injustices of bodt East Indian and African brands of racism.

' Williams ( 199 1) provides a fascinating discussion of supposedly "ethnic" differences in attitudes towards activity in gencrai and work in particular. 'Strreotypes of an ethnic segment's relative emphasis on work sum up these contentions: East Indians work to Iive. Afrïcans work when al1 eIse fails,

101

recognized by people in the village. The older generation has a remaikably ambiguous attitude.

While labelling these people (ofien their own kin) bad bai 'bad boy". ti+man 'thieF. w&h

' unreliable". hisrrutee "lazy" etc.. they simultaneously remark on their stylings " fashion sense"

and are impressed with their a b i l i ~ to m o n 'sweet talk" with women. Men are perhaps more

k l v to tend toward the latter evaluation than are women although this is not unambiguously

the case. The main point is that the acrolect and upper-mesolect is rather heaviiy stratified or

heteroglossic in Bakhtin's ( 1 98 1. 1984) sense. The indexical vahe of forsnal variants in these

upper registers is underspecified and it is oniy through their combination with other kinds of

signs (bodily cornportment, dress etc.) that they become meanïngfuI. These younger men and

women are not targeting the varieties spoken in schools despite the fact that statisticd evidence

might indicate that they are relatively more aaolectal than others h the village. Neither is this

a 'standard" in the sense of a legitimized variety. Rather, the prestige it commands is

essentially coven and its use distinguishes the younger speaker fiom both people considered

rural and bachard (the figures of patrilocal authoritv among them) and from those IabeUed

biggz-ty " arrogantn, English-du& ( i.e. the figures of ins titutional and bureauaatic authoriy).

Any community in Guyana shows some linguistic differentiation according to socio-

economic class (Rickford 1979). From the campus of UG (University of Guyana). to the

market of Georgetown, to the rural villages discussed here. the contrast between relatively

.O acrolectal or mesolectal and relatively basi led varieties is used to signal social differentiation. . *.

and Portuguese and Chinese work only long and hard enough to accumulate suffitient capital to let it work for them" ( 199 157).

102

Such associations between Linguistic varie? and social position do not escape native speaker

awareness. Rather people regularlv comment on the relative appropriateness of a parti&

Linguistic variety and a paaicuiar social identity. Such associations, I have found. are us-

alreadv racialized. Partidar wavs of speaking are not oniy associated with working or d

people but with rurai people of a certain racial group. This amounts to saying that dass, in this

community is not conceptuaiized in a pureiy abstract way. Of course in discussing dass

differentiation within their community people less often raciaüze what they Say because the

ethnic ident i~ fonns the interpretive backdrop of dass discourse. In such cases, as I argue

below. the intersection of gender and dass cornes to the fore. However 1 want to first discuss

the wav in which linguistic varieties are assoaated with particular ratialized dass identities.

The joke (storv) which follows. given to me by one informant but &O well known, provides a

rather suiking illustration of the way linguistic differentiation is linked to race, dass and

gender. I have marked stereo~ped basiilectal items with underlining, aaolectal ones with

double underiining.

BA~w, Hrs WIFE AND THE DOCTOR. Narrated by Gobin Singh

1. Wan dee Babuu tuk eeiing waif tuu di dakta. (w)en Babuu gu nou.. One day Babuu took his<iling wife tu the doctor. When Babuu went

2. dakta aks Babuu Babuu wat iz vor prablim the doctor asked Babuu uBaburr. what is the probh?"

3. see dakta mi na noo mi waif sik Babuu replied uDoctor, I don? know my wfe is si&"

4. see ookee weet am o n a ten tuu hor jos nou n i e cioctor said 'O. K., wait, I am going to attend to hm in a minute. "

S .

6 .

7.

8.

9.

IO.

LI.

12.

13.

14.

1 S.

wen shü taim rüch, dii dakta kaal shii in dakta Song shii, kom ou-di tuu a When it war hm htrn. the domr called hm in, the doctor lirtened to hw insides, and the two 4

dem kom out bak an ii tel Babuu ii see Babuu yor waif want sekshuuwol intokooq thon came out, and he told Babuu uBubuu.your wij& wantr serual interroum.

soo Babuu see Dakta ,@am wa ü waan nou mi bring a m hü fu vuu *am Su Babbuu suid "Doctor, give h a what she wunts. I brougt't hm hen so thntyou couid give hm

mi na noo wa ii wan vuu eiivam nou soo di dakta see what she needed I don 't h o w what she wantr - you give h a what she needs. " So thc doctm suid

Babuu ar vuu shOr wu want mii tuu piv vOr waif sekshvuuwol intorkoors? ii see U8abuu, a n y o u sunyou want me to gipcyour wii senral intniwum?" He raid

3e.r Doctor what @ou) think I'm tellingyou - give hm what she wantr. I will pnp the rnoney?"

eniihou Babuu mek soo ii sit bak an di Dakta kaai di waif an kyer shü in di ntum Anyuy, Babuu mu& a gestun like this and the doctor called kir wiji and t w k hm in the rwm

an ü star tu av seks wid shii soo afta Dakta teek Som taim and he started to have s a with her. So after the doctor took some time,

Babuu (rapping on chair) rap a dii door Dakta a wa y a duu so laang? Babuu rapped on the door and said uDoctor, what is takinggou so long?"

. . Dakta see Babuu weet a minit moor airn -m vor waif sekshvuuwol intokoors The doctor replied '23abuu. wait a minute please I'm giuingyour wije s a a l intercoum. "

ii see oo mi unk vu bina duu di nekst tùig Bnbuu said "00 I thought you w m doing the other thing. "

The story, in its use of code switchhg to convey changes of participation and role, gives a

striking illustration of the way in which language serves as an indirect index of soaal position

and ethnic identitv. 1 sav indirect because these fairly abstract qualities (ethnitity and class) are

constituted by the direct indexid meanings which the variety makes to immediately obvious

personal qualities (such as a lack of intelligence, or ignorance generally).

There is a flipside to this picture, however (cf. Ridcford and Traugott 1 985, Rickford

1 985). Many fokheroes (as opposed to fokstereotypes like Babuu) are crafted in aeole

discourse. Balgobin is the consummate mole-uing, anti-authoritarian folkhero and his stories

are weil known in the communitv.

BALGOBLN GETS THE POINT. Narrated by Deonarine (Papso) Singh

p u no0 lak - p u noo BaalGoobin? brait boy.. You h o w Jack - You h w B a l ' i n ? He? a bn'ght bop

wel Baalgoobin de pan di bla-di tiidül de pan di- mis de pan de blakboord- Weil Balgwbin was tlrere at thc - the teacher was thprr - Mirs w a t h m a t dre bkrckboard

dis a di mis ii get an wan shart sqoort yu noo? wen shü sit dong pan di cheer nia is the miss, she had on a short skiri - you know? Whm she sat d o m on the chair

Baalgoobin de a front ii a bou dong ü hed an ii a -- ii see ii gat a ( ) big (baks) Balgobin wus up a t thefmnt. He would bend d o m and he aa - he said she hm a big box

- a gon gii yu a joook wan wan jos nou -teek it o p I'Il giveyou a couple ofjokes in a minute. T& thut up (an aside tu hic son)

p u noo - an shi star rait pan di blakboord You know? And she started to write on the blackboard.

" v u plos tyuu W a l foor mainos wan you get duii aad t p u tu dat yu get faiv" 'Two plus two epuh four minus oneyou get three udd two to thatyou getfive. "

vu noo an sh-shi tek a tur rong di klaas hi sii huu a pee otenshon you know an nh-she took a tour around the c h s to sec who was paying attmtion

Baalgoobin de a di outsai ii na de insai di ting Bnlgobin i mind was not inside. He 's not al1 there - inride.

shii see Baalgoobin yu get di paint She snid '23algobin. did you get the point?"

se noo mosii paa:n in He said "No, it most have gone in already. ."

12. den mis piip outsai a tuu dankii in a shed n<en Miss peeped outride w h m she saw t h w m two donkqs in the shed

13. di pain gaan in are& Ihc point already went in

14. den mis see oo gaad Baaigobin da iz wa yu duu jos m u ? ï h e Miss said "Oh G d , B a l p h , that is whatyu w m doingjurt now?"

15. se mis ai woz ios luukïn at di vaint Balgubin said uMiss, I wm juct laoking nt the point."

Here Balgobin plays on the word /pa.int/ 'pointn. In line 10. the teacher asks if Balgobin 'gets

the point". For the teacher the point means the 'reason" or the 'idea behind something". Her

usage of this term is both acro led and. within that realn. suongiy associated with the

bureaucracy of education (a domain in whidi examples have singular justifications. in which

problems lead to wrong or right answers etc.). Balgobin achieves miscommunication here bv

taking point in its metaphorical and Creole sense as refening to male genitalia. He thus uses the

question from the teacher and the reference to the point to redirect attention towards cwo

copularing donkevs outside the dassroorn. The fact that this is intentional miscommunication

is clearlv indicated in the 1 s t line where we find Balgobin using the a c r o l e d register (ai woz)

to excuse his misapplication of basiiectal d e s for lexico-semantic interpretation. Many other

stories feature the wise yet imeverent, aeole-speaking Guyanese hero getting the better of an

authority figure. Such stones point to a set of beliefs about Language and other indices of class

position which are at odds with the stereotype of Babuu presented in the first story. If Babuu's

language variety indexes ignorance and a la& of intelligence. that of Balgobin indexes a crafcy,

untrained but highly respected intelligence and resourcefulness. I t is important to note that

Balgobin is bidialectally competent. In iine 1 L .. Balgobin uses the Long aa which. as Rickford

( 198 7b) has noted. is a stereotype of d East Indian speech or broad ta&. In iine 15.

Baigobin displays cornpetence in the aaolectal end of the continuum by his use of subject

pronominal ai (see chapter 5) and idiected copula woz. As I argue in the following chapter, it is

this ab* for bidialeaaüsm that is higNy valued in the community. Basilectal usages. in the

context of known b id ia ied abilities, are &en as expressions. not of ignorance, but of wit,

resourcefulness. cunning etc.. Such contradictory attitudes towards language varieties have

akeady been documented for the Caribbean (Patrick and McElhùuiy 1994) and Guyana in

The connenion becween rthnitiol, dass and language which is part of the popular

Guyanese imagination also became part of Bickerton's ( 1973a) hypothesis regarding

decreolization. Des pite the fact that he provides no statisticai (or implicational) evidence to

show a strong correspondence between race and linguistic variation. he pronounces:

I t is perhaps a regrettable fact that the most important covariable of Linguistic behaviour in Guyana is not age, occupation, education, income, or even urbanization. but ethnic identi~4.e.. whether the speaker belongs to the Afncan or the East Indian communitv..As a result of soaoeconomic forces operating in Guyana over the last centurv O&, the bulk of the African population has moved well away from the basile&, while much of the 1 n d i i population has not.

While I cannot at this time provide empirical evidence to support my position, I think it is a

rnistake to make race the primary correlate of Linguistic differentiation in Guyana. If

agricdtural workers are predominantiy of East Indian descent in Guyana. it is Likely that

Creolese (espetiaily the more basilectal varieties) is associated with this group. However. this

situation has corne about as the result of a complex interaction between race. 'ethnic culture."

m . - colonial ethnic stereosrping and socioeconomic development (cf. espeady ~ i l l i ams ' 1 99 1).

Such a sociolinguistic scenano cannot be undeatood as the direct result of socio-economic

forces working on ethnic or racial comm~nities.~

Whatever the cornplexities invoived in the hirtorical path leading up to the present

situation, there can be h i e doubt that native speakers do make associations between Linguistic

differentiation and raâalized ckss identity. ï h e conneaion between linguistic differentiation

and gendered dass identities is somewhat more complicated. In the foliowing chapter 1 discuss

statistical differentiation in relation to gender, for the moment 1 want to consider only the

ideology comecting ideas about dass, gender with ideas about hguistic differentiation. 1 have

found the recent work of Sheriy Ortner partidariy helpful in thinking through this problem.

In thinking through the wavs in which dass was systernaticaUy mystified in Anglo American

and Iewish American culture. she writes (1991:171):

The panicuiar pattern I want to focus on here is the displacement of class frictions into the discourse and practice of gender and s e d relations. The basic point. which emerged for me more or less accidentally as 1 read a set of cornmunitv studies with initi*. no particular agenda. is t h : gender relations for both rniddlé- class and working-dass Americans (1 have only glanced at &tes to this point) cany an enormous burden of quite antagonistic dass meaning. To tum the point around, dass discourse is submerged within, and spoken through, sexual discourse, taking %exn here in the double English sense of pertaining to both gender and the erotic. And while the general point of displacement holds for both middle-class and working-dass discourse, it works differently in each case.

The kemel that 1 want to take out of Ortner's complex argument is that class is often thought

through (or disguised within) discourses of gender and sexuali. With regards to linguistic

variation in Guyana it is interesting to note that whüe structurai variation is most often

associated with age. what could be broadly labelled 'fimaional variation" is associated with

This b h g ~ w to die central debate in Guyanese social history - the ways in which dass. ethnicity. and race have been CO-constructed in the colonial and post-colonial periods. The interested reader is directed to the work of Jayawardena (1980). D m o n d (1980.198 1 ) . Bartels (1977). Williams ( 199 I ) as well as Despres ( 1967). S M t h ( 1995) and Premdas ( 1995).

108

gender, sexuality and class. So. while Creoiese is assoaated strongly in native speaker ideology

with the oldest generation in the commmiqr. nrrsing and tnlking hard is associated widi a dass of

d, working class people, and particularly with women (cf. chapter 6). At the same tirne. and

this is the important point. these same kinds of speech practices (for example nrrsing) are

associated with stmcturai variants of the language. How does this relate to gender and dass?

My point is that when thuiking about linguistic differentiation according to dass and gender

native speakers make strong associations between these sacial categones and particuiar 'wavs of

speaking" which are taken to exemplify personal characteristics. In the case of dass and

ethnicitv. a co~ection is made between the social category and personal characteristics (such

as ignorance) which are expressed in interaction (such as when Babuu is duped by the Doaor).

In the case of dass and gender. an assoaation is made between a certain goup of socially

situated women. parricular interactive routines (e.g. ncssing) and the penonal characteristics

(lowlv class ongins, ignorance, a lack of intelligence) that they supposedly index The

assoâation between gender, dass. race, on the one hand, and stxuctural/lectal variation. on the

other. onlv cornes at the next indexical level (see figure 4.1).

ethnicity/dass persona1 characteristics structural variation ( ie. ignorance)

(getting duped) INTERACTIONAL ROUTINES (cussing>

gender/class personal characteris tics structural variation (ie. ignorance)

b

Figure 4.1 ASSOCIA~ONS BETWEEN SOCIAL CATEGORES, PERSONAL CHARACTERISTICS AND STRUCTURAL VARLANTS.

109

Another important connection for many speakers also exists between ways of speaking and the

gendered use of space (cf- Chapter 6.7.8). When asked. 'who speaks more aeole men or

women?". many native speakers repiîed that women do because they are removed from a

varie. of public spaces, including school, at a voung age. One informant made this set of

logical relations quite expliat: creole cornes from the "home environment" and a woman, in the

traditional pattern. is resuicted to the home. Therefore women tend to use more aeole. This

notion of an as so~ t ion between lingistic varieties and certain contexts of speaking (home.

school etc) is what I c d the subordinate ideology of stylistic differentiation (see section 2.3.3.).

I t is important to note that such ideas about styüstic variation have effects on the way people

think about social differentiation too. The relations between dass, gender and language are

discwed in more detail in chapters (6,7).

1 attempted to elicit ideas about the relations beween language, dass and gender during

a matched guise test. Conducted with 1 7 speakers, the test revealed suong assotia tions

between linguistic varieties and stereotyped characteristics of the speakers who might produce

The test did not, however, reveal particular1y sharp deavages within the speech

communin/ either dong dass or gender Iines. The results are rather different from those

discussed bv Ridcford ( 1979. 1985) for the plantation communily he studied, and I would

W e s t that such ciifferences are the result of the particular histoxy of socio-econornic relations

in this comrnunig-. The foilowing questions were asked of each respondent (Listed by the

number they appear under in Figures 4.2 and 4.3):

6The test was conducted with 9 women. 8 men. 9 people who belong to the dass of labourïng people and 8 who do not. The recording used is reproduced at the beginning of this chapter.

Huu moor big? (Older) Who is older?

Wich wan a d m ïuayuu go a h f p u bin waan barn mmiii? Which one would you ask if you wanted to bomw money?

H u u p u go waan f o p u barman? Who would make a better boss?

(if d m tuu bina sa&@ wuk) Huu go fnin wok niwr kwik? Who would find work first?

I fyuu gu prablm (an yu waan taak dis ting out) huuguu go taak wid? If you had a problem who would you taik with about it?

Wich wan a d m fuu a duu wiidin'? Which one does weeding?

... a plant gyardin? plants a garden?

... a stu hoom wid shii plknii dm? stavs home with her children?

... a wok in a&? works in an office?

...a wok a di bank? works at the bank?

I l . ... a sel ring pan di peev? a wan striit wenda? seUs things on the street?

huu out a d m tuu go ukt bigitii? wouid act 'biggityn?

.. .go buuz/kwuriI op wid shii neebu? wouîd abuse her neighbour?

...go taak mari nem? wouid talk "mati's" name?

...go kos an taak hard? wodd curse and tak loud?

16. ... noofu kzik? knows how to cook?

1 7. ...go rnek wan gud waif (fo wan man)? would make a good wife?

18. ... alivgudwidpiipl? iives good with people?

The resdts are given below in figures 4.2 and 4.3. Figure 4.2 contrasts the responses of men

and women. Figure 4.3 conuasts those of workuig dass and non-working dass respondents.

Values for the diart were calculated as the percentage of respondents who dioose the b a s i l e d

(more creole speaker) for each question (The samples are reproduced at the beginning of this

chapter). Thus, for example. in the case of question number 1. (huu moor big? "Who is older?")

60% of the non-working group, 67% of the working group. 57% of the women and 70% of the

men choose the more creole speaker. Generaily the responses indicate strong agreement about

the ways in which Linguistic varieties are assoaated with occupation (questions 6- 1 1 ) . There is

less agreement about the personal characteristics of the speaker (questions 12- 18). In this

category we find that alI communiqr members tend to assotiate the les-aeole (more English)

speaker with acting big@ "arrogant" (question 12), but that they disagree about who is Lücelv

to engage in both abuse with a neighbour (question 13) and talking name. The "non-working

class" group (40% for question 13, 25% for question 14) and the women (38% for question 13.

43% for question 14) do not assoaate these characteristics with the creole speaker as

unambiguously as do the men and working dass groups. Women again pattern with the "non-

working class group" in question 17 and question 18 with both groups showing a suong

. . . preference for the basilectal speaker. AU groups consider the basilectal speaker more &ely to

nus and t d k Iiard. Men as a group show the most unarnbiguous responses indicating rhat they

112

hold the strongest stereosrpes ünking lmguage to sociai position and personal characteristics.

ï h u s the male group scores unanimous responses in €ive different categories (Questions

6,8,9,12,15).

18. Who hugmd wüh peoplcl

Evaluations of Builectal and AcmIecta1 Speakers

3. Who w o d 301 wmt f i a m -

Figure 4.2 MEN'S AM) WOMEN'S EVALUA~ONS

Evahfions of Bdecta l and Acmlcdnl $eriers

4 @- Q F r -

Figure 4.3 WORKING AND NONWORKING sVBJECTS' EVALUATIONS

4.3.3 S~~YLISTIC VARIATION

One of the most important characteristics of native Enguistic ideology, as 1 have already

discussed with regards to gender, is the assoaation of stnictural variation and the native

~pology of speech practices or acts which emphasizes purposive hinctionality (cussing ,pc18;ng

ta& ' nice etc. ) . Creole varieties are thus strongly associated with nrrsing and talking hard. In

fact, when I first expressed an interest in Ieaming aeolese I received instruction oniy in

parùdar genres. This induded instruction in calling uf and cussing. Native speakers (both male

and female) often told me that 1 was not speaking hard enough or that 1 needed to be more

assertive with m y interlocutors and more direct with those assumed to be subordinate (most

often women, chiidren and adolescents). S tylis tic variation in the community is thus very much

influenced by this association of stnictural with hctional properties. Because code varieties

are assotiated with practices which realize social relations. shiks between varieties on thc creole

to English continuum become bound up with notions of politeness, deference and solidarity.

Thus. native speakers associate speaking creole with addressees of similar or lower soad

standing and associate speaking more English varieties with addressees of higher standing.

When asked with whom they would adjust their speech, the cornmunity as a whole generates

an irnpücational array of addressees (Table 4.4).

Nancy Dingy Randolf C h a r l e s Shalini Kota S t inka W e W e De0 Marna Z i n i i Moses Seeta Rita Indar Champa

1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C c C C C C C C C C C C C C C C C C E C C C C C C C C C C C E E E C C C C C C C C C C C E E E C C C C C C C C C C M E E E C C C C C C C C C C M E E E C C C C C C C C C E C E E E C C C C C C C M M E M E E E C C C C E E E E E E M E E e C E M E E E E C E E M E E E C C E E E E E E E E E E E E

C=Creolese. E= English. M= h h e d < 1 > ajdajii <2 > mistresdhusband (3 > mother <4> father 4 > sister <6> brother c 72 pandit ~8 2 friend 4 2 respondent's child c 10> god c 1 1 > woman with a stalI in the market < 12> medic c 13> doctor 14> minister in the govermen t.

Table 4.4 IMPLICATIONAL ARRAY OF ADDRESSEES BY SELF-REPORTING

At the top of the array are the working class men who report no adjustment. At the bottom are

the non-working class people for whom the use of Creolere is resuicted to panicdar addressees.

For the higher classes, then, t h i s adjustment to addressee reduces to a simple equation between

acrolectai codes of speaking and poiiteness. When 1 asked whether she thought Creoiese shouid

be used in schools, Champa. certainly the most a a o l e d speaker I i n t e ~ e w e d and a teacher

at a local school, remarked:

Cham: I don 't go for that. When thgr go for rend in the textbuok t h 9 will not undentand a hell what t h 9 are nading. So if we go d o m tu meet thnn and to suit them we are nowlrm and we want to build let-we-say in some tmn an egalitanan soci-p. - ok - SO if we have thmt no w - like Im spreading it in my home we maintain &ain languuge- so when even childmr coma, we have vegetables to sell, t h 9 conte and t h9 have to speak pmperiy. I inrist not that direct& but in a jovial way I put it nicely to t h muybe well you can say t h b so i f t h y come the jn t time ... second tim time ... third tinte they will Say it and whm you comeyou can please say gwd afrrrnoon f i t .

117

Champa here makes a direct association in this passage between speaking properly and a set of

noms for social interaction that embody what is cvpicai (though highly idealized) of polite.

middle class behaviour in Guyana. At the same t h e . most acro1ecta.i speakers rationalize

adjusunent 'dom" the continuum in terms of problems with comprehension. They argue that

thev rnrrct accommodate to the most creole speakers if they want to make themselves

understood but at all other times the addressee is obliged to accommodate "up" to them.

Zinni, a shopkeeper notorious in the village for her s d jokes and hard taking

manner. similarly remarks on the comection between acrolectal styles and politeness.

Jack: If yu gu taak wid wan minista in di govament, nou? u ~ y o u ' w guing to talk tu a ministcr in the govemnrmt, now?"

Zinii: yu gu uuz lil inglish - dis iz rüspektabl püpol - shoo dem id disiplin Y O U will use a littk Engliih - n e t e a n respectable peopk - show a little dkciplinr. "

Onhr the working class men seem to resist this equation of structural variation with politeness.

Thus Nancv. Dingy and Randolf ail remark that they will not adjust for a minister in the

government.

Nan: yu gu gatu spi& seem krüolüz wid am biikaz mi na noo inglish 'You have to speak the rame Creolese with him because I don't know English. "

Ding: wel hou mii noo hi taak - laik hou mi a taak tu yu - jos soo mi a taak tu dem "Weil how I know to talk- like the way I'm t a h g ta you- I tuut to them the rame wv "

In terms of luiguistic ideology these speakers are purists who believe that adherence to

basilectal vanants in such contexts signals solidarity. However, although these people make

e-xplicitlv purist statements. I show in the next chapter that they in fact exploit a wider range of

the symboüc resources avaiiable in the continuum than do other l e s 'purïst" speakers (cf 0 . -

chapter 5).

I t seerns then chat we have uncovered another set of contmdictory attitudes in the

cornmuni?. One group of speakers assoaates the English varieties with poüteness and aeole

varieties with a la& of deference while another group associates the aeole with soiidarity and

English with distance. One way to think about this is in terms of the difference between

positive and negative politeness (cf. Brown and Levinson 198 7). For the highest soaal group,

speaking English implies soaal distance and non-faxniliafity which is highly valued among

membea of a social group who are consciouslv tzying to distance themselves from the

'parochial" character of the village. Social distance and non-famüiarity thus becomes a kind of

negative poüteness. Champa articulates this ideology in the foilowing passage (she is

responding to mv question of whether children start to talk differently when they enter sdiool):

Cham:yes we iwîst that t h 9 speak d@rmztiy. (in what way?) - You ask the child hm name. Ym want the child to respond to what i s your name? The child will tellyou immediately - give them one answer - what is your nume. Instead ofsaying "my name k.. " the child rvill bluntiy say, like in my- our - little, wwhat Lr your name?" the chihi will Say uRajn Singh " well I d o i t want that I wantyou to Say %.y name is Raja Sin@. "

In this passage, Champa makes an association berneen non-acrolectal styles and the reliance on

background knowledge such that the presupposition of intersubjective familiariv becomes an

indirect index of status equivaience. The idea here is that, when speaking creole. people are less

explkit and this reliance on implicit meanings indexes familiarity between the speakers (cf

Bernstein 1964). On the other hand. the expliamess of her English does not imply a relation of

familiari- and, bv extension, a shared background of knowledge. Thus English becomes

associated with a demonsuation of negative politeness - that is, a respect for the interlocutor's

autonomy and privacv. On the other hand, for some of the working class speakers, creole

varieties are assodated with positive politeness - that is, an indication that speaker and hearer

I N

know and approve of each other and that they share a cornmon uworld view" which serves as

interpretive backdrop and against which interaction can take place. This association is not

accidental. Creole varieties are primady associated with iife in small villages where privacy and

the values of indMdualism are rather subordinate to the values of an age-gaded and gendered

solidariw. Aaolectal varieties. on the other hand. are protorypically assoaated with the soadv

and geographically mobile middle classes who are considerably less invohred in the s d

structures of kinship and less susceptible to community d u a t i o n .

These ideologies of adjustment (or non-adjument) to addressee in which language is

taken as an index of social rehtionship (Le. politeness) exist along side another set of ideologies

concerning styhtic variation which take as basic the association between linguistic variety and

the context of speaking. Because this ideology is less often expressed. 1 term it the subordinate

theory of stylistic accommodation. The dominant theos.. with its emphasis on the soaal

position of the addressee. is remarkablv sirnilar to the theories of social and stylistic variation

proposed bv Beli ( 1984). Note that it suffers from some of the same failings insofar as it does

not adequatelv deal wirh the problem of social power and target dioice. If we assume that nvo

speakers corne together to communicate and that they command different. but overlapping,

ranges of a socio-scytistic continuum at some point we need to ask how they negotiate a

*targer". How for instance do Guvanese speakers deade that, when speaking with a minister in

the govemment. it is more appropriate for the working dass. aeole speaker to accommodate up

than it is for the minister to accommodate down? The subordinate ideology answers these

questions with its emphasis not on addressee but on the context of speaking. In this sense it is

more iike a native theory of diglossia (cf. Ferguson 1959. Winford l985a) in which different

120

varieties are considered appropriate in particular places and for certain puiposes. Most cenual

is the belief that Creokse is appropriate for everyday discourse in the home whereas English is

more appropriate for talle in the 'public" sphere of one's life. This view is deariy expressed in

i n t e ~ e w s with native speakers:

Jack:

Mos:

Jack:

Mos:

hou yu gu taak wid yu bodii how wouldyou U A toyour brothcr?

kriioliiz wen yu de hoom Creolese, when you *n at hume

an hou yu gu ta& wid yu mischris? And how worJdyou tulk to your wife?

kriiolüz wen yu de hoom bo if yu de in kompanii in Som pobiik plees yu gatu Creulese when you a n at homey but fyou 're among other people in a public place you have

taak nais a n &sen inglish tu talk nia and d e m t Engfish.

Indar makes a shdar co~ection between socially meanin@ spaces and language varietv:

Jack:

ind:

. Jack

Indar

Yuu noo raa ta& a wa? Do you know what 'raw talk' is?

it rornnfmm the home, home environment, w h m the home, the pamtr especïally speak this way ofcommunicating is a method ofwmmunicating fir thm. t h y get over what t h 9 want tu Say v q , easily - and sometimo this m w tak the child bnngs it tu the schuul situation, it givo a p m b k to the teacherJ because the teacher arhig the question will have to rephrase and go back to the raa taak to get over what he or she ir Qing tu say. It is the home envimnmmt and the cummunity which have a lot ofbearing on this raw t a k

if yu gu a makit nou and yu gu taak to wan Oman wa ga staal de - hou p gu Ifyou go tu market and you 're gving to talk to a woman with a staU there - how would

yu taak tu shi? vou talk to hm?

I will speuk more standard Engluh - How anyou? How much for the oranges? M a t ir th cost of the bananas - my status - peoplc amund miglrt say Iook how this chup aet

jack: an hou yu doz taak tu yu mischris? And how do you tak to your w*?

Indar Same. but sometimrr we mir. it makes some enjogmmt. I f the rnittress doesn't undmtand I break i t up. 1 do spenk it at home. I speak the ra w tulk nt home.

As 1 have mentioned. native speakers use this ideology to think about differences between

men's and women's use of the linguistic continuum. An important aspect of this subordinate

ideology is its recognition of the wav in which linguistic variation is intimately tied to the social

distribution of power. ImpliQt is the understanding that people who do not command more

Engiish varieties will be exduded from parricular contexts of speaking, or at least, in Bourdieu's

( 199 1) words, "reduced to silence".

4.4 ASPECTS OF THE I N T E R A ~ O N BEMEN ~WGUAGE AND CONTEXT IN GC DISCOURSE

So far 1 have tried to show the compiex way in whidi variation is related to social

position and interactive context in native linguistic ideology. In the last section on stvlistic

variation. 1 brieflv considered the use of language in estabiishing interactive contexts and social

relations. I t is co this issue that 1 now tum in the foiiowing discussion of gmnmaticalized

devices for relating speech act partiapants to the context of speaking. 1 retum to the issues of

fonnal variation as pan of a consideration of the way stance and affect are marked in CC

discourse. This issue is also taken up in the next chapter.

4.4.1 DIMENSIONS OF CONTEICT

The grammatical and lexical apparatus of a language typically encodes some aspects of

context and not others. Most commonly encoded cross-hguistically are the dimensions of

122

space. thne and various aspects of paaiapation (induding speaker stance. affect, evidence as

weli as speaker-hearer aiignment) . We have, based on the existing iiterature. few available

descriptions of creole languages from this perspective. Aithough certain relevant areas of creole

gramman have been sub jected to intense scmtiny (su& as tense and modality), most often this

research has been concemed with the grammatical or syntactic behaviour (sometisnes semantic)

of partidar preverbai markers in isolated sentences rather than within n a t d y ocnirring

linguistic practice as such. In the foJlowing 1 sketch out a few areas of GC grammar that are

central ro the pragmatic organization of evexyday talk.

4.4.2 SPATIAL ORGANIZATION O F TALK

As 1 discuss in later chapters, talk is anchored to locally meaningfid spaces. In rural

Guyanese villages the spaces of the house. the yard and the road are particularly important in

this respect. Because people draw authoriry and legitimacy (in short. because space is imbued

with meanings concerning the rights and privileges of spedic persons) speaken attempt to

situate themselves and their interlocuton @ relation to partidar configurations of space. In

fact. rights to speak are largely dependent on where one speaks from - although at times people

may attempt to suategically obscure their situatedness. A major part of this anchoring happens

in the negotiation of HERE. which I de fine as the shiftîng and mutable, vet tacitly agreed upon,

region in which social interactions take place (cf. chapters 5&7). By locating themselves and

othen within or outside the HERE and by fixing its boundaries, speaken orient and position

themselves as thev establish situational relationships of hierarchv, solidarity and autonpmy (cf œ . -

M.H. Coodwin 1990, Keating 1994). The interactional process of fixing the HERE is

123

complicted by the fact that. in any community. space is imbued with multiple meanings and is

divided up by multiple and sometimes overlapping boundaries. Thus a single objective position

in phvsical space may be conceptualized in any number of ways depending on the way in which

it is h e d - the panicular boundaries and temtoriaiities that are foregrounded. Schegloff

( l972:W) remarked that:

(F)or any location to which reference is made, there is a set of tenns each of which. by a correspondence test. is a correct way to refer to i t On any actual occasion of use. however. not any member of the set is 'right'. How is it that on parti& occasions of use some term from the set is selected. and other terms are rejected?

Schegloff argues that each place term (and the frarne it indexes) is drawn from a more inclusive

set and that this process of selection relies on interlocutors extensive lcnowledge of what rnight

be cailed "cultural geographies" ( Schegloff 19 72: 103). Such culturai or commonsense

geographies may be organized in hierarchicai tems (as Schegloff 1 9 7 2: 1 03 argues for

mainstream American understanding of political geography) or in quite different ways (cf.

Chapten 1. 6 and 8 on the commonsense geographies of the village and its paru). I t is this

cultural geography that dows for recognition of place tems in the sense of 'the a b w to

brùig knowledge to bear on them, to categorize. see the relevant significance, to see 'in what

capacity' the name is used." (Schegloff 1972:111).

Because space is organized accordhg to local hames of significance (Halloweli 1954)

and because, in &. selectional process are brought to bear on the great many ways of

conceptuaLizing a given place at a given Ume. deixis always anchors interaction and interactants

to a cultural rather than a purely physical organization of space. As Keating ( 1994) and

Coodwin ( 1 9 90) have demonstrated. cueing the spatial organization of an interactive sequence

can be meaningfd at a number of different levels. For example, because space c m be owned,

124

interactants can make specid ties beween themselves and/or others and specific locations in

interactive space (Goodwin 1990). Sdiegloff ( 1972: 11 7) calls these %, (Relation to Member)

spaces and notes that what sets them off from other types of place formulations is the faa that

people belong to them (and. at the same tirne. they belong to people). Whde Schegloff ( 1972)

suggests that 'home' is espeaaiiy kev here (cf. Levinson ( i983:84) and Fillmore ( 1 9 75:5Off) on

deictic anchoring to a home-base), in fact communities vary in the spe&c focus of this

belonging-to-ness and the deictic organization of spaces may be organized differentiy as it

interacts with other factors induding inheritance. kinship, domestic structure and. of course,

property.

Particularly important for designating relations between particpants and context are,

then, are locationai deictics which, in CC, show most of the distinctions present in other

varieties of English as weli as some innovative forms. CC has deictic forms as locative adverbs

(e.g. yu or hin "here", de or deer 'therem), demonstrative adjectives (e.g. dir ting 'this thing". (d)a

ring ' that h g " ) demonstrative pronouns (e-g. da (a) di man de " that is the man" ) and deictic

verbs (e.g. h m 'corne". gu 'go"). Deictic verbs show many of the same ambiguities as their

English equivalents (Fillmore 19 75, Levinson 1983). Accorciing to Anderson and Keenan's

( 1985) classification. basilectal GC is essentially a tree-tem, distance-from-speaker, wstem

which invohres. somewhat peripherally, a relational feature of visibldnon-visible as a second

dimension of conuast This is iliusuated in the system of locative adverbs.

i deya 'it's here"

i de soolyanda "it's over there" non-immediate - non-visible to me and p u

i de de "it's there" region

Table 4.5 GC THREE-TERM SYSTEM FOR LOCA'ïTVE ADVERBS

Basiled CC also has a system of pst-posed deictic markers (cf Ridcford 1 9 8 7b: 1 7 4 1 75).

These show up especi* dearly and most often in plural noun phrases such as dmr de 'those"

but can also be used. often with affective meaning, with singuiar nominals as in dir tingyu "this

here", dis manya 'this man here". The system of mot and post-posed deictic terms. showing

possible combinations and apparent gaps, is illustrated in Table 4.6.

DENOTAn,lM

region

non-imrnediate - visible to m e and you

dis I - I - 1 &+=a

RELATiON

immediate - visible to me

Table 4.6 THE SYSTEM OF RoOT AND POST-POSED DEICTïCS IN GC

As Danziger ( 1 994) and Hanks ( 1 990) have illustrated, spatial deictics always make

reference to an object or a region by establishing its position relative to the location of some

speech act panicipant(s). This ground for deictic reference is not necessarily the speaker. Manv

languages, Mopan Mayan among them (Danziger 1994). exhibit 2nd person spatial deictics

which locate a region in proximity to the addressee. A smaller number of languages have 3rd

person spatial deictics (again cf- Danziger 1994) which may be marked as visible and invisible.

The very idea of a 3rd person spatial deiaic, discounted by most general understandings of

deixis, suggests that our theoreticai models of communicati~n. which stress the dyadic or,

euphemistidy, the 'face-to-face" nature of speech, may be based on a particular ünguistic

ideology which ultixnately derives units of andysis from the distinctions embedded in grammar

(cf Silverstein 1979). In faa. CC grarnmaticalizes, in the deictic wstem, aspects of interactive

context that are not the focus of many other English systems. In this regard we might note the

parricularlv complex system which locates an animate or inanimate non-participant ( d m or dis

or da) in relation to the position of the speaker (dû. da, ya. de). Such a system diffen from that

desc~ibed by Danziger ( 1994) and Hanks ( 1990) for Mayan languages in which the non-

participant or 3rd person serves as indexicai ground. In the GC system, as in some other

dialects of EngLish (thir-hm etc.), the third penon, or non-participant, is the referential figure of

an arrav of deictic expressions. I discuss the contextualized use of this system in chapter 6 .

4.4.3 TEMP~RAL ORGANIZATION OF TALK

The contextual kame of t& indudes more than the spatial organization of participants.

EquaLiy crucial to ail kinds of interaction, but parcicularly complex in nanative, is the temporal

127

organization of events. Like spatial deixis. temporal deixis anchors partitipants to cultural

relevant formulations of the HERE-NOW. Because New is aiways the point in time occupied by

the current speaker (and therefore daiming a right to speak is aiways a daim to inhabit the

NOW) . interactants frequentlv negotiate its boundaries in talk In Guyana. a frequently heard

inte jection is jus nou ' just now" (or 'in a minute*) which is used to mean 'let me talk and you

will get your chance after." In narrative. speakers orient their unfolding stones vis-a& a

situational NOW of the t e h g . Thus narrative aiways involves at least two temporal frames

distinguished by Jakobson ( L 957) as Es (speech event) and En (narrated event). The

relationship between the two is complex and speakers actively manipulate it to adiieve stylistic

and textual effects. As Hanks ( 1996a) and many others (Banfield 1978, Haviland 1996,

Hickmann 1985, 198 7. 1993, Silverstein 1993, Volosinov [1929] 1973) have noted, multiple

fiames of narrative (and reported speech) involve complex transpositions of the indexid

ground. In reported speech. for instance. deiaic reference is not necessady anchored to the

HEM-NOW of the nanating context. Thus in a report such as:

1. Maw said to John 'don't corne around here anpore."

Said by June to Joe, the 'here" is not necessarilv identical with the place in which June and loe

are located - although it may be. To calculate die referential value of the spatial deictic it is

necessaqr to relate it to the framing report consuuction 'Maiy said to John" which is

temporaliv situated in the past relative to the exchange between June and Joe. The temporal

organization of multiple frames thus dows for the tracking of referents across a number of

situational contexts. 0.

Temporal deictics are important metaprapatic mes. What Siiventein ( 1993) c d s the

128

metapagrnatic calibration of P (the event of speaking from whidi metapragmatic framing is

achieved) and En (the event whidi is narrated) is often the d e n t nie for genres in "oraln

Literature ( proverbs. rituai speech oken being uniquelv framed with what Silverstein ( 1 993 :S2-

53) c a k 'nomic calibration' i.e. as emanating from some thneless or etemal reaim, for eaxrnple,

"it is sai&writtenn). Calibration is also an important area of sqrlistic ma- in narrative.

Reportive calibration (where Es and En do not coincide) is suongiy associated with an explicit

metapragmatics (such as the use of fkaming metalinguistic verbs) , whereas, reflexive dbration

(where F and En coincide) is strongiy associated with an impliat metapragmatics (Le.

contextualization nies. the use of CO-occurring referentid and non-referentid indexid , poetic-

textual structure). But narrators play on this disjunction and may attempt to achieve explicit

metapragmatic framing with reflexive caübration (e-g. ma te+ ... - "1 am telling you ...") or

pseudo-implicit haming of reports which are metaphorically 'brought inton the nanating (Es)

kame. Nanators often use quotatives which focus attention on the interadonai point at which

the message is received rather than when it is produced. A clear example of this is the report

frame, hiir am nou. .. 'heu he- now ...". This foxm recreates the 'originaln voice by

collapsing the audience of the speech event with that of the narrated event Such devices

represent a sophisticated ability to manipulate the habitua1 anchoring of tdk for aesthetic and

other interactional purposes.

Besides Lexicalized rime deictics including adverbs (now/then) and demonstratives

(thidthat) . temporal organization of multiple frames is achieved through the use of tense

marking. In GC. tense is indicated by the interaction of preverbal marking with the inherent

lexical semantics ( Aktionsart) of the verb. AI verbs in GC f d into either stative or non-stative

129

classes (for discussion cf Bickerton 1973a. 1975. Givon 1982, Rickford 1987b, W i o r d 1993).

Unmarked stative stems are interpreted as present. unmarked nowtatives as past (cf. chapter

2). Past is indicated for stative verbs bv preverbal marking with bin so that bin with statives

indicates 'not now" ( i.e. a terminated state) . For non-statives, bin indicates an "anterior" event

Although there is a good deai of disagreement. anterior is genedy taken to mean that the

rvent preceded some other reference point or tense locus which is not necessarily the time-of-

speaking (cf. Chung and Tiiberlake 1985:209 for discussion of such systems) . Rickford

( 1 9 8 7b) argues. after analyshg one text showing partidarly frequent uses of bin. that, with

stative predicates bin is used in cases where the 'no longer' component can be taken as an

implicarure rather than implication. Riddord notes that for the use of &in to be feliatous it is

not necessarv for the state to be terminated in the 'reai worid", rather, felicitous use depends

upon the speaker's attitude towarb the events and their continuing reievance to ongoing

conversation. Rickford ( 198 7b: 13 9) writes:

. . . (T) here is one. ..situation in which Irene's usage of bin would be justified - if she were to tell us that she did not know (or a re ) whether the states in question were still in existence. Intuitively, this is what is crucial for the reader or üstener to infer from her narrative-that whether or not the states in question are s t i U in existence a t the time the narrative is being told, thev were in existence at the point in the past when the evenu referred to in the narrative were taking place.

This suggesu that it might be usefd to Wear the 'no longer" component that Bickerton associates with stative bin as an implicature rather than a strict implication.

For non-statives. the unmarked stem indicates a past action. In this system bin indicates that

the action denoted by the predicate is not only past but also anterior. at least idedly. Rickford

( 1 98 7b: 1 4 1 ) summarizes Bickerton's argument as follows:

Bickerton's ( 1975:47) explanation is that although an action might be "past" simply because it occurs pnor to the moment of speech, it must also be 'anterior" to another

action or actions to take bin. G ~ e n two events that are sequent* ordered in real life as E l and E2, oniy E l wodd 'anterior" and thus eligible for marking with bin. There is an additional qiralification. however. As Bickexton ( 109) notes. a nonstative predicate oniy requires markhg as anterior 'when the speaker inverts normal narrative order (i.e., refers to an earlier event after a later one)." Elsewhere (53). he makes the point even more explicitly: 'even where one [ p s t event] occurred pnor to another, both can be handled by the English simple pst , but (unles the actions are sequent ones in a narrative) aeole will normdy give the [-anterior] one the stem form and the [ + anterior] one the marked-past form".

In his anaiysis of one narrative text Rickford ( 198%) fin& that the absolute condition on out-

of- sequence actions being marked with bin (Le. when narrative sequence and real He events are

not iconic) does not hold at ail thes. He thus concludes ( 198 7b: 143):

The solution to these 'problem" cases is to revise Bidrenon's analysis to provide for optionai marking with bin when anterior E 1 precedes E 2 in narrative sequence. as against the obligatory marking when antenor E l follows E2 in narrative sequence, bin marking on non-anterior E2 wodd still be exduded. The possibilities are these:

E l (bin or O ) . . . E2 (O oniy) E2 (O only . . . E 1 (bin oniy. out-of-sequence antenor event)

We rnight think of the optional inclusion of bin when E I and E2 are iconicailv sequenced as a reflection of the redundancy rhat creole systems. like other fUli-fledged languages . contain.

Perhaps it shouid not surprise us that tense marking is used primady to indicate relevance (or

following Civon 1 982, Poiiard 1 989. Winford L 993 foreground and background information).

Co-presence always potentiaily. by the very faas of contiguity. implies a degree of relevance.

However the same does not hold for events/states that occur at different points in t h e .

Relevance (or kelevance) in these cases must be Uidicated by some expliut means.

4.4-4 PARTICIPANTS AND PERSON DEXMS

Participation is never simply a matter of being present in some situation where

interaction is taking place. Rather, the interactionai hame of participation is continually

negotiated and, at each point, adueved (cf. Goodwin and Goodwin 1990). Like the other

dimensions of context discussed to this point (spatial and temporal), partiapation frames are

established, in part, through the use of deixis. Thus, in some cases, the negotiation of

partiapation may be marked in the text in the use of person-deictics and thus tracked by the

analyst (in other cases a full analysis wili depend on attention to other aspects of the sequential

organization of ta&). Because person-deictîcs in mat, if not dl, languages make a distinction

between participants ( 1st and 2nd person pronoms) and non-participants (3rd person

pronouns), speakers are able to manipulate, in their taik, the structure of partiapation - who is

being talked to and who is being talked about. However, participation is not aiways marked in

the deictic structure of messages. The phenornenon labelled ''indirection" refers to the

establishment of participant f-rames without expiicit marking by deictic or anaphonc elements

(Fisher 1976, Morgan 1991, 1993. 1996. Reisman 1970). Volume, gaze, gesture and a number

of other non-referential indices (induding silence or non-ratification) may thus be used to

communicate information about partiapation frames.

s As Levinson ( 1983, 1988) has noted, however. personal deictics in a given language

only grammaticalize a subset of the participant roles which are distinguishable in actual

interaction. He rem& ( l983:68) :

Although penon deixis is reflected directiy in the grammatical categories of person, it may be argued that we need to develop an independent pragmatic fi-amework of possible paniapant-roles. so that we can then see how. and to what extent, these roles are grammaticdized in different languages. Such a framework would note that the

speaker or spokesman can be distinct h m the source of an utterance, the reapient distinct from the target. and hearers or bystanders distinct addressees or targets. and that sometimes such distinctions are grammaticalized in non-obvious way.

While such an approach is no doubt of considerable analytical interest for the cross-linguistic

study of grammaticaiization (particuiarly grammaticalization of "contextn), the imposition of

such a grid on natufally occurring interaction may obscure. to some extent, the active

manipulation of the ernic, categories that are grammaticalized (cf Hanks 1990: 154) espetriall~

when these emic categories do not coinade with those which may seem analytidy signiscant.

As with other pronominal systems (Siiveatein 1976b. Hank; 1990). CC pronouns can

be ~ a n g e d according to a number of referential hierarchies. Thus plural pronouns d . e r

systematically in their potential leveis of indusiveness (cf Hank 1 990: 1 8 6 ) :

3rd prsn. pl. 1 d m 1 - 1 - +

1 s t pan. pl.

2nd prsn. pl.

The hierardiv of referential categories can also be Uusuated in a reverse manner. Rather than

the inclusiveness of the referential categones, we can arrange categories according to the extent

CO which they diaracterize the referent

abi

ai,m

SPKR

+ ADDR

+ +

OTHER

+ +

Table 4.8 HIERARCHY OF CUA"CTERfZING FEA'LURES OF PRONOUNS

gender

Different types of information are more salient, through the coding features of participant

deictics, in different forms.

case

Looking at the plural pronouns we see one of the most important areas of

grammaticaikation of pragmatic information in GC. The 1st and 2nd person plurals are highly

dual

salient in native speaker awareness (which probably should not surprise us since they are

plurai

+continouslv sepentable. + unavoidable referential cf. Sihrerstein 198 1 and chapter 4). When

asked to identih mkse (/kriio W) . raw talk (/ma ta*), or broad talk (/braad taaW). speakers

neariv dways made reference to these items. Consider the following answers to the question. a

tua ma tnak? "what is raw talk?":

2. laik hou abi a taak. 'Like how we are taking." (Charles)

3. raa taak a laik wen mi an vu a taak - aiyu an abii an soo. 'Raw talk is like

when me and you are talking, 'aiyu' and 'abü' and so on."(Rita)

4. abii an aiyu, "abü' and 'aiyu'"(Stinka)

5. raa ta&? c to G. > wa raa taak man? abii an aiyuu? da di raa taak. "Raw

mk'? What's 'raw talk' man? 'abii' and 'aiyu'? That's the 'raw talk'." (Marna)

iMany other examples could be given to show the salience of these terms as indices of the enrire

idealized (because no one speaks a "puren creolese as defined locally) system of b a s i l e d

creole. Rickford ( 1 9 8 7b: 1 9 7) refers to them as ' highly marked bas i ied elements. " W e this

is no doubt relevant to our understanding of local interpretations of the system, it is also

relevant to our understanding of the deictio themselves. Besides the general cognitive factors

isolated bv Silventein ( 198 1 ), we might also suggest a nwnber of other local and historically

particular factors that make abii and ai@ ppaaicularly salient in relation to the superstrate:

(a). thev differ PHONOLOGICILLY from cognate and comparable aaoiectd t e m s ( a d a

wii, wii, mii, aal aguu, yu)

(b). they differ in tenns of the5 FUNC~ON Dn< - refers to both singuiar

and plural addressee. a 1 9 for plural only.)

( c ) . thev differ in INDEXCAL FUNCîïON (aiyu - computes multiple addressee kom the

indexical ground of the speech signal)

(d). they differ in LMETAPRAGMATiC RTNmON (aiyu - characterizes the speech situation

as involving an explititly multiple addressee)

Considering for a moment the indexical and metapragmatic hinction of these terms. note that

abii and n i p together form an opposition which potentidv divides the speech participants into

nvo mutuaily engaged and mutually co-present groups. In this, the svstem diffen form the

English counterpan which does not differentiate explicitly between singda and plural

addressee (on the historical reasons for this see Silverstein 1985~).

Ir is instructive to compare the historical trajectories of English persona1 deictic plurals,

on the one hand. with the creole forms, on die other. In English. the second person plural

developed into an index of deference and soaal power (cf Brown and Gilman 1960, Silversrein

l98Sc). Silverstein ( l976a) notes:

.At is a shift out of the realm of the singuiar, where an individual is refened to. and into the realm of nonsingular. where, as it were, the summed number of individuals

referred to is greater than one. This makes the addressee count for more than one social individd to his persona acaues the soaal weight of many, as compared with the speaker.

In CC, and other creole languages, the 1s t and 2nd plurals are often used to indicate solidarity

or, at least, commonalitv between individuais - that is, common situation, common

understanding etc. (for an example cf. the passages at the beginning of this chapter fiom the

matched guise. The speaker remarkc -yu noo abii-diiz piwpl la& pientii hat pepa). I t is. then,

si@cant that creoles do not ailow the extension of pluraiity into the indexka1 realm of socid

statuses. It indicates, perhaps, a greater concem for CO-presence and co-identification which

either arresred gramrnaticaiization (if you see a@ developing from English 'AU of youn) or

encouraged regramrnatidization and caiquing (if you sec aipu developing from an Afncan

substrate) .

4.4.5 STANCE

In evew speech comrnunity a great array of devices are used to indicate the speaker's

attitude vis-a-vis some referentiaily encoded message (Besnier 1990. Labov 1984. O h 1992.

1 996b, Ochs and Schieffelin 1989). Ochs ( 1996) and others divide such indices into affective

and episternic dasses (although Besnier 1990 indudes epistemic within his general categoiy of

affective). Such classifications are problematic for a number of reasons. First. they tend to

preclude consideration of what. in grammatical discussions of modality, are termed deontic

rneanings. Deontic meanings have to do with speaker's assessrnent of "obligationn rather than

" possibilitv". In this area CC shows sunüar ambiguities as other varieties of English. Thus:

6 . ii rnos gu. 'Hdshe has to go."

136

can mean either 'Hdshe is obliged to go" or ''Helshe is certain to go." Howwer in past contens

only the epistemic interpretation is possible:

7. ü mosi gaan, 'hdshe must have left."

The deontic past sense is conveyed by 8 or 9:

8. übinmosgu."he/shehadtogu"

9. ü bin hi gu, 'he was supposed to go."

AU natural languages, induding CC. make distinctions of this kind (Palmer 1986). It is further

possible that some languages mark both deontic (obligation) and epistemic (psibility)

through the use of what has been texmed affectiw stance. Certainqr, epistemic stance seems to

be cued through "emotional" involvement in communities where subjective or experiential

assessrnenu of evenrs are taken seriously (Labov 1984). This is certainly the case in Indo-

Guyanese viiiages. In one illusuative case. a woman had been accused, by her own f ' . of

aIlowing her nephew to bring his then girlfiiend to her house for romantic engagements. The

girl. who was under fourteen a t the t h e , upon being questioned by her father. admitted that

she had had sex with the boy and that the act had taken place at Shanka's house. Shanka faced

serious dûapproval from her family and attempted to persuade them that no such thing had

happened. The sincerity of her account and her certainq of its a c m c y is cued by a nurnber of

Peatures. Not the Ieast important of these is her use of markers of affective or emotional

involvement. that is, intensity.

Sh = Shanka AK = Kay (Shanka's mother)

7.

8.

9.

1 O.

I I *

12.

13.

14.

15.

S h:

AK:

Sh:

AK:

S h:

AK:

S h:

AK:

Sh:

AK:

S h:

A. ZC:

Sh:

A.K:

dern na lak op brarnii? dern kech am wee? "Didn 't drey l d up B.? Whm did dirg catch him?"

soo wa-a a wa rilii hapm? wa dern see? s u what realZy happmrd? What did ctiq sq?"

ii see dern kom-(kom) mit ke-kech am in yu mum "He said diat dicy cnmr and m g h t him in your man."

huu dem? kech am in mii room? //huu dem? "Who is 'dim'? l%ey cmyht him in nty r m ? Who i s 'bhem"?

// beebii sing tuu "Baby Singh W."

-00- beebii sing kom he? "Oh - Baby Singh came here?"

beebii sing see shi bin haid onda di bed "Baby Singh said that she was hidinghidden under dre bed."

beebii sing bin hiafo sii da? " B e Singh was here ta see that?"

mii na noo gyal "I don 't know dear. "

wel (a da) wa di poliis na tel mii notin? le shi kom bak hia wid poliis (0.5) dern bina a Well drat's it W y didn 't du policc tell rnc anyhrking? Let h n c m e here with police O rhey were"

mii hous? a wa dern duu a Mi hous? if dern kom hia wa dern duu? wai shi na stap shii '9t my h o w ? What did &y do ut my houce? vbhey cane here what did rhcy do? Wy rlidn 't she stop her"

daata (fu wa shi doz den bin kom out wid a man) "daughw ( ) "

(if shi bin) duu "ifshe &du

WADEMGAFO D W - D W WA?

"WWTDO HAVE TODO?DO WHAT?~

(a hou shi // ) ("How is it. .. "

17. Sh: // an if wa shi duu yestodee nou shi piknii kyan bii takst wid out morsii de =and Bshe do diatyestrrday now hm child can be ben- widhout mer9 t h e ? -

beebii sing waan man shiself wa shi de de shi shart a man shiself B d y Singh wants a man hme@st as she is dim - she Lc short if malc compan-p hme&"

In this short excerpt from a larger exdiange, it is possible to see the major ways in whidi

Shanka defines her stance relative to what she is saying. Shanka's basic strategy is to take the

position of ignorance or not knowing. She immediately ( h e 1) takes the position of the one

who is lacking relevant information and, by extension. the interactional role of questioner. In

iine 3, Shanka asks for general information about what "reaily" happened, thus implymg that

there is a good deal of false information (what didn't r edy happen) avaiiable or in current

circulation. When she fin& herses imrnediately implicated in the story (line 4) she begins

pressing for specific information first about the identitv of the participants (5, 7. 9 ) and then

about the event itself, spe&caiiy the motivation of the main actors (what do they have to

do?). She also questions the procedural fairness of the police ( 1 1) and the motives of the girl's

rnother ( 12). Notice Shanka' use of the discoune marker w 'Oh" to indicate the receipt of new

information in line (7) (cf. Levinson 1983:3 1 1,353). Its use here is obviously sarcastic

indicating a problem with the factual record (i.e. what actually happened) since Shanka knows

that Babv Singh was, in fact. not a t her house. In line 15 S h a h marks her distress and

iricreasing emotional involvement by a marked rise in amplitude. Combined with the question

"What do they have to do?" the increase in volume keys a genre of more public argument

which 1 discuss in the next chapter. I t is partly through her 'emotional" or "affective"

involvernent in the interaction that Shanka constmcts her discourse as "certainn or "accurate."

This involves not ody a certain interactional alignment vis-a-vis other participants (Le.

questioner) and infonnation (i.e. lacking relevant infonnation), but also a certain subjective

state (exasperated. beleaguered) keyed through voice quaiitv as weii as in the sequence of

questions.

In a situation where one h d s a strong comlation between the levek of a language a

person commands dong a continua of aeoleness to standardness and the same person's soaal

dass membership, it is not surprising that switches from one variety to another should be used

to indicate the speaker's attitude towards the message (and for that matter the situation, the

addressee and a whole host of other contextuai features). In fact switches from more to less

creole varieties represent the most salient kind of nomeferential index in Guyanese villages (as I

have discussed earlier). Indications of stance are just one of the effects speakers achieve through

switches between varieties (see above). Of cruad importance here is the fact that viüagen

make a strong association between the levels along a creole continua and levels of education in

the legitimated school system. This means that for certain h d s of knowledge switches to more

standard varieties may indicate greater certainty (along wîth greater authon.). Often,

questions which go unanswered wiU be repeated in a more acrolectal or mesolectal 'voicen as in

the foiiowing example from a rumshop gyafT beoiveen sweral adult men:

Ro = Rohan (Nancy's older brother) Pank = Pank (Member of the patrilocal group whoe owns the

shop where interaction takes place). John = John (Panks' best friend) Nan = Nancy (Rohan's younger brother) Ch = Chiree (Rohan and Nancy's cousin. Friend to Panks and

John)

1. Ro: eepankIemihiirnouyunoomismana "hey Pank, kt me heur - doyou know Miss Manner?"

Pank naworii "Don 't stiut."

John: ( ) mis mana "()Miss M a n m

Pank: ii i i t jringk mis mana "he eats and dn'& Miss Mannm'

Nan: wa tel mi bai? " What are you qying m &il me?"

Pank bota an bred "hum and bnud."

Nan: ai bai wa yu gu tel mi? eeh U H ~ bop what are you going tu &il me?'?

Ro: di man see ii waan hia mo man said he wantcd to hear more..()"

John: naarguubai "Don 't argue boy"

Ro: noo badii na harguu hia - di man na laik Cu hiir laang taim stoorii noobadii na harguu "nobody is arguing h m - dre man dnsn 't Zike hear old-time sary nobody ir arguing

hia Cu fait - yu nao mis mana? aks am if noo huu mis mana? ta make afight -you knm Miss Mannm?Ask him ifhe knuws who ù Miss Mannm?*

John: booloo yu shit op "Bolayau smewed up"

Nan: mis mana a darsii muma or " M i i s Manners is Darsi's mothm or.."

John: ah ha darsii muma "uh ha Darsi's mother"

Ro: eh he we shi bin liv den? "eh he where did she [ive dien?"

Ch: ee? "eh? "

18. John:

19. Ro:

20. Nan:

2 1. John:

22. Ro:

B.--> John:

24. Nan:

shi neem iz beebi na darsii muma "hm name was Baby not Dani m o h "

Ive shi bin liv ? " w h m did shc [ive?"

am nat a m beebii muma "um not Baby modier"

beebi a di muma da-a-a mis mana "Baby is drc mo& that is Miss Mannm"

we shi bin liv? " Where did she Iive"

W E SHI BCN LN? wer shi woz living? " Where did she Iive? Whm did she liwe?"

eh? .

"huh?"

There are a number of interesting aspects to this exchange. In relation to the above discussion

of oo 'oh". it is interesting to note the use of the GC discoune marker eh he in h e 16 which

conventionally indicates the receipt of new and relevant, often unexpected, information. Like 00

"oh" in the previous exchange. eh he is here used sarcastically indicating that the speaker

beiieves that the information received is problematic. Regarding the use of mesolectal 'voices"

to cons titute authoritative stances, note iine 23 in which John repeau the question that had

previouslv been formed (in 16. 19 and 22) as we rhi bin liv 'Where did she liven first with an

increase in amplitude and then as wershi woz living. The exchange is an extended joke on Nan.

Ro and John believe that Nan frequently makes daims to knowledge which he does not

actuallv command. Here they replay a joke which has been used previously to show up his false

daims. "Miss Mannen" is not a real person in the village but a Literary persona who appears in

the locai newspapers. The exchange aiso hinges on the knowledge that literacy levels Vary

across individuais. Ro. knows that, unlüce himself, Nan cannot read particularly weU. He thus

assumes that rather than admit ignorance Nan will assume that ''Miss Mannersw is a local

peson whose whereabouu can be made the topic of discussion.'

But it is not alwavs the case that authonty is constructed through the use of more

English varieties. Certain domains of knowledge are considered knowable only through

sustained engagement in a mole-speaking, rural community (living there, growing up there.

maintaining ties there, etc.). When pronounàng on these domains of knowledge speakers can

use the more aeole varieties to indicate pater certainty. Of course certain objects and

techn.ïques may be considered to occur in more than one domain of knowledge. With respect to

agricultural knowledge. for instance. viUagers tend to show varying degrees of respect for both

school (or 'book*) knowledge and praaical knowledge. I t is in fact not uncornmon for two men

to argue vigorously in a rumshop over such issues. one using the moût b a s i l e d aeole and

drawing on expenences in the rice field and the other using a more standard variev and

drawing on technical knowledge of experts and science. Thus whiie villagers respect science.

school and education to a high degree thev also s a v p kyann tel mi jimii kuk mek ram goot "You

can tell me that Jimmy's cock is a ramgoat." In short, diagea value the knowledge gained

through personal experience and transmitted as local wisdom in Creolere. Local traditions of

knowledge are strong and leave the complete hegemonv of an English equated with a u t h o r i ~

incomplete. This rneans that whiie speakers do mark stance by using different varieties the

specific variecv that they use to do this WU also be determined by the panicuiars of the

7 Rwnshop gyaf between men frequentiy involves just this kind of spaning and joking - as well as 9 . tirateci debatc - over detaiis of seemingly little consequence. Where someone lived. how they died or

how they lived. and particularly their ancestry or geneaiogy become matters around which social relations are built in conversation. Participants emerge as knowledgeable, authoritative, hurnorous in the course of such talk. The rnatter of tdk seems of little importance.

situation.

In framing previoudy occurring discourse, speakers use lexical selection to mark stance.

Thus, as Kiparsky and Kiparsky ( 197 1) iiiustrated. some verbs take a factive complement

whereas others do not. Other more subtle markers of stance may also be found in processes of

lexical selection. Thus &II. although not factive, is essenti* an accomplishment verb

(VanValin and LaPolla nd:29). The inherent lexical semantics of teIl include the notion that the

penon told has been informed. or been made aware of something. Note that infom is faaive

and therefore tell, by implication, contains a weak factivity-

4.5 ETALI AL IN GUIS TIC PREDKCATES AND LINGUISTIC ~DEOLOGY

When reporting and eliciting discourse (see chapter 8) speakers select particular

predicates ro h e the utterance thev are bringing into their own t a k In these processes of

lexical selection, speakers draw on a metalinguistic vocabulary which is organized in particular

wavs. T h i s organization reflects and shapes the way in which users of a parricular language

undentand. interpret and taik about the pragrnatic organiration of grammar. A metalinguistic

vocabularv break. the continuum of different kinds of talk into discrete types. In reponing

speech, a speaker generally selects a particular terni which consuues that token reported as

member of a type. Metalinguistic vocabdary is also embedded in the interactions termed

rlicitauons which I discuss in chapter 8. Metalinguistic vocabulary thus plays an important role

in language acquisition and socialization.

These texms (and their grammaticai properties) are used for thinking through speech.

for talking about speech, for interpreting speech as well as for hming it in conversational

interaction. The terms thus impose a finite ciassification on a field of potentially limitiess

variations and possibiüties. As such. metalinguistic vocabtilaty is never neuval but instead

camies with it ideologicai assurnptions about proto~picai speech swles, events and participants

( cf. Linds trom 1 9 9 2). Thus, despite very si& srnictural and forma1 propertics. two instances

of speech may be considered instances of different generic types if one does not meet the

criteria for inclusion in a part icdu categow. One very common mesure for inclusion in a

parricular categoiy appears to be the gender of the speaker. Speech events such as h i n g

pnmlling, rowing are gendered feminine with native speakers taking the partiapation of one or

more women in the event as a presupposition (cf. Edwarcis 1978). One outcome of this is that

women's conflict is oken written off as some necessary expression of "femaleness" whereas

men's is likely to be taken as indicative of some larger social problem that requires resolution.

Simîiarly. although both men and women "gossip". tnlkng-name is generdy considered a

feminine activiqr. Men's gossip thus tends not to be stigmatized or, often, even recognized in

the same wav as women's.

4.5.1 THE LEXXCO-GRAMMATICAL STRUCTURE OF METAPRAGMA~C DESCRIPTORS:

Looking in more derail at the metaiinguistic predicates of a laquage, we fhd that they

characterize contexts of speaking through both their lexical and grammatical properties. AU

predicates characterize the situations or events to which they refer partiy through the selection

of arguments and the projection of thematic roles upon them. Whiie it is obvious that this is

the case. linguisu have not as yet reached definite conclusions about the number and nature of

these roles (cf. Andrews 1985, Palmer 1994, Vanvalui and LaPoiia n.d.).' I t can be said with

some certainsr however that grammatical roles compose a finite set of distinctions made in the

lexico-grammatical structure of a @en language coding, through a variey of means including

word order and prepositional assignment, the relations of an argument to its predicate. In

English. for example. the grammatical role of Goal is marked by the preposition tu, Beneficiaries

byfor. Agents with by in passive constntaions etc.. Processes such as passivuation affect the

marking of grammatical rules but not their semantic relation to the predicate (the Agent is s d l

the instigator of the action even if it is marked with bp in the passive construction rather than

appearing in sentence initial position).

Metaiinguistic verbs in GC such as km 'curse", kanl "cd", tank 'talk", tel 'teil: gyaf

'chatn. ail take an agentive NP in initial position in active sentences. This is the notional role of

speaker. Thev differ signincantiy in how they characterize the notiond role of addressee or co-

participant. Gyaf typicaliy codes this NP as an Agent using the comitative case-making

preposition wid: ii a gyaf wid a man de 'He's gyaffin with that man."' Tel, ka1 and kos aii appear

'~owever for the purposes of the argument a t hand 1 will employ minimal inventory of roles aithough further investigation may well reveal that further distinctions are necessary. One m u t first of al1 make a distinction between grammatical relations. grammatical roles and notionai roles. Grammatical relations are in many similar to the familiar notions of traditional grarnrnar Subject, Object. Indirect Object e t c They refer to formally marked categories that play important functions in the syntactic processes which are at work in a given language. The marking of grammatid relations is held constant across syntactic processes such as passivization while the meaning changes: i-e. the A of passive and the A of an active construction are formally identical but their grammatical roles (Agent and Patient) difler. Notiond Roles are the parts played by (actual) partiapants in acts denoted by a given predicate. While grammatical roles form a finite set. notionai roles do not since for any given predicate we can identify a spedic, con textual role - ie, for speak - speaker, for taik - talker etc. Notional Roles follow from the inherent lexical semantics of the predicate whereas grammatical roles follow from lexico-grarnrnatical structure. Finally, grarnrnatical roles (in generative theories of grarnmar, sometimes referred to as theta roles) characterize the relation of an argument to its predicate.

9 This grammaticai role has been received attention in discussion of split conjunct and is taken as marking a NP moved from conjoîned subject position: mi an ii a gyaf, mi a gyal wid am.

to code the addressee in a patient iike position but this is somewhat misleading. While kos

obligatorily codes addressee as an unmarked bare NP (Le. direct ob ject) . hl and tel d o w

marking with the Locative preposition pan.

10. 'ikospanmi

1 1. i ka1 pan mi, 'He called me."

12. wel wen sornbadii kom tel lai pan mü. 'Well when somebody cornes and tells

Lies about me. "

Pan is used in CC where action is directed to a surface as in mi dm gu an p h pan da pev '1 used

to go and play on the grave" (Rickford l987-b: 147, Line 346). mi fain am pan a terbl "1 found it

on the table". huu klaim pan am 'Who dtnbed on it?". Such evidence indicates that GC makes

at least a three wav distinction of grammatical relations which code addressee in metalinguistic

constructions:

Table 4.9 LEICICO-GRAMMATICAL CODING OF ADDRESSEE

Gyaf codes Addressee in a M y agentive (cornitative) case whereas kos codes Addressee as a

patient. In the middle are the locative markings assoaated with pan and tur<. Note the apparent

anomalous coding of a Patient-like addressee with uaf as in:

L3. i gyaf op mii,"He chatted me up/ He fooled me"

14. i g.f mii. 'He fmled me."

Such usages in fact suppon the argument that bare NPs are best considered Patients. In cases

such as 13 and l4 the gyafis a kind of manipdative tdk through which the addressee is uicked

into doing something ( s d intercourse, lending some item, handing over money) by

persuasive and flattering words. In such cases the addressee is more directiy affected by the

action denoted by the predicate. MetAinguistic predicates can thus be differentiated in terms of

the way they characterize the role of addressee. Some predicates. which take a patient-like

direct object. characterize the addressee as relatively passive and as something acted upon by

the participant coded as Agent. Others characterize the addressee as more agentive and the

speech event as genuinely reciprocai. I t is likely that ail languages make such distinctions-

Speech communities ciiffer. however, in the way they evaluate each end of this agent-to-patient

continuum and in the meanings attached to particular points. As 1 have noted some genres of

speech are considered parriculariy 'ferninine" in Guyana. Some too are more likely to be used

to describe male advity. A case in point i-s gyaf which, although frequently used to desaibe

interaction between women, is typically associated with exclusively male domains Iüce the

nimshop. There is a degree of contestation here, no doubt This said, it is safe to say that,

comparing patientive km and agentive uaf. native speakers tend to associate the former with

women and the later with men. The description that inheres in the semantic domain of

metalinguistic vocabularv is thus not neutrai. As I show in the following chapter it has wide-

D iangîng consequences for the interpretation of ta& and the everyday lives of native spèaken.

4.6 CONCLUSIONS

In this chapter I have uied to giw an account of the way CC furictiom as a pragmatic

system. I have suggested that the linguistic variation studied by Bickerton ( 1973 1, 1973b.

1 9 75) and Riddord ( I 986b, 1 98 7) is parricularly important for rnarking social relations and

statuses as well as for framing interactive contexts. In constructing texts, and the contexts in

which they occur, speakers also rely on the use of temporal and spatial deictics and well as

various affective kevs. Such devices indiate what relation obtains between participants and a

message and. at the same tirne, help to f m e speech as an instance of a parti& type. Native

categones which typologize t d k into discrete units are revealed in metalinguistic ternUn01ogy

which itself has some unique grammatical and lexical properties as I have described. In the

foilowing chapten, I look at the contextuaiized use of fornial variants and grammaticalized

pragmatic devices in select conversational interactions. This dows m e to further the anaiysis of

particular features sketched here and, at the same t h e , detail their role in the social and

interactional processes which reproduce and contest the dominant ideas about communitv.

CHAPTIERRVE

TAKING NATIVE-SPEAKER AWARENESS SERIOUSLY: MORPHOLOGICAL

VARIATION AND GENDER

5.0 -LE

In this chapter 1 continue the discussion of linguistic variation and gender which 1 have

briefly initiated at several points in the last chapter. I End that the situation resists simple

explanation and forces us. instead, to recognize the way in whicli language and gender are

indirectly and constitutively linked. In conclusion, 1 look at gender and linguistic variation in

relation to the nature of patriîocal group authori~. The explanation for the cornplex

phenornena of gender-based variabüicy. I çuggest, requires attention to the major focus points

in the earlier discussion of the relatively egalitafian community. Speafically, it is necessarv to

ewmine the way in which the patriiocal Nstem of kinship, marriage and domestic organization

differentiates men and women at a general level and at the same t tne dows for a certain

degree of variabili~ which 1 have doçumented in chapter 2. Accordingly. in rhis chapter. I

attend to the particulan of individual positioning within social stmcture while recognizing the

enduring and widespread effects of a sex-gender system which positions men and women quite

differentlv with regard to symbolic and socio-economic resources.

Although Engüsh-leuined, aeole-speaking communities in the Caribbean have received

a great deal of interest fiom soaohguists, verv Little attention has been paid to gender as a

correlate of linguis tic variation. Sophisticated models of gammatical description have been

developed in order to account for the high degree of variation found in these communifes

charaaerized as "creole continuaw (Bickerton 1973a, 1973b, 1975. Rickford 1979. 1987b) and

a vital debate has emerged concerning the social implications of this variation fiom the

perspective of a rather highiy i d e a h d (Le. genderless) speaker capable of movîng 'up" and

'down' the continuum and possibly in other directions as weU (Edwards 1983, Riddord l986b.

1 98 7a. I 98 7b). Still, although researchers have been vey sensitive to the possibSty. even

probability, of multidimensional variation (Le Page and Tabouret-Kek 1985, Rickford 198 Tb,

Gairet 1994). few if any have considercd the use of Linguistic variants as constitutive markers of

gender. In some ways. then, the situation contrats markedly with sotiolinguistic research done

in communities outside the Caribbean over the past thirty years. From its inception,

soaolinguistic research in North Arnerica and Great British has taken linguistic variation which

correlates with "sex" Merences as a primary topic of study (Cheshire 198 1, Labov 197 2,

l98Oa. 1990, Tmdgill 1 W2a. lW2b). However, and as the switch in temiinology is meant to

indicate, this research has bv and large been based on a mode1 whidi sees linguistic variation as

a direct correlate of sex differences (cf. critique in Eckert 1 98 9). Only recently have researchers

begun to question the notion of a direct indexicd ünk between linguistic variation and sex

(Eckert 1989, Eckert and McConneU-Ginet 1992, Friedrich 1989, Gai 1978, 1989, Ochs

1992). In a recent review of the fiterature Eckert wrote that:

Gender differences are exceedingly complex, particular1y in a soaety and era when women have been moving sekowciously into the marketplace and calling traditionai gender roies into question. Gender d e s and ideologies create different ways for men and women to experience iife. d t u r e and society.( ...) Despite inaeasligiy complex data on sex differences in variation, there remains a tendency to seek a single social construction of sex that will explain aii of its correlations with variation. This is reflected in a single coeffiaent for sex effects in variable rule or regression analyses of variation. ntis perspective limitr the kind ofnsults that can be obtained, s ine it ir mtntncted tu

confnning the implüit hpothais ofa single tgpe o f s a I e c t m, worse, tu indicatirtg that thm h no gect at aU. ( Edcert 1 989: 247. Emphasis added.)

One h d s . on reviewing the sparse üterature on gender related variation in Caribbean creole-

speaking comrnunities, that researdiers have often substituted sex for gender and have not

found significant correlations. In the words of Ridcford 'basilectal pronoun usage in Cane Walk

is suongly correlated with social dass but is aimost completely unaMcted by sex membership"

( 199 1 :6 12). Similarly. in the same volume. Escure ( 199 1 :6O4) remarks of the Belizean

community she studied thad

Most of the findings emerging from this study suggest that there is no consistent difference between women's and men's speech pattern: women do not overwhelmhgly use more prestige variants than men; men do not espeaally favour vemacular forms; men and women do not systematicaily differ in the quantiv of speech they produce. Such results are indeed interesring in that they reflect a soaety in which gender roles are less polarized than, Say, in a white middle dass context

Escure's suggestion that a lack of linguistic differentiation correlates. in Belize, with

concomitant lack of polarization in gender roles is questionable on a nurnber of grounds. First,

there is no independent reason for supposing that social role differentiation necessarily Ieads to

Linguistic differentiation (cf. schieffelin 1987). Secondly. the idea that gender ciifferences

should be consistent necessarily leads to the assumption of a one-to-one relation between

Linguistic f o m and soaal identity when, in faa. it is possible for gender based ciifferences to

show up at many different linguistic levels or not at all. Recent work in ethnopphy.

conversation analysis and social theory suggests that gender difference is complex and

multidimensional, and we might expen ünguistic variation to show the same complexiv

1 Gcure ( 199 1:602-604) argues. however. that a certain goup of women (the middle aged group) are developing what she cdls a "bipofar" repertoire which reflects rheir role as brokers between multiple levels and multiple d u e s of the society.

despite the faa that it may not "mirror" role differences at aii times. in ali contexts. in ail

places (Thome 1990. Goodwin 1990. McEUinnv 1995). I t certainiy seems unlikely that

gender roles wiil 'translate" oniy into t h e variables which can be mapped ont0 a simple

standard to non-standard continuum. Third, once we see the ways in wfüdi the very language

use under investigation is constitutive of the gender differences being reported it is a apparent

that there is a kind of circulanty to daims which see direct relations between them (Eckert and

McConnell-Ginet 1992).

In the following I employ ethnographie and statistical models to investigate the ways in

which language use is indirectS, and constitutively related to gender differentiation in d

Guyana. The analysis focuses on the various meanings assdated with variants in a verbal

repertoire and the wav in which these meanings in turn become linked, through various forms

of social practice. to local understandings of men and women as sociaily located actors. I argue

that gender effects have been missed or misinterpreted largelv because the notion of

metalinguistic awareness has not been thoroughly theorized in sociolinguistics. 1 compare the

analvses of Labov ( 1 9 72, 1 990) with Siiventein ( 1 98 L ) . concluding that gender effew cannot

be undentood without taking account of the fact that metahguistic awareness is conditioned

bv the referentiai and indexical attributes of the linguistic elements involved. attributes of

linguistic fom which are deteetable only if one attends to discourse level patteming.

Very Little work has been done on sex Merentiation in language in the Anglophone Caribbean. Most of the primary soaoiinguistic studies of the area. such as Winford ( 1972) or Edwards ( 19 75). confine their sample populations to men. whiie others, such as Akers ( 198 l ) , veat sex-based variation as inadental to other concerns. or rely on rather limited data.

This is a surprising situation given the general adcnowledgrnent in socioiinguistics that

gendedsex-based differences in language-use constitute one of the primary areas of variabiiity

and change in progress (Labov 1980a. TrudgiU 1972a, 1972b). It is undear to this point why

gender has been so radicdy under-theorized in creoüstics - especially since it has figured so

prominentiy in other kinds of soaal-stientific investigation in the Anglophone Caribbean ( i. e.

anthropology. sociology etc - for an overview cf Trouillot 1992). Whüe a few exceptions do

exîst (Escure 1 99 1. McWhorter 1997, Nichols 1983 ) . gender has generaiiy been written off as

insigntficant in accounts of variability within aeole-speaking communities - as evidenced in the

many studies which use, rather unapologeticayl, "male-only" sarnple populations (see the quote

from Winford ( 199 1a:S75) above).' Those studies which have not confhed the sample

populations to men have often reported no sigxuficant differences between men and women-

The classic study of sociohguistic variation in Guyana is Rickfords 1 9 79 the si^.^ Riddord

begins by shorvuig that with singuiar pronoun categories lumped together there seems to be

Little difference bemeen men and women of the same class:

* ~ c ~ l h i n n ~ ( L997) has pointed out that ernpirical gaps of this kind are often generated by androcentric definitions and andyticai categones. In this regard. the commonsense notion that it is the local. male solidarity networks that produce the most authentic creole has had an effect on methodology and research design. There is a certain scholarly "machisrno" operative in aeole studies which takes the imaginrd. "completely other", basilect as the only authentic object of neolistics. That this leads to an almost complete lack of attention to issues of gender is not surprising in that. in rnany cases. gender is expressed in tems of shifts of soaai and styiistic importance (thus compting the imagined homogeneous basilect).

'~ickerton 1973a. 1975 indudes no discussion of gender differentiation in his work

Within the EC, the women are siightly more basilectai than the men (-77 vs. -74). but onlv 4 bv - a very narrow margin. In generd. these statistics provide no support for the suggestion made in chapter four that differences in the socialization patterns of males and females rnight have their cordates in linguistic behaviour.

Riddord ( 19 79: 366) then goes on to break up the categories in order "to demonstrate that the

equivalence of the sexes shown ... is not merelv a function of aggregating the various

subcategories." Although RiMord does find sigtUscant degrees of variation between men and

women within some individual subcategories (to be discussed below) he condudes that

( 1 9 79:3 68) "it is this constant balanhg out of differences - where they do exist - which

produces the overd impression of equivalence between the sexes." Such an argument is based

on the assumption of equivalence between different subcategory variables within a single

grammatical ca tegory ( i.e. pronouns) . However, we may find that different subcategories and

the variants therein actually have different meanings for their users. Given this situation. we

might hvpothesize that men and women are using the creole-to-standard range differently.

iAh.ik maintainhg the idea of a single more-to-less creole dimension we want to investigate the

pecuüarities of each variable thus revealing the various indexicd meanings that are commonlv

a ttached to particular pronominal subcategories.

5.2 THE VARIABLES USED IN THIS STUDY

Pronouns in Guyanese Creole, as AUsopp ( 1958) Bickerton ( 1973a) and Riddord

( 1979) have illustrated, show a kind of robust variation dong a number of dimensions and this

makes them parùcularly well-suited to variation analysis. The focus on pronouns in studies of

vanabiüty in G.C. should not be taken, however, to indicate that other areas of the grammar

are more homogeneous - this is most certainiy not the case. Pronouns have been chosen here

155

because they are easiiy identifiable and analysabte dong both linguistic and s d dimensions

( "copulas", the other grammatical category desaibed in Bickerton's ( 1 9 73 a) classic study, as

Winford ( 1990) has shown. are a rather more problematic case).

Basilectal singuiar pronom contrast morphologicdy with corresponding mesolectal

and acrolectai ones in a nurnber of sub~ategories.~ This is iiiusuated in Table 1 :

SWmcT OBJILCT GmITmB 1st mi mi Basilect

ai mai Acrolect

Yu Basilect yorz Acrolect

am ii Basilect (hl ii Mesolect him hiz Acrolect

am ii/shi Basilect shii Mesolect hor horz Acrolect

ii Basilecr; its Acrolect

Table 5.1 MORPHOLOCICAL BASILECTAL, MESOLECTAL AND ACROLECTAL PRONOUN CONTRASTS

In actual usage. speakers in the sample did not approach the aaolectal end of the continuum in

trinomial subcategories. We are thus lefi with binomial variables in each category. We can also

exclude for our anaiysis those subcategones which show little or no variability despite the

potential for it. Cenitives show variation when the Guyanese population is considered as

4 1 have puposefdly exduded variants which differ at a phonologid Ievel since such cases require levels of explanation which are beyond the scope of this chapter. Rickford (1979, 198 1) gives comprehensive accounts of vowel laxing, h deletion. and t deletion in pronoun forms. Vowel laxing produces the aiternation between forms such as mi and mii, shi and shii.

totalitytY5 However the corpus uxd here is confined to relatively basilectal speakus many of

whom either do not command the acrolectal forms or do not use them with any great

fiequencv. For this reason. 1 have included genitives in the analysis only at select points since

variation is not tobust enough to @ve a detailed anaiysis of this category. We are left with the

variables isolated in table 2.:

3rd/ m a s c

fem

Basilect Acrolect

am Basilect (hl ii Meeolectal

am Basilect shii Mesolectal

Basilect Mesolectal

Table 5.2 MORPHOLOGICAL BASILECTAL AND RELATIVELY ACROLECTAL PRONOUN CONTRASTS ADWSTED TO INCLUDE ONLY RELEVANT CATEGORIES

The analysis that foilows thus focuses on variability in two subcategories; 1st person subjew

and 3rd person objects.

5.3 DIFFERENCES BETINEEN WOMEN AND MEN

.. If we take genitives, 3rd person ob jects and 1 st penon sub jecu together and look for

ciifferences between men and women we h d that men appear to be using slightly more

' There were a few exceptions to this d e . Genitives are induded only in the calcdations which asxss overdl variability for ail pronoun subcategories (Table 5.3) in part to make the results comparable to those reported in Rickford 1979.

basilectal variants than worned

Table 5.3 DIsWUTION OF BASILECTAL StJBfiCTs (IsT), Osncrs (3m) AND GENITIVES BY SEX OF SPEAKER'

men

women

However. if we break up the analysis into pronominal subcategories the resu1t.s are more

complicated. Thus we find that with regard to first penon subjects (altemation between aümi).

% of basiiectai variant

85

75

women appear more basilectal:

me s m p k for the present study consists of workers. a few teachers and shopkeepers (with

total

428

37 1

fairly equal distn bution or men and women). Each token was coded for the dass of the speaker (worker or non-worker). The effects for this parameter were of marginal importance with non-workers showing a slightly grcater trndency for mesolectal varia&. The worker-nonworker divide here is rather artificial. Unlike plantation communities in Guyana (Jayawardena 1963. Rickford 1979. 1986) the village in

varbru! weighting for d e application (basilectal variant)

0.568

0.422

question is not marked by an unambiguou and locally sdient dass division. Patterns of interaction are not so much conditioned by socio-economic dass in this context (in at least one case a shopkeeper and a worker were best friends). There is another problem with the application of straightfonvard dass- anaiysis in this context. As Nichols ( 1983) has noted. occupation (the primary index of socio-rconornic class in most studies) is aimost always aiready a gendered notion. (There may be some exceptions to this generally applicable statement. cf. Rickford 1986 on the "emic" status of dass distinctions used in that study). When soaolinguists w i g n dass groupings aaoss gender groups. for instance to a husband and wife, a potential arises for misidentification since it is often the c a ~ that work conditions differ considerably for rach gendcr (cf. Nichols 1983). For this reason. in order to explain some of the effects of class. age and sex interaction I have anaiysed patterns for individuals rather than as members of particdar categories.

w 8

7 Rrsults were generated using the Goldvarb program designed by David Sankoff. Factor weightings in cxcess of 0.500 indicate a positive effect on the realization of the variant. Significance is usually accepted for weightings of 0.550 or higher.

Table 5.4 DISTRIBUTION OF BASILECTAL 1 ST PERSON SUBJECTS BY SEX OF SPEAKER

This is consistent with the resultc reported in Rickford ( 1979). In the subcategory of objects.

again consistent with RiNord's ( L 979) analysis of C a n e men appear slightiy more

basilectal than women:

-

Table 5.5 DISTRIBUTION OF BASILECTAL RD PERSON OBJECTS BY SWC OF SPEAKER

men

women

The relativeh. minor effect of speaker's sex is magmfied if we consider oniy the most

stigmatized fonns in this group: the use of basilenal marker am for animate objects (cf.

Rickford 19 79) :

-

varbd weighting for d e application ( basilectal variant)

0.554

0.453

- - - - - -- -

men

women

basilectal variant Total I - --

varbrui weigh ring for rule application (basilectal variant)

r r

varbrui weighting for nile application (basilectal variant )

0.3 79

0.663

% of basiIectai variant

92

97

-- -- - -

%ofbasilectalvariant

58

48

women 37 158 0.413

Total

442

25 1

-

Total

178

203

Table 5.6 D I S T R I B ~ O N OF BASILECTAL RD PERSON O B ~ C T S FOR AMMATE REFERENTS O N ~ Y BY SEX OF SPEAKER

I t is w o d noting that the distribution of variants desmbed in this paper seems relatively stable

159

aaoss t h e and geographic location. Rickford's ( 1979) finduigs conceming the distribution of

varianu between male and female groups match. in many respects, those reported here despite

the fact that the data was collected over twenty years ago in a rather different community. In

both my corpus and that reported by Rickford (1 979:367) men show a higher fiequency of

acrolectai variants than wornen in the subcategory of 1st person subjea (ai). Similarly, Ridcford

reports the sarne complex patterns of objea variation as those discussed here (see below and

Riddord 1979r368).

Now supposing that basiiectal variants in each subcategory had similar pragmatic-

indexid values in this communi~ (Le. that saying mi as opposed to ai and saying an as

opposed to ii/shiüit amounted to essentiaily the same thing) we might agree with Rickford

( 19 79) that there is a levelling out of differences which gives the overall impression that gender

is relatively i n s i ~ c a n t . Alternatively, if we accept that different variables within the

pronominal categoiy may be associated with quite distinct indexid meanings we might still

attempt to explain gender differences according to a single coeffiaent for sex dong traditional

Labovian Lines. Thus, we might suppose that women are lagging in the use of variants that are

stigmatized within the Iarger community (stable sociolinguistic variables) while they lead in

changes that cars , local of ove* prestige (changes in progress). But this expianation (the classic

"gender patternn cf Fasold 1990, Labov 1990) is not appropriate either. W e am as 3rd penon

object might classify as a variant stigmatized in the larger community (see below). ai most

defhitelv carries local overt prestige as opposed to mi. Women's behaviour in this community

thus does not conform to what has traditionally, but conuoversially. been labelled the "gender

pattem" (cf aitique of 'gender pattem hypothesis" in Coates and Cameron 1989, Edcen 1989.

5.4 VARIATION WITHIN AND BEMEN CATEGORIES

Why do men and women differ in their use of these variable categones? There are a

number of possible explanations for the observed facts. We might, for instance, suppose that

one of the variables does not correlate with a unidimensional aeole-to-standard continuum,

This possibility has never, so far as 1 know, been w d to explain the kinds of gender-related

facts we are dealing with here. Howwer, the possibility that variation in "aeole continuan is not

resuicted to a single dimension has been suggested on several occasions. Washabaugh ( 1977).

for instance, suggested that variation in Providencia could not be expIained according to a

single dimension. Rather it was necessary to take account of a distinct casual-to-formai

dimension (a separate dimension of styiistic variation). Such a situation seems highly unlikeiv

given what we know about the relationship between social and s~l is t ic variation (Bell 1984)

and Bidcerton ( 1977a) argued convinchgiy that the variation discussed by Washbaugh could

in fact be explained in term of universal phonological processes. More convincing is the

description of several communities in Aftr of Idmtity . Le Page and Tabouret Keiier ( 1 985) argue

in that monograph that, within a single aeole-speaking community. speakers orient themselves

towards several distinct and often ethnicdy marked targets. While thiF is no doubt the case for

communities such as Belize and perhaps Guyana, such arguments do not, in and of themselves,

detract from the persuasiveness of a theory which modeis some variants dong a single more-to-

less creole dimension shaped by both social (Riddord 1979. 198 7b) and language universai

(Bickerton 1973a) factors. As such, we rnight argue that one of the pronominal variants we are

dealing with in the following (aümi. iVshii/it/am ) is actually a 'gender marker" and is off the

unidimensional continuum. Howwer, such an explanation is ultimatdy hadequate for several

reasons. First. as Ecken ( 1989) argues, there is no convincing evidence for a unique gender

marker in other speech communities (despite the fact that the 'iconic" values of parti&

variants such as "question intonation" might be indirealy correlated with gender - see also

Ochs 1 992). Second, and more importantiy, on the basis of native speaker inte~ews. it

appears that the aeole-to-standard dimension is the salient one for the evaluation of both

variables. In order to address this 1 s t kind of evidence, howwer we need to examine the nature

of metahguistic awareness and its treatment in the sociolinguistic iiterature..

5.4.1 METALINGUIS~C AWARENESS:

The problem of what different variants mean to and for their users within a community

has not been thoroughly theorized in variationist sociolinguistics (for an exception see

Lavendera 1 9 7 8). The question can, 1 think, be addressed as two separate but related problems:

awareness and local significance or meaning. This is simphr to say that in order for vaxiants to

have distinct and andyzable indexicai meanings (Le. indicating dass membership. region of

origin etc.) usen must be aware, to some degee, of the potentid for variation within a

partidar categoiy. The dassic statement on the question of awareness in variationist

sociolinguistics cornes from Labov ( 1972) whose variable phonoiogical d e s depended on

equivalence (Le 'different ways of saying the same thing" - cf. Lavendera 1978). More

precisely. phonological variables, of the sort Labov first discussed in Socioiinguistic Pattmrs, are

referentially equivdent but pragmaticdy differentiated. I t was native awareness of this latter

quaiity that Labov used as the basis on which to divide sociohguistic variables into thrre

Some linguistic features (which we will c d indiCaton) show a regular distribution over soao-economic. ethnic. or age groups, but are used by each individual in more or less the same way in any context If the soaal contexts concemed c m be ordered in some kind of hierarchy (like soao-economic or age-pups). these indicators can be said to be stmrified. More highly developed sotiolinguistic variables (which we will cal1 markm) not oniy show social distribution over socio-economic, ethnic. or age groups. but can be ordered dong a single dimension according to the amount of attention paid to speech. so that we have stylirtic as weU as social stratrfiurtwn (Labov l9?2:283).

The third o/pe of linguistic variable Labov c d s a stmeotype and desaibes as follows:

a number of sociolinguistic markers rise to oven soaal consciousness, and become strreotypes. There may or may not be a h d relation between such stereotypes and actual usage ... Most communities have local stereonfpes. such as ' Brooklynese' in New York City which focuses on ' thoity-thoid' for thirty-&rd; in Boston. the fronted broad a in ' cah' and ' pahk' receives a great deal of attention. Speakers of the isoiated Cape Hatteras (North Carolina) dialect are known as ' hoi toiders' because of the backing and rounding of the nudeus in hi& tide. etc ... Such social stereotypes yield a sketchy and unsystematic view of Enguistic structure to Say the l e s t In general. we can assert that overt social comection of speech is exrremely irregular, locusing on the most frequent lexical items, while the actual course of linguistic evolution, which has produced the marked form of these variables, is highiy systernatic (Labov 1972: 292).

The categorization is based. then. on a notion of differing levels of awareness (evidenced in the

effectr of attention paid to speech. i.e. style shifking). Working within a Labovüui model. we

could then distinguish each cype in texms of its position on a feature ma& of the following

sort:

Table 5.7 FEATURE MATRIX FOR INDICATORS, MARKERS AND STEREOTYPES.

indicators

markers

In dividuai Awareness

+

Consaous Social Evaluation

Labov's general theory of sound change in reiation to the s a of the speaker hinges on the

distinction between changes emanating fiom above and below consaousness (Labov IWO).

Thus. women. according to the traditional soaoluiguistic wisdom, favor the incorning prestige

form in cases of change from above and are innovators in change from below (Labov 1990:205-

220). In the first case. the change is most often explained in tems of women's greater

sensitivity to ovext prestige. a greater reliance on symboiic over materiai (political or economic)

power by women. or a tendency. on the part of men. to reject overt expressions of middle-dass

values in favor of working class notions of "mascuiinity" (cf Tmdgdi 1 9 72a). Labov ( 1 990)

interprets the tendency for women to innovate in the case of changes from below in tems of

the asymmeuy of child care respowibilities. Women. having more contact and more often

serving as models, tend to pass on the changes in whkh they are leading to the next generation

more effectivelv than do men.

To the extent that researchers have been attentive to this issue. they have generally

taken Labov's categorization as a s t h g point, with most work unquestioningly accepting that

markers are the variables best suited to soaolinguistic anaiysis (because they show regular

distribution along two dimensions not being affected by speakers' cowcious attitudes toward

speech varieties). The method has then paralleled in some respects, the exclusion of native

speaker int~iitions in Americanist and earlv anthropological linguistics and is potentialiv open to

the sanie critique leveled at those schools (see Silverstein 198 1, Woolard 1 992).

Under the assumption that native speakers are aware that variation is a possibiiiy for a

particular categoiy, and that the variants are accorded srylistic and social significance (Le. that

they are markers or stereotypes), the referential pedarities (lexical, morphological etc.) of the

164

variable are Mt out of the anaiysis (Lavendera 1978). In many cases, this seems to have been

the direct resdt of deaiing with phonologicai rather than syntactic or rnorphological variables.

Discerning the effects of the referential and indexical characteristics of a variable also requires

the analyst to examine variants in their iarger discourse context. However, vaxiationists have

rarely taken the discoune and interactionai functions of variants into account in their analyses

(see Edcert L 996 for an exception).

Rather than ewminuig the referential and pragmatic pedant ies of the variables in

question, most research. then. has àrcumvented the problem of what varianu mean tdfor their

users by positing some kind of standard-tenonstandard intexpetive frame through which

native speaken assess the social and stvlistic significance of particular variants. Induded here

are oppositions beoveen creole and English. basilect and acrolect. vemadar and formal

registea. One confusing aspect of ail these oppositions is the degree to which they are informed

by native (Le. locallv meaningfùi) categories versus categones invented bv the analyst Most

studies do not directly address the degree to which the analytic categories used are part of

native speaker metaiinguistic awareness or discourse (for an exception see Hiii 1987). However.

in order to make a convincing argument for change or variation as a resdt of "stigmatizationn

or "prestigen it is necessary to show that native speakers share, to some extent, the analvric

categories adopted by the andyst.

In fact, Indo-Guyanese creole-speakers do tend to think of variation as aligned dong a

creole-to-standard continuum (cf. Rickford 1 98 7b, see also chapter 3). As I have discussed in

the previous chapter, the polar exuemes are referred to as raw talk (/ras ta&) or Cmlesc

( /kr i ioW) on the one hand and deep EngM (/&p ingiish/) on the other. Native speaken often

refer to more mesolectal varieties as mir-up taLt (/&-op ta*) or as brokm-down language

(/brooken dong Iangwij/). M e n queried as to why the mole end of the continuum should be

cded raw most informanu responded by insisting that it m gat nw inghh in i t '1 t doesn' t have

any English in it." So it would appear that native-speakers do, at least in some cases. evduate

speech production in tems of a unidimensionai continuum.

In the case of the variables discussed here, there is Little doubt that villages evaluate the

variants according to a creole-to-standard dimension and that they do not dirrctly equate either

variable category with men or women. Evidence for the salien. of the creole to standard

continuum cornes from native speaker intuitions about the two poles of linguistic variation in

their community. The variation between ai and mi is thus often exploited in cases where one

penon mimics or mocks another. the use of the acrolectal variants being strongly associated

with acting üke an English duck. Variation in third person objem is evaluated in similar terms.

In this case it is the basilectal marker that is marked. As Bickenon ( 1973a) and RTckford

( 1979) have noted, am usage is associated with country-origins, and East Indian e t h n i c i ~ .

Bickerton ( l973a:659) writes:

[...] nrn is d e n t because it is the only basilectai fom in the system which is not perceived to be present in the acrolectal y t e m (allowing for phonetic altemation[...]). I t therefore becomes stigmatized: it alwavs occurs, for instance, in imitations bv Ahicans of d Indians, even though many of the latter have wholly or panîally abandoned it.

In interviews, people often gBre examples of raw talt which contain an. As one of the few

elements which is not a part of the Engiish system am has become, perhaps, the salient marker

* of the creole varie.. In Labovian terms. it is a stereocvpe. Such observations by native-speakers . . indicate that both variables are evaluated in terms of an ernic aeole-to-English interpretive

166

frame. We thus do not need to postdate the existence of distinct local, and non-referential

indefical meanings. The explanation for differences in usage is rather to be found in the

panicularïties of each variable a t the referential and pragmatic lwel. If we look dosely at the

way these items function in discourse it is possible to account for differences in usage by men

and women in terms of the categories themsehres and in terms of male and female strategies for

text-construction. Such an explanation is based on the reaiïzation that the variables are caught

up in a number of huictionai subsystems simdtaneously. Thus. they are locatable in terms of a

standard to non-standard continuum and, at the same time. function as referential items in

actual utterances. This leads us to adopt a rather different perspective on metalinguistic

awareness from that proposed by Labov, one whïch takes into account the suggestions made by

Siiverstein ( 19 8 1 :2):

For the native speaker, the ease or difficuity of acwate metapragmatic characterization of the use of the foms of his [sic] own language seems to depend on certain general semiotic propenies of the use in question. That is. the basic evidence we have for awareness of the pragmatic dimension of language use. susceptibiiity to consaous native testimony, is univendy bounded by certain characteristics of the form and contextuaily-dependent funaion of the pragmatic markers in speech.

While the three fanon Silverstein isolated in that stud? do not reved signifiant dimensions

of contrast for the variables discwed here, his suggestion that metalinguistic awareness is

"universaiiy bounded by certain characteristics of the form" is, nevenheless, a powerful and

useW supplement to Labov's argument conceming the influence of regularity, systematicitv

and ubiquity.

Traditional linguistic terrninology, which lumps into a single category 1st. 2nd and

ilv verste in ( 198 1 ) isolates the following factors: unavoidable referentiality. continuous segmentability, relative presuppositional quality.

167

penon refening forms, conce* an important difference between deictic and nondeictic ldcai

items. As has been pointed out many Urnes before (Jakobson 1 WO[ 19571). so called 1st and

2nd person 'pronouns" differ from the 3rd peaon forms in so far as they make reference to an

entity (usually a person) only by indexhg a shifüng role in the speech event 3rd penon

referring forms do not on the other hand rely on such speech event variables in making

referentid sense. The implication is that the use of 1 st and 2nd person forms (and their

varianu) is more incielibly bound to the way in which social actors ocmpy speech event roles

whereas 3rd person forms are more bound to the way in which speakers make reference to

entities 'outside" the speech event As it turns out, this difference has effects on the wav in

which speech is gendered in this speech cornmunity.

5.5 SUBJECTS - THE PRAGMAnC VALUES OF 1 ST PERSON SUBJECTS

Because fùrt person reference is necessdy self-referential (referring, that is. to the

s hifting role of speaker), a a o l e d variants take on special. or foregrounded. meanings in this

context. Spe&caUy, the prestige meanings associated with acrolectal variants (ai) are

predicated of the speaker in this case. Ai usage thus involves not oniy an assertion of acrolectai

cornpetence, as do other elements of the ideaüzed aaolectal code. but sirnultaneouslv a

foregrounding of the assumed identitv of the speaker.

The explanation for clifferences here, then, mav involve the way in which men and

women differ in interactional style generally - speàficaliy the way they assume positions as

authoritative speakers. For men the use of ai is associated with the presentation of self as

respectable and removed from the rural or working-class lifestyle. Often this ocntrj in cases

where the speaker is attempting to manipulate a hearer who might be sensitive to such foms of

social distinction. The manipulation depends on assuming a position of respectability. Consider

the following (eliated) example of a prayer to god.

NS : i f m i gu sit dong in mi alta nou -mi gu see mi noo fu taak tuu gad

if I go an rit down in my a h r nmo - I will say I kmw how to talk tu god

- mi (mol s kom mi (ma) s see oo gad r i -(ch) yuu protek mii, ai waa

- I have to corne I mort say god I want you to pmtect me, I want

yuu giv mii h e l t an strengt, ai w r ~ yuu protek mi hous, protek m i i

you to giue me heulth und sfengïh, I wantyou to pmtcet my houce, pmtect my

p i k n i i i den, yuu noo? dern kaina wee da.

chiIdm you know? that kind of th hg

Women also use ai with manipulative predicates in request-type acts. Consider the foilowing

example in which a aunt is requesting her nephew (3 years) to t a k

SS : ri ~ r ~ t yuu taak yuu m o s taak le mi h i i r hou i a taak

I wnnt you to ta& you most talk l a m e hem how he taih

Men also characteristicaliy use ai when calling-off wornen on the road. Usually the woman is

someone not completely farniliar with the speaker. The following is a reported example:

TS : wel yu gu star prooch shi- yu noo- kyeer wan- kyeer i n g l i s h langwidj

well you wiil start to appmach her - you know? - use a - ure the rnglish laquage

t u u s h i - see heloo. r i lov yuu brmbii w a t s op- yu noo- ai lrik di

with hm - say "hello I love you baby what's up? -you know? I iike the

wee yuu wak-

way p u walk

The particular effect of the aaolectal pronom is, in this case, linked to its pragmatic and

referential value. Because it is necessarily self-referential the assurned acroleaal. cosmopolitan

identitv of speaker is foregrounded. Feliatous usages of ai index authority. respectabiliy and

middle-dass values genedy . EspenayI important for men, 1 believe, is the way in which such

uses are associated with ideaIized middle-dass seegender relations based in romantic loveW9

Women generally avoid the use of ai because such presentations of self are, for them, more

Likely to backfire. Women wül often be called on their uses, and an audience may contest the

assumption of such an obviously non-rtiraI identiq. In one case that 1 recorded in my field

notes, a young women who was taking basic secretarial lessons in the viüage called out to a

friend aaoss the fence:

girl : hai darling, ri goin in, ri gon sii yu l e e t a

hi darling l'mgoing in, l will see you la*

A group of young men overheard the cal1 and proceeded to modc the girl. The boys d k d out

Irai hai ai goin in ai goin in fo leson ai komin out k m ni, hi I'rn going in I'm going in for my lessons, Im

coming out Inter. MM~ impression is that women, as opposed to men, are much more likely to be

calied on such usages if people suspect that they might be assuming a role beyond their countq

on*. Similarly. women are more iikely to be labeiled biggiy in 'talking-name" sessions both

bv men and other women. Although men are also subjeaed to this ridicule it is less k l y to

have damning effects for their reputation - and may go on for some t h e before they even

become aware of it. Furthemore, young men often use ai in situations where the key is fairly

playful. When calling off a girl on the road with ai l o v p '1 love you," a young man cannot

e-xpecr to be taken seriously. There are too many CO-occurring signs (such as dress, the location

of the interaction) which indicate that he is not what he is pretending to be and that his

'The idea of romantic "loven betwecn a man and woman - a kind of equal partnership based on mutual respect - is not part of the dominant village and aeole idiom. More often gender relations between married aduit men and women are conceived of as a power struggle and a hierarchid ordering (the "proper" hierarchy is to have the man as boss but this is frequently not the case - men who cannot control their wives are the butt of a great many jokes between other men).

intentions are les than completely sincere. In this sense the usage is perhaps playîüily

manipulative - the challenge is to see if one can tri& the girl or woman into thinking that the

speaker is other than a local. working-dass male and, of course, this is rarely successful. Women

ivho use ai are more ükely than men to be interpreted as self-elevating - hearers will often

assume that these women actually believe themselves to be a member of the dass who

habituallv use ai.

5 -6 THIRD PERSON OB= VARIATION:

As alreadv noted. women. as compared to men. appear to be more basiiectai in their use

of first person subjects. whereas, in the subcategory of third person objects, they appear more

mesolectal. Looking carefullv at the variants involved, the third peaon objects present us with a

rather complicated problem. As the basileaal pronominal system gives way to more mesolectal

patterns. case distinctions are lost and animacy and gender distinctions are added (Bickerton

19 73a. Rickford L 9 79). In using the mesolectal terms. then, speakers must asses the referent's

relative animacv and its gender (for animates). Not surprisingly then. one hds that both the

gender and the relative animacv of the referent has an effect on the frequencv of variant

pronouns usages.

50 the relative position of each variant on a more-to-less aeole dimension is tied to the

kinds of sernantic and referential distinctions introduced. Thus am is most basilectal and is

most infrequent when it is used as a referring term for ferninine referents (see below). Again the

peculiarities of the variable seem to have an effect on the way it is interpreted in terms of more- e . .

to-less creole intrrpretive h m e . Specificallv, there are more and less favoured environrnents for

the use of am and these depend eruciaiiy on characteristics of the referent

171

As it turns out. although we h d men and women using am at ody marginay. different

fiequenties (Table 5.5). women overwheimingly avoid using am in its most creole and

stigmatized conte- (for animate referents. Table 5.6). Another way to put this is to sav that

women lead in the change by which am becomes hctionally speciaiized for reference to

inanimate referents. The changes in the category of third person object involve the introduction

of important semantic distinctions (cf Bickerton 1 9 75). Here the salient hinaional dimension

is referential rather than pragmatic (the focus is on non-participants to the speech event rather

than on the values and qualities assoaated with any of the penons occupying participant

roles) .

Speaking of the variable use of am, Ridcford ( 1979:359-360) remarks that there is

some indication that it is acquiruig (or has acquired) a certain degree of specialization as an appropnate form for neuter (or non-human) objects ... am is still a more stigmatized form than any of the corresponding subject or possessive foxms in the basiiect. but speakers seem to operate with a d e that says: if vou are going to use it. use it more often for neuter objects than anv other.

The data here confhms such a tendency to reserve am for inanimate referents. This is dear from

the results of a binary cociing:

varbrul weighting for d e application (basileactal variant)

% basilectai variant

inanimate 75 93 0.722

Total

Table 5.8 EFFECTS OF AMMACY ON THE VARIABïLïïY OF RD PERSON OBJECTS. (46 of am usage)

However the effect of animacy of the referent is rather more complicated than this

binaxv distinction suggesu since it interacts with gender-of-referent effects. In fact. if one codes

for the gender of the referent according to a tripartite scheme whidi distinguishes male, femaie

and neuter entities one fin& that neuter (or inanimate) and male referents overwheimingky

favor basilectal marking but femaie referents do not

Table 5.9 EFFECTS OF REFERENT GENDER ON THE VAlUABIUTY OF RD PERSON OBJECTS.

referen t

fernale

male

neuter

It is apparent that female referents favor die use of a mesolectal variant (shiilii) while male and

inanimate favor marking with basilectal am. What accounts for this strong dispreference for

% of basilectal variant

17

80

75

basilectal variants in the feminine subcategones? One possible explanation is that native

speakers are more aware of the non-standardness or creole nature of am when it is used for

Totd

147

110

IO9

female referents. Nthough he does not say this explicitly we rnight expect this to be the

va. brui wieghting for d e application (basilectal variant)

O. 155

0.78 1

0.73 1 -

position taken by Bidcerton. According to Bickerton change within the aeole continuum is

the universal prinuples of partial seIectiun and h t gort (cf 1973a: 644). App

these principles Bickerton suggesu that change proceeds in an orderly fashion and he attempts

to show that diachronie change and synchronic variation are nvo sides of the same coin: the

implicational relationships of synchronic variation mirroring the way in whidi change has

spread thughout a subsystem like pronouns. He argues that the basiiectal grammar has the

foollowing pronominal forms in the third person:

STEP I subject genitive ob ject

masc. fem. neut- masc- fem. masc. fem. neut.

i z z i i am am am

and that change occurs first in the ferninine subject, @ h g the foilowing:

STEP 2 subject genitive ob ject

masc. fem. neut, masc. fem. masc, fem. neut

i shz i i i am am am

The n e z mesolectal form introduced is the neuter object it:

STEP 3 sub ject genitive

masc. fern. neut. masc. fem,

shi i 1 z z

ob ject

masc. fem. neut.

am am it

B ickerton then suggests:

At this stage gender differentiation is interrupted, as the change process 'recognizes' the salience of am ... the prinaple of Ieast effort ensures (a) that replacement will precede rather than accompany gender differentiation. and (b) that a form in the system will constitute the replacement. The svstem is now:

STEP 4 subject genitive objen

masc, fem. neut. masc. fem. masc, fem. neut,

shi i i 1 I z 1 it

Gender differentiation of the objen is then completed by the tramference of another rxisting form from the subject class:

STEP 5 sub ject g e ~ t i v e object

masc, fem. neut. masc. fem. masc. fem. neut.

1 shi i i i i shi it

The suength of Bickerton's analysis is in its apparent parsirnony and the overall neatness with

which it seemingly accounts for a wide range of facts. However, in this case at least. his

explanation is neither completely convinkg nor comprehensive. Two questions remain

unanswered Fint, no explanation is given for why gender differentiation should proceed by the

introduction of feminine rather than mascuüne fums (step 2). Secondly. it is unclear why

animacy distinctions precede gender distinctions in the subcategory of objects. More

problematic, 1 think, is the statisticai evidence which both Riddord ( 1979) and the present

study bring to bear. In fact the ordering of changes shouid, on the basis of statistical

disuibutions. go in just the opposite direction from what Bidcerton suggests given the retatively

infrequent occurrence of ii and it in discourse. l0 In step three, Bickerton suggests that ii and it

actuaiiv precede the introduction of shii. Given statisticai evidence which has shown a very

strong dispreference for basiiectal variants in the €exninine subcategories, this situation seems

RicMord ( 1979). discussing the problem of gender differentiation in third person

subjects (and suggesting that the same reasoning might be appiied to the objects), attempts to

explain the dispreference of basilectai variants in feminine subcategones:

we c m infer that native speakers avoid the genderless foms not only because they are obviously non-standard. but also because they pose potential problems of reference

1 O Bickerton's implicationai s d e s are d e with "nile conflicts" and "deviances" in this area (cf Bickerton 1973x660-66 1). See Pavaone ( 1980) for a discussion some problerns with implicational scaling.

and understanding of a particularly acute type [...] if one speaker's hi is always masculine whiie the other's may be either masculine or feminine. the potentiul for getting characters mked up is vexy great ( 1979:36 1).

According to Riddord. then. referential ambiguiy is resolved by a generai statistical tendencv

such that, in otherwise unmarked contexts, am refers to male and inanimate referents whereas

shii is the unique referential tenn for feminine referents. WMe this seems to fi t the obsenred

facu of statistical distribution, it does not explain the preference for mesolectal marking in the

feminine subcategory." The referential ambiguity could just as effitiendy be resolved by a

statistical tendency to take am as referring to feminine and neutedinanimates and ii as referring

to male referents. In such a case we would expect to h d the distributions reversed with male

referents suongly favoring mesolectal marking over feminine and inanimate ones. So Riddord's

suggestion that the dispreference for basiiectai variants in the feminine subcategories is the

result of pressures to be referentially unambiguous does not explain the original selection of the

femuiine subcategow as the one to be differentiated.

Perhaps, then. the explanation shouid be sought in the interaction of anima. and

gender effects and sex of speaker effects. We have aiready seen that the variation is affected bv

the animacv of the referent and that women are leading the change by which a m acquires a

specialized h c t i o n as an appropriate referring rem for inanhates only (Tables 5.8 and 5.9).

Women avoid the use of nm when referring to animate referents. At the same time women and

men favor mesolectal marking in the ferninine subcategory (Table 5.9). This cornplex pattern of

sex-based variability seems resüient to simple explanation. However we might suppose that

"~ickford ( L 979:359) himself remarks that he is unsure whether his suggestions should be taken as an explanation.

women have set a community nom here. Women. more than men. tend to rcserve am for

inanimates and prefer gender-marking pronouns for animates. If women Iead in this general

process it is not too hard to imagine that they set a pattern in whidi mesolectal marking is

preferred for ferninine subcategones in panicular.

Some support for this suggestion cornes from the way in which men and women M e r in

their use of am and the mesolectal variants. Although basileaal variants are strongly

dispreferred for both men and women in the feminine subcategories, men are more likely than

women to use ii for male referents while women are more likefy than men to use shii for fernale

referents. Such patterns seem to indicate an m a of contestation between men and women.

When men talk about other men they use the tem marked for gender and animacy more often

than women do. When women tak about other women they use a term marked for gender and

animacv more often than men do.

neu t u

F i p r e 5.1 EFFECTS OF THE GENDER OF REFERENT FOR MW AND FEMALE SPEAKERS FE RD PERSON OBJECTS).~ - -

p p p p p p p p p - - - - - - - - - - -

Thus. figure S. 1 shows that the influence of gender and anirnacy of referent effecu is partiaiiv

çoiiapsed for men as opposed to women with differences between the categones being relativelv

conuacted for the male group. For men. the change seems to be between more and less creole

(or basiiectal) ways of saying the same thing. This is to Say that. for men. the aitemation

benveen am and its mesolectal variants is d e t e h e d more by the social. than the referential.

significance attached to each varîant. As noted above. and in much other work on the problem.

am is suongly involved in the expression of working dass solidarity and local values. Social and

stvlistic factors no doubt affect women's usage of the variants too. However. for women

another, referential, dimension becornes ~aninilarly salient In avoiding the most stignatized

uses of am in reference to animates an association between this form and inanimates is set,

Furthemore, women show a suong preference for mesolectai. that is expliat gender and

animacv. marking for members of their own sex-based category. This a m to highlight or

foregound the animacy of the referent in discourse. I t appears then that women and man are

using the variabüity of the pronominal system quite differently. For women, more than men.

characteristics of the referent, in addition to the socia i and styiistic factors, play a major role in

determining patterns of variabiliv. l2

The foliowing is an example showhg the general preference for mesolectai marking for

ferninine referents and basüectal rnarking for neutedinanimate referents. S is complaining to

her sister-in-law (Sh) that a young girl who Lnres next door was dimbing on her fowl pen to pick

plums and subsequentlv broke it but would not tell her mother (Miss):

S . : s h i a pik p l o m . wen m i aks .hi yestodee see s h i na tel mii se shii

she wnr picking plum. Wken I &d heryestmlay she didn't tell me that she

brook rnr dong sh i na ansa -- shi kla im pon di fou1 pen an brook am

bmke it I w n she - she didn 't answer -- she cùiimed m the fou1 pen and bmke it down

dong- .shi aks mi at hou (inaudible) den mi se shi na b i g iinof fu

down.. she ask me if() t h I arked ifshe wasn't big enougft to

tel Mis da shi klaim pon-

tell hIiss that she had climbed on

Sh: -00 yea

-oh yeah

'* In accounting for such a situtaion we might suppose that either. men are not attending to the relative animacy of the referent, or. that they are purposefuily characterizing al1 referents unremarkable with regard CO animacy.

S.: - am brook am- mi na gu t e l ihii

- uhm bmkr it. I'M not going to ~ 1 1 hm

koz s h i gu koe. shi don bina kos dis maarnin mi en noo fo w a

because she will m m She WU already cicning thir momingfor I dm't h o w what

The next example illustrates the alternation between the mesoleaal term for ferninine referents

(shü) and basilectal term for masculine ones (am). Kavita (Kav) is t e h g me how she, her

mother. and her aunt had been playing with a litde giri Mando (3 years) who was also present

at the tirne this report was @en. During the course of the teiiing, Kavita's brother Kumar (Ku)

asks Mando to repeat her performance hom the day before. Kavita tells Mando to curse Kumar

(Kavita's brother). switching to the basilectal pronoun (am) in douig so:13

K a v :

Ku :

Kav :

yestodee Mamii te l antii Beebii fu te l m h i i toch m i i an shi toch 6hi

yesteràay Marna told Aunty Baby tu tell h a to 'touch me" and she touched hm

so an mi te l ihi toch m i so an shi toch m i abii tel t e l .hi don no0

like th3 and I told her 'touch me" like this and she touched me we told to tell hm uright-onn

hiir ahii mii don noo claughs>

this ir what she said - Iright-on"

toch m i de Mandoo

touch me there Mando

sok yu teet pon am kos am tel i le i gu beed i d o t i skin

s z~kyour teeth on him - turne him - tell him to go a wash his d i 9 skin!

S . 7 I NTERPRETATIONS OF COMPLEX GENDER-BASED PATTERNS OF VAR.IABIllTY

To summarize the argument to this point, 1 have demowtrated that the two variables

" ~ t is likely that am in this example is rnarking affect. The interactionai routine which the children refer to here as roch mi and don noo are phatic and themselves affect marked. When one friend says to another tuch mi the two make fisü and touch lightly. The significance of the gesture is highly contextuai but usuaily indicates agreement or approval of something said immediately prior to the routine being initiated. The expression don nm (lit. "known") often accompanies such a gesture. It also marks agreement. approval andor interactional alliance.

180

disnissed show significant but seemingly inconsistent effects for the sex of the speaker, arguig

that these inconsistenaes are not due to the effects of a mdtidimensionai continuum (i.e.

neither of these variables directiy index gender or any other soaai category) . Instead they tend

to be evaluated bv native speakers in terms of their position on a single aeole-to-standard

continuum. To explain differences in their use by men and women I have suggested that it is

necessaxy to look at the intemal stmcuire of each variable, pdcuLarfy the way in which

speakers reach metalinguistic awareness through an understanding of pragmatic and referential

qualities. In the case of a h i altemation, 1 have argued that differences between men and

women are a result of different te-building strategies. Specifically, because these are deictic

forms, indexing speech act variables, they have rather spedic indexid entailments: ai usage

foregrounds an assumed ident&y for the speaker - one that is relatively more difficult for

women, as opposed to men, to sustain. In the case of altemation between am and its mesolectal

variants, 1 have suggested that women show a rather different kind of sensitivitv to the

referential or semantic distinctions involved. in addition to social and stviistic factors. Patterns

of use with ferninine, masculine and neuter referents (figure 1 .) indicate that men and women

emplov rather different suategies in marking animacy through the use of pronominal variants.

Having discussed the importance of the referential and pragnatic pecuiiarities of each variable 1

now consider the general implications of sex differentiation in relation to recent work on the

relationship benveen speakex's gender and linguistic variation.

In the situation I have reviewed here it is not possible to characterize women as either

sigdicantiy more standard or more aeole than men. Both of these explanation, however, 0 . .

might have been offered given either emic or etic intuitions. In the former case. dthough in

181

manv cases people refuse to even postdate. native speakers wül sometimes remark that women

are more basilectal (more aeole-speaking) than men. A variety of reasons are suggested for whv

this is the case ranging kom the moralistic insistence on the effects of teenage pregnancy and

rernoval from the educationai svstem to normative ideas about the dominance of male

members in rural Guyanese households.

On the other hand it might be sunnised that women are more sensitive to overt prestige

and men to covea forms especially @en the importance of male solidarity networks and aews

which has been weii documented (Abrahams 1983, Wilson 1969, 1995[1973], chapter 1).

Wilson ( 1 969). for example, argued that men in Providenaa weie much more oriented towards

Local reputation as opposed to supra-local values of respectabiüty. Men's networks tend to be

more embedded in informal assoaations between villagers rather than in the institutional

structures of church and school. Although Besson ( 1993) has criticized this work for its one-

sided interpretation of institutions iike the diurdi in the Caribbean, few ethnographers have

challenged Wilson's centrai argument conceming the pivotai role of idormal and local

associations in the organization of men's daily activities in the Canbbean. Though women are

involved in local social networks such associations tend to show less cornmitment to the covert

values attached to speaking ueole. Creole speech, associated as it is with unruly behavior

(Abrahams 1983) and ce- kinds of peer relations, seems to be more strongly tied to the

male networks than to woments.

Since, however, both explanatiow posit d o m effem for the sex of the speaker

neither is sustainable given the facts of variant distribution (cf. Ecken 1989). This leads me to

suggest an altemate explanation. one that I think explains the distxibution of variants as a

182

resuit of sotial and cultural forces at work in the community. As I have argued earlier. women

seem to be more open to charges of acting biggity and therefore are more Iücelv to avoid forms

such as ai which foreground the speaker. At the same time women are also hequently

aeoie speech is often taken as a marker of such personai qualities.14 Women are thus like1y to

avoid a m and other highly stigmatized markers. For men the extreme basileaal end of the

continuum is used a sign of solidarity and degiance to an oppositional value system (women

can also participate in this but face the possibility of misimrpretation by others) (cf. Riddord

Traugott 1985). But it should be noted that for men the real svmbolic payoff cornes from an

abiütv for 'bilectaüsm", that is a display of cornpetence in both basilectal and mesolectal or

acrolectal varieties (see chapter 4). For the competent bilectal speaker, aeole (basilectal) usages

(what community members recognize as real raw talk) are interpreted as strategic choices rather

the necessaxy expression of a la& of education and resdtant iüiteracy.

This becomes clear once we look at the patterns for individual speakers. In general.

speakers can be grouped. on the basis of their usage of both variables. into one of four

categories: acrolectal, mesolectal. basüectal and rnixed code usen.

ACROLECTAL: 5% or higher rate of non-basiled usage in subject 50% or higher rate of non-basilectal usage in ob ject

MESOLECTAL: 5% or lower rate or rate of non-basilectal usage in subject 50% or higher rate of non-basiiectai usage in ob ject

BASILECTAL: 5% or lower rate of non-basilectal usage in subject 50% or lower rate of non-basilectal usage in object

''of course al1 of these are false attributions. In general. and in this sample. women in the village have as much forma1 education as men.

MnaD-CODE: 5% or higher rate of non-basilectal usage in subject 50% or lower rate of non-basiled usage in object

Whüe the dividing points seem somewhat arbitrary, the M o n of speakers in this way into

relatively focwed ( three categones: acroleaal, rnesolectal. basilectal) and relatively difhised

(one categorv: mixed code users) groups does seem to correlate with saiient soaal categories in

the village.

ACROLECTAL Moses Gigi

MESOLEC~AL Shanka Ri ta Zinii Kavi ta Kurnar

MED CODE WcWe Kota Ashan Cobin Nancv

AGE

- 24 32

41 38 35 9 7

30 33

17 47 41 34 37 -

OCCUPATiON

teacher worker

worker/housewife teacher shopkeeper none none

SUBJECE3 % of non- basilectal variant

75 15

housewi fe housewife

Table 5.10 CATEGORIZATION OF SPEAKERS ACCORDING TO FREQUENCY OF NON-BASILECTAL VARXANT IN 'IWO SUBCATEGORIES'~

OBlEcrS % of non- basilectal variant

67 63

O 4 O O O

worker baker shopkeeper wor ker worker

Notice that aii the women. with one exception, fd into the focussed categones dong with the

82 67 56 70 70

O 3

"I have included only those speakers for whom a sufficient number of tokens was available to get a sense of their patterns as individuais rather than as member of sut-based groups.

17 45

40 18 14 8 6

50 50 44 40 33

chüdren. We should note the apparent anomaiy of the female worker Gigi falüng uito the

acrolectd focussed group. Gigi is well uaveled and in fact currently lives outside the community

in question. Her husband is kquentiy cded away to work for extended periods and at this

time Gigi traveis both to town and to visit her family who reside in the village. Her rather

acrolectal patterns reflect this extensive travel, especdy her famihrïty with the capital which

is unmatched by anyone else discussed here with the possible exception of Ashan. As an aside,

it is instructive to note that villagers who had known Gigi since she was a s m d girl often

suggested that she had forgotten where she came from and who she really was. Moses. the

other member of the this aaolectal categoxy. however, is highly respected and considered to

exceptionally intelligent by many of bis fellow viiiagen.

The basilectal users - Marna and Seeta - are both married and living with their

husband's patdocal group. AS such. they face the most severe restrictions on their movement

(patrilocal group 'honoi being partiaiiy dependent on thek "reptation". see chapter 1).

Another set of women are in the mesolectal group. Obvious soaal differences distinguish these

women from those in the basilectal group. First. mesolectai speakers Rita and Zinü are

emploved and thus not restncted to the house in the same way as other women (i.e. Mama and

Seeta). Secondly, neither lives with their husband's patrilocal group. The other mature woman

in this group. Shanka, is not permanently employed. However. she no longer Lives wirh her

husband and maintains a good deal of penonai autonomy with respect to her patdocal group.

Uniike other poor and working dass women in the cornmunity, Shanka is free, in many

contexts, from the surveillance of men and the soda1 sanctions which can be imposed by a

patriiocal group.

185

In conuast d the miwed-code users. with one exception. are male. I t is revealing to note

that the one exception is an unxnanied shopkeeper who is the undisputed head of her family of

four resident sisten. their mother and one child (see mv desaiption of the Paknr in chapter 2).

(Note that men in the village frequently commented on ways in whidi Ashan was

"unwomanly" and, to them. unattractive sexdiy) . She has a wide range of contacts and has

uaveled extensively (to England and also aii over Guyana). Kota, a baker, who seils his product

right up and down the road from Rosignol to Mahaicony is also weU traveled both l o d y (with

his evexyday work) and region* (hawig worked in Brazil. Suriname and Trinidad).

To some extent then language use seerns to correlate with actuai and potentiai

opportunity for travel. However such correlations breakdown when we consider the bu& of the

mixed code users - the male workers. With little experience of any other place than the village

and its surrounding environs. their usage does not seem to correlate with extensive uavel.

Although not restricted to the home as are their wiws. they have neither the resources nor the

opportunity for extensive travel. What unites them with the other mixed code users is their

position within their own nudear kin-groups and domestic units. Ali of these people are the

generaliv acknowledged pubiic heads for their domestic unit They thus have the author i~. at

least in some contexts. necessary for the felicitous use of ai. They &O have local standing or

reputation enough for the feliatous use of am. Even the youngest member of thh group

( WeWe). who is ody 17. is the on* emploved member of his nudear famiiy which includes his

grandmother and two of his cousins. With the exception of the auolectal group members. ail

the focussed users, including the chiidren in the mesolectal group. are relatively disempowered

wvithin the context of theK own families in cornparison with the memben of the mixed-code

While I do not want to overstate the dominance of men in the symbolic marketplace.

since many women do use both the basilectai and reiatively acrolectal varieties in some

contexts. statistical analysis does seem to indicate a situation in which women are more tightly

consorauied in their use of the symbolic resources potentiaily available. As such. the situation

resembles that described by Hill ( 1987) for Nahuatl speaking women. Hill ( 1987: 158) remarks

of the Nahuatl women in her sample:

I t seems likely that women may experience active exclusion from male patterns and that thîs is the reason they are at once both less Mexicano and l e s Spanish than men[ ...] I might propose that women in Mexicano communities are. in fact, marginal to major social arenas dominated by men: the system of community offices. with which power coding is associated. and the system of compadrazgo and male hiendships with which the solidarity code is associated[ ...] They (women) are dearly sensitive both to stigmatizatons that emanate from the n o m of speaking casteiiano and to stigmatizations that emanate from solidarity-code purism about speaking mexicano [...] Rather than think of the speech norms of women as marginal to a core of male norms. we might instead think of women's speech as highly constrained within a narrotv range of possibilities, at the same time less Mexicano and less Spanish than men's speech. whereas men are able to use the full range of code variation.

On the other hand. the situation 1 discussed here contrasts rather sharply with that for Detroit

High School students desmbed by Eckert ( 1989). Edcert ( 1989:256) argued that 'deprived of

power. women must satisfy themselves with status. 1 t wouid be more appropriate to Say that

women are more status-bound than men." She goes on to suggest that ( 1989:256): "(w)hereas

men can jus* and define their status on the basis of their accomplishments. possessions. or

institutional status. women must jus* and define theirs on the basis of their overd

character." According to Eckert ( L989:256). then, "symbolic capital is the o d y kind that

187

women accumulate without impunitv." The evidence 1 have discwed here. particdariy in the

light of findings by both Hiil ( 198 7) and Ecken ( 1 989). thus points to some interesting points

of cross-cultural variabiiicy:

1. The distinction b e ~ e e n doing and speaking, between power and prestige, between

power and the symbolism of power (symbolic capital) is a cultural one and is thus

susceptible to a variety of local interpretations. We would expea that some

communities will be more attentive to the action in talking and wiU therefore not show

the same kinds of gender Merences in this regard as that described for Detroit

teenagers.

2. The distinction between action and talk is iikely to collapse in communities which

have experienced large-sale disenfranchisement. In a corn- where mqbody is

relatively disempowered vis-a-vis a larger metxopolitan center or colonial meuopole.

symbolic distinction, instantiated in part through h g u a g e use, wi l i take on

increased sipficance and may, in fact. be dominated or jealously guarded bv men. In

such a situation symbolic resources cannor be accumulated with impunity.

I 1 have argued that the soaolinguistic situation in t h i s rural Indo-Guyanese comrnuniy

should force us to re-examine previous sociolinguistic descriptions of Anglophone Caribbean

Creole communities and explanations of gender-based ciifference in generai. The situation

described, in terms of the statisticai distribution of variant foms across male and female

groups. is remarkably simiiar to the one described by Rickford ( 1979). This leads me to

condude that we are deaüng with fairly stable patterns probably common in many

predominantly East Indian communities in xurai Guyana. 1 have attempted to explain the

ciifferences in terms of enduring gender d e s and the partidarities of each variable. Once the

partïcuiars of the variables themsehres are taken into account, I think it is possible to see a

generd pattern emerging; men (particuiarIy a certain subset of working-ùass men) tend to use a

geater range of the continuum with regard to pronouns. Women on the other hand seem to be

more restricted to a focussed area of the continuum. At the sarne t h e , women and men differ

in their use of the foms themsehres. In the case of 1st person subjects this entails a favoring of

the basilectal variant whereas in the case of the 3rd person objects it entails that the change

they lead in (towards a speaaiization of am for inanimates only) takes a rather definite

direction (i.e. a preference for mesolectal marking in the ferninine subcategory) . In both cases

women and men show a very complex (vet sometimes quite different) kind of variabiiity that

indicates a great sensitivity to the values of partidar referents dong a number of different

sotial, referential and pragmatic dimensions.

The situation 1 have discussed is significantly different from that described for Detroit

teenagers by Eckert ( 1989). In that investigation speakers, both male and fernale. seem to have

fairly equal access to the symbolic resources embedded in language. This situation no doubt

reflects the common ideologicai separation of 'practical" and linguistic or symboiic action in

many western speech communities (the distinction between language and action. power and

prestige etc. - cf. Rumsey 1 990). In the community discussed here language and linguistic s g e s

seem to rnuch more dosely associated with 'real" power and are thus more jealously guarded.

In this case 'impunity" is no more guaranteed in the symbolic r e a h than in the econornic or

pofitid ones. Given this situation. women restnct their speech pductiom to a focwed

variesr in the continuum. presumably one they feel most suits theK standing in the local

community. Men. on the other hand, are more iike to range over several styles, drawing on the

symbolic markers thus avaiiable. At the same. there is evidence that, whiie leading in the

change which brings in the less 'soiidarity" marked basilectal variants, and therefore perhaps in

some ways exduded h m expression of local identity. women are perpetuating a linguistic

categorization that foregrounds their own animacy in th& use of 3rd person object pronouns.

We should be aware. too, of the possibil$r that a restriction to a certain speech style.

indicated by a relatively nmow use of the variation available, may itself indicate a concem for

die accumulation of s p b o l i c capital. I t is not, in fact. possible to equate a gceater range of

linguistic variation directly with a greater concern for, or greater investment in, the vagaries of

the symbolic marketplace. In some situations. there may be a greater symbolic payoff in using a

nmow range. In faa, 1 have argued that women in the f i g e considered here ARE motivated.

in their Linguistic behavior. by a concem to present an appropnate image of themselves. This

motivates both their rejection of the most stigmatized variants and the most prestigious ones -

thev are walking the thin. and rektively safe. midde ground. This shodd not be taken to mean

that men are not concemed with the moral person they present through various symbolic

practices, induding speech production. They too are motivated by a concern to display

themselves symbolidy. In this case. however, i t is the men who have more invested in

displaying multiple group or commuity memberships through their use of linguistic vanants. 1

have suggested that this is in part the resdt of specific political and economic factors that have

affected rural people in the Anglophone Caribbean (Le. a situation of relative disempowement

vis-a-vis a metrople).

In the vülage there is a generallv held belief that women should spend more time than

men in home and in many ways their movement is resuicted. However to take this as the

reason for sociolinguistic Merences would be a mistake. La& of exposure m o t be used to

explain the dispreference for a variant Like ai which is widely known in this community.

Furthennore, to equate restrictions on movement with sociolinguistic patterns is to miss the

ethnographie generaüzation that women must be carefd both in how they move through the

cornmuni. (and who they interact with etc.) and how thqr talk - both concems emanating

kom a common communitybased constmction of guider d e s . As Eckert ( 1989) has reminded

us. ultirnatel~ it is power whidi motivates and constrains linguistic usage and in the case

discussed here this power takes very Local fonns. Wide ranges of linguistic usages. such as those

evidenced by the people 1 have labeled mixed-code users, require that the speaker occupy a

certain social position to be effective. The generaiiy obsewed pattern in which women are more

focwed on a particular variety reflects the exclusion, with some exceptions, of women korn

such positions of local, kinbased authoritv.

Plate L . Tvpical working class wooden house

Plate 2. The vard, the road and ovo roadside houses

Plate 4. Bottomhouse

Plate 3. The yard

Plate 5- Children by the road

Plate 6. Brothers of a single pamlocal group retuming f'rom reIigious work

Plate 7. Crandniother and qandson

C ~ R SIX

CUSSINC UP AND TALKING HARD: AFFECTZVE DISPIAYS O F U ~ ~ ~ ~ ~ ~ ~ " AS

CHALLENGES TO GENDERED TERRlTORIALITY

Coolie Mother

jasrnattie live in bnd- D o m hut big Iikr Buta shoe-box, Beat chthes, weedyard, chop wood,jêedfDwl For th& body and that body and e v ~ blartni body, Fetch watm, a11 dagJitdi watrr Iike i f t . wholr - Whole slow-pOwing Canje river God mate Jurt jhr she one own bu&

Till shefoot-bottom crack and she hand mt up And m e swann fmm she mouth likc nd-antr And she cou& bbod on the gmund but rnash it in: Because jCtsrnattie heart hard, she mind set hard.

To hurtlc rave she one-one slow penny, Because one-one dutty make dam m s s the Canje And she son Hanlall got to school in Georgetown, Mut Wear chan starch pants, or thcg go laugh at he, Stmp leuthm on he foot, and he m w t mad book, Leam ta& proper, take m m , go to England univmity, Nat trrm out like he mm-suckm chantar dadee.

David Dabydeen (p. 62 Book of Cunbbean P o c S )

"Polite language is a symbol of high status and ...p eople of high status and aspirants to ic refrain(ed) from adopting the customary mode of conducting disputes. Thus abuse and assault are characteristicallv low-status methods of disputhg (Jaywardena 1963:89) ."

In rhis chapter. t again take up the discussion of language and gender relations, this ùme

in an extended analysis of a partidar suetch of conflict talk In attempting to discern the

extent and the limits of patriiocal authority, as weU as the strategies women develop for

contesting it. 1 am again drawn to the investigation of spatial organization and bodily

cornportment I find that. in contesting the way in which domestic and social relations are

swctured in the village. women simultaneously question the gendering of space and bodily

cornportment. Performances of passion are discursive sites in which women are diowed certain

righu to speak. Occupying such a speaking position is itself, however. a risky undertaking since.

in such cases. the speaker is parriculady dependent on audience ratification

1 have suggested in chapter 4 that. when makirtg spatial descriptions, participants draw

on the background knowledge of a cornmonseme geography in giving shape to the interactional

and nlrated spaces designated in a convenationai sequence. One aspect of such conversational

place formulation that has not been extensively discwed in the iiterature is the role it plavs in

verbal conflict. Because the regdation of bodies in space is so intimately tied to the exercise of

social power in the village I am discussing and elsewhere (see chapter 2 and Duranti 1994.

1997. Foucault 1979, Keating 1994) place formulation in conversation often becomes the

object of heated dispute. Although some mention has been made of the wav taik and

participants are related to property spaces (Goodwin 1990. Levinson 1983. Schegloff 1972)

place formulations have generally not been theorized as sites of contestation in discourse-

However, the fact that participants work with an at least partidy shared cornmonsense

geography does not entail perfect agreement about the 'rightness" of paaicular place

formulating moves in conversation.

In the analysis that follows 1 attempt to danfy two aspects of the relation between

verbal dispute and place fonndation. First, because space is closeiy linked with social power.

place formulations are the site of a grcat deal of negotiation and contestation in discourse. As

such. I am concemed to show the way in which, while participants sometimes ratdjr one

another's place formulating moves, they do not always agree on particular characterizations of

locatable spaces. Instead. partiapants tend consme space strategically to suit th& own ends

and achîeve their own interactional goais. Second. spatial descriptions and place fomulations

are resources whidi members draw on for expressing certain kinds of social relationships. As

such. the analysis that foiIows pays dose attention to the role of place formulations in the

collaborative construction of oppositional frameworks (see Goodwin and Goodwin 1990).

In the particular transmpt that 1 present a woman engages in an affective display of

passion (/pashon/ "a state of uncontroiiable and intense emotional disquiet") during which she

contests the wavs space and gender are co-constructed. In her talk, she confronts her husband

and his mother and charges that they unduly restrict her movement But her t a k presents a

challenge bevond the referential content of her words. Through an increase in volume and a

shrill intonation - which can be heard from a great distance - she establishes a participation

framework which reaches beyond the delimited spaces of the house and yard (see chapter 1). In

ashg this register she brings into the interaction neighbours and b~standers who would

othenvüe be non-partiapants and, in this way. recontextuaiizes her own taik as public

discourse. She then uses spatial and persona1 deictics to realign these people into larger

participant structures and to anchor herself to a place removed from the house. The public and

perfonnative nature of the taUc is also cued through nurnerous affective keys induding

199

interrogatives, parallelism. fronting and lexical items locally considered nuse words. Her talk

then breaches, and thus challenges. the boundarïes that delimit the gendered spaces of the

house. vard and road. In the exchange the woman and her mother-in-law collaborate to

produce an oppositional alignment central to which are their conflicting constructions of the

space within which the interaction is taking place. In condusion, 1 discuss the problematic

outcome of the event For while the challenge was effective insofar as it publidy humiliated her

husband and e v e n t d y forced him to d o w her more freedom, it also inmased her

marginalkation within the comrnunity. When a woman engages in such hard t d k it is generallv

taken as a sign of her immodity and indecency. F d y and fiends comrnented afterwards

that cussing and fighting of this son was not proper and that it shamed the filIllillv with which it

was associated,

6.1.1 WOMEN AND DISPUTING IN THE GUYANESE CONTEXT

The interaction that 1 discuss is a dispute between two women. Given the gender of the

participants. and the potential for reproducing existing stereotvpes of men and women. it is

necessaxy to say a few words about the soaal foundations of interpersonal contlict in this

community before moving to an analysis of the transcript In Guyana. as in many other places

(see for example Kulidc 1992, 1994), women have a wideiy accepted reputation for being

argumentative. divisive and prone to verbal dispute. Such a stereotype exem considerable

influence both on native accounts of verbal interaction (Sidnell n.d.4) and scholarly writings

( for an example see Edwards 1978). Despite the tendency to psvchologize and gender

argument in this way (ueating it as the inevitable expression of an essential femininity), it is. in

200

fact, apparent that certain sotial forces tend to give shape both to the production of confIict

and local interpretations of its verbal manifestation. The first thing to note is the way in which

women, particularlv mamied and junior women (i.e. daughters) , often h d themsehes h g

under the authoritv of a senior man. As 1 have described in chapter 1, residence in this

communitv is primanhr patrilocal so that manied women tend to live with their husband's

families. Patrilocal units, consisting of several nudear families occupying a plot of pauilineally

inherited land. have a very public face. Others tend to judge an individual's actions in tenns of

the overall character of the group and this Ieads group members to vigorously defend their

reputation or name against scandal. For a number of reasons that cannot be adequately

described here, ' the result is a situation in which women, more than men, h d it necessarv to

cultivate moral penons that are 'respectable* and "proper". In the face of a very active gossip-

svstem women are hequendy called upon to publidy defend themselves and their actions. This

then leads women to engage in certain kinds of dispute which men generally avoid.'

The other major implication of the patriiocal soaal system which is relevant to what

foilows has to do with the organization of-space and the regdation of movement which 1 have

discussed in chapter 1. Within the vülage gossip-wstern. evidence of immoral or disrespectable

behaviour is often aeatively generated. Thus, seeing someone on the road and headed in a

certain direction is ofcen taken as an indication that that penon is making an iKat visit to a

lover. Cùcumstantial evidence of this kind can be very damaging to the personal reputation of

L . .. ce Wilson 1969. 1973 and Besson 1993.

e en do sornetirnrs engage in conflicts about "respectability" and moral character. 1 am making a statemént about a statistid tendency which has given rise to a leminization of verbal dispute in ideologicai terms.

20 1

both a woman and her husband- Husbands who are considered to have UnfaitMd W e s are the

objects of much derision. Added to this is the widely held belief among men thar women

cannot be uusted and must therefore be controlied by senior men (i.e. a father or a husband).

In an attempt to preserve and protect the3 own reputation and 'gwd name", then, men make

suenuous efforts to keep women at home under the watchful eye of their own pavilocal group.

Married women must always have a reason and an intended destination when going out

Policing the movement of bodies through space thus becomes a primary arena for the exercïse

of male authoricv and power. Women. however. do not always submit to the regdations of

senior men. Thev frequentiy devise reasow for leaving the house and engage in visiting on a

regular b a h despite their husbands', sometimes f i d y physical, reprisais. Women are often

forced to defend their rights to movement and their general personal autonomy against the

impositions of their affinal patrilocal group. This tension between wife and patriiocal group.

rxpressed in tems of rights to movement, thus also leads to a greater invohement bv women in

certain kinds of verbal dispute.

Girls are also more iikely than boys to have their rnarriages arranged by their parents according to the traditional Hindu custom, and keeping them closer to the home helps prevent dismption of such phns and to increase theïr atuactiveness as potentid maniage parmen. At Hindu wedding receptions, the women usuaily remain upstairs talking and fussing around the bride. whüe the men drink and dance among themselves downstairs ( Rickford I 9 79: 140- 14 1 ) .

As I have described in chapter 2, the spaces of the house, the yard and the road are very much

meanin@ to people in the village. Part of this meaning arises out of a gendering. But gender in

202

this and in other cases interacu with other locdy meaningfui categories. In particular, it is

apparent that spaces become associated with certain activities induding ways of speaking, ways

of holding the body, and ways of interacting with others more generally. This CO-construction is

pdcuiarly evident in the way the road and the house become sites of heterosenial exchange as

RicUord's commenu above suggest*

According to Wilson ( 1969, 1973 and Abrahams 1983) the road in the Anglophone

Caribbean is associated with male reputation for virility, peer p u p s , loose mords and sexuai

activiv. In the village these attitudes are expressed in ways of holding the body, and ways of

speaking on the road. Men usually sit or stand with some dear view of what is going by. They

may be drinking, playing car& or gyafing. When a o u n g woman passes, the young men Say

heloo lov! ai lovyu! 'helio love - 1 love vou" or waa gyal,yu shain op "what gyal! you dress up!"

Such remarks draw attention to the woman's appearance. and to the possibüity of heterosexual

relations. Women, on the other hand, do not usuaily stand by the road- Instead thev walk

along i t at a reasonable speed, u s d y with an umbrella or something to keep the sun off their

face and well covered by Sari or western style clothing.

If somebodv "big" addresses them politely the woman is evpected to caii back politely.

If a boy caiis them off thev may cut-ges and increase their p a ~ e . ~ Rarely do 'propernHindu or

Moslem women enter prolonged engagements with men in these situations - even if thev are on

hiendlv relations kvith them (the exception is when they have some joint business to attend to -

Glt-cyc involves a look to the face and then down and across the body. I t may be combined with suc&-&th or stchuups in which air is sucked into the mouth through the teeth and lip to produce a unrnistakeable ideophone. Both gestures are comrnonly raken to express disgust or disrespect. See Rickford and Rickford ( 1976).

203

school etc.) So the appropriate mponse would seem to depend on the age and status of the

man învolved. If it is a young boy who is mannlrlr ('full of himsell") the best thing to do is to

cut-rye and keep moving. If he is older and Weil respected then a polite 'thank-you d e " would

be more k l y . If the man is on roughly equai standing with the woman a quidc exchange may

ensue in which recognition is adcnowledged.

During an i n t e ~ e w which probed native speaker attitudes towards language use. I

asked the foilowing question:

1. an fp sii wan bai/&pal wayu laik pan di m d h o u p gu taak tu ram - inglish or moor

knioliiz? (if you see a boy/giri that you lüce on the road how would you talk to

himher? English or more aeolese?)

While men usuallv suggested that they would use more English in this situation, women, on the

whoie. began bv stating that the boy would have to tdk to them kt. One woman said:

2 . wel bai gn kom taak tu mi fors - if mi star taak I gu see - wa shi a p l è e , p ssii da gyal ga kiip

shi plees ( 'Well. the boy has to corne and talk to me first - if 1 start taking ( to him) he

will sav 'What is she up to? That gïrI, you see. must keep her place.")

Conversational sequencing in such situations can, in this way, be suongly gendered. There is a

4 The question then becomes what constitutes "proper" behaviour. A nurnber of women in the village do cultivate moral, public persons that do not fit the usual mode1 of "proper". Unfortunately discussion of the entire range of responses which are possible for women in Indo-Guyanese villages is beyond the scope of this chapter. This said i t should be noted that a woman's situatedness in these comrnunities is not simply a matter of Ianguage. ideology, economy or other structural factors but instead is determincd by a cornplex interaction of these factors as well as each woman's way of responding to them - based on the set of enduring dispositions which have developed through her life- course (Bourdieu 1977). In the lollowing. I locus on one woman in particdar whose plight and responsrs seem. in many ways, suggestive of a larger g o u p of Hindu women who live in the rural community where 1 conduct fielchvork and who are married to agrïcdtural labourers. But many other women in the village do not confonn to the image that Dabydeen (see the poem at the beginning of this chapter) draws at d l .

204

set of expectatiom which in practice are variably instituted. Most people atuibute su&

expectations to the "traditional" way. They may ate as evidence of the traditional way the

avoidance d e that. big bodii kyaan tuak tu lil boujii 'big brother should not talk to littie brother's

wife". The maditional way speafies appropriate modes of compomnent for women in pubüc

places such as the road. Although it exerts considerable influence it aLso contradicts, in many

ways. expectations and ideas about the nature of soaal relations in 'modem" or 'urban"

settings. Some women thus align themselves with 'modernity," opposing this to the backward

traditions of the village men and women.

If the road is auociated with illiat heterosexual relations, the house symbolizes the

sanctitv of the proper Hîndu marriage (cf. Smith and layawardena 1959 on Hindu marriage in

Guyana). The house most often has its origin with the nudear unit breaking off from the

extended group and sentiment reflects this. The greatest insult of the PNC regime from a male.

East Indian perspective was the way it rhreatened the sanctitv of the house and famdy. Black

policemen could. at any tirne. corne into one's house, take what they wanted and abuse one's

wvife.' The sacred 'Indianness" of the house and the heterosexual union is synbolized in the

house altar. Everv Hindu home 1 visited from the srnailest backdarn one-beboom to modem

roadside d w e h g s contained an a i t , at which prayea would be offered daily. These aitars are

made up of a shelf holding numerous religious pictures whidi depict Hindu deities (Krishna,

Latchmi, Canish. Anuman etc.) Also contained in the altar are vases and flowers dong with

' The PNC (People's National Congres) controlled local. regional and national governmrnt .. under the leadership of Burnharn and Hoyt up until 1992. The partywas marked by a predorninantly

Afro-Guyanese membership. I t h a been implicated in a number of rigged elections and crimes against the local population. Most Ekt Indians in Guyana considered the regime oppressive. (see Premdas 1995)

brass plates upon which offerings of bumùig incense, camphor. Ieaves and fresh flowers are

made. The care and use of this altar is the responsibility of ail family members and the practices

surrounding it. in the division of responsibiüw and distribution of authority, ritualize

heterosexual and family social relations.

"Traditional" wisdom is somewhat contraciictory, containing seemingiy incornmensurate

ideological positions on the gendering of talk For while there is certaidy an image of the docile

"traditional" wife, this is balanced by an image of the hard-takittg cussing woman. The hard-

caiking woman. as a stereotvpe. is often conmsted with the polite. soft-spoken yet confident-

self-assured middle dass, modem, wornen of Georgetown (and elsewhere cf. Edwards 1978 and

my conduchg chapter). ThCi is a pervasive stereotype and one that women themselves draw on

in dailv negotiations with others (as I show below). Note that the same woman who is quoted

above answered the question, huu a km mom man o r homan? ('Who a m e s more men or

women? ") quicklv with the following:

3. rvel nnchori1ii homan d m see hornnn strengt in d m mout sw d m doz kos moor - "weU

naturally women. They Say that a woman's strength is in her mouth so ihev curse

more.

Tlie self-assertiveness assoaated with curring (cf. Sicineil forthcoming) is at odds with the

stereowpe of woman as subordhate to men (espeaaily husbands). The act of m i n g also

involves a level of (public) performance rvhich is not traditiondy expected of women. Such

performances problematize women's relegation to the delirnited space of the house insofar as

they expand participation to include people located ouuide the space of the house and yard 0 . .

(i.e. the domestic group of the patdocal unit).

6.3 BACKGROUND TO THE T R A N S C m

Pria and Raja had established a smail roadside stand about four months before thc

events discwed here took place. At this stand. rather ironicdy named the "Queen Bee", they

sold saLreeoo with sauni (a local snack made with flour and water), biscuits, and cigarettes to

people passing by as weii as to young bovs and girls, adolescents and addts who were, for one

reason or another. on the road. Pria ran the Queen Bee as her own project - it had been her

idea, she made the salseeo, she kept the Little account book of debts. Most often she and one or

two of the chiidren would sit at the littie stand and serve anyone who was buying.

Most of the patrons were men between 15 and 25 years old who lived in the immediate

vicinity. They would corne to sit on the bench, gyaf, smoke agarettes and occasionaily to play

car&. The Queen Bee, located near a number of locally important homes, became a popuIar

spot to meet hiends and lime 'hang out". Often one found ten to fifteen people lined dong the

road. Manv of the men and boys were Raja's nephews who Lived in the same and neighbouring

var&. Another group of about 4-5 young men (ranghg kom 20 to 25 vears in age) also met

regularlv at the Queen Bee usually around six or seven o'dodc. They were better off

economicaliv than most - one owned a car, another had a stake in a large coconut estate.

another was a teacher. One of them, Lennie, whose father was the owner of various properties

including a hotel. became quite good hiends with Pria. They would sit next to each other and

gyaf and joke. Rumour that the two were carryïng on an iiiicit affair did not take long to spread.

Soon evervbodv seemed suspicious. Raja got word of this and became much more protecrive

and involved in the Queen Bee. He would corne out earlv, teil Pria to go inside the home and

seil the goods himself. In various way he let Lennie know that he was no longer welcorne. The

207

gossip became more intense however and when Raja went to his sister Gigi's for an ovemight

stay he returned to find Pria invohred in a new story. Apparently Pria had signalled from her

veranda ro Lennie who was riding his bike dong the road at dusk Raja's nephews had

apparentlv watched her as she indicated that Raja was out of the house and that he should

enter through the badcdam. After hearing this Raja starred drinking. Eady on in the evening he

and Pria began quarrehg. They were Ioud enough so that people at the roadside. about sucty-

seventv feet away, could hear quite clearly what they were saying. The fight ended with PM, bv

this Ume bleeding kom the top of her head, leaving to go to her father's house with her

children. That night Raja broke up the Queen Bee with great fdare. He then went into the

parure (the sea-side of the road - i.e. away h m his own house) and aied outside his esvanged

brother's house. The incident was greatiy ernbarrassing to Raja's pauilocal unit. The brothen

and their wives remarked that Pria's hard-ta& made the familv look low dass and d i q . One of

the \vives sent her husband - Raja's big brother - to ta& to Pria. She was told she m u t come

back home. and stop quarrehg, cussing and making public displays of this kind as it belittled

the family and was unfitting of their status. Although they didn't confront her diredy. the?

also told me that thev believed it was m e that she had been carrying on an affair with Lennie.

Ail the brothers and their wives agreed that the problem was with Pria's upbringing. She was

backdarn people, cane-cutting people, people who didn't know any better. Raja, who took his

suspicions of his wife's infideüty to be confimed. intensified his watch. He demanded that she

stav home. He demanded that she ask for permission to leave the house when she had

somewhere to go. When she stayed too long he would come and coilect her. Other members of

the extended patriiocal unit also intensified the watch. This induded A.K. Raja's mother, who

208

iived in the main house in the same vard with Raja and Pria. A widow, Kav enjoyed certain

kinds of freedoms not usudy associated with younger women. As a woman of 65 years she was

removed from the heterosexual market place and was therefore dowed to walk out when she

pleased. In fact she was known as a wdker. This caused some embarrassrnent to the farnilv too.

According to them she was often dxunk and bathed infrequentiy.

Just previous to the interaction rhat 1 present Pria6 had corne in from the road and was

situated near the gated enuance to the vard. Kay. Pria's mother-in-law. was standing in hont of

her house. As the interaction progressed. Pria moved to the back of the yard and into the smail

house that she shares with Raja. Raja entered the yard from the road just after Pria. walked to

the back. and retired in the bottom house hammodc. Kay remained in h n t of her house. These

spatial relations are s

Figure 6.1 PART~CIPANTS' MOVEMENTS THROUGH THE SPACES OF ROAD, YARD AND HOUSE.

Names have been changed to protect the anonyiuty of the participants.

The rnovements of the particpants help to differentiate major units of ta& in the overd

organization of the sequence (on the use of the body in differentiating talk units see C.

Godwin 1 984. on overall organization see Levinson 1 983:308-3 18). As will become dear in

the foilowing analysis. the interaction can be broken into four units . In the openhg sequence

(lines 1-9). Pria and Kay indicate their major concems and their reasons for engaging in the

dispute. For Pria, this has to do with the over-zealous poücing of her movements by Kay. Raja

and the rest of the pavilocal unit Kay's concems dearly revolve around the relations between

her house and Pna's nudear famih/. This section ends with Kay adopting the position which

Pria daims has driven her to engage in this confljict (speafically, she suggests that Pria is 'not

proper," that she is immoral). Kay thus offen Pria a conversational opening for the airing of her

concems (i.e. the opportuniSr for disagreement and direct confrontation). This section

correlates with Pria moving into the yard space from the road and taking her position facing

Kay. The second unit contains Pna's major daims regardhg the unwarranted actions of the

patrilocal group. This is an extended turn at talk which Pria produces facing her principle

addressee (Kav). In the third Mt. which is a transition point in the interaction. Pria makes a

convenational move that is remarkably sirnilar to Kafs earlier one. Thus. when no one replies

Co Pria's complaints, she changes the topic to that whidi Kay has introduced in the opening

sequence. She thus introduces her opponent's talk just as Kay did for her earlier. This section

ends with Pria moving badc to her own house. in the finai section. Kay makes her cornplaints

regarding the actions of Pria and Raja. By this time. Pria and Raja have retreated and Kay is

occupying the space to the side of her house. The combination of bodily movement through

space with precisely timed and formulated conversational moves thus differentiates the most

salient units in the overd organization of the interaction. For the analyst such coordinated

action is sviking evidence of the skillhil collaboration which goes into the production of conflict

tak and oppositional formatsi

6-3.1 THETRANSCrUPr

The manscript s t a r t s with Raja retuming fiom the road where he has been out looking

for Pria. In the tirne he has been out. Pria has returned from her father's house. Kay began

quarrelling upon her return. diarging her with infidelity. immorality, indecency and warning

her that Raja has been out looking for her and is now retuming.

(6.1)

G y = Kay (Pria's rnother-in-law. Raja's mother. Midiael's wife Pria= Pria (Kay's daughter-in-law, the in-marrying affine) Raja= Raja (the youngest son of Midiael) Shan =Shanka (Raja's sisiter, Kay's daughter)

2. Pria: a gud i kom a koata gu gatu tu tel am wan ting "its gooA he's c m c Kota wdl have ta tell him sornething"

'Although 1 raise the issue of collaboration later in the paper the topic desemes separatr and full-length treatment. A survey of the literature on discourse in the creole-speaking Caribbean reveals notions of "noise" (Abrahams 1983, Reisman 1970). "busin" (Edwards 1978) etc. ail pointing to the absence of cooperative collaboration in talk In Reisman's widely uted ( 1974) paper, he argues that Antiguan discourse is characterized by an enrerne individual assertiveness. on the one hand, and an almost pathologicd scanning, on the other. A number of scholars have suggested that such "Conventions" of disorder arise in opposition (or as resistance) to European or metropolitian. that is "polite". modes of discoursr. However. i t should be apparent that the dispiay of opposition, conflict or twen "disorder" in conversational interaction is aiways a product of the collaboration of the participants involved.

3. Ra j a:

4. Kay:

S. Pria:

6. Shan:

7. Kay:

9. Kay:

ho01 di batrn a di geet "hold rhc bottu?n of the gae"

aa- AIYU S K O N T NA .MAIN MI WAN DEE - AIYU NA MI GI NOTIN (0.5) GUD "d2 4ym cun ts hawen 't lmked afrrr me a singk day -you don 't give me anything

na gu fait huu na beta // man* "(1 don'tjght) one who is no bethor ahan mc"

// noo noo* fait yu tuu ( 1.5) udun'tfightyac two "

yu skont na praa ( )- insaid a hee (4.0) 3opr cunt (pu 're) nut proper - inside of h m "

yu skont (5.0) uvm j e u cuntn

As 1 turned on m y recorde?, Kay had just finished waming Pria that Raja was coming into the

vard h m the road (figure 6.1 ) .9 She also dairned that Pria's infidelities, wanderings and

constant quarrelling with Raja had shamed the good narne of her husband's family. She

concludes with a resounding and rhetoricai 'good" in line I ( i.e. '1 am glad you understand me

and r a t - what 1 am saving. '). Despite this apparent attempt to close the interaction early on

( cf. Schegloff and Sacks 1 9 73), the two quiddy establish an oppositionai format in the

following lines. In line 2, Pria uses third person pronouns (Le. non-participant deictics) to

creativelv establish a participation framework in which Kay, and not Raja, is the p M u p d

addressee. At the same tirne as this establishes a speaker-hearer dyad (Pria and Kay) it ~ O W S

8 1 had been recording that day at Raja's house. Raja's three children had been telling stories for me about a hdf-hour before this inàdent ocnured. My recorder was dose by so when things started to

get heated 1 gathered up rny bag quietly and went to the front of the yard. Everyone who was present knows me well and considered me, at the time. to be a member of Pria and Raja's family. Shanka Raja's sister. was one of my primary informants in the Reid.

9 Although, as an older widowed woman. Kay was ailowed privileges to speak, she was frrquently not believed. She had a reptation for talking name and people w d to Say. shii lm h( de in yong piiopl stmrii 'She loves to get herself in to young people's business.'

2 12

Pria to individuate herself in relation to Raja. But in line 4. Kay refuses to ra* this attempt at

individuation and instead refers to P h in the plural (azyu). Pria and Kay are thus atternpting to

establish different oppositional formats in participant structure:

PRLA'S OPPOS1710NAL FORMAT:

Pria -- vs -- Kay and Raja (The patrilocal unit)

&Y'S OPPOSITIONAL FORMAT:

Kay -- vs .--- Pria and Raja (The nudear family)

Table 6.1 CONFLICTING O P P O S ~ T ~ O N ~ FORMATS

In terms of the establishment of a participation framework then. opposition is aeated on at

least two different levels. At the first level of opposition, both Pria and Kay construct

opposition through the use of pronominal fums or partiapant deixis (see Table 6.1 ). The

second level of opposition is consmcted through the non-ratification, by JSay. of Pria's attempt

to individuate henelf (i.e. in conversational sequen~ing) .

The oppositional formats established in participant structure also have a spatial

dimension to them. In each case an individual is opposed to a goup of some importance to

local social orgarkation (the nuclear famüy and the patriiocal unit). These groups correlate

with the salient social spaces of the house and the yard as 1 have discwed in the previous

section. This initial spatial opposition reappears in a number of guises throughout the

interaction. particuiarly the last section in which Kay speaks (lines 20-24).

There is another level at which opposition is coiiaboratively produced in these first few

213

iines. In line 2. Pria d a i m s that Kota. the big brother in the patrilocal group. will have to 'teil"

Raja something. In Indo-Guyanese villages one common foxm of dispute resolution is giving

cornplaint. This involves an offended party asking a superior (a big brother. a father etc.) of the

offending party to use their authoriy to bring that offending party in h e . Pria is threatening

such an action in line 2. The threat is somewhat ambiguous however. A few days earlier. Kota

had been called in to taik to Kay about her behaviour. Although the two had reached an

agreement, Kay's behaviour in this interaction is quite obviously transgressive. Pria is thus

implying that Kota could aiso be brought into control Kay. ffiy is in fact attending to this

implication and, as I show below, responds to it in line 4. Kay's response is suuaured as a

parallel construction (Jakobson 1966):

aiyu skont na main mi wan dee 1 1

'All-pu haven't looked after me a single day' 1 1

ajru na mi gi notiît

'Ail-you don't give me a thing'

Kay here equates general care and concem expressed by the verb mind (/main/) with giving- The

verb mind implies not ody providing economically but also emotionaily. To Say that you mind

someone in this communin/ means that you are responsible for them. Adults mind chiidren.

Kav's paraiielism takes giving (money etc.) as an index of mùidùig. The absence of one entails

the absence of the other. This constitutes a response to Pria's threat insofar as rninding is a

relationship of mutual and asymmeuical obligation. Thus. someone who min& you is also

someone that vou owe somethllig to and whose authority you are obllged to acknowledge. If a

woman wishes to repudiate a man's control over her she will Say i na main mii 'He doesn't look

after or control me.' If a man says this about another man the imputed/disputed relation

2 14

implied is one between father and son. big-brother and Iittle-brother. Thus, Kay implies that

Pria's threat is empty. Since they do not mind her. they have no right to tell her what she can

and cannot do, where she can go, or how she must behave.

Retuming to our discussion of participant structure, notice the way in which the

oppositional formats deiineated above (Table 6.1) match the s o a a l relations Kay is drawing on

in iine 4. In each case. the oppositionai format opposes an individual to a supra-individual unit

which is more than the sum of its parts (Le. a family, a paari ld unit). Kay's remarks in line 4

draw on her construction of Pria and Raja as constituting a nudear farniiy which is a single

domestic unit, has its own intemal structure and has relations, as a single unit, with othen. The

implication here is not only that the two houses (that of Kay's deceased husband and the one

belonging to Pria and Raja) have completely independent domestic economies but also that

there are no relations of solidarity. obligation and respect based on asymmetric kinship

structures between them.1° If this were m e , it wodd mean a constant state of non-cooperation

between nvo households in the same yard.

Continuing the analysis of the oppositional format in the first few lines, we can note

that in iine 5 Pria explicitlv ratifies an oppositional alignrnent between henelf and Kav. This is

a complex move by Pria, for while she acknowledges the emcrgence of opposition (i.e. a

'fight"), she daims that she will not engage in such an activity. The reason given for this

purported non-engagement is (moral) equivalence between herself and Kay. Pria daims

specificallv that she d not fight with someone who is no better than herself. In fact, dus

statement. far h m removing her from the opposition. funhem her earlier threat to inform Kota

'OThis is m e insofar as villagers tend to take a very practical (as opposed to official) perspective on kin relations (the distinction is from Bourdieu 1977).

of Kav's behaviour. Pria's suggestion of an equivalence between herself and Kay, while

disguised as a withdrawal from conflict, points to the possibüities of a counter attack, one in

which Kav's own transgressions of patrilocal authority are held up to smtlliy. Kav maintains

the oppositional format in üne 7 by responding to Pria's assessment of equivalence with an

urunitigated and dispreferred second. Denying their sùnilanty, Kay daims that Pria is not

'proper". Not inconsequentiaily Kay ünks Pria's immoral behaviour to expectations for proper

conduct in spe&c places. She ends her assessment of Pria's behaviour with insuid hee 'inside

here' thus making the connection between place and ideas about cornportment expiicit There

is an ambiguitv here which, in this case, is probablv an artifact of the recording process as

much as anvthing else. It is unclear whether Kay means:

a. vou must be proper inside of here

b. you are not even proper here. inside the yard

In either case. however. Kay's remarks, given that the locally and contextuallv saiient space

which the speaker and addressee could be said to be 'inside" is the vard, serve to invoke

patrilocal authority and ' respectabili~" and. at the same tirne, to constme the interactivelv

negotiated boundaries of 'here" in terms of corporate group p ropeq (Le. the space of the

vard) .

Even in these few lines, then, the interlocutors have introduced discourse entities

(participants, social groups) which have readüv recognitable relations to spatial organization.

Note k t the nonobvious way in which Pria situates herseifwithin the yard space but at a

remove from Raja's s m d house at the back rhrough the use of deictic k m 'corne". Her spatial 9 , *.

placement. as she constructs it, thus supports both her daim to indMduation (because she

2 16

separates herself from the family's house) and the defence of her own actions (she is not now

on the road - she has returned) . When she says a gud i km.. she uses her own position in

interactive space as indexical origo (cf Fillmore 1973, Klein 1983). Her presupposed position in

relation to sociah meaningful spatiai arrangements of the houe and the yard thus becomes

part of the background knowledge on which she draws iater in the interaction. Kay is also

consuuing the spaces of the house and yard in a partidar way. As 1 have noted. she stresses

the independence of the two households. This analytid separation of the house from its

surroundings becomes the focus of her atm& iater in the exchange.

M e r behg m s e d by &y, Pria launches into a monologue of her own. This is delivered

in a high-pitched and markedly louder voice than her previous utterances. Most people 1 talked

to took this stretch as the pivotal moment in the exchange and suggested that thh was an

eauample of a person speaking out of passion. Pria uses a number of affect marking features to

index rhis emotional state induding an increased arnpiitude. higher pitch. curse words and

rhvthmic. pardel structures. Since people who get their passion are usually not held responsible

for the outcomes of their actions in the same way as those who are cool and caldating there

are good reasons why Pria would use such an affect-marked register (on affect marking see Ochs

and Schieffelin 1 98 9 ) .

Note that. in terms of the participant structure, Pria is again attempting to distance

herseif from her husband Raja in her use of a mi "In vs. ii "hen conuast( h e s 2, 10- 16). Kav.

likewise. again refuses to r a t . this individuation and iwtead addresses Pria in the plural (aiyu

and don) repeatedly (lines 4, also 20. 23, ). Kay is unwüling in this way to take Pria as a

participant on her own and aims her attack a t the nuclear unit Aiyu contrasts directly in the

paradigrnatic set of pronouns not only w i t h p 'vou' but also with abi 'a-of-US' and the

implication of a speaker plus others is suong here even though Kay uses mi 'me' in referring to

henelf. This then creates an opposition beween the nvo houes and between the patriiocal

unit - (abii) - and Raja and Pria - (aiyu). Pria then takes the flmc

(6.2) 10. Pria: (shi kyan gu out) an mi kyaan gu noowee (0.5)

"(she can go out) and I can 't go anyvohcre

1 I . ii mos tel mii wat a klak mi mos gu (huu) wat a klak mi rnos kom bak (0.7) he hm to dl me what time 1 rnostgo (hm) what time I mast.com buck

12. mi +an gu a rood noowe yu mos kom an luk €0 rnii an saarch fo mii mi no I can 'tgo on the road anywhmyou musc anne and lwkfor me and se& for me I don?

noo wa skont yu a iuk fo mii an saarch fo rnii mi na noo ( 1 .O) knmv what dufuakym Imk/or me and searchfir me I d m 't k n m

- Mt1 DON TEL YU IF MI WAN TEEK MAN f a YU kyaan stap mii (7.0) - I'vt dready mïd you - i f 1 want m have anothrr m m (rus) ym can't srop me

rnii a big homan €0 miself an mii mos noo rang an mii mos noo rait 1 nni my own aùult w m a n and I have to know what is wrong and knmv what ir right

so mi na frigin onastan wa mek ii ga saarch €0 mii (5 .0) h t ' t rtnderstand why he musc search fBr me

Pria's attack here is not ody dirrcted at Raja but at the pauilocai unit as weil. This is

particuiarly apparent in Pria's digrunent of the participants. Note that she refers to both Kav

and Raja using 3rd person. nonparticipant deictio (Kay in iine 10 and Raja in Line 1 I ). So to

whom is rhis segment addressed? Shanka is close by but the volume with which Pria deLivers

this sue tch of taik indicates that she wants her words to reach other ears. The audience, it

appean. is essentialiy anybody who wili k e n . %y this time a number of people were gathered

outside the fence which marks the extent of the yard. Her talk in this way breaches an

important boundarv which divides the space of the patrilocal group from that of the village as a

218

whole (ie. the yard fencing). At the end of line 1 1. having made her public announcement, Pria

takes Raja as her principal addressee. The content of her speech here focuses directly on the

social meanings of spaces and the movement of individuab within them. She uses modal kyan

'can' and buan 'can't' to diaracterize a realm in which social actors' use of, and movement

through, space is lirnited by conditions of possibilïty and obligation. '' Kay is allowed to walk

around but Pria's movement is restricted by Raja's supemision. This irreaiis world that Pria

consmcts reaeates the spatial arrangements of the house and the road. Thus Pria uses the

refening expression n mod ('on the road') dong with the deictic verbsgu 'go' and b m 'come' in

orienting participants to this hypothetical space modelied after the one they are inhabiting.

Note that the indexid ground on which these verbs rely for thek interpretation is not, in this

case. the position of the speaker at the t h e of speaking. Modals effect a transposition of this

ground and substitute for it a hypothetical one, thus detadiing the description of narrared

space from the interactional space wirhin which it is made. Although when Pria uses p 'go'

( üne 10.1 1 ) the irrealis and redis grounds match, when she uses kom 'come' the ground is

uansposed to somewhere outside the spaces of the house and yard. So Pria is not only creating

a hvpothe tical world in which she is situated outside these boundaries but is dso taking as the

ground for her utterance the place at which Raja will (or-does) find her. There is an ambiguitv

again here. Pria is daiming that Raja's behaviour is unwaffanted and shamefd because she onlv

socs to visit friends and family. At the same tirne she implies that this place where Raja wili

" 1 think the sense - epistemic or deontic - we give to these auxiliaries in this context depends on the interpretation vis-a-vis a surrounding co-text. Fria codd be indexing a world of impossibility (an spisternic interpretation) when she says mi kyaun gu nmwe if we take this as an outcorne of the social world (1 cari't go anywhere - 1 have too much work to do). The deontic meaning is perhaps more direct (1 can't go anywhere because other people stop me).

219

find her codd be the site of a uansgression of the heterosenial union. The latter interpretation

seems consistent with the next line in which Pria inaeases her volume to tell Raja that she can

take another man if she wants to and he can't stop her. But this threat is mitigated when she

claims status as a big-woman. Most importantly. Pria is arguing that she is a mature adult and

shodd be granted respect and rights to autonomy that go dong with su& status. She condudes

with a daim about the fundamental moral autonomy of the mature and capable person.12

Pna's remarks here are met with silence from the other participants despite the long

pauses in lines 13, 14, 16. These are obviously potential tum-transition points but Pna's

interlocutoa refuse to engage here and thus do not ratifjr what she is saying. This absence of a

reply helps make Pria Look unbalanced, irrational and perhaps &O passionnte insofar as the

silence is taken as an indication that Pria has failed to engage an addressee. Faltering in this

wav. Pria makes a rather surprising move in line 17. Now she uses the plural deictic in self-

reference aiigning herself with Raja and asserting the relevance of the nudear unit. She thus

now ratifies the participant framework that Kay had tried to establish in the first few lines.

Funhermore, Pria now brings the taik back to the topic that ffiy had originaiiy inuoduced -

that is, the economic relations between the two houses. In a sviking instance of collaboration

~vithin the midst of conflict this concession suddenly brings Kay back into the exchange. Kay

teasseru the oppositional structure in line 18 through an obviously aggressive interruption,

121t is interesting to note the altemation between modal kyu(a)n *can('t)' and modal mos 'must'. Whereas kyan is somewhat ambiguous between epistcmic and deontic meanings, mus is (in this case) unarnbiguously deontic predicating as i t does a domain of obligation. In line LO (ii nios...) the obligation is predicated of Raja - (Le. Raja is obliged by his own sense of male authority to corne look for her). This same source of obligation can be inferred from the instances of mm in lines 11 and 12. In line 15 however the source of obligation shifts from Raja's moral authori ty to a more general social ethic. In line 15 Pria is evoking a world of rights and obligations deriving from soaal and community noms (Le. a big-woman is responsible for her own moral actions).

220

upgrading her contribution so as to drown out Pria. Pria's remarks in h e 17 are a report of a

question which has been asked previously, spedcally an inquixy into a loan that has not been

repaid. This reporting of an interaction whidi has alreadv taken place has the effect of focusing

attention not on the monev that is owed but on the social relations that have broken down.

Kay's response to this is a highly dispreferred one. Not only does she interrupt Pria and refuse

to. once again. respond to the question of the money she aiso attempts to suspend the tum-

taking system with the pre-announcement Ic mi t e l y somting 'Let me teil you something. (see

Levinson 1983:345-364). Sequences iike rhis provide fairiy incontrovertible evidence that.

despite the appearance of disorder in conflict talc, participants collaboratively produce

O pposi tional alignment structures through dose attention to the content, fonn and sequencing

of pnor t a and their interlocutoa positioning (a point made persuasively by Goodwin and

( 4 - 3 ) 17. Pria: abi aks wa kom pan tuu twenti eet p u see wan faïv // shi gu kompleen ( ) gu gu luk fo mi

we a k e d what was lefrfiorn two-twenty eightym saidjwe // "

19. Pria: agulukfa* lmkin for

// LE MI TEL YU SOMTEVG yu mos tink a //let me tellyau sornethzngp most think this is

20. &y: Raja daadii prapatii A MI1 PRAPATlI DIS an mi hozban prapatii a( 1.7) gud Raju's fadtcr's proper-p - thic is my p r o p e q and my husbund's property g d ! "

Once again we find that participants use space and spatial organization as a symbolic field on

which to create an oppositional alignment. Here Kay uses what Schegloff ( 1 9 72) called an 4

e - * expression: 'Another sort of term can be abbreviated as q, for "relation to members." Such

foms as "lohn's place," 'Al's house." 'Dr. Brown's office" are among those intended"

( Schegloff 1 9 72: 9 7). Such expressions point to a "belongingness" relationship of person to

place. The specific daim Kay is making goes deeper than this however. She not only daims that

she belongs to this space but also contests Raja's (and. by extension, Pria's) rights to any part of

it. She uses the proximal demonstrative dis as post-position (a mii prupatii dis 'This is my

propeq') to reference a space which is within her immediate perceptual field. The effect is not

oniy to characterize the space of the yard as without intemal divisions but aiso to situate

henelf unarnbiguously within i t The referent of dis is dearly the yard as a whole (which she

presents ostensively to the audience) and the fom relies, for its interpretation, on the notion of

a unit of land tied to the patrilocal unit as indicated in the smunding CO-tex- (a mii hodan

prupatii 'it is mv husband's property' line 20). So Kay hert daims to own the entire yard and at

the same time chamcterizes this space as u n d ~ d e d and homogenous (so that du wodd in this

contrasiive set refer to something located outside the perimeter of the yard). In fact Kay's

daims to propertv rights of this sort do not match local expectations or d e s of inheritance

(and thus she also challenges these too). She goes on:

NONBADII KYAN NOT TEEK DEM Al PAS MI1 TODEE AN DEM BLIN A

"nobody can rakr ahtir g and pars me oday and d-y belime thar it i s

DEM HOUS A MAIKt PRAPATiI K Y M TEEK DEM AI PAS MI1 GU3 ( ) k i r home - it is Midtaeii p o p e 9 - they can't takr k i r y e and pars mii - gwd ()

&VA AND RAIA WAiF

Raja and Raja wift. "

WAN KRAS DE [NSAI DIS HOUS NO? A KRAS A KRAS LAIK APN SKONT KYAAN nitre is a "crossn imidt fhis house, no? a "cross *a **mm" it seem us dtaugh au qfyou cunk can *t

DON TAAK (NOTW KYAN) YUU RiCH AZ DA KANKRIIT BATMHOUS AN SO - A KRAS jnish tdkin (notiting can 't) ymr 're us Nh as that mnneiz botbmho~se and so on - i t i a "moss "

222

Kay continues to make daims to property in h e s 20-22 asserthg that Pria and Raja's actions.

given that she is owner of the property on which they live, is tantamount to ve pars (/ai pa i

'some expression or behaviow taken to indicate a la& of respectn cf. Jayawardena 1963). In

h e s 23 and 24 Kay effects another constnia of the spaces of the house and yard. In line 23

she uses proximal demonsmtive dis 'this" to tak about Raja's house: wan kiar de inmi dis home

no? 'There is a 'cross' inside this house no?' Krar is both Raja's local falsename and a term for

some string of bad lu& What is particularly interesting is the way she uses d b "thisn to refer to

the house thus bringing it into the same proximal frame of the yard. Remember that this is the

space over which she has daimed ownership through the authority of her dead husband (line

20). lmmediately after this she refers to the conaete bottomhouse using the distal

demonstrative da "thatn in line 24. This situates Raja, who is seated in the bottomhouse.

outside the space associated with the pauilocal unit Moreover, the opposition of pro19rnal to

distal spatial kame establishes a boundary between what Kay claims as her propercy (induding

the yard and Raja's house) and what she consmct's as Raja's property (the concrete bottom

house). '' Although dic and da do prototypicaliy make reference to physically proximal versus

physically distai locations, this usual interpretation can be suspended in certain contem. The

house (referred to as d ~ r ) and the bottomhouse (referred to as da) are at exactly equal distances

from Kav at the time of speaking. Kav uses the default values of these demonstratives to

characterize space in tenns of social. rather than physical. relations. As Levinson ( 19943357)

'%e insult is meant to go further than sirnply saying Raja is not the owner of his own housc and yard. Kay is charging that Raja must live by the diarity of others. About a week and h d f earlier I had given Raja the money to lay concrete over the daub floor of the bottomhouse. The two of us along wïth one other spent a couple of days laying i t out and lilling it in with rand. rock and conaete. Everybody in the yard knew what we were doing and who had financed i t

223

notes 'There are many sociological aspects to [...] deictic dimensions, e.g., whether to describe

some space as 'here' or 'there' mav depend on whether one thinks of it as near 'us' or near

' them. ' this being sociologically defined. "

6.4 AFFECTIVE DISPJAYS OF PASSION:

1 have argued that, in the interaction presented above, Pria dearly indexes an affective

state of passion in her speech. She uses a variety of cdturallv recognizable cues tu achieve this

metamessage. Most obvious is the inaease in volume and the occurrence of locaily recognized

curse words but one rnight also note the use of rhetoncal dependent intemogatives (mi na noo wa

skontyu a luk fo mii) and parallelism (lines c 10- l l >). Rather than seeing this display of

"emotion" as a somehow natural outcome of fwtxating Qrcumstances. it is perhaps more useful

to undentand it in context as a strategy ( Luu and Abu-Lughod 1990). Pria is. in her speech.

chailenging suongly held (hegemonic) notions about rights to movement within the village. At

the same t h e she is challenging the authonty of her husband and his patrilocal unit Such

transgressions are often met with firm, sometimes physical, repris* in the form of beatings and

h t h e r restrictions on movement However. in assessing what is to be done with people who

have broken some aspect of the local village (or patrilocal) code of conduct, persons may take

mto consideration their emotional state at the time when the transgression was committed. A

person who is cold and calculating (i.e. a thief) wül usuaiiy be punished more severely than a

penon who acts out of passion. Passion is considered an involuntary state. In fact unlike other

emotions it seems to be conceptualized as part of the extra-penonal context. Thus although a

penon mav be said to be "passionaten it is more usual to Say something Like. ii de in wan pnshon

224

'hehhe is in a passion." Such locative constructio~is seem ahivays to be w d when the state is

brought about by something beyond the control of the experiencer. Thus one may Say, ii de in

rvan prublm "hekhe is in trouble." The impiication is that the situation is now beyond his or her

conuol. Villagers also use possessive constnictions to ta& about passion as in, wen mi get mi

pashon 'when 1 get my passion." Again the implication is that the 'emotion" somehow exku

outside the core of the person. While not assuming to understand the intricaties of local

theories of emotion a t this point 1 wouid venture to Say that Pria uses an affective display of

passion hem so as to avoid some of the problematic repercussions of her verbal attack on Raja

and the patrilocal unit There is the chance that she will be excused from blame if people are

wüling to accept this as a genuine display of passion. When making such assessments people

usudy take into consideration the events leading up to the diçpiay and Pria is ais0 carefui to

make this context dear in her discourse. Unfortunately for Pria people did not generdy accept

this as a genuine display of uncontroUable passion. My impression from interviews is that such

judgements of legitimacy are more often made of men than of women (i-e. rights to feel passion

are unequaiiy disuibuted by gender). Women are more ofien stigmatized for affective

behaviour of this sort. In faa. as 1 have noted (cf. chapter S), busing, cussing and other foms

of arguments are generdy feminized and a woman's engagement in such conflict talk is

hrpicaiiy taken as the inevitable expression of her fernininity rather than as an index of some

deeper social problem which requires resolution. Women's exclusion kom power. authox-ity and

"public discoursen is thus naturalized in local ideologies of ta& which Iuik particular genres to

underlying (i.e. essential) psychologid tendencies and from there to gender.

While the audience recognized that Pria was indexing a state of passion. most remarked

225

that this was not genuine. They suggested that the frequency of such outbums indicated that

they were more theavical than uncontroliable. Severai people I talked to afker the inadent

occuned remarked shii a pke se shi hed na gud bo shii get nofsem, 'She pretends that she has lost

her head but she has plenty of sens4 she is srnad she knows what she is doing."

Members of the extended patrilocal group believe that Kay is a diunk and therefore do

not hold her responsible for her verbal outbusts. Pria. howcver, cannot daim to be under the

uifluence of dcohol. Rather. she contextuaiizes her speech as affective and occurzing as the

result of an emotional state beyond her immediate contd. Besnier ( 1990) remarks:

In many societies, women can oniy express affect (and sometimes have a public voice at all) in specially bracketed situations or through the use of particdar genres. Thus 'veiledn and 'ambiguousn genres such as chanting, weeping, or speaking in tongues, the perfomances of which often invohe dtered States of consaousness, are frequently but not always, gender pokrized. One characteristic does seem to apply cross- culturall~~ the low social evaluation of such genres on the part of either men or the entire group.

In the case 1 have examined above, it appears that the speaker uses contextual mes to bracket

her speech as an instance of a paaicdar affect-marked type. Pria gains a certain critical voice

on gender relations through a culturallv d . interpretable (altered) state of consciousness even if

local people refused to recognize its legitimacy.

As 1 have mentioned. most people did not accept Pria's speech as a genuine display of

passion. Instead, they suggested that it was indicative of her low-dass origins, her malevolent-

divisive nature and her self-interested inteiiigence (her "sense" in the local varieo/). Alhugh a t 0 . 0.

k t . Pria seemed to have gained some ground with her affines because Raja was forced to ailow

226

her more freedom in order to prevent another embarrassing display. in the end she was M e r

stipatized and lost face with neighbours and the affinal patrilocal unit People avoided her

and gossiped that she was low and not proper. Pria's use of this cultufauy interpretable r e e t e r

thus seemed to ba&e a t l e s t in part because the interpretive community would not

recognize its legitimacy. As Tom Stopparci's PIayer from Romcran& and Guildnztmi are Dead

remarki. "audiences know what to expect. and that is all they are prepared to believe in"

( 1 9 68: act MO. p.6 2). Ultimately. in the case under discussion, it was. as the Player suggests.

the audience and the audience's expectations that were most important in detennining the

meaning and efficacy of Pria's display. Howwer, the example 1 have discussed points to the

possibilities for contesting the local construction of gender and domestitity through the

organization of space and expectations for b d y cornportment As 1 have discussed in chapter

1. in-marrving wives who iive with their husband's pauilocai gmup are in a partinilady

disempowered position. As such, despite the reaction of the Local community, it is important

that we recognize the parriai efficacy of Pria's performance of passion. As 1 discuss in my

çoncluding chapter. such strategies of resistance as disnissed here are but one option. Other

wornen in other socio-economic positions develop other strategies which are sometimes more

successtiil in contesting the CO-construction f space gender and authority.

In terms of the impiications for the study of spatial description and conflict talk,

evidence of the kind examined here suggests that spatial description and place formulation in

verbal dispute is both a major area for the expression of conflict and. a t the same time at least

under certain social conditions, an important object of conflict itself. Spatial orientation, while

relying on a background of partidy shared knowiedge (i-e. a commonsense geography) is aiso a

227

resounie for partiapants and. as such, can be strategicdy manipulated to achieve interactional

goais such as the production of oppositional formats and paxticipant structures. In tems of the

analysis of verbal dispute it is M e r apparent that inhabited spaces are important areas for the

negotiation of identity and digrunent in convenationai interaction. Participants to an

interaction orient themselves to space not only as a set of identifiable locations but as field

imbued with social meaning (Hanks 1990) whidi is thus an important resource for sitmting

oneself in relation to others and in relation to an emergent interactional text. We have seen

that such interactional ends are accompüshed collaboratively through the use of deïxis, through

the description of spaces which are related to important social groupings and through the

phvsical movement of bodies in space. Verbal disputes of this kind thus provide one kind of

evidence that spatial conception in discourse is achieved through the joint actions of

participants.

CHAPTER SEVEN

DEVEL~PING A HABITUS FOR EVERYDAY L m : THE CONTEXTS

OF CHILDHOOD AND ADOLESCENCE

7.0 INTRODUCTION

In attempring to understand language use within a broader sociocultural framework, it is

useful to consider the way in which the practice of taik invotws not oniy the expression and

transmission of referential messages, but also a mode of self-cornportment - a way of holding

oneself that ir appropriate to the social situation. Consider what is involved in this latter aspect

of talk The participant must be able to sort through all the possible information available in

what could be loosely understood as the physical context, as well as in the CO-text, to corne up

with what is relevant to the ongoing interactive sequence. 1 have touched on these issues bnefIy

in chapter 3 and reexamine them in the next chapter through a consideration of eiiatation

sua tegies. The generd concem here is to show the ways in whidi talk and context are active-

CO-consuucted; that context. even at its seemingly most basic spatial level. is creativelv

consvucted in tak By way of introduction to this set of complex problems, in the present

chapter 1 consider the contexts of early childhood in a more general sense. My main concem is

to show the way communicative bodies are consuucted through socio-cultural prinaples.

FoUowing Ochs ( l988), 1 consider the use of affective svategies as basic to this process. Three

moments in this affective shaping of the communicative bodies of chiidren are discwed. In the

fin t case, Long before children acquire Ianguage. caregivers begin to incdcate a certain way of

understanding the body. The values that 1 described as autonomy and age-gradedlgendered

solidaritv in the introduction to this work are embedded and passed on through the very

practices of holding and caring for Mdren. In the second moment, with emergkig linpuistic

skills cornes exposure to range of caregiver verbal strategies which more preciseiy direct children

in community-based moral and soad education. Guyanese strategies, like those of Samoans

(Ochs 1988). can be usefdly contrasted with those most familiar to white middle dass North

Amencans insofar as thev are less Md-centred and more child-decenuing. In the find moment

1 consider hem, chiidren move out of the position of dependents. This happens comparauvely

early for a i i children but has different consequences for boys and girls. AU diildren deveiop

peer-pups but for girIs the development of such relations tends to be restricted by the fact

that early on they are expected to contribute to the household economy. For some Lhis means a

v e y rapid transition From dependent (up till age four) to caregiver (by 7 they may be caring for

their siblings). For others it means conaibuthg to household labour in other ways - for instance

cooking and cleaning. Bovs on the other hand are not expected to work in the same way.

Rather, thev are encouraged to engage in various semi-productive activities like fishing and

birdhg. Between the ages of 7 to as old as 18 they experience a kind of personal autonomy and

freedom that is unrivalied in the village. Although they may be called on by superordinate kh

to fuifil short tem task most of their time is spent with a peer group. Boys quickly recognïze

that although they are given considerable freedom to roam with their peer-group, when cded

upon by higher ranking kin they must comply or face senous punishrnent Girls on the other

hand do not experience the same kinds of freedom to engage in peer activity. Early on they are

expected to contribute to the household economy and by the time they enter the heterosexual

economy (with the onset of puberty) , they face the sarne restrictions, though u s d y less

severely enforced, as do their mothen (see Chapter 6). These socialization processes have the

230

effect of laying down or indcating what 1 texm the mole-habitus by which I mean a l o c ~

managed, generative schema for seifkomportment, perception and interpretation. Embedded in

this set of schemes for action are the saiient cultural prinaples of autonomy and age-graded,

gendered soiidaritv which are vital to the reproduction of the social group.

7.1 THE PLACE OF CHI~DREN IN THE WORLD OF ADULTS.

If, after reading the literature on socialization and language sociaiization one were to

make a single cross-cultural generalization i t would probabty be this - communities differ in the

degree to which caregivers regularly adjust interactionai routines to suit the Enguistic and

cognitive abiJTties of the M d . On the one hand we k d communities, iike the white midde-

class in North Arnerica. in which interactive sequences are routinely adjusted to the perspective

and abüities of the M d or infant- Such child-centeredness in caregiver-child interactionai

organization is usually associated with a set of folk-theones regardhg individual intention and

the autonomv of mind (cf. Ochs 1982). On the other hand we h d communities in which the

child is expected to adjust to adult perspectives. interactional abiiities and generally f i t itself

into an existing social milieu that is often highly stratified, with age nearly always figuring as

one of the primary organizing principles of this stratification. Such strategies of child rearing are

often associated with a folk-theoiy of mind whidi is more dialogic, which sees individual actors

as lrss than completely autonomous individual agents. and which emphasizes, instead. the

unfinalized ( Bakhtin 1 986) nature of thought and utterance and the ultirnately impenetrable

nature of mind. Whereas in Samoa, the decentring is evident in the absence of a specialized

register (Le. babytalk), the emphasis on intermediaries ( providing basic instruction in status

23 1

hierardues), and in the use of eliatation routines, among the egditarian soaeties of the Kaluli

of Papua New Guinea (Schieffelin 1990). decentring is most obvious in the emphasis placed on

assertive speech a m or "hard words". In an egalitarian soae. where social interaction is based

on the principles of reciproaty and obligation. ianguage is undentood as a tool for

manipulation and the attainment of persona1 goals. As su&. Kaluli caregivers make an effort. in

part through the EW-elititation routine (cf. Schieffelin IWO), to instruct chiidren in the uses

of language as an instrument of assertion. In terms of verbai strategies. O& ( 1982) has

suggested that the chiid-centered pattem is canonicaily associated with the expansion sequence

whereas the child-decentring pattern is associated with the eliatation.

THEORY OF MIND

adjusted to the child - ideology of egaiitarian social relations

in ten tional-autonomous individds

child adjusts to addt - ideology of stratified (Samoan) /age- graded (Kaiuli) social relations 1 dialogic, in ten tionality downplayed, cornplex (or binary) rnodels of self

Table 7.1 CHILD-CENTERED AND CHILD-DECENTRING CAREGIVER STRATECIES

VERBAL STRATEG IES expansion. baby-tal k

A central question for the present study, then, is how Indo-Guyanese suategies in the viUage fit

into this range of cross-cultural variabüity. The problematic nature of the question when

appiied to a case like Guyana. 1 think highligh~ something important about Md-rearing

practices, something that may have been rnissing in many studies of soaalization. Although

child-rearing pradces do reflect (or reproduce) saiient cultural principles (Ochs 1 982, 1 992). it

elicitation

232

is important to recognite diat, like other dtural practices, these routines are embedded in

ongoing socio-econornic processes.' ï'hus. for instance. chiid-decentring seems to be related to a

certain social situation in whidi autonomy and individual achievement is relatively subordinate

either to, on the one hand. local systems of kin-based stratification (and the politics of heage - as in Samoa) or, on the orher, to economic adaptions which are organized cenually around

reciprocitv and obligation (Kaldi). At the other extreme, it is not difficult to see the wav in

which diild-centering strategies, so famüiar to the Anglo-Amencan micide dass, are linked to

the emphasis on individualism and individual adiievement in the economic marketplace. In

cases sudi as the Noah Arnerican one, where one h d s a major preoccupation with

maintainhg or even raising dass position across generations. it is Likely that caregiven WU

place an emphasis on the individual M d ' s presumed talents, abilities or skills which are the

key to her "success" (e.g. 'Oh, she's going to be a doctod" - said when the duld bandages a

dolI) .

Child rearing practices thus become indices of dass just as the presumed 'results" of

such practices do - i.e. "decent" "respectable" and 'respectful" children. thieving children.

scandalous children, etc.. In Guyana, there is a fairiy clear split between those practices of child

r e a ~ g that are associated with 'poiite" culture and those assoaated with d working d a s s

culture and the socio-economic relations of the village. In a general way this dimension of

variabilitv maps ont0 the variability described in the chart above. This is to say that, within the

broader cultural framework of an official or legitimated 'Guyanese" cultural tradition (which is

' 1 have heard from a number of sources that lane Collier at Stanford is working on the problem of socialization in relation to changing socio-econornic cirnimstances. Unfortunarely 1 have not been able to access this work.

233

most dosely as~c ia ted with urban middle and upper dass of Georgetown), practices whidi are

" child-centred" are highly evaiuated. This indudes baby- té& doting on smd-dllldren,

interpreting their early movements and voc;ilizations and generally placing them at the centre

of the soaai interactions where they are present2 At the same time, within the more locaüzed.

more h l y embedded, culture of village life sudi behaviour. especidy from mothen. is

considered excessive and. to a certain extent, seif-elevating. Thus. in the village context, the

chiid is more iikelv to be rewarded and praised for behaviour that most resembles adult

practices such as msing and falling out (see the next chapter), independence and physical

strength. The 'childlike" qualities of the diild are rarely praised or encouraged (such as

children's ways of speaking, ayuig, dependence on the mother or other caregiver) . Now any

caregiver is likely to draw on both sets of noms (Md-centering/child-decentririg) depending

on the situation, the audience, and a number of other contextuai parameters. This said, the

latter framework (the one in which the chiid is decentred) tends to play the dominant role in

most households - especiaily those households which maintain a strongly working-class

domes tic econornv. However. assuming that the historical shift towards wage labour and away

tiom patrilocal group authority (which 1 documented in chapter 2) continues, we wodd expect

to see chiid caregiver strategies gradudv and consistently move towards the &id-centered pole

af the tvpology laid out in table 7.1.

However, in working-dass households where the male addt is a labourer in the rice

fields and the female adult is at home, little extra tirne is available to devote to specialized

interaction with inf'ts and children. Mothers are busy (or exhausted) for a good part of the

Representations of such behaviour frequenrly find their way onto local TV sets in commercial advertising.

234

dav (see Chapter 6). This does not mean that mothen and diildren do not interact, it only

means that such interaction is not modeiled as a speciaüzed genre of 'interaction with

chüdren." Children, when they are home from school, are incorporated into the daily routine of

domestic labour. When incorporation of &&en into the everyday actMties of ad& is not

possible (or when they are actively exduded from add t actfvity such as drinking), duldren tend

to aeate highly independent and self-replatkg peer n e t ~ o r k s . ~ Aithough opportuniùes for

peer-networking are iimited by the exigences of a gendered economy, a great deal of education

goes on here. Peer-network practices of this kind tend to be removed fiom the economy of

'caregivingn to the extent that the children themselves are removed from it.

As it stands, whereas chiid decentring fits tongue-and-pve with existing and deeply

embedded systems of age-grading. gendered patterns of movement and socio-economic

relations of obligation based on kinship, chiid-centred suategies ohen clash with expectations

based on these p ~ c i p l e s . For instance a doting and "overly " attentive mother faces possible

social sanction for misdirecting her attentiveness especially if chiid care start..~ to interfere with

her abilitv to defer and serve people who are accorded greater status - Like a husband's older

brother or father. Such doting is likely to be taken as self-elevating in these contexu indicating

rhat a mother beiieves that she and her children are of greater importance than high-status

guests. Generallv, then. chiid-decentring strategies are more often adopted in everyday

interaction. At the same tirne eveqrbody is aware that in certain situations Md-cenued

O . .. ]The language of peers and caregivers has in the p s t hidden the kinds of labour that q e done by children and 1 do not which to continue this tradition (Schlegel 1996). In section 6.4.1 (below) 1 riiscuss the ways in which daughters (and sometimes sons) become caregivers at an early age. In stressing the importance of peer networks however 1 want to bring to light the degree to which chiidren crrate their own. often transitory. social worlds sometimes independent from adulu.

235

strategies appear more 'sophisticated" and are more highiy evduated. Some variation is thus

apparent in this as in ail comm~nities.~ One t h g that should be noted about the case I discuss

here is the way in which the naturai dependence of children on others potentially contradicts

the ideai of autonomy which I have suggested is centrai to addt sociai organization (see

Chapter 1). In light of this it is perhaps not surprishg to find that the emphasis o n kin-

obligation. spe&dy the rehtiomhip whereby one person mind 'cares for" another - becomes

centrai to the wav people thtik about relations between parents and chil.dren. The relation

between parent and chiid is the protot~pical example of non-autonomy and dependence. The

move out of this condition. the tirne when a person becomes an adult is metaphoridy

extended to relations of statu between non-kin - thus 'big" refen bodi to age and to status.'

7.2 CHILDREN'S BODIES - AN EDUCATiON BEFORE THE ONSET OF LANGUAGE

The value of autonomy and its limits in the face of age-gradedgendered solidarity are

embedded in the wav a chiid (or any penon for that matter) knows its own body. Long before

thev are engaged in verbal routines, children leam about the social milieu through the wav they

are handled by caregivers, peers and others. This kind of education is pax-tidarly important for

the preverbal infant but soaalization by su& means continues throughout the Me course.

Thus, the restrictions on movement that I have documented for aduit women shouid be

undentood as a continuhg form of socialization that works in part through the body. The

1 And of course it is cruaal that we don't equate child-centred strategies with 44affection". "care" or "love". Decentred strategies are embedded in an elaborate pedagogy just as are child-centred strategies. The ciifference is not in the quantity or quality of the attention given to the child In fact child decentred strategies often involve elaborate routines some of which 1 discuss below.

'on the metaphorid extension of mind to relations between adults see Chapter 6.

236

education that the chiid receives through the handling practices of others reflects culturai

schemes that the caregivea have for organizing ideas about the chiid's body and how one. as an

adult. should interact with i t Caregiven employ a number of terms to desaibe the bodies

both of children and addts. One descriptor that, when used in reference to children. is usuaiiv

cause for some concern i s j k meaning thin and genera3ly weak Caregiven often assoaate

beingFne with sickness and may make this association explicit by the use of the term maaga

which refen to a thinning out of the body as the d k a result of sidcness, thus people Say ii get

maaga when a M d is losing weight due to vomiting or malnutrition. Generally, at least with

regard to children. heness is negatively evaluated (and associated with illness) whereas being

"fat" is positivelv evaluated and assoaated with health. The preference for robust bodies is

camed on into the aesthetics of later life aithough it may corne into conflict with other

(perhaps foreign) notions of beauty. Children are encouraged to eat when food is available.

Parents worw less about a d d d who is a good-eater even if they cornplain occasionaily that ii

belii na gat batm "His beliy is a bottomless pit".

The general concern to produce, both though reproductive and cultural processes.

children that are robust is part of the more general concem to impart a sewe of independence

eariy in life. A child who is always carried by its mother, who can't be put down at all and who

is prone to sickness is both a liabiIity and sign that something may be wrong with the iineages

that produced i t As soon as an infant is capable of holding up its head it starts to spend a good

deal of its time phvsicaily detached from the mother. When at home the infant is often placed

in a hammock and rocked when agitated. M e r a tirne caregivers will take the dllld out and

aUow it to roll around on a burlap nce sa& or on the wooden floor of a raised house. By 4 to 6

23 7

months the M d is well accustomed to sucking on pieces of cake or biscuit which supplement

breastfeeding. Various 'teas" are also used as supplemenu to breastrnilk, especiaIly by mothers

who spend tirne out of the house working or who h m problems lactating. Often, the

responsibüity of 'Iooking" after a diüd (making sure it stays in its mat or in the hammock, that

it does not cry too much. that it is relatively dxy) falls on an older sibling, usuaüy a girl. AU

caregiven, chitdren and addts alike. u y to foster the independence of the diild and will not

hold it for extended penods unless it is si& By the t h e the M d can crawl. it is allowed to

roam around the house and its perirneter with minimal interference from others. If there are

other children in the yard, the moment the child starts walking it enters a new set of relations

with its peen. Parents encourage sibiings to chase and entertain one another and generally

refuse to involve themselves in disputes.

The use of the hamrnodc and the floor by caregivers has the effea of producing, in the

child. a certain independence with regard to the use of space. In addition, caregivers are also

procective of theh own bodiiy space and, in fact. police it quite diligendy. Children who c h g to

their mothers are regularly told na k m kana mi 'Don't get to close to me/get me into a corner."

A mother mav cornplain that a child a hombog mi 'is bugging me/restricting me/ getting in die

wav." The cultural need to defend one's personai space, stemming from particdar ideas about

the nature of the bodv and pmrileges to move about unhindered. is of course exacerbated bv the

high d a m e temperatures and the faU1y epidemic incidence of various kinds of skui mites and

head lice. There is usually both a literal and metaphoncal sense to the expression mi skin a

sbach mi "my skin is itchy," which mothers frequentiy use as a rationale for not ailowing their

children to touch them.

AU this should not be taken as an indication that people, induding parents and

chüdren, do not touch each other or that they maintain some impermeable cushion of air

around them. This is. of course far tkom the case. In fact, if communities vaxy in terms of the

noms for toudllng and physicai contact, the Indo-Guyanese village is no doubt towards the

more tactile end of the scale. Two boys or two men will often share a harnmodc or a bed and

generally parfners 'male fkiends" have a fairiy physical relations hip expressed in everything h m

nonsenous s p a h g to affectionate hugs and holds. At weddings, when barxiating, men dance

together for hours in a style that seems distinct$ homoerotic with one man "winding" dom

the other's leg. Women who are fnends also hug, although the toudiing and handling of one

another seems more perfunctory and ntualized. Men and women who are not related may also,

without any implication of a suaial relationship, touch each other under certain conditions. If

the woman is older and has taken on a fictive kin role (like aunty) then it is acceptable. If they

are ntualiy attached to each other as platonic friends thev mav also engage in phvsical

exchanges without any implication of a senial relationship. One such ntualized relationship is

"rakü" in which woman take male friend as fictive "brothersn.

I t will be noticed that aU the relations discussed so far are between addts or between

adolescent peen. In the case of relations between adults and cMdren physical shows of

affection, in fact physical displays generally, become inflected with status. Whereas diildren are

often directed to distance themseives from caregivers, adulu reserve a right to manipulate in

various ways the child's body. Every day, caregivers in most households (espeaaily durhg

school session) comb their children's hair and pi& nits and head lice. Children generally resist

at first and are forced to submit sometimes with a firm hold on their axms or a leg around theù

midsection. A great deal of tears are shed over such daily routines.

Force is also exercised on dllldren's bodies in other ways and in other contexu. Boys of

about four vears and older often have a special 'unde". a dose hiend or relative of their father.

This man WU ofsen induige them with sweets and cenain kinds of attention the child is

unlikely to get elsewhere. Ohen, this uncle will also stimulate the boy's genitals to produce an

erection. In most cases. the chilci stmggles at first but eventually gives in. Such forced

stimulation, it is suggested, encourages the growth of the p a i s and the development of a

healthv (Le heterosexual) appetite. The lesson for the Wd. i t seems, is in the futility of

resisting the WU, which poses as benevotence, of adult men who hold respected position in the

household.

Although the dùld spends a good ded of time wandering about the yard relatively

unhuidered by others. the earlv expenence with adults provides a lesson on the restrictions that

an age-graded and gendered soiidaritv imposes on autonomv. The manipulation of the a d ' s

bodv from an early age, when combined with the aggressive defence of the pesonal autonomy

of the caregiver. provides the child with an early mode1 for relating s t a t u , autonomy and

obligation. These models serve as basic frames for understanding later interactionai routines.

7.3 AFFECTANDSOCIALCONTROL

One of the most important means that caregivers in ail comrnunities use to control the

behaviour of children is affect or the expression of 'feeling" and 'emotion" (cf Ochs 1988. and

articles in Schieffelin and Ochs 1986). Caregivers both index their own affect and attempt to .

arouse particular emotions in children. W e the development of aHeaive cornpetence in

240

children, that is the ab* to encode and decode Enguistic expression of cuitUray. parti&

understandings of human emotion, is in and of itself an interesting topic. in the following I am

more concemed with the way in which the indexing of affect by caregivers in interaction

faditates the uansmission of the broader socio-cultural values of autonomy and age-

gradedgendered solidarity which I have argued are cenual to the reproduction of a reiative

egalitarianism in the village. 1 have argued that before the onset of language, children are to

some extent educated -through the handling of their bodies- in both the importance of

individual autonomv and the iimits that are imposed on it by sociai dations of kinship and

status. With the emergence of linguistic and interactionai competence the child's education in

this svstem of vaiues moves to a different level, The child's role in instnictional routines

becornes much more that of an active participant. Mereas the eariy. pre-linguistic education

which operates seemingly diredy on the body takes the M d as a manipulable and passive

entitv. the later verbal routines to which 1 now tum aiways engage the chiid as an active

participant in some sense. I f we are to take Vygotsky ( 1978. see Cole 1985. Weasch 1985) at

lus word, we would expect that such learning which relies on the active engagement of the

leamer is generailv more effective. At the same time the preiinguistic education selves a vital

role in incuicating the most basic outhes of a dispositional habitus (cf Bourdieu 1977).

7.3. I DIRECTIVES - INDEXING AGE GRADED AND GENDERED SOLIDARITY

Directives not oniy direct an addressee to do certain things in a certain way ( perfom an

action) thev also by virtue of their form estabiish or index various kinds of soaal relationships

between participants to an interaction (cf Brown and Levinson 1978, E M n Tripp 1976,

Goodwin 1990. Silverstein 1987). Caregivers in the village give many more dixectives to

children than chüdren do to caregivers. Directive use indexes status and. because age is a

primanr measure of status in the village. directives generaily flow from older to younger

participanu. Children occasionaliy issue ckectives of their own to caregivers but these almost

alwavs rnitigated. The most common form of rnitigation is parti& intonation contour and

tag na 'nown.

1. ta& tu (r) am, na m a? "Talk to him now, ma"

-- 2. gufa (r) a m n a?"gogetitnow."

Children are notorious for h a h g hard ean - that is to say. for not listeni- and, by extension,

not heeding the directives of their parents. As such, parents often feel it is necessq to use

vocatives of the sort 'ee bai", 'ee gyal" and directional directives like 'wach mün as in the

follovving example.

LM: ee bai \va& mii wa bodoo neem "Hey boy, look at me. What's Bodo's name?"

Such directives are important injunctions of the type Bourdieu ( 1977) discussed in his

considerarion of habitus. These sequences of directive-response (where the diild's preferred

response from the chiid is essentiaq non-verbal. i.e complying to the directive by adjusting

bodv orientation and location) have the effect of instantiating age-graded relationships through

control of the subordinate's body. This form of education prefigures a set of operations through

whih power is exercised and social relations based on gender, age and status are displayed.

7.3.2 THREATs, FEAR AND PHYSICAL VIOLENCE

When directives are insuffiaently persuasive. caregivers often relv on the use of threats

of physical violence (cf Ochs 1988). Most often these thmats take one of two forms. The k t is

an explicitly performarive type: ma wanrnyu 'I'm warning you". The utterance is e f i p t i d

implyùig impending action on the part of the caregiver. The second type is a rather p e d a r

utterance. Ma biitprr translates as '1 am beating you" or '1 beat you (habit-). " Caregivers

never threaten ma gu b i i t p or mi gu b i i t p or mi dor bittyou. Imperfective n in CC is usually

interpreted as having progressive or habitua1 meanings (see the appendiv to this work).

However neither of these meanings fits the context here. Rather the impedect marker here

signals an impending aaion. This leads us either to rethink our interpretation of the imperfect

marker or allow for the possibility that in this context (and the usage of a with this sense does

seem to be restricted to this context) the imperfect marker has certain metaphoncal extensions.

Threats of the sort discussed are partidarly important to the inculcation of the values

of age-graded and gendered solidariry and in fact are used as indices of asymmetrical kinship

relations throughout much of the kifespan. A relatively small uncle (i.e. not the big Chacha for

his patrilocal group) will threaten and "wam" his nephews until they reach the age of maturity

(when they are "big" and have their own families) and occasiondy even after this. OccasionaUv

conflicts arise between undes and nephews. In one such case that 1 recorded a nephew ( 17

Fars old). pushed to anger by what he felt were actions aimed ro tnke aduantngr on him,

threatened to hit his unde. He was thrown out of the yard by the uncle. Looking at the

w c r i p t below. it is possible to see the way a dispute over ownership of a bicycle qui*

transfoms into one about the relative soad positions of the partiapants. Specificaily, Stinka's

attempt to stand up to Ben (Le. to assert his personal autonomy) fies in the face of

expectations based on an age-graded solidarity.

Stink=Stinka (Ben's nephew, Gy 's grandson) Ben = Ben (Stink's uncle, Kay's son) Kay= Kay

I l .

Stink huu bai baisikl ting iaik daa? who bought the bicycie parts Lîke drat?

Ben: wa raang? what? m g ?

Stin k mi want am put am an mi baisiici I want them to pt on my bicyck

Ben: fo waa? for what?

Stink: da baisikl da da rnii o o n da mi oon da ï l a t bicycle then, tlrarS mine, that my one ahere < laugh ter >

Ben: bai luk a de da bina (greesiz) yu kyeer // de* b q look a t drm diat was ( ) ym 're uzqing // B m

// da* mii baik da. mii bin bai dima put am an de //that ïs my bike there I baught npCctors and put rhem on

Ben: egoes for S.> kom le mi slap yu //skont* corne let me slap you // nrnt

Ben:

// Ai:' /@Y

huu bai da? who bought that?

Stink: M I B A I D A P U T A M D E M I A B A I U A W A D E A N D E A A L A A L

I bought drat and put it there I baught ail that Lr on dtere - ail - all

12. Ben: teek out (Bramii) mowe out of the wuy

13. K: NA BIiT AM EE EE dot1 't beat hirn h y h q

14. Ben: ( ) skont mm? () cunt eh?

IS. K: O I - E E - E E - W A D I S K O N T W B I ~ A M F A W A Y U B ~ ~ ~ ~ F A ? W A D I S K O ~ ' N N A K ~ \ M F A ? hey h q - what diefick didyou beat him/or? whatyuu beat him for? what t h e w y u u knock hirn for?

16. Ben.: ee? ee? eh? eh?

1 7. S tink (yu) wan kom fait mii? ( S tinka picks up coconut grater) yotl want to corne andFght RU?

18. Ben: jos da yu moda skont a uuz that alom you (skont) win use (b beat me)

Living in the same vard, uncles and nephews may have long standoffs of this sort trading

threats und the vounger one b a h off. Here the conilict arose over disputed ownership of a

bicvcle. Although participants often begin to argue about righu and privileges to material

propeq. the contlict most oken moves to issues of age-graded respect and kin-based

obligation. In this exchange, Stinka's behaviour obviously contradiw expectations based on

age-graded soiidaritv, first. in his rehisal to accept Ben's authoricy and decision regarding the

bicvcle (evidenced in his pleading in lines 5 . 7, 1 1 ) and, then. in his unwiliingness to retreat

ivhen the dispute becomes physical (in line 17 Stinka pi& up a heaw flat iron rod used for

grating coconuts and threatens Ben with it). Ben's attempt to assert a superordinate position

(appropriate given his position as uncle in a kin-based system of stratification) is apparent not

oniv in his aggressivelv physical role ( h e 8) but also in his attempt to occupy the interactional

role of questioner (a role he usurps from Stinka in iines 2 and 4). Stinka, on the other hand,

245

despite bis assenion of autonomy and relative economic independence (when he says that he

bought the things on the bicyde), seems to ratify Ben's understanding of kin-based authority

pretisely by generating a rationale for his daims to the bicyde (mi bai da put a m de). thus

expressing an acceptance of the assertion that in normal circumstances he would not have any

right to it. Stinka delivers this passage with exaggerated prosody thus framing it as a public plea

for 'justicen. The unde need not justjfy his daims to property in these tenns. Rather his daim

is baseci solely on bis superior social position. As the conflict progresses. then, the ob j ea of

disagreement shifu from the realm of propeq to that of soaal obligation and respect (couched

in ternis of physical supremacy).

Although, in later Me, such eudianges tend to be associated with men. within the

nudear unit, the mother is often the one that issues threats and wamings of this sort. I t is a

part of her domestic responsibiüty. at least in many households, to enforce a certain amount of

discipline through physical punishment In fact a great deal of variation is apparent in wav this

responsibility is divided between the parents, the seriousness with which it is approached and

the degree to which threau are matched with actual bearuigs. In some cases, mothen are in

charge of corporal punishment. while, in others. it is the father who takes on this responsibility.

The parent who is not charged with the responsibility of corporeal punishment often takes an

e.xt.remely indulgent role and may even side with the diilchen in characterizing the threatening

parent as unreasonable and tyrannical. This parent often adopts more Md-cenued strategies

including babytalk and various odier forms of adjustment to the perspective of the M d .

As 1 mentioned earlier, variation in this regard seems at least partially conditioned bv a

class orientation. This became apparent when 1 asked the foilowing question in the coune of an

246

interview ij*p hiiryu piknii kos w n p gu diru? "If p u hear your child m e , what would vou

do?" h e a to this question were of three 'ypes. The most common answer was an immediate

ba t t h m . The rational for qui& punitive action is usuallv based on a notion of respect. Cussing

offends the relations of asymmeuical obligation and respect that are centrai to the reproduction

of age-graded and gendered solidarity. A number of people gave what appears, on the basis of

ethnographie observation. to be the more honest answer - that they would do nothing. Some

respondents danfied that although they wodd usuah/ ignore such behaviour they might

punish the child in some circumstances panidarly if the cussing took place in the presence of

respectable people. In fact nrrsing is reguiarly elicited from smaii duldren and generaily caregivers

admire the kinds of sesassertion it displays. Finally, a few respondents told me that they would

quietiv take the child aside and explain that this behaviour was not appropnate. Thev thus

suggested that they would foiiow a chiid-centred strategy in which they took the perspective of

the child into account when de%g a suitable instmctional routine. Such behaviour seems

remarkablv rare. The responses to the question then. despite the fact that thev do not mirror in

anv accurate wav actual instructional practices. reveal the range of options potentiaiiv avdable

for caregivea in terms of social conuol.

While threats. like directives, seern obviously assoaated with the teaching of age-

graded and gendered solidariv (and are often reproduced in the conversational discourse of

husband to wife) they can ais0 become part of the teaching of peaonal autonomy. Father and

uncles ( partïcularly Chaclra =Father's brother") are often very (playfùlly?) physical and

codrontational with theu srnall boys. Repeated &eau like kom let mi b m o k p boonr "Kom let

me break your bones" or simply JW wan ded na bai "You want to die now boyn are

247

cornmonplace in these kinds of relationships. While the teadung of social place in the solidarity

system seems to be the kt effect of such threats. a later effect is the assertion of autonomy.

Thus in one case a smaii bov ( 1 year Srnos.) was continuaily threatened by an unde who lnred

in the same yard. H e developed a cowering posture when the unde approached and would run

away whenmr he heard the man's voice. At the same t h e he was receivhg instruction from

his father (a much younger but weU-known rival of the d e ) in showing power (Le. displaying

musde, hitting and managing pain). At about two years the little boy's attitude towards the

threatening uncle changed - while he would s t ü l mn when the man approached he would k t

make a display of fearlessness (showing musde and picking anythuig that was amund that

could serve as a weapon) untii at the last minute he retreated.

7.3.3 TEASING AND SELF ASSERTION

Caregivers often engage in extended "teasing" routines with Chizdren in which some

threat is made (Le. take away some item belonging to the chiid) in order to elicit a defiant

response h m the M d . Often another addt coaches the M d in making a correct response to

the rnock "threat". Such routines have the obvious effect of encouraging self-assertion in the

chiid and caregivers typically make what it is they want the M d to Say quite dear through

clicitations. I t should be noted that teasing reiies on the chü& growing understanding of meta-

communicative devices (specifkally voice quality) and that such routines index a range of k e p

(e.g. mock : senous, Hymes 197458). More than just a f o m of soaal conuol, teasing is a

"presentation of techniques, procedures and modes of interpretation" ( Schieffelin 1986: 166.

see also Eisenberg 1986. Miller 1986). This complexity is pdcdarly apparent in the way

248

teasing routines encourage alliance making and self assertion through the intrrpretive process.

The foiiowing example illustrates some of the compkxities of teasing. Here an aunt first (Pria)

threatens to eat a candy the s m d bov (Sha) had been @en. The boy starts to ay. This is a

dturally dispreferred response as it is taken as an index of dependence on others. When the

boy does not respond in the preferred way (which is to assert autonomy) the caregiver (Pria)

@es expliat verbal insuuction in the way of an &citation (tel am no). When the boy does not

respond to this, the other participants (specificaily hiF cousin 9 year old Kavita) up the stake of

the teasing session - first hatening to kül his father and then to carry away his Little brother.

Findy Raja (the -de) eliats an explicit framing strategy from the boy (tel am a j o u k p mk):

Pria:

Sha:

Pria:

Sha:

Pria:

M:

Kav:

Pria= Pria (Shamir's aunt)

pi mi piis swiitii bai gi mi piis yu (g)u gi mi? gtve n e a piecc of the sweetie boy give me a piece myou going tu gïve me some?

teek ad? ad? le mi iit ad (shall 1) ~e al1 fit? Let me eat the whok dring

c cries >

yea le mi iit ad. ( ) wa mek? yes kt me eut it al1 ( ) whp mt?

tel (am) bodoo neem -(wa bodoo neem) tel him yau brother's name (what Bodoo's nrrmc?)

Dodii we mi slipoorz? Dudii when a n my slippers?

8. M: ee bai wach rnii wa bodoo neem h q boy - [ook at me/pay amntim - what is Bodoi namr

9. M: (i na gu iit am han rnii han rnii) ke won 't eat itgive it to me

10. Pria: kom gi mi piis ( ) na wori wid shii kom c m givc me a piece - don 't worry about ktr mmc

4 Baby goes to get a piece>

f 1. Pria: tel am noo tell ller hm

12. Kav: gimiipiis give me a piecc

13. Pria; tel shi noo tell l ier "no"

14. Kav: wach ma gu ici1 Tukii nou watrh - I'm going tu M Tukii now

15. Sha: a-acnégatives a-a

16. Pria: shi (g)o kil Tukii ( ) haia pon shii ske wilI kill Tukii - hofZer at hm

17. Mama:kaalshileshi() cal1 her let her

18. Pria: ka lsh ikalsh i ( ) knd her kaal her

1 9. Raja: (we yu gu) Qeri Bodoo. luk Kavta gu where are you gozng? Tu caT awuy Bodo? Look Knvta Lr going m

20. Raja: kyeri Bodoo. tel shi a jook yu mek curry Boduw. TctZ her it wus just a joke you made!

W'hat is paxticular1y interesthg about the exchange, besicles the emphasis that caregivea put on

e . *e assertion of individual autonomy and the protection of privilege, is the way in which,

through the incorporation of elicitation routines, the boy is encouraged to, himself. frame the

250

ongoing interaction as an event of a particular type - at fint as a defence of himself (te2 am no

line 1 1) and then as a plea for mercy (tel am a j o o k p mek line 20). 1 discuss such elicitations in

more detail in the next chapter.

T h e lesson about penonal autonomy and the defence of pmrilege that Mdren learn

through these routines is an important one. As 1 have argued, the counter-force of age-graded

solidaritv often makes the defence of oneself and one's privileges extremely difficullt My own

reaiization of this came about in the loilowing way. At s w e d times I fdt 1 wanted to do

something for lower status and younger members of the patrilocal group with which 1 was

associated through fictive kinship. In one case I gave gifts to two of the lowest ranking members

of the group. One was of exceptiondy low status because his mother had had him out of

wedlock. Both she and the father married other people and the boy was never integrated into

either nudear group. Thus iilegitimate. he depended on the charity of his matemal

grandmother and the extended group. The other dùld was low status primady because his

mother was poor and also without a husband. He could therefore not caii on the protection of a

father when bullied by his various cousins. Although he did some times threaten to teil his

mother of a wrong-douig, given the structure of authority within the patrilocal group. this did

not carry the same force. Within days (in one case less than a few hours) the gifts that I gave to

these people (in one case a chiid's toy tvhich no one else had any use for) had been taken away

bv older cousins. These items were never seen again. The children whom 1 had given them to,

while acknowledging that it was bound to happen, seemed more defeated and subordinate than

before they had received the &. The importance that caregiven attach to the teaching of

penonal autonomy is thus not trivial.

Teasing routines with smaii children as targets are reiated to the shaming practices

which all people experience later in life. Teasing provides instruction for Mdren in ways of

managing agressive verbal attacks from others and this is vital to the maintenance of a public

face in later Me. Caregivers are weli aware of t h need to inculcate a 'thick skin".

7.4 BOYS AND GIRLS - THE SOCIAL C O N S T R U ~ O N OF GENDER

The social category of gender becomes partidarly salient in later childhood. As in other

relativelv impoverished communities children are incorporated into domestic and even market

econornies at a vew earlv age. %y the time they reach h e years diildren are accustomed to

perforxnhg daiiy chores such as sweeping and washing out the house, picking up, and cariying

water. Bv the time thev reach nine. girls are generally capable of deaning out the entire house.

washing wares and a few can eveii cook roti. In some homes young female dllldren are charged

with most of the childcare for their sibiings. Childcare duties can begin as early as five or six

vean old. This movement into a gendered world of work and soaal obligation is the topic of

the foiiowing section. 1 focus especiaily on the way oppositional notions of such categones are

constructed through the practices of caregivers.

7.4.1 EXPECTATIONS O F GENDERED BEHAVIOWR IN LATER CHILDHOOD AND ADOLESCENCE

Children's peergroups up until the ages of ten or eleven show little evidence of polarized

notions of gender. AU children engage in games such as marbles. bumpa or uicket, as weii as

role playing games in which they fashion their own ephemeral soaal roles and statuses. In such

activities there is Little evidence of polarization in tenns of linguistic patterns. Both boys and

252

girls use a great number of unmitigated directives and the emphasis on hierarchy vs. aiiiance or

cooperation does not seem parricularly pronounced in one group over another. In later Me, as I

have mentioned, it is the boys' (rather than girls') alliances that are partidariy obvious and

displaved through soaal action (although such alliances often involve a hierarchical

dimension). Play among the younger children (up to 12 or 13), w M e highly age graded. is

often conducted in rnixed gender groups. Howwer, from the ages of about seven to ten diildren

experience a number of radical changes. Girls. as 1 have mentioned, become more involved in

the domestic economy and the reproduction of the household unit and are thus removed h m

the peer groups to large extent Boys. on the other hand, carry on with heaw investments in

theù peer groups, playing cards. riding bicycles, birding, tishing, playing cricket and marbles.

Gender and linguistic practice are actively CO-constmcted in this process (Goodwin 1990,

Goodwin and Goodwin 1990). Girls are more involved in the domestic economy and are iikelv

to be on the receiving end of directive-response sequences. In contrast, the boys are more

involved in the flexible and contested relations of the peer group.

7.4.2 SHAME AND SHAMING

A number of anthropologists have pointed out the central role that notions of shame

play in maintaining soaal control in small social groups (see Ochs 1988, and the articles in

Peristianv 1966). Ideas about shame are often compared with the more f d a r notions of g d t

(more f d a r to the middle class in North America) and most writers have noted that while

@t is centred around an understanding of personal conscience. shame is more hdamentallv

comected to public evaluation. When 1 first &ed in the village. and in fact throughout my

253

stay, I found the emphasis on shame as opposed to guüt quite troubling. 1 took it for granted

that appeals to personal conscience (being a comrnon. human idea of what is the right and

proper or fair thing to do) would find sympathizen and be an effective strategy for bringing

people around to my way of thinking. However, in makùig moral decisions. people were

invariably more concemed with the question of who would fkd out than with questions about

the intxinsic nght or wrongness of the action. This is not to imply that Indo-Guyanese vilkgers

(or any other people for that matter) are without some abstract moral code. Rather, I question,

in the kt place, the degree to which action, in any community, is actually based on this code

and, secondly, I would question the cornmon assumption that such an abstraa code is

somehow inherently "bettef or more advanced because i t is seemingly disinterested or

" impartiai" (on such notions see the brilliant aitique in Young 1 WO:98- 1 16). In Guyana

persona1 involvement is not so unequivocdy associated with bias or daimess as it is in other

communities. More often than not, people involve themsehres only in those disputes where

thev feel thev have something at stake. The importance of moral evaluation and public scruti.nv

is also apparent in the soaal institutions of eye -ps and pmving s ~ ~ . Eye-pass ïs a red or

imaglied affront to one's human dignity (cf Jayawacdena 1963, W i m s 199 L ). Charges of

this sort appeal to local notions of equality and basic rights to be treated with respect. More

importantly, in the soaal practice of making 9e-pas charges and. just as importantly, defending

oneself agaiwt such charges. villagers hold each other's actions up to public sautiny and

evaluation. Proving storiei occur when a dispute reaches an impasse. Both offending parties wiil

be caiied together dong with any penpheral participants to the dispute who may have made

claims in the coune of the resolution process. The idea is that the tmth should be hashed out

254

when evexybody is present because people are more like1y to teii lies when the person they are

talking about is absent (a masonable assumption). A number of disputes were brought to partial

resolution in this way during the course of my fieldwok A proving story creates its own public

and participants are careful to stock the "jury' with their own supporters. More to the point.

such dispute resolution procedures provide ample opportunity for shaming those who have

acted in ways that are considered inappropriate or unacceptable (see Brenneis 1984. and

Morgan 1 99 7. on similar procedures in Fi jian and Afncan-Amencan cornmunities respectivelv) .

One of the ways in which shame lingers and continues to have a lasting effect in ternis

of social control is through the practice of gMng falmames. AU East Indian people în the d a g e

have what is considered to be their bookname. Deonan'ne, Raja. Baal, Siiram. Pankaiee, Pria.

Kosilla . Isha . Ashan, Shamnmon are tvpical examples. However these names are not hequently

used and, in fact, a number of people are h o s t never called by their real book name except in

the context of religious rituals. This is much more often the case for men than it is for women.

although there are a number of Unportant exceptions. False names often make reference to

some kev point in a story about the individuai to whom they are asaibed and th% then indexes

a whole set of assumptions about the person so named. The names are not dwavs used to

shame. Sometimes, rather, they valorize some presurned characteristic of the referent. Thus

occasiondy a man's name indexes his Mrility - e-g. Shaf 'Shaft." However, most names either

have no directlv discemable connotations or they have strictiy shaming implications. In the

first categorv are names iike Faalin, WeWe, Nancy, Shanka. Such names make very tenuous

?ssociations between the people so named and whatever else f d s within their referentijd scope. 9

However a number of names do describe the referent. This class includes names like K m

255

'cross/bad luck". u<nt "hck-up -(Hindi) ", ffikman 'Cockman". Pongman 'Poundmann. Stinkrr

'Stinker" which are deemed appropriate because the referent is considered to uthibit the

qualities desaibed by the lexical item (see Kay's use of Knu in the chapter 5). These names can

be used to shame the referent in the right conte-

Other names are more expikit and more unambiguously shaming. One little girl, who by

seven was the prïmazy caregiver for her six month old brother. eamed the n i h a m e Luy. In

another case. 14 year-old boy who held a full time but low paying job at the rice factory was

labelled Eut and walk wak. A man who was considered to be both a dnrnk and misedy (a very

unpopular combination) was given the name Dinncr-mm because he was rumoured to carry

home a Little half-bottle of nim evev night to drink by himself. A number of young men who

had in common the tendency to refuse work when it was offered were called Dundee Millionaire.

Such narnes may last a day or a Life One.

Even more descriptive names are used when a story is just breaking, aithough these

narnes mav also stick for a long t h e . A man who had stolen gold jeweiiery from his neighbours

and subsequently been found out was given the moniker Goidman. The young men who refused

to do rice work (see above - Dundee Miflionain) often called the set men employed as tomb-

makers and grave-diggers. Diggar (possibly denved from niggm and obviousiy contemptuous).'

When a group of older boys broke into my house and stole a cassette player they collectively

eamed the name Jack(D) Tape (tape is the word commody used for a cassette player) and

lack('s) ZIing(s) which was a source of great embarrassrnent for their families, particularlv their

6 ~ t h o u g h the narnes disnisxd here are used in a number of different contexts. it is most insulting to use them as cd-nanies. In the case of D i p . the young men used this name to cal1 off the other group across a crowded road Eventually it led to a confrontation and the narne-callers, who were younger and srnaller, badced down

father with whom 1 was weii acquainted at the cime.

Shaming through name givllig and address is thus a very common form of social control

in the village. Reviewing the cases 1 have discwed the reader will notice that. for the most

part, thev invohre men and that most of the shame involves notions of laziness or thieverv

(two strongly assoaated concepts in the village - people often remark that a young man who

doesn't work is bound to end up a thief). M e n women are shamed in this way the name more

often indexes assumptions about their sexuality and licentiousness. A case in point is

Roadrunner. This was one of the names whidi just about everybody in the viilage knew and used

in reference (white, in address, thev tended to use her other fake narne Chinee). But the use of

shaming in con t rohg presumed "aberrant" behaviour went much further than fabenamer in

this case. People feh compelled on many occasions to push the young wornan till she responded

- thev taunted her both with the name and with other insults. Because she refused to conform

to the dominant mode1 which rekted gendered categories of persons t o social spaces she was

siibjected to intense and regular shaming. In the foilowing example a little girl, Baby notices

Chinee nuining by in the yard of the next house and cal& out to Chinee addressing her as

Roadrunner. Baby's aunt, Gigi, then instructs her to insult Chinee in other ways. Note the way

in which others get involved after Baby initiates the routine. I t is in such a way that shaming

becomes "public'.

- -

Baby= Baby (Gobin's daugh ter) Gob=Gobin (Gigi's brother. Baby's father) RR= Roadnuiner Jack=Jack (the researcher) Gigi =Ci@ (Gobin's sister. Baby's aun t)

c calling out > Roodrona Roadnrnner

Gob: Roodrona wan Roodrona tuu Roodrona nak dong di ool man shuu- Roadncnner one R d m n w two Roudruntter knack duwn the oid man's s h a

moda skont -rnother skunt

Gob: Roodrona wan- Roadrunner one

Gigi :

Jack: -wa shi see? wa shi see? what did she say? what dul she say?

Ci@:

Gob:

yu moda skont your muthm skun t

EE GYAL A HUU YU TAAK TU? EE? CET BAK YU SKONT YU NOO? h q girl - who are you talkïng in? get back hmyou shn~puu know?

Gigi: (yu hiir dem) big word laik shi smook doop yar hear those big w d - shr acts like she smokes dupe

Jack: arait arait alrait na fait op ~lr ight , ~Znght don 'tjghr

Bab-

Gob: wen shi pas bak hii when she passes back here aguin

Gigi: (yu teepin) claughs> ... shi mad da gyd gatu smook doop biliiv yu () gatu Cyou taphg?) she ii aazy diat @ri m u t smoke dope (1 believe)yu rnust

see shi smook doop tell her she smokes dope

shi dadii korn (nou) wen i sii-shi na gu de a rood ad dee her daddy came home when he corne she won 't be on the road al1 day

Jack:

Baby:

Jack:

yes shi dadii a kom yes hm fadrer camc home

-shi dadii de hoom? is her daddy at home now?

18. Cigi: -hiir Beebii Beebii mos tel shii doz smook doop hrar - Baby Bubyym have to maR hm durt sh smokes dope

19. Baby shi dadii de (an) her daddy i s -

20. Ci@: tel shi se smook doop (wen shi paas bak) see smook doop rait? aa- d l her that she smokes dope -

21. wa yu gu tel shi? what myac going to &il hm?

23- Gigi: ai da di shat yes that's tfght

24. Seet: -a wahapm? what3 going on >

25. g i : -( ) srnook doop smoke dope

26. See: a wa? whut's guing un?

27. Gigi: Roodrona Roahnner

28. See: wa shi tel am? what did shc suy ta him?

29. Gigi: yu moda skon t your mutirer s h n t

30. See: a da shii a duu thatk what she dm

3 1. Cigi: s l a u g h s ~

In the exchange Gigi encourages Baby to teil Roadnrnner that she smokes dope (refening to any

narcotic but most commonly marijuana). According to Gigi, drug use is the only possible

explanation for Roadrunner's behaviour. In addition to the her reputation for s e x u a l impropriety

and her disregard for community noms regarding the gendered use of space, Gigi reaches the

condusion that Roadninner is a dmg-user on the basis of her disrespect for male authority

figures. When RoadruMer mutters moda skont in line 3, Raja interprets it as a nrss addressed

directly to him. Under this interpretation. Roadrunner's utterance breaks a nurnber of stronglv

held mies for social interaction between vounger and older and between femaie and male

villagers. Gigi's response is to instigate a public shaming by having Baby tell her that she

smokes dope. Smoking dope in this East-Indian community is strongiy associated with crime,

stigrnatized and lowiy ongins and o h madness (there is ais0 an association with neighbouxing

Afro-Guyanese communities). The choice of Baby, a four year old, as the mouthpiece (or

animator - cf Levinson 1988) is strategic. Gigî does not want to impücate henelf too deeply in

the disputes of the village and since shaming does not require an authoritative speaker but

rather a svrnpathetic audience.. Baby's voice is ideal.

7.5 ADOLESCENCE AND THE HETEROSEXUAL MARKET

7.5.1 SPACE, SOCIO-ECONOMIC RELATIONS AND THE TRANSITION FROM CHILDHOOD

ï h e stories of Roadmnner, Pria (chapter 6) and Rohan's girls (chapter 2) highlight the

wavs in which men and women, boys and girls are differently caught up in the h e t e r o s e d

market and the social organization of space in the community. This CO-constxuction of

sexualiry. space and gender is partidariy important in the transition fkom childhood to

adolescence. Male and female chiidren are generaily aiiowed the sarne kinds of freedoms with

regards ta movemenr through space although, practicdy speaking, femaie chiidren may be

s-omewhat restricted by the fact that they are expected to work and must therefor s tay close to O

the house. In adolescence things change drarnatically in terms of rights to movement. Girls are

260

genedy sent out of the house only for immediately practicai reasons. They may be sent to the

shop, to the Mandir, to seil greens etc.. Otherwise they are kept close to the house. They may

be removed from school if they show littie academic potentid andlor an increasing interest in

boys. Roadmnnm was scomed because she did not make this txansition from childhood to

adolescence in terms of movement Boys, on the other hand. are sent out of the house so as to

keep it unduttered. There is a general dispreference for a boy who stays at home all day - a

houseboy. Boys are encouraged to walk-out uith their fiiends, to go tishing, to go birding, to piay

car& and they rnay even be @en bicycles to achieve these ends. In later life it is the boys with

cars, txucks or some other motorized vehide that tend to eam the highest standing both among

their peers and among the adult community. Knowing the area beyond the village, being able

to handle oneself with all sorts of people. knowing the possibiiities of town-üfe and being

competent in its vemacular are ail qualities that are highly evaluated for young men (compare

Eckert and McConne~-Gi.net 1 9 95). Most of the older boys in the village tiy to cultivate such

an image of themselves. t t should be noted that a few women also cultivate "male" syles. They

adopt the mode of dress (baggy pants, expensive brand narne Amencan jerseys or T-shirts.

kango hat etc), sqries of speaking, and modes of self-compomnent (see Chapter 3). They are or

daim to be weil uaveiled. A few such women escape public evaluation because the manner of

self-cornportment is tied to an occupation for which they must travel and look after themselves

such as selling goads dong the road. Most. however, are considered immoral and not pmper.

They may work f ~ r three months in one of the Chinese restaurants and then move on

somewhere else. They seem never to be integrated into the cornmuniîy. Women who adopt

such strategies are veiy much in the minority and tend to be marginaiized.

26 1

Generdv, by adolescence, the major characteristics of gender differentiation in the

village are beghnhg to be adopted This indudes mannea and styles of speaking of course &o.

Boys, in addition to cultivating tom-ta& are deweloping modes of authoritative grandstanding

similar to the older men. Some young women and older girls adopt more demure and

deferentiai speech styles. W e some become exceptionaily quiet, othen are brash and jokey.

Manv more switch betwcen mdtiple "womui's voices" depending on the c o n t a

Girls must learn to balance the expectations of respectabiüty, emanating primarih, from

their pauilocai group but also from the community a t large. with the world of a rapidly

changing vouth culture. Girls do not iive in a world separate forxn boys despite v ~ o u s

restrictions on the nature of thek sotial relationships. In their interactions with boys their own

age the girls do not always adopt the stvles of speaking and behaviour which are normativelv

expected of them - demure, deferential, respectable. The heterosexual market contains wirhin it

the conmdiction between rnodemity and traditionaüsm. Thus girls may orient to either the

urban "cool" (and, with it. the styles typically associated with the young men of Goergetown)

or the rural traditionalisrn of the viilage (see my discussion of ambiguity in the traditional

pattern in chapter 6). The second option is, however, much more common.

For the bovs, adolescence is a time of temtorializing impulse (Deleuze and Gaattari

1 98 7). Bovs stake daims ro places where they can lime "hang out" fieely without interference

from others particularly senior members of their patrilocal group. In one complicated case a

group of cousins came to occupy their grandmother's house. The pauiarch had died and aii the

sons were now married and had established their own houses. The sons threatened to take over

the house and look after their mother (a poor family, the sow would have Iiked to seii the

262

house). The old woman. fiercely defensive of her own autonomv, gradually ailowed a number of

her adolescent grandsons, most of whom had been in trouble with their fathen. to move uito

the house. This resulted in a number of tensions. F i m it undercut the authoriq of the fathen

(the old woman's sons) and made it impossible for them to conuol the behaviour of the voung

men. Secondly, it provided some defence of the house's assets by the old woman. Despite a

n o m of patrilineal inheritance, the sons were unable to secure the property and it remained in

the hands of the grandmother. The young men. and this is rny main point, made a great show

of occupying the house. They held the balcony iike a defensive position. Someone was always

looking out A number of scuffies ensued over their unndy behaviour in the house which,

people remarked. shamed the memory of the pauiarch and the group generdy. Arguments also

arose over the use of p ropeq and the relative right over the house and the yard within which it

was tocated.

Later temtorializing projects (for example, building a house) are related to c o m m u n i ~

sanctioned transitions such as mamage. But before maniage. and until late adolescence, boys

move in peer groups attemptuig to occupy and defend places within which they can claim

autonomy. In contrast to the behaviour of the boys who are perpetually defending these spaces

against various forms of intrusion, girls seem motivated, bv the time they reach adolescence. by

a "detemtorializing* impulse (again see Dekuze and Guattari 1987). Rather than securing

space in which they can assert their autonomy, girls and women are more concemed to resist.

mitigate and othenvise challenge the way in which gendered bodies are resvicted to particular

spaces. They attempt through various strategies to erase the lines of temtory that hem them in.

In chapter 6 1 discussed one woman's use of passion and affect marking. I have &O discussed

263

the story of Roadnmnm- Generally women plot and scheme to get away from the circumscribed

space of house and yard- They organize and participate in community sanctioned daytnps.

They jedously protect their privilege to go to market in part through theïr, at some times

vigowous, defence of an authoritative discourse regardhg knowledge of food, wares and the

exchange market Women are often in charge of the domestic economy and use this work to

jus* various outings. Once h e d in the market, women spend much of the time socializing.

Young women often accompany their mother on such expeditions and may even be dowed to

venture out on their own when they reach a certain age. The construction of gender roles in

the village thus results in a situation in which women must maintain economic relations (with

market selles, with various women who can be counted on to hold goods and give them on

credit) in order to have soaai ones. Because women's soaal relations are in large part justified

through economic relations. the maintenance of such economic relations becomes vital to the

maintenance of their entire soaal network outside the home. Rather than depending on soaal

relations to fùfdl economic neeb , women rely on economic relations (and economic

justifications) to maintain their social network This revend of an economistic logic points to

the various ways in which women consuuct their own - partially independent - social worlds

within parameters largely set by male interests. Here, as elsewhere, women use the dominant

stereo-pe of appropriate ferninine conduct (in this case, the stereotype of the woman who is

hardworking and responsible for domestic reproduction) to achieve their own soad ends (see

the conclusion to this work).

7.6 SOCIALIZATION AND LOCAL THEORES OF LANGUAGE AND MEANiNG.

264

The work of Ochs ( 1982, 1986a, 1986b. 1988, 1992. l9!?6), Schieffelin ( 1986, 1990)

among others (see Demuth 1986, Peters and Boggs 1986. Schieffelin and Ochs 1986, Wauon-

Gegeo and Gegeo 1986) has been instrumental in demonstrating the way in which language

socialization is organized through local theones of meaning and language. Aithough aii children

do leam a language, they do not leam that language in the same way or through the same

interactional mechanisms in all communities. Language-Iearning is deeply embedded in the

processes by which communities and ideologies are reproduced.

With this in mind. we should note that a great deal has beai written about ambiguity in

Caribbean language and culture (see Fisher 1976, Kodunan 1986, Morgan 199 1. 1993. 1996,

Reisman 1970, 1974) but Little has been said about the way in which such ambiguiqr is

managed in everyciay interpretive practices - specifidy those which attempt to attach

re ferentially and pragmaticallv unarn biguous significance to utterance segments. Although

people in the village. as in other Caribbean communities. play with ambiguity and the open-

endedness of language they &O, at times, attempt to reach unarnbiguous interpretations. There

is alwavs. however, a general recognition of the potential for ambiguity. Further, people are well

aware of the fact that appearances and surfaces significance can be deceiving. The cenuaiity of

ambiguiw wkhin the communicative ideology of this comrnunity thus extends to interpretative

practice. Language. and other behaviour, is dways approached with an awareness that what

people mean is not alwavs shply derivable from what they Say. As Duranti and Bremeis

( 1 986) following Bakhtin ( 1 98 1 ) and Vvgotsky ( 1 9 78) remind us. the audience is a co-author

and play a necessary role in relating significance to the fonnal shapes of verbal messages. This m . *.

gap benveen what is said and what is meant is negotiated differendy in different communities.

265

To begin with. different communities assign "directness" and 'indirectness" dBexent values and

often associate each style with paninilar kinds of soaal actors (Keenan 1974). Furthemore,

the kinds of background information which are taken as relevant to the interpretive process

v q across Merent comxrtunities (Morgan 1 99 1 ) .

People in the viUage are weii aware that the meaning of what is said is &O not

constrained by a speaker's professed or imputed 'intentions." The familiar notion of epe-parz is a

good example of this extension of meaning beyond the pureiy referential content of verbal

production. "To eyepms someone is to ... belittle and humiliate hun. to ignore his rights and

claims. The notion of towering a man's dignity and prestige by repudiating what is jus* his due

is the essence of this term (Jayawardena 1 963 : 72) ." There are no conventionalized "force

indicating devices" (Levinson 1983) for v e - p a s . As Jayawardena ( 1963:76) notes:

an gre-pass dispute can be generated by a variety of circumstances. Objeaive evidence, such as the acquisition of supenor social status symbols or conspimous use of wealth is not necessary. A subjective impression that some slight has been offered, and an accumulation of such slights lead e v e n t w to a vioknt dimu the antecedents of which are often obscure.

Eye-pms is a breach of the moral code of egaiitarianisrn such that the offendeiJs unjustified daim

to sotial supenority or conuol over another peaon lowers the prestige of the victim. I t is

rssential to note that eye-pass is often not purposive action. "The intention to eye-pass may be

present or absent Or it may be deduced by inference" (Jayawardena 1963:76). This being the

case, an act becomes eye-pacz oniy when it is interpreted as such by the victirn. We mi& sav,

then, that eye-pars consists of IWO intenelated (or didogic) acts or moments. The k t is the

of ye-pars, the second, the recognition and interpretation of it as such. Note that

although the fint moment is temporaiiy pnor, the second defines the first retrospedvely.

Along with the possibilities of indirection and ambiguity go the possibiüties of

insinceri. Speakers may not oniy resort to indirection but also to lies and deceit A popdar

expression in the village isyu kyaan tel mc jintii kat rnek ram pt 'You can't tell me Jimmy's codc

is a ram goat-" The philosophy behind this proverb is one that recognizes Ianguage as a

potentiai resource for iiiusion. persuasion and confusion. Other proverbs also emphasize the

deceptive d e of language, for instance, people often remark dag wa bark p h t i i mn wm taim kom

fi fuit 'A dog that barks a lot ~ n s when the time cornes to figh+"

People in the viliage thus tend to think of other's behaviours more as part of a well

thought-out performance than as the expression of some deeply buried and essential identitv.

Thus people often remark on the disjuncture between appearance and reality by saying that

someone is playing big or that they pke se d m hed na gwd 'They're pretending they're mad."

The proverbs and the expenence of ye-pars charges laid and received instiU a general

recognition that the addressee (or the interpreter genedy) to any interaction must sort the

uuth lrom the unuuth. People have little sympathy for those who are incapable of this. S m q

Bill " Smart BU" and Stupzty Bill 'Stupid BU" are both folk-heroes in children's stones. but i t is

alwavs Sensey Bill that comes out on top. Remember ais0 the story of Babuu and his wife at

the doctor's office h m chapter 4. A great importance is attached to the kinds of intelligence

ivhich ailow one to make appropriate assessments of the people one comes into contact with.

Such local theories about the way in which discourse worh explain a number of

preferred strategies in language soaaiization. If speakers and their messages c m o t aiways be

taken at face value (because of the potential for insinceriîy and deceit) and if messages are open

to a number of interpretations above and beyond those which their speakers intend. there is a

267

necess.. when speaking, to make as expiiat as possible both the referential content and the

contextual frame of one's talk In the course of eiiatations children may be encouraged to

danfy or reframe their utterances when misinterpretations occur (e-g. tLI am j o o k p me&). When

children are instructed to convey messages. céuegivezs invariably check that the messenger has

undentood the message correctly. Chiîdren are also sometimes encouraged not ody to convey a

message but also. if necessary. to defend this message against the charge that it is untme.

At a more general b e l one should note that early language education focuses on

particular speech genres which are relativdy halized in Bakhtin's ( 1982) texms - that is, ones

in which the potential for ambiguity, rnisinterpretation and deceit is relatively minimal. In

cases of nrcsing the choice between intcrpretations is essentia. binary - between serious and

non-senous. In caliirzg again the choice is binary with calh u s q being interpreted as Criendly

but in highh. marked contexts &O construable as insdting (cf. the discussion of d i p above).

Similarlv, invitations to o a f present chiidren with relatively transparent frames for speaking

(see the extended discussion of elicitations and metalinguistic tenns in the foUowing chapter).

Rather than instruction in ambiguity, the most frequently discussed feature of

Caribbean discourse. children receive instruction in its management Being able to deliver

unambiguous messages and being capable of sorting through the potential ambiguities in the

ivords of others are important skib in the viUage and caregivers show recognition of this in the

Npes of instruction they offer to novices.

7.7 CONCLUSIONS

In this chapter 1 have discussed some of the ways in which infants and diüdren are

268

socialized in the village. Consistent with the programmatic statements by Schiefklin and Ochs

ed. ( 1986) and SchieffeIin and Ochs ( 1986). 1 have attempted to show both the way chiidren

are socialized to use language and the way they are soaaüzed through language use. In the

discussion. I have suggested that a varies. of Linguistic practices are instrumental in incdcating

an understanding of the competing values of autonomy and age-graded soIidaxitv. 1 have also

suggested that an important aspect of this indcation process is the way it differentiates along

the iines of gender. The consequences of thiç difkentiation are far reachg. The sociaiïzation

process reproduces, through this differentiation. the positionhg of men and women in a

domestic economy and ideas about gendered rights to spaces and movement. As such.

sociaiization practices play an important d e in the reproduction of a hegemonic ideology of

gender relations. This is evident ais0 in the way young men are aliowed the fxeedom. and often

encouraged. to roam the village and pursue their interests, whereas young women are q u i c e

integrated into the dominant domestic and h e t e r o s d economies of the village. In the last

section of the chapter, 1 discussed. in general terms, the way in which children are socialized to

use language in a wav consistent with local expectations and ideologies of communication and

meaning. In the next chapter, 1 take up this issue in more detail and attempt. in the process. to

in tepte the major themes of gender, space. soaalization and linguistic practice into a single

argument.

CWWER EIGK~

LEARNING SOCIAL AND SPATIAL -TION THROUGH ELICITA~ONS

8.0 INTRODUCTION - ELICITA~ONS, SPATIAL ORIENTATION AND SOCW REIATXONS:

Caregiven in many communities fkequently direct diildren to Say something to someone

else. In studies of language soealization such routines have corne to be known as eliatations

(Briggs 1 986. Ochs 1982. Ochs and Schieffelin 1984). Eliatations provide novices with

detailed contextual scaffolding upon which they are expected to construct their own utterances.

At the same tirne. they expose children to a great deal of information about the social worid in

which thev he. EliQtations indicate who should be spoken too. in what manner and what

should be said. In short, an eh-tion models a projected scene in which the child is construed

as taiking to a definable social other.

These routines can dso provide the analyst with a rich area for the investigation of

native metapragrnatic knowledge (Silverstein l976a. 1993) - that is. knowledge about the uses

and hinaions of language. The m o d e h g of speech contexts in elicitations is achieved bv

consuuing speech as an instance of a paaicular type or genre which is indicated by the framing

metalinguistic predicate (for example: tell* m n e , ask etc.). Like other predicates. the verbs used

in elicitations are open to analysis bv standard linguistic techniques which reveal differences in

uansitivitv. as well as lexical and grammatical meaning (Sihrerstein 1985a). Metalinguistic

predicates are combined in elicitation routines with directives such as Iwk and watch which

indicate the manner of perceptual focus and deictics such as here, thm, corne and go whi. locate m . -

participants in relation to locally meaningful spaces. Co-ocauring linguistic elements of this

270

sort h c t i o n to provide detailed characterizations of acts of speaking induding information

about the wavs in which bodies should be coordinated with soQalhl recognized spaces. Thus

rach type or genre, as modelled in elicitations, has its own pecurianties regarding the s m d .

spatial. and temporal organization of participants.

M y main concem in the following is to show that knowledge about social location and

spatial location is rnutuallv embedded and that ta& is one of the major insuuments for the

development and transmission of this kind of complu: cuiturai knowledge. Con- to

traditional interpretations of linguistic interaction, such as those of Saussure ( 1983[ 19 151) fig

8.1. which characterize speech as the uansfer of referential messages between disembodied and

absuacted individuals, the eliatations routines I discuss cannot be understood without

reference to the social and spatial contexts which thqr simultaneously presuppose and

instantiate or entail.

Figure 8.1 SAUSSURE'S EXCHANGE C m m .

The manner of instntction in such eiicitation routines is always sensitive to the importance of

gender and relations based on age and kingroup membership. One could Say. then. that this is

insvuction in the social practice of language - that is, tdk punded in cornmunitv-based - ideologies and sotial relations. Eücitations, insofar as they direct novices in appropriate ways of

inhabiting Lived spaces. are Lijunctiow of the type Bourdieu ( 1977) and Mauss ( 1932) fomd

so important in the inculcation of a bodv h& and habitus. This ability to inhabit space in a

culturally interpretable manner, which is instiUed in part through eIicitations, invohres an

orientation of oneself and others to important boundaries and borden which mark off one

space from another. Partiapation in any act of speaking invohres, among other things, the

establishment of some emergent and interactionaily negotiated sense of what counts as HERE as

opposed to THERE.~ These emergent spaces are not defined according to purely objective aitena

such as actual, physid proximity or visibility- Rather. the interactional HERE of a partidar act

of speaking is always constmed relative to some sense of the way in whidi space is p d t i o n e d

into IocaIlv recognized subspaces such as houses, yards. rooms etc. and within the frame of

some activity (be it primady verbal or othenvise) which delimits the range of possible denotata

for deitic elements in speech (cf. Sihrentein 1993).* There is always a mapping, through the

mediation of an actMty hame, kom a local sense of geography to an emergent spatial context

of speaking. Now insofar as the endurhg local geography embodies or realizes important soaal

distinctions, such as those between men and women or between members and nonrnembers of

particular kin-groups. these same distinctions are likely to be of importance in the way

' ~ a n k s ( 19905 16) writes: "'Here-novu' is never a sheer physically reality to which we can meaningfuily apply objective measures. As the ground and by-product of communicative practices. i t is inevitably a lived space made up of perspectival subspaccs, costnictured with the corporeal fields of human acrors. and located within a broader soaocultural frarne space. Frarne spaces provide the field of possibilities from which the actual conditions of Iinguistic practice are denved and against which the present is cornrnonsensically understoodn

'"...many of the spatial. temporal. and objectural divisions signdled by indeXicals are predefined by things like architecture, activity spaces, calenders. work rhythm, and the soaocultural values of objectsw (Hankr 1996x222). At the same time the range of "predefined" things or regions to which such indexicals could refer is potentidly infinite. This then accounts for the necessity of some level of metapragmatic regimentation or delimitation - that is, some sense of what i t is that participants are engaged in. what silverstein ( 1993) calls an interactional text

272

emergent contexts of speaking are construed in elicitations and in other verbal routines (on

socio-spatial organitation and s o d relations see Bourdieu 1977; on the Linguistic coding of

spatial distinctions see Levinson L 992, 1996). In what foIIows 1 fht give some details regarding

the role of the body in talk and then move to discuss the ways in whidi the local geography of

the vülage in question embodies salient soad distinctions between men and women and

between kin and nonkin. I then illustrate some of the ways in which these distinctions are

mapped into the emergent, interaaionally negotiated, spatial contexts of eliatation routines.

8.1 BODY SCHEMA, SPACE AND HABITUS:

In his theon, of W ~ S , loosely dehed as a set of durable dispositions which inciine

agents to act and react, perceive and interpret in certain ways, Bourdieu accords a spetiai

importance to the ways in which the body is uained so as to inhabit space in particular ways in

panicular contexts. According to Bourdieu ( L977:94). the body is always made or

rnanufactured through soad process:

... ueating the body as a memory. [communities] entrust to it in abbreviated and practical, i.e. mnemonic, fonn the fundamental principles of the arbitrary content of culture. The prineiples em-bodied in this way are placed beyond the grasp of consciousness, and hence cannot be touched by voluntaxy, deliberate transformation, cannot even be made explicit; nothing seems more ineffable ... than the d u e s given the bodv ... through injunctions as insignificant as 'stand up straight' or 'don't hold your knifk in your left hand. *

The inculcation results in the production of a durable BODY-HEXS which generates an infinite

series of appropriate b o d y postures, gestms and ways of speaking for a particular context.

Such a shaping of the body and, parridarly. its way of inhabiting h e d spaces, is indcated

fiom a young age. Highly arbiuaiy aspects of a culture thus become part of the seemingly

273

natural order of things. What we tend to forget, perhaps, is that manners of speaking no less

than eating or wallang, are inevitably corporeal and are thus subject to similar foms of

injunction and inculcation. As Hanks ( 1 996) notes "what distinguishes different genres and

different fields is not whether they engage the body but how they engage it" Certain

orientations of bodies to space are in f a a presupposed by paaicular acts of speaking. For an

obvious case of this one rnight consider a telephone conversation. or a greeting, a secret gossip

session, a 'talking-tom, a proposal of marnage. AU such activities presuppose some particular

organization of interactive space. In speaking, we unthinkingiy orient our bodies so as to more

or less match these expectations of how a typical exchange of such and such a type should be

played out - we may stand up, projea our voice M e r , change the direction of our gaze etc.

Typical wavs of speaking - part of their unique contextuaiization - is, then, ingrained in the

ways our bodies inhabit interactive space.

Marcel M a w had, of course, introduced the concept of "techniques of the body" and

habitus to anthropology as eariy as 1935. He suggested in 'Techniques of the Body' that

seemingiv natural bodily techniques such as walking, swimming and nwsing had their own

cultural histoy and were in fact dways irnbued with value. But the body hexk organizes more

than just the obvious and readily observable aaivities mentioned by M a w . More hindamenral

are the wavs in which apperceptive schemas and ways of orïenting oneself to interaction

become ingrained in the individual through the process of socialization. As Hdoweli

( 1955: 184) noted. coordinated action aiways depends, in the first case, on the ability to orient

oneself and CO-interactants to emergent spatiai frames:

The human individuai is always provided with some cuituraily constituted means that are among the conditions whidi enable him [sic] to partiapate with his [sic] fellows in a

world whose spatial attributes are. in part, conceptuaüzed and expressed in common terms. Ontogeneti*, self-orientation. object-orientation. and spatio-temporal orientation are concomitantly developed during the process of socialuation.

In the foliowing i take orientation to be the ab- of the self-aware actor to locate his or

herself in relation to the boundaries and spaces laid out within a taatly agreed upon local

geographv. This ab- is obviously reflexive in that it relies on the capaaty to perceive oneself

as process, rather than a h e d point, in space. Much of everyday ianguage production and

interpretation is dependent on this refluive c a p a q for orientation. For instance, the referent

of deictic words such as 'heren or 'there* is dways made relative to some GROUND and in most

cases it is the body of the speaker which senreç this purpose. However, it is not the purely

physical body but always some Local understanding of it that is mobilized in such referential

praaices. This involves, according to Hanks ( 1990, 1 996a). a localiy ofganized and generative

corporeal schema duough which self-aware actors assess the feasibüity of potential modes of

perceptual access. Part of this self-awareness must inevitably involve considerations of the wav

in which the actor is undentood in relation to salient social categories. Participants to an

interaction take into account not only their spatial location but also their sociai position vis-a-

vis others. In fact, social engagement dways involves an interaction between the participant's

awareness of both sociai and spatial positionality. Thus parties to an interaction can show

deference or elevation in interactive routines by occupying spaces in certain ways (Duranti

1994, Keating 1994). At the same tirne. partiapants, when inhabithg interactive space must

be aware of the way in which certain moves may presuppose a certain social position. and.

therefore. how they are partidy restricted in the way they comport themselves. - . *.

8.2 h W M T G TO KAAL ACROSS SPATIAL BOUïUDARXES:

Femaie caregivers often encourage female chiIdren and infants to h l to somebody

passing by on the road but men and smd boys rareiy if ever engage in such behaviour. Most

often the reapient of the h l is some s d y important other - an aunt, a grandmother, a

brother, a sister etc.- so kaal eliatations carry a lot of information about the ways in which

linguistic interaction is sensitive to social relations based on obligation, kinship and status

differences. But h l eiiatations also cany with them a good deal of infornation about the

spatial organization of ta& and about the ways in which bodies should be oriented to one

another. Kual presupposes an addressee who is located somewhere outside the immediate HERE

of an ongoing interaction. Thus in elicitations caregivers never say km1 mii but instead ka1

am/ii/shii. This restriction to third person pronominal. as opposed to partiapant deictic,

indicates that kaal is associated with a movement from outside to inside a shared interactional

space. As it happens, this shared interactional space seems to be aiways homologous with the

bounded spaces of the house or the yard. The following ewmple illustrates this mapping fkom

local geogaphy to interactionai space. Shanica, a mother of 12. is encouraging her youngest,

Alicia (8 mons) to kaal to Ranü who is a farnily hiend. She first asserts that Alioa is competent

in such routines ( 1 .) and then encourages her to show Seeta (Shanka's sister-in-law) and

mvself. There is some conhision though as Seeta is undear as to who Aiicia h a b i t d y caiis -

Ra& (a fnend of Seeta's) or W. What is revealing about this confusion, 1 think, is that Rakii

had been waiking dong the road just prior to this eli~tation routine. Seeta immediately

assumes, based on the spatial arrangement of possible participants, that Rakii is the expected

target Shanka clarifies that it is in faa Ranü not W. She then props up f i a . who had been

276

lying in her m s . and faces her out towarb the road bouncing her slightly so as to get her

attention. With body onented correctly to space - so that the prospective speaker is faced out

from the house toward the road - she directs Alica to kaal c fig 8.2 >.

Figure 8.2 SPAML ORGANIZATïON OF THE FIRST KAAL-ELICITAIION R O ~ (Transcript 8.1, fines 1- 12)

5Seeta:

6.S han ka:

--

Shanka=Shanka (Alicia and Taiga's mother) Ranii=Ranii (Shanka's neighbour) Seeta=Seeta (Shanka's sister in law)

NOTE- Shanka is visiting Seeta at Seeta's house in Shanka's family's yard. At the time the interaction took place. Shanka., Aliaa and Taiga were all away from their own home but in their patrilocal group yard

ya -shi a kaal Taiga Ftanii wa Ranii see bad bad Ranii doz alweez yes -she mZL to Tiger. Ranii - whut- Runii sag she is bad ba(d) h i i d w q s

se shi bad yu bad ( ) shi a gu duu kaal fo Ranii say she is bad -you *re bad ( ) she is going t~ canfor Ranii

Rakii shi a see? is it URakii" she says?

Ranii "Ranii"

kaal Ranii call Ranii

shi a kaal Taiga pleen Taiga she calk Tiger clearZy -tigerm

kaal Ranii call Runii

The distance fiom the bottomhouse, where we were seated, to the road is about sixty or seventy

feet. Although the road is raised so as to form a kind of stage from the perspective of the

houses. it is often difficult from this distance to make out who it is, exactly, one sees. Shanka is

aware of this potential dif fd ty and when Aliaa does not respond to the first elicitation

attempt, she redefines the task so that it relies less on the ability for long-distance perception

and recognition. First, she gives verbal information that is meant to substitute for the visual

cues. She tells Alicia that Ranii is coming (line 8). She also supplements the manipulation of

Alicia's body with verbal cues wach "watch" meant to direct her attention and indicate the

mode of access to perceptual information.

(8.2 8.Shanka: wa Ranii see? kaal antii Ranii luk Ranii kaal shii wach Ranii a kom

what doa Ranii say? call mnty Rann - h k Rani i- cd hm - watt& Ranii ~3 m i n g

9.Seeta: wach wardt

1O.Shanka: wachRanii<iaughs>SWAAkaalRanii wu& Ranii watdz cal1 Ranii

1 1 .Seeta- kaal shi cal2 her

Unable to get Alica to àaal to Ranü. Shanka encourages the little giri to change her perceptual

focus. Ranii had been outside the perimeter of the yard on the road and Shanka had directed

Alïàa to look out from the house. Now, Shanka encourages Akia to look out h m the house

but to a place within the space of the yard (iine 14. cfig. 8.3 >).

Figure 8.3 SPATIAL ORCAMZATION OF 'IHE SECOND KAAL-ELKïTATION ROOTINE (Transcnpt 8.2, lines 13- L5)

279

This hl-routine, then. takes as the salient boun- the one that divides space of the nuclear

family, whidi is the house, from the space of the patrilocal unit, which is the yard. It therefore

contmsts with the earlier eiiatation which was primarily onented toward the boundary which

separates the yard from the road. This diange in spatial orientation is reflected in the change in

the target participant. Ranü is not a family member and thus the salient boundary for her is the

one that marks membership in the patrilocal p u p from non-membership - i-e. the yard space.

However in the foilowing Aiicia is encouraged to kani her brothu Taiga who is a member of the

extended patrilocal unit. For Taiga the salient boundary is not the yard perimeter - he is free to

cross it at any time and. in fact is constantlv coming and going. instead the boundary of

particular importance for the orientation of M a a and Taiga is the house perimeter. Aiiaa is

within this perimeter whiîe Taiga is outside i t Further Taiga is not a member of the nuclear

family so he does not have f d pmvileges with regard to thû space - he can be told to Ieave by

an adult family member at any tirne.

(8-3) 13.Seeta: kaal Taiga den

cal1 Tigcr thn

14.Shanka: wadi Taiga (a) kom wach Taiga a kom kaal Taiga -1uk i de wach wat& Tiger is mi>y - watdi Tiger ïs wming - d Tiger - look he? dinc watoh

15-Setita: kaai Taiga c d Tiger

Duruig this exchange Shanka begins gesturing at Taiga thus using her body as signalling

medium. She also uses the deictic verb kum to consvuct an alignment between participants and

an orientation to space. Because Shanka and Aiica are essentially sharing the same space ( M u a

on her lap) the ground for this deictic reference includes their shared location and perceptuai

orientation. But the actual placement of partiapanu at the tirne, with Taiga in the yard but

outside the space of the house, and the four of us seated arounû the bottomhouse table,

indicates that the deiaic ground here is more extensive and, in faa, indudes the shared

interactional space whidi the proximate paniapants are cutrently inhabiting - i.e. the

bottomhouse. This consvual of space is reinforceci in line 14, when Shanka specifies a location

THERE which Taiga is inhabiting against the same interactional ground.

As I noted earlier, while fernale cafegivers frequently eliat kaLr primariiy from female

children. I have never witnessed addt men engaging in such routines with children of either

s e x It might be surmised, then, that women in this community have developed communicative

practices such as kahg routines which partially rnitigate. if not challenge, the ways in which

their movement through the spaces laid out in a 1 0 4 geography is resuicted. In kaal routines.

women seize on the opportunity for social interaction that presents itself when a hiend or

4

family member is passing by. Such routines. &en. transform mundane and instrumental uses of

the road (Le. getthg from a to b) into a form of social engagement As sudi, kaling of(and

visiting exchanges which are thereby precipitated) is an important means bv which women can

circumvenr restrictions on their movement and, by extension, restrictions on the quality and

quantiv of their soaai relationships. Fmm this perspective, kal-eiicitation routines open up

possibilities for visiting and the dissemination of knowIedge between women. Notice that adult

women are not only soaaiizing novices in this way. but are also using them strategically to

achieve their own ends, that is c a k g in a friend to a i t

8.3 LEARNING TO KOS AND CLAIMLNC RIGHTS TO INHABIT INTERACTIVE SPACE.

In the next ewmple, a young girl, Mando. (3.3) is encouraged to kos by an older girl

named Kavita (9). î h e reapient of the kos in this projected frame is Kavira's little sister and

Mando's distant cousin. four yearsld Baby (4.9). The children had been playhg in the

bottomhouse of 1Gwit.a and Baby's home for some tirne with Mando coming and going

periodicaliv. Mando lives with her parents in her great-grandmother's house which is in the

same yard. She therefore has an ambiguous rehtionship to the emergent play area in

cornparison with Baby and Kavita.

1. Baby:

2.

3. Mando:

4. Bab.

S . Kavi ta:

6. Kumar :

Baby=Baby (Kavita and Kumar's younger sister, Mando's cousin) Kavita=Kavita (Baby and Kurnar's older sister. Mando's cousin) Kurnar=Kumar ( Baby and Kavita's brother, Mando's cousin) Mando = Mando (cousin to Baby, Kumar, Kavita)

EE GYAL (mw frarn deer) WA Y U KOM HIA FA? yu wan mi kyeer yu bak a hey girl. muvejhm, thme, what didyau m e heref..? ShaR I mnyyou back m ]

Linda dem? Linda S house?

wel guwan yu wee well go - on ymr way

tel-kos shi tell - curse her

riimuv from deer. riirnuv from deer (mandoo) nrnovefmm hem rernazfrmn hm Mando

( ) from hee? rimuv from he ( ) jhm here? removefiom here

rimuv an stee ( ) pikcha remove and stuy ( ) picturc

10-Kavita:

1 1.Kavïta:

12. Mando:

13.Kavita:

14.Mando:

1 S .Baby:

16.Kavitz

1 7. Baby

18.Ibvita:

1 9.Baby:

20. Kavi ta:

2 1 .Bab-

22. Ihvi ta:

23.Baby:

24. Kavi ta:

25.Baby

kos shii curse her

tel shi. kos shi tell her curse her

yu moâa. yu moda your motheryour moifrer

az shi kosh yu kosh shi bak as shc rursesyau - m m hrr back

fok af

kyaan duu y u notin kos shi bak (she) c m 't hann/troub[e yau nrrse her back

fok af Jitck

kosh shi bak nrrse her back

kosh shi bak na mek plee fu ko(r)s shi kosh shi bak eune lier back - don'rfwl around - mrse her - mrse her back

(yu kosh shi Manda) <whisper> yard curse her Munda

(yu noo) fu kos? yu laam? yon know h m to cune?you leam?

26.Kavita:

27.Baby:

28. Kavi ta:

kosh shi bak mrse hm back

(wee til shi dadii kom mi gu tel am) wuit till h n daàdy cornes - I am going ta td him

p u na kos shii? sh- shi na ga kors bak yu? did you nor cune hm? mustn 't she Lurseyou back?

In order to understand why Kavita instnicts Mando to bs speaticaily (not some other

verbal acfviw) and to understand the kinds of information that Mando is there Sy exposed to

we need to examine the place of the eiiatation within a smunding context which indudes

both preceding utterances and the background assumptions of contextuai relevance that are

partiaiiv shared bv participants. We might begin by noting that &os is conceived of as an aimost

physical, as opposed to verbal, action: the linguistic component of a locally recognized kos being

combined in practice with threatening postures and qui& movements of the hand and head.

The occurrence of the lexical items locaily dassified as m e wordc is thus a necessarv but not

sufficient criterion for a given utterance to be considered an act of kosing.

Given this ernphasis on physicalitv, it is not surprising that kos is suongly associated

with. and hence indexes, a certain structuring of interactional space. in the practice of kosing,

interacting bodies are mutually engaged by h e s of perception and close proximity. Because kos

as a practice is so strongly assoaated with interaction between highly p r o h a t e participants.

the metalinguistic predicate bs in eiiatations presupposes this proximity for the projected

context. Looking at the transcript we find that the small girl Mando is invited to kos in iine ( 1 0)

only after her antagonist(s) has h a t e n e d to move her from the immediate space of ongoing

interaction ( h e s 1-2, 4, 6 , 8 ) . A more detailed examination of the openllig iine reveals that

Baby is using a number of different indexicals to characterize her relationship to Mando and

284

Mando's relationship to the emergenr play area. From the s t m she uses an adcires strategy

which indicates an age-graded relationship. Such a c d to attention - EE GYAL Is typical of

parenu to chiidren espeaaiiy when combined as it is here with an increase in volume. From

there Baby goes on to use spatial deictics to establish the contested nature of the interactive

space. Babv first establishes an opposition between an unmarked and a marked W R E

which is coterminous with Mando's location (rnuvf im der). This makes Mando's position

cognitively salient. Baby then characterizes HERE as a space which encompasses both Mando

and herseif ( w a p k m hia fa) and opposes this with the space occupied by Linda hS3 As Linda

and her extended f a d y iive in the neighbouring yard Baby's opposition dearly establishes the

"here" of the emergent interactive space dong lines defined by property and inheritance

through kinship within the LOCAL GEOGRAPHY. The fact that Mando's young parent's have no

property of their own whereas Baby's f d y may daim to own a house and part of the yard is

h i g w relevant and foms an important interpretive ba&&op to the interaction. These spatial

relations are schematired in figure 8.4.

Figure 8.4 SPATIAL ARRANGEMENTS AND PAR~CI~ANTS (Transcript 2 , Iine 1)

Baby funher establishes the relatimship between Mando and henelf by suggesting that Mando

ma ab^ is aiternating between abrorbtive and non-absorbtive readings of HE= When she says "muv frarn deer" m E R E and HERE are aiiped exdusively with each participant inhabiting a distinctly bounded and seperate space. When she says "wa yu kom hia fan the udusive relation has collapsed and now HERE includes the region that was formerly undertood as M E R E .

285

should have a purpose in coming to Baby's famiiy's house - thus Baby asks w a p kmn hia FA?

This indicates that Mando is not a member of the immediate household and patrilocal

domestic unit since household members need not have a reason for ousupying the space of the

their own house. Findy Baby suggests that she kyem Mando back to Linda's house. Kger in

Guyanese Creole is not exactiy synonymous with English carry aithough it can indicate an

action whereby an agent manudy transports an object such as an infant. But kyeer can also

mean somethuig like English smrf - as in ma gu kpeerp a S u u Daik 'I am going to take you to

Suuz D a k " I t indicates that one person is more capable or agentive than the other.

So Baby is doing a lot of indexid work here to establish a context wherein Mando must

respect her authority. comply to her directions and recognize her own iimited priMlege vis-a-vis

a now estabüshed interactive space. When Kavita elicits Mando to kos in iine (5.) she is

consuuing a projected context in which Mando defends her pmrilege to occupy the same

interactional space as Baby. While Baby's contextuaibation relied on a variety of indexid

signals - a number of which 1 haven't discussed here - JSavîta's projected frarne uses lexico-

grammatical structure and the lexical-semantics of a verb which typifies speech interaction. As 1

mentioned above kos refers to an action that is conducted between highly proximate

participants and it thus presupposes this for the projected frame which Kavita is establishing.

Furrherrnore of aiI the verbs used in elicitations, hs , in terms of its lexico-grammatical

structure. rnost characterizes the argument which codes addressee as a patient, that is as a

directZy affeenrd entiv (see chapter 3). As such ffivita's elicitation to km entails a defence by

Mando of her privilege to occupy the space within which the interaction is taking place. Now a

defence of this sort must draw on some understanding of the way in which people can be

286

Iegitimatelv related to definable places (cf. Goodwin 1990). As 1 have aiready suggested, a

nurnber of criteria are used in assessing such multiple relations. In this case, Baby is deariy

prioritizing a notion of the inalienable right of a nudear family to occupy the space of their

own house. Kavita. on the other hand. seems to invoke two notions. While foregrounding the

fundamental autonomy of the individuai and assoaated privileges to go about one's daily

activities unhindered bv othen. she simuitaneously invokes the importance of patrilocal group

membeahip which. in some cases, can ovemde nudear group membership as the salient

cultural criteria relating people to spaces.

In peer-focused activities such as the one disnissed here involving eliatations to kos.

chiidren learn wavs of negotiatïng body and interactive space which are vital to the protection

of their individual autonomy. Such kos-dicitation routines seem not to be so highly gendered as

ones involving h l . Since both bovs and girls mut dwelop abiiities to defend their prideges to

occupy interactive space, cointidïng in many cases with patrilocal or nuclear group land-

holdings. this lack of gender differentiation is perhaps not surprising.

8.4 CONCLUSIONS:

Evidence of the kind presented here suggests that complex forms of spatial orientation

in communicative practice do not denve from a basic dyadic mode1 of the son suggested in

Saussure's diagram reproduced at the beginning of this chapter. Rather. spatial orientation

involves an a n c h o ~ g to local spaces and an abiüty to shift one's perceptuai focus in ways

- appropriate to an idhite variery of situations. Instruction here proceeds by modelling.linguistic

practice as an instance of a particular type for which contextual parameters, such as the relative

287

location. directionality and perceptud f ~ c u s of participants, can be set The knowledge derived

from elicitation routines speaki not only to the Md's developing linguistic and communicative

competence but also the M d ' s emergùig understanding of the way space is partitioned into

srnaller units. In addition, these subspaces such as yards, houses, and the road are always

understood and negotiated relative to one's social location. Thus children learn both a local

geography and spatial orientation for communicative practice in and through the categories of

gender and kin-group membership. To condude. it is apparent that, through elicitation

routines. children are leaming important lessons not just about the uses and functions of

l a n p g e but also about the nature of the sociai space within which everyday life d e s place.

By extension children also Ieam about the relevant social categones of kin-group membership,

gender and age ranks. The analysis of elicitations. then, highlights the way in which knowledge

about social categones. spatial arrangements and bodily cornportment in talk is embedded

within complex interactional routines. As such, they are an important resource for analysts and

language leamen alike.

At various points in the preceding pages, 1 have suggested that traditional ideas about

women and gender difference play an impottant role in the organization and interpretation of

talk Thus, in chapter 5,1 suggested that women's avoidance of the acrolectal pronoun variant

ai was motivated by a need to protect oneseif from charges of acting bWp. At the same time,

women tend to avoid solidaritv-coded am because such usages can potentiaily mark them as

c o u n q , stupid and ignorant ( e s p e d y when am is used in its least favoured context - for

reference to animates). In terms of discourse Level p a t t e h g and self-cornportment, 1 argued in

chapter 6 that the interpretation of women's role in verbal dispute is strongly infhenced by

dominant ideas about the relation between language and gender. The suong comection

between argument and femininity is captured in the proverb, man strengt de in ii han, homan wan

de in shi mout "A man's suength is in his hand, a woman's is in her mouth." The idea that bwing

(a name given to women's argument in Guyana) is essentially ferninine (and resuicted to the

working class) has also iduenced scholarly writing. Thus Edwards ( 1977: 196) working in the

framework of the ethnography of speakingl writes:

Busin interchanges are engaged in by working dass women in Guyana. The usual social purpose is to setde an q e - p a s dispute between soaaily equal women. Each contestant tries to re-establish socid equalify in the face of what is perceived as an attempt by her

1 Edwards. embracing the tools of ethnography of speaking, does not recognize the ideoiogical and intrrpretive nature of metaiinguistic terrninology ("named speech events"). Speech practices and their objective characteristics are thus equated with the terms and definitions native speakers give to them in the process of interpretation.

adversarv to assert soaal dominance.

The restriction of busing to working dass women is made even more expliat in a long f m o t e

in which Ed\vards ( 1977:2 1 1-2 12) suggests:

Middle class women do not engage in burin contests since among middle dass people busin is a strongly stigmatued working dass behaviour pattern [...] In conflict situations (induding those invohring p - p a r ) middle-class people in Guyana generally defend their social rights and points of view with reasoned argumentation - fkquently, but not necessarily. in the presence of others who act as arbiters. These altercations. however, never become as loud or as publidy onented or as personah/ abusive as buriru [...] Because men do not burc each other in Guyana, one frequendy finds cases where women bure against men. secure in the knowiedge that men are prwented by the social rules from responding in kind. If a man in such a situation attempts to buce, he opens himself ro the insult of being labelled an anti-man (Le. a homosexuai) .

I t is, however, hard to reconde these assertions with Edward's own data. In the example

Edwards presents, Johnson, the curen t mate of one woman and the former mate of the other.

involves himself in the dispute from the s t a h His invoivement. however, goes completelv

without comment in the analvsis. In the midst of what Edwards calls a "classic" busin session,

Johnstone is most certainlv invoived. The following is an euierpt from the uanscript presented

in Edwards ' ( 1 9 7 7:2OO-20 1 ) paper.

lohnson Jean Johnson:

Millicen t Joan: Millicent Joan:

Look wornan, you're disturbing me. You are supposed to pay some money to the court What happen? Look woman. why you don't stop pestering me? I don't want hear anything from you about any money. If you want your money, why you don't go to the court? 1 don't have any business with you. Why you don't drop dead? Go dong about your business. eh! (Johnson leaves the window, door being shook violently.) (to Joan) Why you don't go away and stop bothering people. I didn't come to you. 1 corne to Johnson. Well if yu come to he. yu come to me too. How you come in this? Why yu don't keep out? (standing now on Johnson's mother's badc steps). Keep out? Keep out? He is me husband and whatwer involve he, involve me. And he ent want see you. (Johnson cornes back to the window) Why you don't keep out. yu red whore yu

(Johnson rushes to door. Releases bolts. Rushes out onto platform and cuffs Joan furiously about her face and head Joan falls. Battering continues. Millicent rushes out and pulls Johnson away. Johnson stops pwiching and looks up. The yard is filled with relatives and neighbours. Neighbours are looking from doors and windows. Johnson returns inside and doses door).

Remarkably. Edwards seems quite oblivious to the way in which participation is negotiated at

the beginning of this excerpt (with Joan attempting to address Johnson and Millicent daiming

to speak for hm). Even more remarkable is the way Johnson's violent outbunt is completely

ignored in the analysis. I t is as though the local ideology of speech events makes the native

observedresearcher unable or unwilling to recognize male involvement in routines that are

classified as buring. I t seems, then, that the linguistic ideology which assoaates disputes of this

kind with the expression of an essential femininity is powerful enough to shape sdiolarly

discoune on the sub ject.' Outside of scholarly discourse, at the more local village lwel. the

effea of this ideology is to frame women's disputes as inevitable (being the expression of what

is fundamentaily female) and divert attention away from the social causes of the c o n n i a One

might ask whether this ideology, implicated in a metalinguistic teminology which separates out

women's argument and accords it speaal maunent (Le. lexicalization) , is, in some sense

accurate. If, then, one were to compare verbal disputes which have wornen as their main

?lese stereotypes of creole wornen and argument were already present in colonial writings. In the collection edited by Abrahams and S m e d ( 1983) we ftnd several cases where colonial writers make a connection between women, dispute, irrationality and disoderliness. The remarks of Charles Wilhn Day ( 1852: vol II. 1 1 1 - 1 14) are typical:

A negroe market place presents some droll scenes. and a stroll through one never fails to repay a looker on. Billingsgate must hide its diminished head, m u t be silent, at the vituperative vocabulary of a couple of infuriated negresses. So excitable are these people. that a fourth of an hou never elapses without a 'scene'. Such gesticulation, such pantomime. such a roi1 out of unintelligible phrases. making it difficult to recognize one's own language.

participants with those involving men. would it possible to find signifiant differences in

conversationai structure? That is to Say. is the distinction made in metalinguistic tenninology

based on observable differences in the stmcture of the arguments themselves? In table 9.1,I

have compared four disputes which differ dong two dimensions: the gender of the participants

and the relatively public nature of the event itself. The latter is measured in terms of the

number of non-participants who were involved as overhearers.

-- - - -

Example 6.1 Chapter 6 women's pubiic dispute

Example 7.1 Chapter 7 men's public dispute

Example 4.1 Chapter 4 women 's non-public dispute

h p l e 4.2 Chapter 4 men's non-public dispute

overlap

aggessive upgradi ng interruption

non-ratification

suspension of turn-taking through pre- announcement

3rd party mediation

other initiated rrpaid clarification

Table 9.1 SOME FORMAL FEATURES OF DISPUTES OVER FOUR INSTANCES

In ternis of the f o n d features of conversational structure, the disputes seem to be just as m . .

obviousiv differentiated in terms of their public or nonpublic character as they are by the

gender of the participants. However, the ideological nature of m e t a h ~ t i c terminology is

seen in the wav it singles out women's arguments as a panicular type (despite the la& of

O bvious distinguis hing formal features) , focuses attention on them, and characterizes them in a

certain wav. This gendering of language and self-cornportment is hegemonic in so far as it has

become a part of the resources people use in making sense of the worid (i.e. traditional wisdom

and ucommonsense"). Such mtegies are, as 1 have demonstxated in chapter 6, fhdy

embedded in the eveqrday pranices through whidi people are gendered.

1 have pursued these questions. and raise them again here. in order to emphasize the

way in which the notion of speech community is inflected by power. I have suggested that

people invoke notions of the egalitarian speech community in talle In doing so they

altematively emphasize notions of autonomy or age-graded and gendered solidarity. Sùnüarly,

partiapants to an interaction invoke altemate notions of gender in order to frame talle. I now

turn to a consideration of the way in which challenges to the hegemonic gendering of self-

cornportment are embedded in the multiple ideologies of the egalitarian c~mmunity.

9.1 MODES OF RESISTANCE

9.1.1 THE "TRADITIONAL" PATTERN

While women's role in dispute processes tends to be interpreted as the expression of

something essentidy ferninine, this does not make traditional strategies of hurd taik necessarily

or cornpletely ineffective. Although, in the example 1 discussed in chapter 6,I have emphasized

die way in which Pria was interpreted as low-dass (and was therefore stipatized) such

strategies are. in faa. an important means by which women use traditional stereotypes to

mitigate forms of gender inequaliq in the village. Because women are considered prone to such

293

affect marked and public display, because they are women. they are expected to engage in

them. This, then. is a discursive space for women within the traditional pattern. During public

displays of this sort. women shame husbands (and other opponents) by exposing aspects of

their behaviour (or imputed behaviour) that would othennrise not be held up to public scmthy.

Women, for instance. frequently suggest that their husbands do not live up to "traditional*

notions of masdinity. Thev mav charge that their husbands are unable to provide for their

families. that they are socidy impotent (Le. that they are bossed around by weryone else -

that they can't stick up for themselves), or that they are s d y 'inadequate" (that they are

a u n p m uhomosexuai*). This kind of strategy uses dominant stereotypes of women (as prone

to verbal dispute) and men (as ideally "masculine") to publidy shame husbands. But such

suategies also work at a more basic level in so far as the wife's agressive role in the contlict taik

is itself an icon of the husband's faüing authority. Sudi verbai strategies thus draw on and. at

the same time critique, the ideology of age-graded and gendered solida* whidi presupposes a

relation of asvmmevical obligation between wife and husband. In making this critique women

often draw on notions of autonomy as the constitutive feature of the egaiitarian sociey. Thus,

in chapter 6. we saw that Pria daims status as a 'big womann who m u t know wrong and right

for herself. Women, then, in contesting the traditional gender roles manipulate both aspects of

the ideology of egalitarianism. However , whiie contesting the details of panicular ins tantia tions

of the traditionai gender pattern (such as in a par t idar marriage). these strategies.

nevertheiess, seem to reproduce the hegemonic gendering of talk In su& performances women

appear divisive, prone to argument, excessively emotional and even irrational. Such strategies

do not. therefore. seem capable of deconstructing the dominant notions of gender

differentiation and inequalitv in the vükge.

9- 1.2 OTHER FoRMS OF RESISTANCE

In conuast, the move of some women into what are traditionaily male roies does seem

to deconstnict hegemonic notions of gender Merence. Women may occupy both working dass

and more upwardly mobiie 'male" roles. In the fkst case, women who, for vaious reasons,

become the sole economic providers for their households often dt ivate public faces that are

parricularly assertive. Most often, these women eschew puformances of passion iike the ones

noted above and instead prefer more 'controiied" argument and debate. Although deft in the

serious. decision making genres which are typicaily dominated by men, these women also

engage in gyaf which is punauated with sunial humour. Although these women do nor move

freely in male spaces such as the rumshop, they do often hold their own private drinking

sessions during which they sing love songs in which wornen figure as the sexual and emotionai

aggressor. Such occasions also set up contexts for otherwise demure women (who may be

uivited) to explore other foms of gender performance. As Godwin ( 1988, 1990) has

consistentlv demonstrated the meaning and significance of gender ciifference emerges with

parcicuiar contexts of activity. Earlier, 1 mentioned that widowed women, many of whom fail

into this categoxy now under d i s w i o n , ofien h d themselves in a very tenuous position facing

penodic shortage more frequently than othen. The domestic authon- which these women

hold thus does not mirror their economic standing and, veiy often. they are some of the poorest

people in the comrnunity. Despite moving into many male roies, femaie household heads are

paid women's. not men's, wages. This said, the same women are, for the most part, removed

295

from the scanda1 network and find it l e s necessary to cultivate 'respectable" representations of

themselves. In fact they seem to explia* reject such concems for respectability in their s e d

jo king.

Women may &O occupy more upwardly mobile socbeconomic positions which are

.pically associated with men. As 1 mentioned in chapter 2, Ashan and her sisters. the Pakar.

became the ownen of a very successful shop in the vülage. Whiie the domestic structure in this

case was quite obviously age-graded, with the senior sister acting as famüy head it was not

gendered in the usual way. The brother was 'kept" by the sisters and was subordinate to them.

Ashan and her sisters expüarly rejected the idea that men were necessary to the ninning of a

successfui business. Aithough male workers were kquently employed on a day-to-day basis. ail

executive decisions were made by Ashan in consultation with the next oldest sister. Women Like

Ashan upset the tx-aditional hegemonic gendering of social roles and economic order. Their

movement into male d e s both in domestic structure and in the economic marketplace

effectivelv deconstnicts traditional notions of gender difference. The resultant uneasines which

resulted among others (both male and female) points to the importance of such strategies in

promoting social change. In interactions with othen. Ashan draws on both cypicdly male and

typicallv female speech styles. 1 have already Uustrated the way in which Ashan's ability to

frame interactions in t ems of both her positions at the head of the domestic unit (within an

age-graded and gendered solidarity) and as an autonomous individual is illustrated in her

variant pronominal usage.

9.2 S O C ~ T I O N , GENDER AND PowER

296

Women's important role in soaalization also points to possibiüties for resistance to the

hegemonic genderïng of s o a d roles and practices. However. in most cases, socialization

practices seem to reproduce rather than to challenge the traditional pattern. Thus. women are

frequentlv involved in the policïng of other women's movement. and in the shaming practices

which 1 documented in chapter 7 and which play into notions of 'respectability." Soaaiization

practices cleady reproduce both the ideologies of egalitarianism - as autonomy and as age-

graded and gendered solidarity. Such practices also seem to reproduce traditionai forms of

gender difrence and inequaiity. AU caregivers soQalize @ris and boys differentiy. Girls are

dressed up in pretty dresses and an attempt is made to keep them neat and dean. Boys are

generally aiiowed to nui freely. first naked and then in boktn "underpants." Male caregiven are

more iikely than women to engage in 'fear" routines with young boys (see chapter 6) but

generally the ciifferences between boys. on the one hand. and girls. on the other. is evident with

both male and fernale caregivers. So whüe caregiver strategies do not seem to deconsuuct

dominant ideologies of gender difference (in fact. as caregiver strategies. expücitlv oriented

towards reproducing the "wav things are" from the participant's point of view. they are more

iikelv to be motivated by a desire to inculcate such difference). they do include instmction in

routines which rnitigate the constraints on women's social lives. 1 have made this case in the

discussion of hl-eliatatiow and it should be noted that women frequently instruct srnaii girls

in nrcsing and hard talk too. As 1 have argued earlier. however, such strategies do not contest

difference per se but do offer partial solutions to those whose iives such Merence and inequaiitv

cons train. 0 . *

Socialization practices seem more devoted to the reproduction than the contestation of

e-xisting power relations. At the same tirne. such practices provide expliat training in

manipulating the multiple notions of community whidi are potentiaüy avaüable at any

interactional juncture. As such, soaalization practices, whüe reproducing the power relations

between men and women. are also vital to the reproduction of the means by whidi such

relations of inequaiity are mitigated and contested in t&

Rickford writes ( 1988:SS):

Onlv a handfd of studies within creolistics have deah with pragmatics, conversation and'discourse analvçis. or the analysis of speech acts or events, although these aspects have become mo& centrai within sociohguistics over the past decade. This is not in itseif a bad thing, since it could be argued that these kinds of approaches are less relevant in theoretical and practical terms to the kinds of communities in which pidgins and creoles are spoken.

WMe I agree with Rickford ( 198855) when he goes ont0 to argue that 'some aeolists are

revening to variation-bhd and context insensitive descriptions. whether in old-svle

suucturaList or new-style generativist molds". 1 find the analytical separation of udiscoune" and

"variation" troubhg. As a number of studies have demonstrated, m q variable features of

laquage are controlled bv discoune phenornena (for examples see Ochs 1986a, Eckert and

~McComeii Ginet 1995). As I have argued in Chapter 4, some aspects of pronominal variation

in G.C. require a discourse-based explmation. The generai point here is that qualitative and

qualitative perspectives, rather than being incompatible. can be mutually illuminacing.

The situation has changed significantly in the ten yean since Rickford made these

remarks. Increasinglv cornmon are fusions of mainstream Luiguistic perspectives (which has

rediscovered variation through optimality theorv) with socioiinguistic techniques and

mediodology (for example Sells. Rickford and Wasow 1996). At the same t h e . some

vanationists are moving closer to discourse-based approaches (which draw on recent work in

the study of conte-xt in linguistic anthropology). It appears that divisions are once again

asserting themselves between dosely related disaplines through new alliances and contingently

retiprocai perspectives. Meanwhile. linguistic anthropologists have been reassessing techniques

for linguistic tvpology and description in an attempt to darify the functionalit- of some of the

more troubling areas of linguistic structure. Finally, linguisticallygrounded social theorists and

conversational anaiysts are operationaiizing theoreticai perspectives on hguage that can

adequatelv deal with the complex CO-construction of gender. soaal location and activity in

linguis tic practice.

So where does creolistics fit into this changing landscape? One way to answer this

question is to ask what we don't know at this point. When 1 ask this question 1 am nor so much

concerned with knowledge in ternis of disaplinary divisions or theoretical schools. We certainlv

don't know much, for instance, about discoune markers in any creole variety (and the relation

of such elements to processes of grammaticalization). However the answer to the question is

not. I think, to send a hundred and one CA specialists out with tape recorder in hand.

S p e w g the precise meanings of eh eh in an inhite variety of contexts. whiie of potential

sigruficance to the development of a culturaliy sensitive CA. would not. it seems, be of much

devance to the vast majonty of creole-speakers in Guyana.

This is not to Say that CA or discourse based approaches should not be applied to

creole-speaking communities. On the contrary they may tum out to be indispensable. For when

we ask what we don't know it is the people and the communities that should be considered. not

abstract notions of a "language" or "the range of variability."

As such. one m e r to the question of what we don't lcnow would invohre the relation

of gender to language. As 1 showed in düipter 5. there are vexy few variationist studies of creole-

speaking communities that consider the import of gender in any depth. M e variationist

techniques have been very usefui in illuminating the relevance of socio-economic dass, we have

to admit that they have largeiy failed in adducùig the relevance of gender in the context of

creole-speaking comrnunities of the Caribbean. In chapter 5,I suggested that this situation

resulted not from any la& of diffexentiation in the speech community but from a general

misconception of the wav in whid, gender becomes relevant to social actors. Such arguments

have become increasingly common in gender and variationist studies. As it turns out. the

ùnponance of gender may only be revealed by perspectives that combine qualitative and

quaIita tive. variationist and discourse approaches.

ïhere is another reason why discourse and pragmatically based approaches to language

are relevant to the comrnunities studied. As a number of scholars have shown, this level of

linguistic organization is not neutral. Rather ideologies of language have the power not onlv to

shape the way people tdk, and interact generaily. but also to naturdize relations of power and

privilege. I have argued that the conceptual organization of contlict talk in the village

naturalizes relations of power between women and men. This ideology. embedded in ways of

thinking about language, rnakes women seem Urational. prone to argument and trivialues the

kinds of oppression that they face daily. The application of m e n t CA and anthropological

approaches is drasticaily needed in order to arrest the process by whidi these stereotypes are

recreated in the academic and appiied literatwe. Creolistics and aeole studies would thus. I

300

think, oniy benefit from greater cooperation between variationists. theoretical and descriptive

iinguists. conversational analysts, anthropologists and those working in the areas of soaal. post-

colonial and feminist theory.

APPENDIX 1

A BREF SKETCH OF THE VARIER

0.0 INTRODUCTION

The following is a brief sketch of some aspects of CC phonology, grammar and le&.

This account is intended to senre as guide to the tramaipu pmvided in the text and as a

gened refemece tool in the discussion of grammatical features and morphosyntactic

variab*. The reader is directed to the sources ated in the text form more detailed accounts.

1.0 PHONOLOGY

The set of phonological contrasts employed in CC and some of the allophonic and

sociolinguistic variation found in more creole varieties is discussed by Wells ( 1982) and in

SicineII (n.d.5). CC shares a number of weli known phonological features with other

Anglophone creoles. Most characteristic is the neuualization of the opposition between /th/ and

/t/ and between /dh / and /ci/. Thus, father is pronounced /fada/ and thing is pronounced /tin#.

Also. following WeUs ( 1982). we rnight note systematic reduction of consonant dusten as

another characteristic GC shares with the Anglophone aeoles generaiiy. As Weiis ( 1 982:566 )

notes, the "phonotactics of consonant c l u s t e ~ g is sociolinguisticaUy sensitive" and people

generailv varv soàally and stylistically in the degree to which they reaiize consonant clustea.

However in some cases, such as /sen/ for send and /hooV for hold an ahost invariant use of the

reduced fom may well indicate an underlying form which does not contain the final consonant. O . 0-

In formal speech styles, then, h a 1 dusters must be reconstructed and added to "reduced"

underlying foxms thus giving rise to the possibîiitv for hypercorrection (cf Rickford 198 7b.

302

Patrick and McEihinny 1994). Another feature that GC shares with JC (Patrick 1996) is glide

insertion. The palatal glide appears between velar consonants and a foflowing open vowel (cf

Weils L 98 2:s 69). Thus cat is realized as /Qat/, gnf as Igd, car as mad, can as Man/ etc.. The

glide does not appear in cases where the vowel is derived from a badced vowel in the l e d e r

language (Engiish). So wrn is not /kyaani/ but instead /kaam/ and c@ is not /kyafii/ but

instead /kafii/. In CC word final /d is variably realized. I t occurs most frequently in

monosyilabic worcis (Le. when it is stressed) such as car kyd. However. several items seem to

have been 1-ed as r-less and rarely occur in anything but r-less form /hia/ or /ya/ for h m .

/niia/ for near, and /de/ or /ded for thn. It is very infrequent in wtressed. word h a 1 positions

( teacher. fa ther, mother) .

The vowel system of GC is strikingly sunilar to that of JC (Weils 1 98258 1 ) . Exduding

rare acrolectal variants. the vowel system indudes the following phonologid contrasu:

Table t VOWEL INVENTORY FOR GC

The vowel /d is marked with an arrow to indicate the possibüiry for badonrard movement.

Certain lexical items, such as corne, are vaxiablv realized with a backed vowel as in h m / or

/hmrn/. Such variation is relatively rare however and seems to be restricted to particuiar lexical

items. Note chat in the basilectal system a nurnber of the low vowels and diphthongs are

merged. Thus the /d of other dialects (Ge& pot) is realited as /a/ as is the initial vowel in the

diphthong in boy - /ûy/ becomes /ai/. This makes for some rather amusing homonyms such as

l a v e r and linr /laid. The long / ~ d of other varieties of Engüsh (in RP. cloth, and tulk) is

303

redkced as /aa/ in basilectal creole varicties. This last feature appears to be a stereotype, in the

Labovian ( L 972) sense. of East Indian and rurai speech in Guyana (cf Ridcford 1987b). The

vowels of face /fees/ and place /plees/ are long monophthongs. There are a number of

phonolo@cal processes that have affected the shape of individual words. These indude

metathesis as in ask: and vowel insertion as in a m /aram/.

As the terms "English Lex.edw and 'Anglophone creolen indicate, the lexicon of GC

and related varieties is derived mainiy from the English dialem of Great Britain. In

predominantly East Indian viiiages however a number of semantic domauis are covered by

Hindi terminology. This is espeuaily the case for the relatively specialized vocabuhies of

kinship, cuisine, and religion. The Hindi component combines with an eariier African one to

produce a rather distinctive lexicon. In addition, as with other Caribbean Creoles, the CC

Iexicon includes a number of elements from Portuguese, Spanish and also from the English

working class düilects of the eighteenth and nineteenth centuries.

OLder people often remark that the Hindi kinship system is no longer undentood or

used properly by the younger generation. Nevertheless there are a number of Hindi terms in

evervday use in the village. They are listed below with the genealogical defmitions:

aja a ji

nana nani

chacha chachi

bonai bau j ii saroo

GENEALOGICAL RELATION --

fa. fa. fa. mo.

mo. fa. mo. mo.

1 fa. br. fa. br. wi.

, m. si. hus. mo. si.

mo. br. mo, br. wi.

si. hu br. wi. wi. si. hu.

grandfathet grandmother

grandfa ther grandmother

brother in Iaw sister in law brother in law

Table 2. HINDI KLN TERMS IN COMMON USE IN TME VILIAGE

Kinship terms are marked morphologically dong two salient dimensions: gender of the referent.

and nature of relation (father or mothef s side). Kin terms refexring to females have suffixl-il,

those referring to males have suf f ix / -a / . Root terms for kin on the mothen side begin with a

nasal segment (either /n/ or /rd) whiie terms for kin on the father's side do not (cf Rampaul

1978).

The set of Hindi kinship terms in the village is sornewhat convacted when compared

with non-diaspora East Indian communities (Rampad 1978). The contraction is not randorn.

rather it is the terms for ego's juniors which have been abandoned. In the village the very use of

a Hindi term is taken, in itself, as an addressee or referent honorific insofar as it indexes a

'traditional' contextual frame of respect The terms. as a paradigrnatic set, then, have taken on

this second-order indexicaüty (cf Silventein 1996). The use of Hindi terms for juniors (such as

beti "daughterm and beta "son") is therefore is rare and somewhat contradictory.

305

The sematic domain of cooking is similarly populated with many Hindi lexical items.

Thus greens are refened to as baigan or balanjii 'eggp1antn. bajii "calalodspinach*. uluu

' potaton. Curry is cooked in a kahani 'a round pot" and roti on the tawa 'a flat, thick pan".

When cooking the curry one mut let it bunfee 'let it cook covered, simmering in it's own

water". For daal 'spht peas cooked with water, garlic and oil" it is necessq to first chonkze the

masala 'add rnixed spice to hot oïl." The roiiing out of roti is refexred to as beeke. One popular

local Song goes; 'mamii beelee rotii. da& chonkee daal, mamü beelee rotii. bodoo üt out ad."

ReIigious prafüce is also commody taJked about using Hindi rather than English words

although people are quite cornfortable in exchanging church for rnundir or mosque. Esoteric

religious knowledge invoives a number of Indian place-names as weU as the names and qualities

of deities (Hanuman, Ram, Ganish, Ladunü etc-). Everybody in the village can desmbe what is

invoived in jhandil and various tvpes of Puja2.

Besicles the Hindi element in the lexicon, people ofien note the presence of what they

rem "creolesen words (cf. Bullen-McKenzie 1978). As opposed to the Hindi texms discussed

above these words are considered to have been borrowed from MeGuyanese populations or,

altematively, are reasoned to be the spontaneous aeations of East Indian immigrants. In fact

ma- of the lexical items and phrases can be show to have some comection (either ttuough

bonowing, calquing or loan translating) to African sources. Vihgers classify the following as

particularly ' creolesen words.

1 1hand.i *The religious ceremony of thanksgiving to Hanuman. performed by a pandit; it is held on a Tuesday or a Saturday."(Allsopp 1 W6:3 12)

* Puja "An act of payer and worship of a Hindu deity either by an individual or collectively in a ri tua1 seremony." (Allsopp l996:454)

- j d

1.

expression indicating something doesn't suit l YOU

to push-up, to alter the position of sornething

I

11 mannish 1 someone who engages in antisocial behaviour

*as

Table 3 WORDS CONSWERED PROTO~YPICAUY "CREOLESE"

a curse word. iiterdy - buttodcs

sosososUU

An important word of uncertain origin is "matü". This is used in a variety of contexts but

gossip or talking seaetively

usudv denotes a person who is of similar standing, status and age (cf. Wiliiams 199 1). One's

'mati-man" is one's equal. I t is aiso used as a reaprocal pronoun in expressions iike dm gii mati

jringk "Thev gave each other the drink "

Singular p r o n o m are discwed in some detail in chapter 4. The plurai 1st and 2nd

pronouns (nbi and niyu) are discwed in chapter 5 and see also chapter 6. Here 1 jusr sketch out

the svstem for interrogative. relative and reflexive pronominal reference.

Interrogative pronouns in GC share a number of characteristics with those in other

varieties. Bickerton ( 198 1) and Lefebvre ( 1993) have both used interrogative pronouns as

sources of evidence in their respective daims conceming creole genesis. The interrogative w . .

pronouns can be classified into morphologidy complex and simple groups.

I I - -- -.

huu "whon

SIMPLE

wa

wich

Table 4 INTERROGATIVE PRONOUNS

A basic mo-way distinction governs the choice of relative pronouns. Wa is used for

embedding ail clauses except those that refer to locations and some human referents. In the

former case onh we or wiraid is appropriate. In the latter case wa altemates with huu (on other

cornplex sentences see 1.6- 1.7).

1. di plees we yu a liv 'The place where p u he."

2. di gyai wa/huu yu laik "The girl that you me."

As Rickford. notes some speakers also d o w O-marking of relative clauses in some contexts

( Rickford 1 98 7b: 1 50- 15 1 . Bickerton 1 98 1 :62). Reflexive pronouns usually Lake the form of

corn pounds on non-reflexive foms. thus: iiself; shiisey etc..

3.2 DEIMS:

The system of spatial and temporal deixis is discussed in chapter 4.

"rvhat"

"whichn

3.3 AR~CLES:

The anide system which marks an noun phrase as definitdgeneridspeatic has received

some attention in the literature on CC. Bickerton uses this systern in his arguments for a

language bioprogam ( l977b. 198 1). The GC system is schematized alongside English in

Bickerton 1 98 1 . The examples are from Bickerton ( 1 9 7 7b:S8) :

COMPLEX

wa mek

wisaid (wee)

'whyn

" where "

mi bai di buk, "1 bought the [presupposed known to speaker] book"

mi bai wan buk. '1 bought a [presupposed not known to speaker] bookn

mi go bai buk, "1 s h d buy a book or books [even speaker does not know

which]. "

In fact, the system is mu& less monolithic than Bidceaon suggests (cf. Muhene 1992 on the

notion of non-monolithic grammars). This is due in part, no doubt, to a constant interfaence

with English which accounts for the frequent occurrence of the undenpecified aaide a in

discourse (cf. the "aeolen sarnple reproduced a t the beginning of chapter 3 where a serves as a

generalized article in ail contucts). The form is often used rather that wan as an indefinite

specific anicle as in S.E.. However in many cases n occurs as a reduced form of demonstx-ative

da. In addition a is used in many cases where both definite and indefinite senses are possible as

i n , p gu lait op n faia. " p u will Iight a/the fire.".

4-0 THE SIMPLE VERB PHRASE

4.1 COPULAR AND ATCRIBUTIVE PREDICATION

Bickerton's ( 1973a) analysis of the CC creole continua isolated a set of stnictures in the

basilect which he considered copular on the basis that comparable English constructions

emploved the verb "ben. This set induded the following types:

NP ( +cop) + Possessive

NP ( + cop) + Attributive

NP (+cop) + Locative Wan aligeeta de a bak 'There is an alligator Locative 1 back there."

' Rajish a di bai fada l

Di net a mi wan

Table 5 U C ~ ~ ~ ~ ~ " STRUCTURES IN GC

Bidcerton argued that the basileaal copular system distinguished between equative and

-- - --

'Rajish is the boy's father. "

"The net is mine-"

possessive (marked with a), attributive (marked 0) and locative (marked with de) and that this

-

Equative

Possessive

system slowlv gave way to the English one using "to be" (Bickerton 1973a). Winford

( 1990,1993) and others (Alleyne 1980) have questioned the assumption that one can isolate a

unified category of copular constructions in the basiiect. Winford ( 1990, 1993) shows that a

and de are used in a wider range of contexts than Bickerton originaw reported. For example de

is used not only in locative expressions but also in assertions of absolute existence like the

7. work na de. "There is no work."

De is also used in existential structures where English uses t h m + i s :

8. miit de in de Thele's meat in i t "

9. piipol de a tap de 'There are people up there."

Fina& de is also used in certain attributive structures:

10. di piknii de smaal yet 'The child is stiU smalLn

1 1 . di piknii de in CM yiir 'The chiid is three years old."

Winford's reanalysis of attributive structures shows the degree to which basüecral (and

probablv mesolectal) gramman differ from aaolectal ones, providing strong evidence that the

' polvlectal" and "seamless " nature of the continuum has been overstated ( Bickerton 1973 b) .

Widord shows that, unlike adjectives in more standard varieties of Engiish. attributives in GC

and simüar varieties combine with various T(ense) M(oda1ity) and A(spect) categories

(examples 12 & 13 from Winford 1993: 185):

12. di piknü gd don/ bin/ mos sik 'The duld wiU be/is akeady/waslmust be si&"

13. di waata g d don/ bin/ mos blak 'The water wiU b d i already/was/must be

black "

Winford argues that the sud, sentences 'demonstrate that items like sik. big. b u . ool. etc. in

fact fundon lüce intransitive verbs predicating a certain characteristic or q d t y of the subject

NP ( 1993: 185)." Winford goes ont0 show that the behaviour of such predicates with regard to

TMA categones is not uniform. Rather only the attributive predicators which Winford labels

Pwsrcu P R O P E R ~ can combine with progressive a marking (ewmples 14- 16 are from

Winford 1993: 186):

14. di piknii a sik 'The child is getting si&"

L 5. di waata a hat 'The water is getting hot."

16. di daag a ded 'The dog is dying."

Other attributive predicates combine only with k m 'corne", gct 'get" or ton 'turn" (cf Winford

1 9 93 : 18 7). nie predicates that can combine with progressive markling share a retatively non-

s tative , dpamic character. Winford therefore suggests that one can divide attributive

predicates into processual and stative subclasses. The evidence of possible transitive uses of

3 11

some attributive predicates which is presented by AUeyne is also relevant (cf Aiieyne 1980,

1984, 1987. Winford L 993: 195). AUeyne argued that 'adjectival" uses of sik 'si&" and raip

'ripe" were in fact passives derived from aaive sentences. The sentences in (a. ) would then be

derived fiom active counterpaw listed as (b.) .

17a. di piknü sik "The child is s i k m

1 7b. di waata sik di pilaùi "The water made the M d si&"

18a. di mango raip 'The mango is ripe. "

18b. di son raip di mango 'The sun ripened the mango."

Whether we agree with AUqne's analysis or not, the transitive uses of attributive predicates is

further indication of their processual and dynamic (rather than stative) character.

The discussion of 'copular" stmctures and attributive predication specifically highlighu

the fact that certain areas of the basilectal grammar diverge from S.E. and other aaolectai

varieties in terms of underlying grammatical categories (cf. Bickerton 1 9 75. Riddord 1 98 7b).

Such grammatical focussing results in a polarization of basilectai and acrolectal varieties which

is in some wavs similar or parallel to the kinds of polarization one sees at the level of language

use ( cf. Rickford's 1 9 8 6 discussion of dass polarization in the use of pronouns) . .

4.2 T(ENSE), M(ODAL~TY), A(SPECT) CATEGORIES

The system of tense, m d t y and aspect has been the subject of much debate in the

Literature on GC in parricular and creole laquages generally (Bickerton 1975. Civon 1982,

?$divene 1984. Poilard 1989, Singler 1990). For the moment 1 want only to discuss the 9

general character of each category in GC and give some sense of the way they combine. I

discuss past marking in tems of temporal de- in chapter 5.

As is weii known, CC makes use of a distinction between stative and nonstative verb

Npes (cf. Bickerton 1975, Winford 1993). The interpretation of unmarked stem fonns

depends on whether the verb is stative or non-stative. Unmarked verb stems which are dynamic

or non-stative are past and compieted.

19. mironuIran"

20. ü iit "he ate

Stative verbal roots which are unrnarked refer to States presendy in existence:

21. minoouIknow

22. mi waan wan '1 want one"

TMA markers. foilowing Winford are outüned below, for GC:

TENSE

MODALS

past future prospectual - - --

progressive habituai perfective completive

deontic epis temic

bin gu a gu ( W/ non-statives onlv)

a ( non-statives only) ddoz O don

mos. kyan. kyaan. fu mos. b a n

Table 6 BASIC -GS FOR TMA MARK~ERS (adapted from Winford 1993: 99)

With dvnamic verbs. TMA markers combine to give the foilowing specifications for the

m(i) a iit "I'm eating"

mi don ut " I'm finished eating"

mi bin üt "1 had eatenn3

mi gu iit '1 wilYwould eatn

mi fu iit " I am to eat"

mi kyaan iit "1 can't eat"

mi mos iit "1 m u t eat"

mi bin a iit '1 was eating"

mi bin gu iit "1 was going to eatn

mi bin fu iit "1 was to eat."

m(i) a gu iit '1 going to eat"

mi bin kyaan iit '1 couldn't eatn

This gives the reader some idea of the way in which TMA elements combine in simple

clauses. A great deai has been written about the rneanings and ordexings of TMA markers and

cannot be repeated here. The interested reader is directed to the excellent o v e ~ e w in Winford

( 1993) or the works of individual authors (Bickerton 1975, Givon 1982. Mufwene 1984,

Poiiard 1 989. Singler 1 990).

5.0 PASSIVES

GC passives resemble in many respects the passives of related varieties. As Winford

Bin requires a more complex analysir and is not really amenable to English glosses of this kind (cf. chapter 3)

3 14

notes ( 199 Ib:ïS6), most studies on mole passives begin by pointing out that many of the

features of the 'European canonid passive are not to be found in aeole languages." However.

as Winford goes on to argue, drawing on the work of Sierierska ( 1984). passives do not

g e n e d y share a great deal of features aoss-linguisticaUy (and in some respects the European

canonical passive is rather peniliar). W i o r d adopts Siewerska's characterization of the passive

as a construction in which:

(a). There is a corresponding active. the subjen of which does not fùnction as the

passive sub jea;

(b). The event or action expressed in the passive is brought about by some person or

thing which is not the passive subject, but the subject of the correspondhg active;

(c). the person or thing if not overt is at l e s t strongly implied.

According to such a definition CC has a nurnber of different passives including the

following:

36. waeva ting Noel gu niid kyan put wansaid 'Anything that Noel wiii need can be

put on the side."

3 7. di rais don kuk 'The rice is hished cooking. "

38. aal di ting dem sel "All the things were sold."

Note that in CC the passive interpretation is not indicated by verbal morphology as in S.E..

Rather a passive interpretation is derived h m the interaction of three factors; 1. word order. 2.

the inherent lexical semantics of the verb. 3. the relative animacy of the NP in subject position.

3 15

Craig ( 19 753) cornes to simikr conclusions when he argues that there are t h e conditions on

passivization in JC:

(a). only activip-do-transitive vehs can passivize

(b). no stative verbs can passivize

(c) . the sub ject of the verb must be incompatible with its inherent activity

Echoing Craig's ( 19753) condition (c), AUeyne ( 1987) argued that sentences such as (39) with

animate subjects are iikely to be interpreted as active rather than passive since human subjects

are more ükely to be agents than patients in this context (or at least this is the default

interpretation for this syntactic context)

3 9. 'di pücnü-dem kis aredi

However a general consvaint on non-animate subjects does not appear to hold for GC passives.

Thus (40. from Winford 199 1b:26 1) is acceptable:

40. vu stab "You've been stabbed."

As Winford ( 1 99 1 : 2 7 1 ) illustrates passive construction are formaUy indistinpis hable fiom

anticausatives in GC and related varieties. Anticausaive constructions are fonned by the

*samen svntactic process which makes a transitive object the subject of an intransitive.

4 la. dem bai brook di few 'The boys broke the fence."

4 1b. di fens brook "The fence was brokerfïhe fence broke-"

In many Creole varieties induding GC such consuuctions are actudy ambiguous between

passive readings, in which an agent is implied (S.E. 'The fence was broken") . and tme

3 16

anticausative readings in which the event or action seems to occur spontaneously (S.E. "The

fence broken).

CC marks complementation with se or& (other foxms of complementation induding

those with causative and immediate perception type mauix predicates are discwed in Winford

1993). Alrhough both types of complementation have been the subject of a good deal of debate

it is possible to gîve a purely desmptive account without becoming enmeshed in the theoretical

issues. As Winford ( l993:29 1 )notes:

Certain subdasses of CTP's (complement taking predicate) are assoaated with specific complement types in CEC. Thus, predicators which assen, report or express uuth-value judgemenu about their complement tend by and large to have indicative type complements with ITR (independent thne reference). By contrast. CTP's which express commands. reques ts. desires etc. qpically have non-indicative type complements with DTR (dependent t h e reference) .

Generallv. se htroduces S-like complements with I T R That is to Say that 'the time

reference of the complement is in no way logically bound by the time reference of the ma&

predicate" (Winford 1993: 1 90 ciùng Noonan 1985). Such indicative complement types

contrast with the non-indicative type which do not ailow overt TMA specification. are qicailv

in irrealis mode. mav or may not be S-like and are introduced byfu. It is possible to divide the

predicates with which se combines into a number of distinct dasses. The foliowing classes

incorporate the verbs most commonly combining with se. ( Winford 1 993:292-293 mentions a 0 . .

number of other classes of verbs.)

3 17

A. UTTERAN& PWICATORS: In such constructions se introduces a complement of reponed

speech. The construction may report the speech directly (43.46) or indirectiy (42,44.45,47). In

indirect constructions we find a shift of deictia so as to characterize contextual pazameters in

terms of the reporthg rather than the reported ment In both cases the se is optionai.

Predicates counted as memben indude both generalized verbs of speaking (td, tank) as well as

predicates whidi more precisely identifjr the Uocutionary force of speaking (kWan& kos, beg nkr,

Shi na tel mii se shi brook am "She didn't tell me that she broke i t n

gyal a taak se maa yu goo a tap yu üt brekfos? "The gïrfs said 'Ma, you went

upstairs, did vou eat breakfast?'" (Riddord l987b:229 iine 1 132)

Shi a bar i l se -wel se da Mis- wa mek Mis na wan maxiid "She q u e l s saying

-weU saying- saying that Miss- why doesn't Miss want to get mamed."

daadii i a kos se aiw düz a ta* out shi 'Daddy, she is cursing saying that al1 of

y u are talking about her."

an den mi tel am -aa- den mi beg am see i na büt mi soo oo bai oo bai 'And

then I told him -aa- then I begged him not to beat me saying '00 boy oo boy'"

Rohan a dout se i na get di juuwol 'Rohan was arguing that he didn' t have the

jewel. "

B. PROPOSITIONAL A ~ E PREDICATORS: In example (48) the se follows (biliiv) a verb which

expresses "an attitude regarding the vuth of the proposition expressed as [...] complementn

(Noonan 1985: 1 13, cited in Winford 1993:292). In many cases the "attitude" indicated for

the subject diverges from that of the speaker (as in 48).

48. ii biliiv se ii tuu big fu kom "He believes he is too important to corne."

C. PREDICATORS OF KNOWLEDGE AND ACQUISmON OF KNOWLEDGE: Here what is desuibed is

the "state or manner of the acquisition of knowledge" (49.50) are examples with predicates

ondustan "undentand" and fain 'fmdn.

49. shi na onciastan se da a di bai laik di gyal 'She doesn't understand that it's the

boy who likes the girl."

50. (a)n yu faui se(e) yu d m si dong a di step an i gun staat taak "And you sit

down on the step and he would start to talk" (Riddord L987bA78, iine 687)

D. PRETENSE E'REDICATES: Here the speaker indicates a disaepancy between the default

interpretation of some action (induding speaking) and the 'reality behind it" (a discrepanqr,

that is, between appearances and reality.) Predicates in this dass indude pfee upIay" andfuul

" fool".

5 1. na plee se wel yu greet fu see eniiting to mi 'Don't act like you are to good to

Say anything to me. "

6.2 COMPLEMENTS WITH F U

In contrast with se complements which are S-like and have K R (independent tirne

reference), complements tofu are generally reduced and have DTR (dependent time reference).

The analysis of* complements is extremely complicated and scholan disape on a number of

3 19

points induding the categorid status offu (Winford 1993) and the nature of the complement

inuoduced (Le. whether i t is infinite, finite or neither cf. Bickerton 1 98 1, Mufwene 1 98 9,

Washabaugh 1975.1977. W i o r d 1985b). For our purposes it is necessary only to give some

indication of the tvpes of smaures in wtùchfu is used to introduce a complement Some

examples are @en below (52-54 from Winford 1 993:309):

52. Jan staat (fu) mek monü 7oh.n started to make money."

53. Jan tel Meerii fu rnek monii "Iohn told Mary to make money."

54. Jan wok haard fi^ mek monü 'John worked hard to make money."

& Winford ( L985b. 1993) has noted the complements in s u c h c a x s desuibe something

"hypotheticd or unrealized". a possible future and it is in this sewe thatfu's functions as

complementizer and as modal aU)CiLiary are semanticaily related. Winford classes the predicates

which take complements introduced byfu in the following way:

A. ASPECKJAL OR IN CE^: such as s t a a t ' s tar tn do not have the ixreaiis mode of other types.

Winford ( 1985b) argues that such instances offu are modal arurüiaries not infinithl

complemenWers (as suggested by Bickerton ( 198 1) and Washabaugh ( 1977)) and that the

inceptive matrix verb takes a VI? complement Su& an analysis is supported by the fact that

tionting such complements is unacceptable as it is for other VPs.

55. *a Wfu mek moni Jan staat

Winford h d s fur<her support for his position in cases wheref;/f. altemates with other

auxdiaries such as imperfective a (56-57 from Winford 1993:3 14):

56. JanstaatfuWU~ohnsta~edtocry."

57. Jan staat a krai 'John started aying. "

F W y as Winford remarks his exphnation better accounts for the aiternation betweenfüfk and

O in such structures.

B. MODAL^ PREDICATORS: Predicates sudi as gat 'got" and bong 'bound" take a sub jecdessfu

complement as in:

58. hou moch touwil yu gafu get wen yu de a grong? 'How many towels do you

need to get when you are on the ground."

C. D E S I D E R A ~ AND ACHIEVEME~ ~'REDIcATEs: Predicates such as waan 'want" and hoop

"hopen express desire that the complement proposition be realized. Winford suggests that these

can be dnnded into two dasses. CLASS A have experiencer subjects and aiiow a subjectiessfu

complement onlv if the undeatood subject is coreferential with that of the CTP (complement

taking predicate) as in the case for agni, hwp,fzü, pmrnb. These resemble the dass of

achievement predicates (laam. MO. manij chrai) which also takefu complements only in cases of

agent-agent coreferentialitv. CLASS B indudes predicates such as waun, ehpek, raada with which

sub ject-subject coreferentiality is optional.

, 59. ii na nü hi gu 'He doesn't want to go." (Class A)

60. yu onciastan wa a chrai hi shoo yu? 'Do you undentand what 1 'm trying to

show you?" (Achievement)

6 1. mi noo hi taak tu gad '1 know how to t& to Cod." (Achievement)

62. di pikni waan fu slüp 'The child wants to sleep." (Class B)

63. di piknii waan slüp "nie chiid wants to sleep." (Class B)4

63. di piknü waan Meerii fu kom bak tunait 'The duld wants Mary to corne badc

tonight. " ( Class 8) (example from W i o r d 1 9 93 :3 1 8)

Fu is optional in cases of agent-agent coreferentiality, e.g. "Jan waan (fu) slüp". The basic

finding after many years of research in this area is thatfu-complements cannot be equated with

SE structures like to-complements (Bidwon 198 L) orfor-to-complements (Washabaugh 1975.

1977). A geat deal of evidence has been marshalled in support of such a uniqueness position

with regard to these structures. 1/ In demeolized varietiesfu is often replaced (in some contexts,

e.g. with tel) not by to but by mos (another auxiüary); "replacement by mur is in keeping with

the typical strategy. adopted by the CEC speakers, of irnporting SE forms whidi appear to

them to be hinctionaily equivdent to partidar aeole forms, and of r e p l a ~ g the latter by the

former without anv immediate diange in CEC grammar itselt " ( W i o r d l985b:604, cf. also

Roberts 1980). 21 Despite Bickerton's assertions to the con- (and Winford's apparent

acceptance that changes of such a kind are taking place in basilectal CECI5 "fi does not occur as

a pure infinitive marker as SE to does in the tvpe of sentence Like To err is human." (Roberts

1980:30). 31 Uniike for-to complements in SE (For Mary to seU them wodd be difficult) thefù

4 Aithough as Winford ( 19933 18) notesfi is optional here. there is a potential semantic ciifference. Waan often occurs in more metaphoric c o n t a where it is roughly equivdent to "needs" as in di wii l waan ail "The wheel needs oil", or ii pan w a m rob "The pants need to washed" In such cases the subject which appears to be doing the "desiring" is not animate. The "desiderative" sense of w a n appears to be suspended in such contacts. Note thatfu is not acceptable here, thus *ii pan waanfu wmh. So the sentence di piknii w w sliip is arnbiguous and indicating either that "The child wants to sleep" or "The child needs to sIeep."

9 . - Bickerton ( 198 1 ) allows as grammatical the constmctionsfi kdz tu&a na bin iizi "To catch a

tiger wasn't ç q " Winford ( 1985b:606ff) provisionally dlms such sentences and provides an altemate explantion to that of Bickerton's based on extraposition.

structures in roughly equivdent i haadfi Meeni& sel d m so cannot be moved to subject

position because it is not a constituent (cf Winford 198%:609) .6

7.0 VER^ SERIAL~ZA~ON

Serial verb C O I I S U U ~ ~ ~ O ~ S (SVC) are found aaoss a great variety of language groups

induding those of West Afnca and Papua New Guinea (Foley and Olson 1985). 'Zhey are also

well attested in the aeole languages of the Caribbean. Typid examples are (b W i o r d

1993):

64. Mieri waak go a maakit " M q waiked to the marketn

65. Jan bring moni gi shi "John brought money for her."

66. Di pikni tail paas mi "The chiid is taller than me."

The examples Uustrate the various functions associated with the second verb in a serial

construction. Grammatical relations (directional in (64), benefaaive in (65). comparative in

(66 ) .) are marked bv a verb where in other languages they are marked by prepositions or

complementizen. However, the second verb does more than "mark" relations between verbs

and their arguments and as such serial constructions are not solely "valence inaeasing

stntegiesn although this seems to be one of their functions. As Lord ( 1973) remarks the second

verb "is always in some sense a M e r development, resdt or goal" of the first. The degree to

which the two verbs are semanticaiiy integrated depends. in paxt, on the level at which the

predicates axe "serialized" (see below) .

As Winford ( 198513) notes such structures rhould not be confused with instances of front fowsing which involve the fronting of constituent as in a@ mek moni lan km. For an exphnation of why such structures are grammatical see Winford ( l985b, 1993).

7 - I SVC'S AND OTHER MULTI-PREDICATE CONSTRUCTiONS

Sebba ( 1987 cited in Winford 1 W3:2 12) has suggested the following defining

characteristics of a serial verb construction:

in a sequence of the form V, ... V,

(a). Bath V, and V, must be lexical verbs. i.e. m u t be capable of appearing as the

only verb in a single sentence.

(b). If it is possible to conceive of V, and V, as denoting separate actions at all. then

both V, and V2 must be interpreted as having the same tense and aspect. Thus, for

example V, may not be interpreted as "past" if V, is interpreted as "futuren.

(c) . There m u t not be an ascertainable clause boundary between V, and V2, i.e.,

they must be within the same dause.

(d). No conjunction should separate the verbs in sequence.

Citing Foley and Olson ( L984), Winford adds that WC's "are heavîiy constrained so that some

core arguments are shared by aU the verbs in the seriesn (Foley and Olson l984:X). The most

common restriction is that aii vehs in the constmction m u t share the same subject.

Based on the above criteria it is possible to distinguish SVC's from other forms of verbal

linking such as coordination and parataxis. Distinguishing cases of coordination from cases of

serialization is fairly straight forward. Winford @es ( 1993:2 14) the following examples:

67. Jan bai wan buk an gi di pikni "John bought a book and gave (it to) the child."

68. lan bai wan buk gi di p h i *John bought a book for the child."

69. Jan ron an kom ina di hous "John ran and entered the house"

70. Jan ron kom ina di hous " j o b ran into the house."

As the gloses indicate the two structures are semanticaIly&stinct. Whereas in the coordinate

sentences (67.69) the meaning of the second verb depends entidy on its independent status.

in the SVC (68. 70) the second verb takes on a specific fundon related to its role in the

partidar constmction. In addition to semantic differences SVC's, unüke subordhate

structures, never contain a coordinating conjunction. Finah, SVC's never d o w independent

tense/aspea marking of the serialized verbs. This contrasts with coordinate constmctions as the

following example illustrates ( from Winford 1 W3:2 14):

7 1. Jan bai wan buk an go gi di püaii tumara "John bought a book and will give it

to the chdd tomorrow."

SVC's must aiso be distinguished hom paratactic constnictions. In both paratacric structures

and SVC's we find a "subjea NP foliowed by a series of verb phrases, without overt marken of

coordinationn (Winford l993:Z 16). The two constructions cliffer in a number of respects. First

paratactic constructions ailow independent tenshpect marking of the linked verbs. According

to Noonan this correlates "a aucid semantic difference. namely that paratactic constructions

contain two assenions, ie, each dause is separately asserted. whereas serial consmaions

contain just one, encompassing the entire constructionn (Noonan 1985: 77 cited in Winford

1993:2 16). Furthemore paratactic constructions are subject to Ross's (1 967) coordinate

structure consuaint which prohibits extraction from a coordinate stnicture. Winford iuusuates

this with the following example:

72. di haak kedi di diikin iit it. The hawk caught the chicken and ate i t"

73. '(a) wa di haak kech - üt (it) ?

8.0 FRONT FOCUSSING AND COPMNG STRUCTC~RES

AU languages have devices which alter the syntactic packaging of information in the

clause to achieve pragmatic effects such as bringing paxticular referents into focus. The it- and

Wh-defts dong with cases of left and right dislocation are examples of such syntactic devices

which operate in S.E. (passives. dative movement aiso invoive information packaging in this

sense cf Foley and Van Valin 1985). Typidly. such constructions move a dausal constituent

doser to the begimüng (or end in the case of nght dislocation) of the sentence. Foley and Van

Valin (1985:355-358) give the following examples for S.E..

74. I t was a sandwich that Ron ate. (it-cleft)

75. 1 t was Ron who ate a sandwich. (it-cleft)

76. M a t Ron ate was a sandwich (wh-deft)

77. The one who ate a sandwich was Ron (wh-cleft)

78. John's knife, 1 acàdentdy cut myseif with it (left dislocation)

C.G. and the Caribbean Engiish Creoles generally are well known for their fronting capaaty

( Markey and Fodale 1 983). In his discussion of focus in JC and BC (Barbadian Creole) ,

Roberts ( 1 99 1 ) isolates two devices used in JC which are also common in CC. The first he

compares to the English i t -dek Examples follow (79-80):

79. a di mango Jan iit "It was the mango that John ate."

80. a da shii a duu "That's what she does,"

Like S.E. it-de& the pragmatic focus on di mango is essentially contrastive (it was the mango -

326

not something else - that John ate). Note the use of the panide a in sentence initial position.

While some authors gloss this as a F ~ C W particle the fact thas in more mesolectal varieties. it

is replaced by iz indicates that it has more than just a superficiai affinity with equative a

( Rickford 198 7b: 184). Such preposed focus particies ais0 occur optionally in intemogatives

(8 1 ) and may be used to indicate contrastive focus for NPs that have not undergone honting

(82)-

8 1. a wa den duu a mi hous? 'What did they do at my house?"

8 2. a mi gu bring di touwil chuuzdee ' I'm going to bring the towels on Tuesday."

Fronting also occurs in the absence of a. The foiiowing example is h m Rickford ( 1987b) and.

as he notes, gives a sewe of the range of elements that can be topicalized:

83. ya yu gu b ~ g Nana? 'Here you'U bring Nana?" (Rickford 1987b:241 iine

1247)

In th is case the adverbial deicticya is topicalized through fronting. More common is NP

fronting as in the following wample:

84. foor viirz ii get 'He is four years old."

Rickford ( 1 98 7b:MO) remark; . " (O) ne other area that seems to require m e r research is the

details of GC topicalization, in panicuiar, the question of what kin6 of elements are eligible

for sentence fronting and topicalization with and without a or u." The following example

indicates that fronting can operate on elliptical stnictures:

85. a bina siks chiken a oonIii chrii m a sü 'There were six chicken ( ) but onlv three

1 see."

The other syntactic device discussed by Roberts ( 199 1) which is used to bring a particular

327

constituent into focus is copying. As Roberts ( 199 1:29 1 ) notes of similar constructions in jC,

copying has no structural parallel in S.E. but does in some West Afncan languages.

86. a pelt ü pelt a brook ii h u t 'A pelting that's what broke its foot"

87. a de ü de 'There, that's where he is."

88. na chupit di gyal chupit? 'Stupid, isn't that what the girl is?"(Bidwon

1973a:648)

Roberts ( 199 1 :29 1) compares such copying to SE left dislocation. remarking that the sentence

' c m be treated as a left dislocation in that it has a repeated elernent within the dause but the

repeated element is not a proform as it is in left dislocations in English. In addition, the

sentence has the copula element that characterizes it-deft sentences, but it is not an it-defk

sentence, not only because it does not have an it equivalent but also because the syntax of the

sentence will not d o w it. This method of focussing on the verb - by putting it to the front and

repeating it - is regarded as a typical Caribbean creole structure not only because it occws in a

number of creoles but aiso because it diffefs from European language structure." Like other

cases of front focussing with a, predicate copying of this kind involves contrastive focus. Thus in

the examples given above it is. in the first case (86). asserted that the foot was broken through

a particuiar kind of action (perpetrated by an intentional agent - also brought into focus - i.e.

this was not a accident). In the second case (89) the device signals a contrastive focus on de. a

region somewhere within the interactants range of perceptuai access (cf- chapter 5). as opposed

to some place beyond the immediately perceivable. The point here is that the participants to

the interaction should watch what they Say as the subject of the locative expression is dose by.

Abrahams. EL ( L 983) nit Man-of- Won& in the Wrst Indicsr Prrformance and the Emrrgmct of C d CuIture. Baltimore: John Hopkins University Press.

Abrahams, R and J. Szwed ( 1 983) A* Afnca. New Haven: Yale University Press. Aken, G. ( 198 1) Phonologicul Variation in the Jamiacan Continuum. Ann Arbor: Karoma. AUevne, M. ( 1 980) Comparative Afi-American. Ann Arbor. Karoma AUebe, M. ( 1984) 'Perfect and Passive." Paper presented at the 5th Biennial Confmce of flic

Society fm Caribbean Linguistics , University of the West Indies. Jamaica. Alleyne. M. ( 1985) "A linguistic perspective on the Caribbeann in Minrz and Price e h .

Canibean Contoun, Baltimore: John Hopkins University Press. m e . M. ( 1987) 'Predicate structures in Saramaccan" in M. AUeyne ed. Studia in

Saramaccan Language Structure. Foiklore Studies Roject, University of the West Indies. Jamaica.

AUsopp. R ( 1958) "The English Language in British Guiana.' Englirh Language Teaching 12(2):59-66.

AUsopp. R ( 1 996) Dictionuv of Gnbbeun Englirh usage. Oxford: Oxford Universiq Press. Anderson. B. ( 1 983) Imangincd Communities: RCfl.m-m on the Chgin and Spread of Nationalim.

London: Verso. Anderson, S. and E.L. Keenan ( 1985) "DekisW in T. Shopen Ed. Language Typology and

Sgntacrif Desmmption. Vol. III Grammatical Categories and the Len'wn . Cambridge: Cambridge University Press.

Andrews. Avery ( 1985) 'The major functions of the noun phrase." in T. Shopen Ed. Language Typology and Syntactic Description. Vol. I Clame Stnrcf~re. Cambridge: Cambridge University Press.

Bakhtin, M. M. i 198 1[1936]) "Discoune in the Novel." in M. Holquist ed. n i e Diabec Imagination. Austin: University of Texas Press.

Bakhtin. M. M. ( 1984) Pmblnnr ofDost0mkp"s Potrtks, tram. Cary1 Emerson. Minneapolis: Univeni- of Minnesota Press.

Bakhtin. M.M. (1986[1979]) "The Problem of Speech Genres." in Speech Genres and Othm La& Essnys , Austin: University of Texas Press.

Banfield, A. ( 1 9 78) " Where epistemology, style. and gramrnar meet literary history: the development of represented speech and thought." Nnv Litmrry History 9 (3), pp. 4 15- 54. Reprinted in Lucy ed. ( 1 993)

Bar-HiiIeI, Y. ( 1954) 'Indexical Expressions." Mind, 63.359-376. Banels, Dennis ( 1977) "Class Conflict and Racist Ideology in the Formation of Modem

Guyanese Soaew." Canadian Rer>iew of SociuZo~ and Anthmpology ( 14)4,396-405. Baugh, J. ( 1983) B l d Street Sperch: i t s histmy. stnrcture and rtrrvivnl. Austin: University of

Texas Press. Bauman. R and J . Sherzer eds. ( 1974) Erplorations in the Ethnography of Speaking, Cambridge:

C.U.P.. Bell, A. ( 1984) "Language Çtvle as Audience Design." in Language in Society 13: 145-204.

Berleant-Schiller, R and W. Maurer ( 1993) "Women's place is every place; rnerping dom* and women's roles in Barbuda and Dominica." in J. Momsen ed. Wm~i and Change in thc Caribbean . B Ioomington: Indiana Univexsity Press.

Bernstein. B. ( 1964) 'Elaborated and resvicted codes: Their social origins and some consequences." in nir Ethnography ofCommunication. J Gumpen and D. Hymes eds. American Anthmpologirt 66. 6. pt II:S5-69.

Besnier, N. ( 1990) "Language and Affect." Annual Review ofAnthmpolu~ 19: 4 19-5 1. Besson. 1. ( 1993) 'Reputation and Respectability Reconsidered: A new perspective on Afrm

Caribbean peasant women." in J. Momsen ed. Womm and a n g e in the Caribbcan. Bloomington: Indiana University Press.

Bickexton. D. ( 1973a) T h e Nature of a Creole Continuum." L a n p g e 49:640-69. Bickerton. D. ( L 9 73b) "The Structure of Polyleaal Grammars." in RShuy ed. Pmceedings of the

Twmty- l2ird Annual Round Ta& Meeting mi Lunguage and Lhpisticr, pp. 17-42. Monograph Senes 25. Washington D.C.: Georgetown Uninrersity Press.

Bickerton. D. (1975) Dynamics ufa Cmok @sstmr. Cambridge: C.U.P.. Bickerton. D. ( 1977a) 'Putting Badc the Clodc in Variation Studies." Languuge 53(2): 353-360. Bickerton. D. ( 1 9 7 7b) "Pidginization and Creolization: Language Acquisition and Language

Universals." in A. VaIdrnan ed. Pidgin and Ceok Luig<istia. Bloomington: Indiana Universis. Press.

Bickerton. D. ( 198 1) nte Rwts oftanpage. AM Arbor. Karoma. Bisnauth, Dale ( 1995) 'Snobbery" Sunday Mimr. Nov. 5.. p. 5. Horn and Gumpen ( 1972) "Social Meaning in Linguistic Stmcture:Code-Switdiing in

Nonvav." Gumpen and Hymes ed. Dit~cti0n.s in Suciolinguistia: the Efhnography 4 ~ommMinr t ion . New York Holt Reinhart.

Bourdieu, Pierre ( 1 9 7 7) Outline of a T h e o ~ af Practice. trans. Richard Nice. Cambridge: Cambridge Universiv Press.

Bourdieu, P. ( 199 1 ) Langziage and Sy>rbolic Power. Cambridge: Harvard University Press. Briggs. Charles ( 1986) Learning How To A k A sociolinguistic appraisal of the interview in social

science reseurch. Cambridge: Cambridge University Press. Briggs, C ( 1992) "Linguistic ideologies and the naturalizaùon of power in Warao discourse."

Prngmntics 2:3, pp. 387-404 Briggs. C and R Baurnan ( 1992) "Genre. Intenextuafity. and S o a d Power." luunial oflinguistic

Anthmpology 2(2): 13 1- 172. Brown, R and A. Güman ( 1960) 'The pronouns of power and solidariv." in T. Sebeok ed.

Style ni Language. Cambridge. Mass.: MIT Press. Brown. P. and S. Levinson ( 1978) "Universals in language use: politeness phenornena." In E.

Goodv ed. Questions and Politeness: Strategies in Social Interaction. Cambridge: Cam bridge ~nivéni ty press.

Brown, P. and S. Levinson ( 1979) "Social structure, groups and interaction." in K. Scherer and H. Giles eds. Social Marken in Speech. Cambridge: Cambridge University Press.

Brown. P. and S. Levinson ( 198 7) Politeness: Some Universak in Lnnguage Use. Cambridge: C.U.P..

Bden-McKenzie. K. ( 1978) 'Some Creolese Words in Common Use." in J. Rickford ed. A Festival of Guyanese Wordr . 2nd. ed. Georgetown: University of Guyana.

Cassidv. F.G. ( 1 9 7 1 [ 196 1 1) Iamaica Talk Thnc Hundnd Yenn 4 the Engiish Language in Jamaica. 2nd ed. London: Maani&n.

Cheshire, j . ( 1 98 1 ) Variation in the use of ain 't in an h a n British dialect. knguage in Society 10:365-382.

Cheshire, 1 ed. ( 1 99 1 ) Englidi Amund the WorId: Suciolinguistic Perspectives. Cambndge: Cambridge University Press.

ChomsQ. N. ( 1965) Aspects ofa Theory of Syntar. Cambndge. Mass.: MIT Chung, S. and A. Timberlake ( L 985) "Tense, aspect and mod-" in T. Shopen Ed. Languagi?

Typology und Sytncric Description. VOL III Grammatical Catepries and thc Laicon. Cambridge: Cambridge University Press.

Clifford. J. ( 1 983) 'On Ethnographie Authority." Rrpmentatiotts 1, no.2: 1 18-44 Coates. J. and D. Cameron eds. ( L 989) Womm in Their Spccch Conununities: NOP perspectives on

language and s a . New York: Longman. Cole, M. ( 1 985) "The zone of proximal development Where Culture and Cognition mate

each other.' in Wertsch ed. Culture, Gmimunimtion und Cognition, Cambridge: C.U.P.. Comrie. B . ( 1976) Aspect; An intmduction to the study o f v d u l aspect and reiated pmblmü.

Cambndge: Cambridge University Press. Craig, D. ( 1975) 'A Creole English continuum and the theory of gramrnar- " Unpubiished ms. Danziger, E. ( 1994) 'Out of sight, out of mind: person. perception and hinaion in Mopan

Maya spatial de*." ].B. Haviland and S.C. Levinson eds. S p a a in Mayan Languages. Linguistics 32 (4/5).

Deleuze. G. and F. Guattaii (1987) A ïlourand Piutlaus. Minneapolis: University of Minnesota Press.

Demuth. Katherine ( 1986) "Prompting routines as cultural innuences upon language acquisition. * In Language Son'uIization Amss Cultures, eds. B. Schieffelin and E. O&. Cambridge: Cambridge University Press.

Despres. Leo ( 1967) Cultural pluralbm and nntionalïst politifs in Britiih Guiana. Chicago: Rand McNaiiv.

Despres, L. ( 1 970) Différentiai adaptations and micro-culturai wolution in Guyana. In Afr- American anthmpolugy; contemporary perspectives. Ed. N. E. Whitten and J.F. Smed. New York: Free Press.

Despres, L. ( 1975a) Ethnicity and resource cornpetition in Guyanese society. In Ethnicity and resoum cornpetition in plural societies. Ed. L. Despres. The Hague: Mouton.

Despres, L. ( 197%) Ethnicity and ethnic group relations in Guyana. in n i e n m ethnicïty: p e r ~ p e ~ v t s f m n i ethnokb~. Ed. John Bennet. St Paul: West.

Douglas, Charles E. ( 1 930) Report on the cultivation, trratmen t and prospects of rice in British Gtiirrna. London: HMSO. (From Richardson 1974)

Drevfus, H. and P. Rabinow ( 1 982) Michel Foucault: Beyond Structuralisrn and Hmeneutics. Chicago: Universi~y of Chicago Press.

Dnimmond, L. ( 1980) "The C u l t d Continuum: a theory of intersysterns." Man 15:352-374. Dnimmond. L. ( 198 1) "Ethnicity, " e t h n i n u w and culture theory." Man 693-696. Duranti, A ( 1986) "The audience as co-author: An introduction." Tat 6(3) 239-47.

Duranti, A. ( 1 992) 'Language in context and language as context: the Samoan respect vocabulq. " in A. Duranti and C . Goodwin eds. Rethinking Confert: Language as an interadive phnomoion. Cambridge: Cambridge Universiv Press

Duranti. A ( 1994) Fmm Grnmmar to Politia: Linguirtic Anthmpobgy in a Watem Samoan Village. Berkeley University of California Press.

Duranti. A. ( 1997) 'Indexid speech aaacs Samoan communities." Ammcan Anthropologist 99(2): 342-354.

Eckert. P. ( 1389) 'The whole woman: Sex and gender differences in variation." hngauge variation and Change 1 :245-267.

Eckert, P. and S. Mdome&Ginet ( 1992) 'Think practicaüy and look locally: Language and gender as community-based practice." Annual Review of Anthropology 2 1:46 1-490.

Eckert, P. and S. McConne&Gi.net ( 1995) "Construaing meaning, constmcting sehres: Snapshou of language, gender and ckss from Belten High." in K Hall and M. Bucholtz ed. Gender Artinrlated: Language and dic So&l& Constnrcted Se* New York Routledge.

Edwards, Walter ( 1975) Suciolingrirtic Behaviour in Ruml and Urban Cinvmtrznces in Guyana. D. Phii thesis, University of York

Edwards. Walter ( 1 977) " Tantahin and Burin in Guyana." Anthmpobgical Linguistics , Vol. 20, No. 5.

Edwards. Walter ( 1983) Code Selection and Shifting in Guyana." Languugt! in Soci-p l2(3). pp. 295-3 1 1.

Edwards. Walter ( 1984) "A Community-Based Approach to the Provenance of Urban Guyanese Creole." in Sebba and Todd. eds9apcnfmm rhe York C'ok C o n f m ,

Edwards. Walter ( L 994) "The Grammars of Phonologicd and Morphological Choices in Rural and Urban Ideolects in Guyana." paper presented at ? l e tmth biennial meeting $the SociepfOr Caribbean Lingristia. Guyana.

Eisenberg, A. (1986) Teasing: verbal play in two Mexicano homes." in B. Sclüeffelin and E. Ochs eds. Language Sociulïzation amss Cultures, Cambridge: C.U.P..

Er--Tripp, S. ( 1 9 76) "1s Sybil There?': The structure of some American English directives. " Language in Society 5:25-67.

Escure, G. ( 199 1) 'Gender d e s and linguistic variation in the Belizian Creole cornmtmiy* in j. Cheshire ed. English Amund the Wurld: Socialinguistic Perspectives. Cambridge: Cambridge University Press.

Fasold, R ( 1 990) The Socioiinpistics of Language. Oxford: Basil BlachveU. Ferguson. C (1959) "Diglossia." Word 15:325340. Reprïnted in P. Giglioli (1972) ed. Lrnguage

and Social Gmtat. Haxmondsworth: Penguin Fillmore. C. ( 1973) "May we come in?" Smiotica. 9. 97- 1 16 Fülmore, C (1975) Santa C m Lectures on Deixis, 1971. Mimeo. Indiana University Linguistics

Club. Fiilmore. C. ( 1977) "The case for case reopened." in P. Cole and J.M. Sadock, eds. S p t u and

Seman tics 8: Grammatical Relations. New York Academic Press. Fillmore. C. ( 1982) 'Frame Semantics." in Linguistic Society of Korea ed. Linguistics in the

Morning Calm, Seoul: Hanshin Publishing Co. Fisher, L. ( 1976) "Dropping remark and the Barbadian audience " Ammicm EthnoIa@t 3 ( 2 ) .

227-242.

Fa.. W. and M. Olson ( 1 98 5) " Clausehood and verb serialization." in J. Nichols and k Woodbury eds. Grammar fnsidc and Outitide the CIaure. Cambridge: Cambridge Universiq Press.

Foley. W. and R Van Valin ( 1 985) 'Information packaging in the dause." in T. Shopen ed. Langunge Typloogg and Sytuctic D e m i p h . VOL I C7aure Stmrture. Cambndge: Cambridge University Press.

Foucault, M. ( 1979) Dikcipline and Punish: Thc Birtlr of the Pnion. New York Vintage Book;. Friedrich. P ( 1 9 89) "Language, ideology and politicai economy." A m d a n Anthmpologst. 9 1. Cal. S. ( 1978) 'Peasant men can't get wives: Language change and sex roles in a biiingual

community. " Language in Society 7: 1 - 1 7. Gd. S. ( 1989) "Language and Political Economy." Annual Review ofAnthmpologp 1 8: 345-367. Gambhir. S. ( 1 98 1 ) n e East Indian Speech Cmmunity in Guyana: A Sociolinguistic Study with

Special RCference to Koinc Formation. Ph.D. dissertation, Universi~ of Pennsyhania (UMI dissertation services.)

Gambhir, S. ( 1 983) "Diglossia in Dying hguages: A case study of Guyanese Bhojpuri and Standard Hindi." Anthmpo[ogical Linguistics (25) 1 :28-38.

Garrett. P ( 1994) 'Conceptualizing Linguistic Variation in Creole Environments: Beyond the Continuum." Paper presented at the Tmth Bimned Confnnce of the Society for Curibbean Lingitirtics. Georgetown. Guyana, August 2428.

Givon. T. ( 1 982) 'Tense-Aspect-Modality The mole proto-cype and beyond. " in P.J. Hopper ed. T m Aspect: Between Sentantics and Prugmatia, pp. 1 15-63. Philadelphia: John Benjamins.

Goffman. E. ( 1974) Frame Analysir: An Essay on the Organizatiun ofErpt~?ence. New York Harper Row.

Goodwin. C ( 1984) 'Notes on story stxucture and the organization of participation." in J.M. Atkinson & J. Heritage eds. Sm<mcres of Social Action: Shidicr in Conversation rlnalvsis.Cambrïdge: Cambridge Universisr Press.

Goodwin. C. and A. Duranti ( 1992) 'Rethiriking Context: An Introduction." in k Duranti and C. Goodwin e h . Reihinking on^: Lunguuge as an Intmuctiw Phenomenon. Cambndge: Cambndge University Press.

Goodwin. C. and M.H Goodwin (1990) "Interstitial Argumentn in A.D. Grirnshaw ed. Cunflict Tnlk Sociolirrgt istic Investigations of arguments in con verration. Cambridge: Cambridge University Press.

Goodwin. M. H. ( 1 990) He-Said-She-Said: Taik as Social &pnization among Blnek Qiildmi. Bloomington: Indiana University Press.

Gnce. H.P. ( 19 75) " Logic and Conversation." in P. Cole and J.L. Morgan eds. Syntar and Semantics 3: Speech Acts. New York Academic Press.

Gumpen. 1. ( 1 962) "Types of Linguistic communities. " Anthropologiral Linpistics 4. 1 :28-40. Gumpen. J. ( 1968) 'The Speech CornmUNty." Intemational Encyclopedia ofthe Social Sciences.

New York: Macmillan. Reprinted in P. Giglioli ( 1972) ed. Lunpage and Social Contert. Harrnondsworth: Penguin.

0

kaeri, N. ( 1987) Maldfernale differences in speech: An alternative interpretation. In k ~ . Denning e t.al. ed. Variation in Langauge; NWA V-XV at Stanfrd. Stadord: Stanford Universitv. Deparunent of Linguis tics.

HaUday. M. (1976) "Anti-Languages." Arnmmuzn hthropologist 78. 178-183. HalloweU, A.. IMng ( 1955) Cultural Factors in Spatial Orientation. In Cultzin and Erperiiaia.

Phiiadelphia: University of Pennsylvania P~ss. Hanks, Wiam ( 1990) kfierential Pructices: Lungunge and Limd Spaa among the Muyu. Chicago:

Chicago University Press. Hanks. W. ( 1 992) 'The indexid ground of deictic reference. " in A. Duranti and C. Goodwin

eds. Rethinking Corrtntr Language as an intmactive phmornenon. Cambridge: Cambridge University Press

Hanks. W. ( 1993) 'Metalanpage and pragmatics of de*." in J. Lucy e d Ref%xiwe Language: Reported Speech and Mefapgamatics. Cambridge: Cambridge University Ress.

Hanks. W. ( 1 996a) Lnnguuge and Communicative Pructices. Boulder. CO: WesNiew heu. Hanks, W. ( 1 W6b) 'Language form and communicative pradces." in Gumperz and Levinson

eds. Rcthinkug Luiguistic Rehtivitp. Cambridge: Cambridge University Press. Hanley, Eric ( 1975) 'Rice. politics and development in Guyana." In B e p d the sociohqp of

developmmt: eeonomy and society in Latin Ameriuc and Afica. Ed. Ivar Oxaal etal. Boston. Massachusetts: Routledge and Kegan Paul.

Hadey, Eric ( 19 79) 'Mechanized rice cultivation: the experience of an East Indian community in Guyana." In Peasantc, Plantations and Rural Comrnunitiw in the Canbbcan. Ed. A. Mauss and M. Cross. Guildford: Universisr of Surrey.

Hanley. Eric ( 198 7) 'The Guyanese rice industry and development planning: The wrong solution for the wrong problem." in J. Besson and J. Momsen eds. Land and dtwelopmmt in the Cnnbbean . London: Macmillan

Harvey. D. ( 1990) The Condition of Posbnodemity, Oxford: Basil Biadcwell. Haviland, J. ( 1 986) "'Con Buenos Chiies': Tak. targets and teasing in Zinacantan. " Tcxt . 6 ( 3 ) . Haviland. 1. ( 1 9 96) "Projections. transpositions, and relativiv. " in J. Gumperz and S .C.

Levinson eds. Rethinking Linguistic Relativity. Cambridge: Cambridge Univeni. Press. Hickmann. M. ( 1985) " Metapragmatics in M d language." in E. Mertz and R Parmentier eds.

Smiotic Med iation . Orlando: Academic Press. Hickmann, M. ( 1987) "The pragnatics of reference in chiid language: some issues in

developmental theoxy." in M. Hidanann ed. Social and Functional Appmacha tu Languuge and ntougtit. Orlando: Academic Press.

Hickmann. M. ( 1993) "The boundaries of reported speech in narrativce discourse: Some developmental aspects." in 1. Lucy ed. Rejkxive Lunguage: Reported Speech and Metupragmatics, Cambridge: CambndgeUniversity Press.

Hill, J. ( L 98 7) 'Women's speech in Modem Mexicano." in S.U. Phîlips, S. Steele and C. Tanz eds. Lanpage, Gender and S a in Comparative Penpective. Cambridge: Cambridge Universitv Press.

Hymes. D. ( 1 96 j ) " Models of the interaction of languge and social Me." Journal of Social Isszles. 23.2:8-28.

Hvmes, D. ( 1968 [ 196 21) "The Ethnography of Speakingn in J. Fishman ed. Readings in the Suciology of Language. The Hague: Mouton. pp. 99- 13 8.

Hvmes . D. ( 1 9 74) Foundations in Soeiolinguirtics . Philadelphia: University of Pennsylvania Press.

J acobs-Huey, Lanita ( 1 9 9 7) '1s there an authentic African-ArneBcan speech communiy Carla revisited?" in nit proceedings of NWAVE. 1996. Penn Working Papers in Linguistics

Jakobson, R ( 1 95 7) 'Shifters, verbal categones, and the Russian verb. " in Selcctcd Wmkr II: Word and Lunguage, pp. 130- 147. The Hague: Mouton.

Jakobson. R ( 1 966) "Grammatical parallehm and its Russian facet " Lnnguage 42: 399-429. jakobson. R ( 1990[1957]) 'Shifters and Verbai Categoiies" in L. Waugh and M. MonviUe-

Burston eds. On Languuge. Cambridge, Mass.: Haxvard Universisr Press. Jayawardena, C. ( 1963) Conflct and Solidarity on Guynme Plantation. London: Althone Press. Jayawardena. C. ( 1 968) 'Ideology and coriflict in lower class communities." Comparative

Studies in Society and Hktory 1 O(4) :4 13-446. layawardena, C . ( 1980) "Culture and Ethnicity in Guyana and Fiji." Man 15:430-450. lesperson. 0. ( 1965[ 19241) The Philosophy of Gmmmr. London: m e n and U n a . Keating, Elizabeth ( L 994) Language. gender. rank and social space: Honorifics in Pohnpei,

Micronesia. In Culfural Prrfmmun~~~: Pmceeding 4tk thcird Benkeb Wonrai and Lnnguage C o n f m . Berkeley: Berkeley Women and Language Press.

Keenan. E. ( 1974) 'Nom-Makers, Nom-Breakers: Uses of speech by men and women in a Maiagasy cornmunity." in R Bauman and J. Sherzer, eds. Exploratiom in the Ethnography ofspeaking, pp. 1 10- 124. Cambridge: Cambridge University Press.

Kiparskv, P. and C. Kiparsly ( 197 1) 'Fan" in D. Steinberg and L. Jakobovits eds. Smrantia: Interdisciplinary Reader in Philosophy, Linguûtia and Psychology. Cambridge: Cambridge

University Press. Klein. W. ( 1 983) " Deixis and spatial orientation in route directions." in H.L. Pi& and L.P.

Acredolo eds., Spatial Onpntution: 7%eoy, Rerearrh and Application. New York: Plenum Press.

Kochman, T ( 1986) "Strategic ambiguity in black speech genres: Cross-cultural interference in particpant-obsewation research." Tèxt 6.2: 153- 170.

I<dick, D. ( 1992) Language Sh$ and Cultuml Repmduciion: Socialüation, Seyand Spcretism in a P a p a New Gztinean ViiZage, Cambridge: C.U. P..

Kulick, 0. ( 1 993) ' Speaking as a Woman: Stxucture and gender in domestic arguments in a New Guinea Vidage." Cultural Anrhmpolun 8 (4):s 10-54 1.

Labov. W. ( 1 966) T h e Social Stratifcation of Englirh in New York Qs>. Washington, D.C.: Centre for Applied Linguistics

Labov. W. ( 1968) 'The reflection of sotial processes in linguistic structures." in J . Fishman ed. Readings in the Sociolbgy of Language. The Hague: Mouton. pp 240-25 1.

Labov. W. ( 1969) 'Contraction, deletion and the inherent variability of the English Copula." Language, 45(4):7 15-62.

Labov, W. ( 1 9 7 2) Socidinguistic Patrmrs . Philadelphia: University of Pemsylvania Press. Labov, W. ( 1 980a) "The social origins of sound change." In W. Labov ed. Lofating

Langutxge in Time and Space. New York Academic Press. Labov. W. ( 1980b) '1s there a aeole speech comrnunity?" In Valclman and Highfield eds.

Theoretical Orientations in Creole Studies. New York: Academic Press. Labov, W. ( 1 984) 'Intensity." in D. SdùffM ed. Meanin6 Fonn and Use in Conttxt: Linguistic

Applications, GURT '84. Washington, D .C . : Georgetown University Press.

Labov. W. ( 1990) 'The intersection of sex and social dass in the course of linguistic diange." Languarge Variation und Change, 2:205-254.

Labov. W. and W. Harris ( 1986) "De facto segregation of Bladc and White vemaculars." in D. Sankoff ed. DiPrni9 and Diuchmy. Amsterdam: John Benjamins.

Lambert. W. ( 1967) 'A social psychology of biiinpuaiism." in V. Maaiamara ed. P m b h of B ilingualism, B e Iournnl4 Social Issua. vol. 23.

Lavemdera, B. ( 1978) "Where does the sociohguistic variable stop?" Languuge in Society 7(2):17 1- 182.

Lefebvre. C. ( 1993) 'The role of re ldca t ion and syntactic reanafysis in Hatian Creole: Methodologid aspects of a research program." in S. Mufwene ed. qfnain ism in Afm- Amencan Language Vurieties. Athens: University of Georgia Press.

Le Page. R ( 1978) 'Projection. Focussing, DiMision. or steps Towards a SoQolinguistic Theory of Language." Society for Caribbean Linguistics Occasional Paper 9. S t Augustine: Society for Caribbean Linguistics.

Le Page, R B. and Tabouret-Keller k ( 1 985) Acts ofldmtity: C'k-Bsed Apppmaches to Language and Ethnicitp. London: Cambridge University Press.

Levinson, S. ( 1979) "ActMty types and language." Linguistics 1 7 pp. 365-399. R e p ~ t e d in P. Drew and J. Heritage ed. Taik at W o k Interaction in institutional SWngs. Cambridge: Cambridge University Press.

Levinson, S. ( 1 983) Pragnzatia. Cambridge: Cambridge University Press. Levinson. S. ( 1 988) " Puthg hguistics on a proper footing: explorations in Goffman's concepts

of participation." in P. Drew and A. Wotton eds. Erving Goffnran: Erploring the Interaction Order. Cambridge: Cambridge Universiv Press.

Levinson. S. ( 1992) "Primer for the field investigation of spatial description and conception." Pmgmntia. 2 (1). 5-47.

Levinson, S. ( 1994) "Deiw." In Encyclodpedia ofLanguage and Linguistics, ed. RE Asher. 2:853- 85 7. Oxford: Pergamon

Levinson. S. ( 1 9 96) " Frames of Reference and Molyneux's question: Crosslinguistic Evidence. " in Paul Bloom e t al. eds. hnguage and Spaa. Cambridge. Mass.: M.I.T. press

Levinson, S. ( 1996) Langauge and Space. Annaal RNinv ofAnthropobgy 25353-82. Lindstrom. L. ( 1 992) "Context Contests: Debatable txuth statements on Tanna (Vanuatu). " in

A. Duranti and C. Goodwin e h . Rethinking Context: Languuge as an Interactive Phnommon. Cambridge: Cambridge Univenisr Press

Lord, C. ( 1973) "Evidence for syntactic reanalysis: h m verb to complementizer in Kwa." in S. Steever e t d . eds. Papmfmm Parasasion ofDiachmnic Syntar Chicago: Chicago Linguistic Society.

Lucy. 1. ( 1 993) "Reflexive Language and the Human Disciplinesn in J.Lucy ed. ReJhive Lnnguage: Reportcd Speech und Metaprngmatics. Cambridge: C.U. P..

Luu, C. and L. Abu-Lughod eds. (1990) Language and the politia ofemotions. Cambridge: C.U.P. Lvons. J. ( 19 77) Semantia, vol 1 &2. Cambridge: Cambridge University Press. Marcus, G.E. ( 1970) Incornplete transformation: socÿù change in a Guyanese d communi~ .

Caribbean Studia, vo1.9. no. 4. Markev, T. and P. Fodaie ( 1 983) 'Lexical diathesis. focal shifts and pssivization: the creole

wice." Enghh Worid-Wide 4, 1.: 69-84.

Mauss, Marcel ( 1973 [ 19351) Techniques of the Body. Ewnumy and soncty 2( 1): 70-88. McElliinny. B. ( 1995) 'Challenging hegemonic masculinities: Female and male police officers

handling domestic violence." in K. Haii and M. Budioltz eds. &der Articulated. London: Routledge.

McWhorter. J. ( to appear) ' B ~ g i n g Gender into Creole Studies. " Pmceedings of the Fourth Bmkeky Women and knguage Gnfmnce.

Miller, P. ( 1986) "Teasing as language socialization and verbal play in a white working dass cornmunity." in Scheffelin and Ochs eh .

Momsen, janet H. Ed. ( 1993) Womm and change in thc Canbbun. Bloomington: Indiana University Press.

Moore. B. ( 1995) Cultural power, mirtance and pfuralimi: Gbnial Guyana 1838-1 900. Montreal: McGill-Queen's University Press.

Morgan, M. ( 1 99 1 ) 'Indùectness and interpretation in Afncan American Women's discourse. " Pragmatics l(4): 42 1-5 1.

Morgan. M. ( 1 993) 'The Africanness of counterlanguage among Afro Americans. "in S. Mufivene ed. Aficankms in A@-Amen Lanpge Varittir. Athew: University of Georgia Press.

Morgan. M. ( 1 996) "Conversational Signifjhg: Grammar and indirectness among Afncan American women." in E. O h . E. Schegloff and S. Thompson eds. Interaction and Gramrnar. Cambridge: Cambridge University Press.

Mufwene. S. ( 1984) "Obsenrations on time reference in Jamaican and Cuvanese aeoles." Engiish World- Wide 4:2. 199-229.

MuhKene. S. ( 1989) "Equivical structures in some G d a h complex sentences." Anrerican Speech 64.4.

Mufwene. S. (1992) "Why grammars are not monolithic." in D. Brentari etal. eds. The Joy of Grammm A fest~chnifor James D. McCawky. Chicago: UNiversity of Chicago Press.

Muhlhausler. P. ( 1980) 'Structural expansion and the process of aeokation." in A. Valdman and A. Highfield eds. Theoretical Onentafions in Creole Studies. New York: Academic Press.

Nichois. P ( 1 983) "Linguistic Options and Choices for Black Women in the Rural South. " in B. T'home. C, Kramarae, and N. Henley eds. Language Gender and Society. Cambridge, Mass.: Newbwy House.

Noonan. ( 1985) %omplementation," in T. Shopen ed. knguage Typology and Sgntactïc Deso+ption i l II C o r n p h Conrtnrctionr. Cambndge: Cambridge U niversis. Press.

Ochs, Elinor ( 1982) 'Taiking to diüdren in Western Samoa." Langaage in Society, I 1. 77- 104. Ochs, E h o r ( 1986) "Introduction." in Language Socialization Ams Cultum. eds. B. Schieffelin

and E. Ochs. Cambridge: Cambndge University Press. Ochs, ELinor ( L 986) 'From feelings tu grammar. a Srnoan case study." In Lnnguage Sodafizution

Anoss Cultures, eds. B. Schieffelin and E. O h . Cambridge: Cambndge University Press.

O h , EIinor ( 1988) Culture and Language Datelopment: Language Acquisition and Language Socialkation in a Samoan Village. Cambridge: Cambndge University Press

Ochs, Elinor ( 1992) 'Indexhg gendern in A. Duranti and C. Goodwin eds. Rdinking Contatr Language as an interactive p h m m m m . Cambridge: Cambridge University Press

O h . E h o r ( 1996) 'Linguistic resources for soaalizing humanity." in Gumpen and Levinson eds. Rethinking Linguistic Relativity. Cambridge: C.U.P..

Ochs, E h o r and Bambi Schieffelin e h . ( 19 79) Develapmental Pragmatia. New York Academic Press.

Ochs. Elinor and Bambi SchieEe1.n ( 1984) ''Language acquisition and socialuation: Three developmental stories and their implications." in R Shweder 6r R Levine, eds.. Culture ï l e o ~ : essays in rnind, sey and motion. New York Cambridge University Press, pp. 2 76- 320.

Ochs. Elinor and Bambi SchieEelin ( 1989) "Language has a h e m " Tat 9:7-25. Ortner, Sherry ( 199 1) 'Reading America." in. R Fox ed. RecapturingAnthmpology: Working in the

Present. Santa Fe: School of American Research Press. Palmer, F. R ( 1986) Moud and Modulity. Cambridge: Cambridge University Press. Paimer. F. R ( 1994) Grammatical Roln and Relntiom. Cambridge: Cambridge University

Press. Parmentier, R J. ( 1994) "Peirce Divested for Nonintiates." and "Peirce's concept of Semiotic

Mediation." in Sigm in So-, Bloomington: Indianna University Press. Patrick. P. ( 1996) 'The ubanization of creole phonology: Variation and diange in Jamaican

(KYA)." in G. Guy e t al. eds. Towardr a Social Scimce ufhngungpr Papen in honour of William Labov. Amsterdam: John Benjamins.

Patrick. P. and B. McElliinny ( 1994) 'Speakin' and Spokin' in Jamiaca: Conflin andconcensus in Socioiinguis tics. " Proceedings of the B&ky Linguistics Sociq 1 9.

Pavone, J. ( 1980) Implicational ~ c n l r c and EnglLcii Dialectology. Ph.D. dissertation. Indiana Universitv.

Peirce, C. S. ( 1 955) "Logic as Semiotic: The theory of signs." in Philasophical Writings Charles Sander Peirce. Dover.

Peristiany. J .G. ed. ( 1966) Honour and Shame: nie Values of Meditmanean Society. Chicago: Univenitv of Chicago Press.

Peters. A and S . * B O ~ ~ S ( 1986) "Interactional routines as cultural influences on language acquisition." in Sdueffelin and Odis ed..

Poilarci, V. ( 1989) 'The partide en in ~arnhcan Creole: A discourse related account." Englirh World-Wide 10: 1, 55-68.

Potter. Lesley. M. ( 1 9 75) Interna1 Migration and Resettiement on East Indians in Guyana, 1870- 1920. Ph.D. Dissertation McGili University. (avaiiable From the National L i b r q . Canada)

Potter, Lesley. M. ( 1 982) 'The pst-indenture experience of East Indians in Guvana. 1873- 192 1 ." In East Indian in the Caribbean; mlonialisrn and the rtnrgglefor identi&: ppapm presented tu a symposium on East Indium in the Caribbean, n i e Univmity of the West Indies. Jzlne, 1975. Miliwood, New York: Kraus International-

Premdas, R. ( 1 995) Ethnie Confiet and Devebpmmt: n e Case of Guy ana. Aldenhot: Avebuy. Putnam, H. ( 1 9 75) " 1s semantics possible?" in Mind, Langauagc, and Reality: Plrilosophical

Papm.. vol 1 . Cambndge: Cambridge University Press. . &mpaul. S. ( 1978) "Hindi kinship terms in Guyana." in J. Rickford ed. A Festival of~kanese

WorrLr. 2nd. ed. Georgetown: University of Guyana.

Reisman, K. (1970) 'Cultural and iinguistic ambiguity in a West Indian village." in N.F. Whitten, Jr. and J. F. Szwed, eds. ~fro-~m&u& Anthropobgy, pp. 129-44. New York Free Press.

Reisman, K. ( 1974) "Contrapuntal conversations in an Antiguan village." in R Bauman and J. Sherzer, eds. Explomtii~ns in the Ethnogaphy of Speaking, pp. 1 10- 1 24. Cambridge: Cambridge University Press.

Ridaord, John ( 1979) Vuruiation in a Cnuk Continuum: qrurntitative and implicational apprvaches. Ph. D. dissertation. Depamnent of Linguistics: University of Pennsytvania.

Riddord J. ( 198 1) "A variable d e for a aeole continuum." in D. Sankoff and H. Cedergren e b . Vunation Omnibus. Carbondale: Linguistic Research.

RiJcford, J. ( 1985) "Standard and Non-standard attitudues in a Creole Continuum." inN. Wolfson and J. Manes eds. Languuge uffnqualitp, New York: Mouton.

Ridcford, J. ( 1 986a) "Concord and contrast in the characterization of the speech community." Sh@U Wonhg Papen in Linpistio. 3.

Riddord, J. ( 1986b) "The need for new approaches to soaal dass anaiysis in sociolinguistia." in Languiage and Communiaitimr, 6.3: 215-221.

Ridcford. l. ( 1987a)"The Haves and the Have Nots: Socioünguistic S w e y s and the Assessrnent of Speaker Cornpetence." Language in Society 16(2): 149-77.

Rickford, J . ( 19870) Dimnrtionr ofa CE& Continuum, S d o r d : Stanford University Press. Rickford, J . ( 1 98 8) "Connections between sd~linguistics and pidgin-aeo1e studies. "

International Journal of the Sociology of Languap. 7 1 : 5 1 -5 7. Ridcford. J . ( 1 9 9 1 ) ' Soaolinguistic variation in Cane W a k a quantitative case study. " in J .

Cheshire ed. Riddord. 1. ( 199 7) "Unequai partnership: Sotiolinguistics and the Afncan Amencan speech

community." Langrrage and Somty 26(2). 161-198. Rickford. J. and A. Ridcford ( 1976) 'Cut-eye and suck-teeth: Afiican words and gestures in new

world guise". Journal ofArnerican Folklore, vol 89, no. 353, pp. 294-309. Rickford. J, and E. C. Traugott ( 1985) "Symbol of powerlessness and degeneracy, or symbol of

solidarity and tnith? Paradoxical attitudes toward pidgins and aeoles." in S. Greenbaum, ed., 7?ze Englih Language Today, Odord: Pergarnon Press.

Roberts, P. ( 1980) 'The adequacy of certain theories in accounting for important gammatical relationships in a aeole language." in R Day ed. Issua in Englirh Creoles. Heidelberg: J ulius Groos Verlag.

Roberts, P. ( 199 1) "Focus and assertion in Jamiacan and Barbadian speech." in W. Edwards and D. Winford eds. V d Phrase Patunrc in Bhck Englirh and Creale. Detroit: Wayne State Universiv Press.

Rodney, W. ( 1981) A Hirtory ofthe Guyanre Wmkng Peopk, 1881-1 905, Baltimore: John Hopkins University Press.

Rodwav, James ( 1 8 9 7) " Ruin. " TimehB', vol. 9 j new series). ~osald;>, M. ( 1982) "The things we do with worb: Ilongot speech acts and speech act theoiy in

philosophy." Languup in Society, 11:203-137. Ross, J.R ( 1967) Constraints un Variablrr in Sptax. Ph.D. Dissertation MIT.

Rumsey, A. ( 1990) 'Wordùig. rneaning and linguistic ideology." Amduzn Anthmpobgist. 9 W )

Russell, William( 1886) Rice. Tirnehn, vol. 5 (&ci series). 104 105. Sa&, H. ( 1984) "Notes on Methodology." in 1.M. Atkinson & J. Heritage eds. Structures of

Social Action: Shzdics in Conmatirni Analysfi. Cambridge: Cambridge Universiq Press. Sawure, Ferdinand de ( 1983[1915]) Corn in General Linpistia. La Salle, Ihois : Open

Court Schegloff, E. ( 1 9 72) 'Notes on a conversational practice: Formulating place." in P.P. Giglioli

ed. LAnguage and Social Contmt, Hamiondswonh: Penguin. Schegloff. E. and H. Sacks (1973) 'Opening up dosings." Semwtira, 7.4,289-327. Scheper-Hughes, N. ( 1992) Death Withmt Wecpingr me Violence ojf Etwyday L@ in BraziZ.

Berkeley: University of Califomia Press. Schieffeh. B. ( 1979) "Getting it Together: An Ethnographie Approach to the Study of the

Development of Communicative Competence." in O h , E. and B. Sdiieffelin eds. SchieffeIin, B. ( 1986) "Teasing and Shaming in Kaluli dddren's interactions." in Scheffelin and

Ochs eds. Schieffeiin. B.B. ( 198 7) 'Do different worlds mean different words?"in S.U. Phüips. S. Steele

and C. Tanz eds. Language, Gender and Sex in Comparative Penpective. Cambridge: Cambridge University Press.

SchieEeIin, Bambi ( 1990) n i e Give and Take ofEmyday L@: Language Socializatim ojffiluli Childm. Cambndge:C.U.P

Schieffelin, B. and E. Ochs ( 1986) "Language Socialization." in Annual RNicw tfAnthmpologp. Schieffelin. B. and E. Ochs eds. ( 1986) Lnnguage Socialization a m s Cultures, Cambridge: C.U.P.. Schiegel. 1. ( 1 996) "The notion of peers in language soaalization studies. " paper presented at

the 95 th Annual Meetings of the Amdcan Anthmpologjcal Association. Schornburgk, Robert ( 1970[1840]) A desmption ofBritish Guiana. London: Frank Cass & CO..

Sebba, M. ( 1987) n i e S p t m ofS&l Vnfs. Amsterdam: John Benjamins. SeiIs. P. et. al. ( 1996) "An optimaLity theoretic approach to variation in negative inversion in

AAVE." Natural Language and Linguirtic %y. 14(3) 59 1-627. Sidnell Jack ( forchcoming) " Eliatations and Ideologies of Community in an Indo-Guyanese

Village." in Proccedings of the 5th Annual Symposium About Language and Society at Austin (S.A. L.S.A.) Taas Lingristic Forum 36.

Sidneil. Jack (n.d. 1) 'Formulating Place in Verbal Dispute: Examples from an Indo-Guyanese Village." under review at Prapatia: Thr par*& journal of the Intentational Pragmatia Association .

Sidneil. Jack (n.d.2) "Leaming Social and Spatial Location through EIiatations in an Indo- Guvanese Village. " under review at Iounial of Lingub tic hthmpob&v.

Sidnell. la& (n.d.3)'Cwing up and Taiking Hard: Affective displays of 'passion' as challenges to a gendered temtoriality," to be presented at the 96th Ammimn AnthropoZu@cal Arsociation Annual Meeting in Warhinglon, D. C..

SidneU, Jack (n.d.4) "The ideology and practice of verbal dispute in Guyanese Creole: Conversational sequence and colonial stereotype." manuscript in preparation.

Sidnell, lad< (ad. 5) 'Prelimùiary Report on the Segmental and Suprasegmental structure of Guyanese Creole." paper wxitten for H. Rogers', Phonetics Class 1994. University of Toronto.

Siewierska, A. ( 1984) Thc pussive: A mpamtiwe linpistic adysis. London: Croom Helm. Sihrerman, M. ( 1976) T h e role of factiodim in political encapsulation: East Indian vülagers

in Guyana." In Ethnicitg in thcAm&xz,s. E d Frances Henry. The Hague: Mouton. Silverman. M. ( 1978) 'Viiiage coundand factionalism: defintional and contextuai issues. " In A

houce divided? Anthmpologiuzl studies of factionalism. Ed. M. Silverman and RF. Saiisbury. The Institute of Sociai and Economic Research. Mernorial Unkersiçy of Newfoundland.

Sihrexman. M. ( 1979) "Dependency, mediation and class formation in nrral Guyana." American Ethnobgirt

Silverman, Marylin ( 1980) Rich pwpk and nk: fclctjonal poli& in mral Guyana. Leiden. Netheriands: Briil.

Silverstein, M. ( 19 76a) "Shifters. Linguistic Categories. and Culturai Description." in Selby and Basso ed. Meaning in Anthrvpob&v, ~lbu~uërque: University of New Mexico Press. -

Silventein, M. ( 1 W6b) 'Hierarchy of featms and ergatnrif~." in RDixon ed. Grammatical Cntegmiicr in Aurtralian Languages. Canberra: Australian Ins titute of Aboriginal Studies.

Sitverstein. M. ( 1979) "Language structure and linguistic ideology," in Clyne. Hank;, and Hofbauer. eds. Thc Elmimtr: A Parasessimr on Lingrirtia Units and LePeLr. Chicago: Chicago Linguistic Soaety.

Silverstein, M. ( 198 1) nit limits of awamess, Sociohguistic Working Paper 84, Austin Texas: Southwest Educational Developrnent Laboratoiy.

Silventein. M. ( 1 98Sa) "The Culture of Language in Chinookan Narrative Texts: or, On saying that . .in Chinookan," in J . Nichols and A W o o d b q eds. Gmmar Imide and Chrtside the Clam. Cambridge: C.U.P..

Silveatein, M. ( 1985b) "nie Functional Stratification of Language in Ontogenesis." in Wertsch ed. Culture, Cognition and Commwiiczztion, Cam bridge: Cam bridge University Press.

Silverstein, M. ( 1985~) 'Language and the culture of gender: At the intersection of suucture. usage and ideology. " in E. Meru and R Parmentier eds. Semiotic Media tion . Orlando: Academic Press.

Silventein, M. ( 1987) "The Three Faces of Function." in M. Hidanann ed. Social and Functtonul Approacher to Languuge and 'Tnou&t, Orlando: Academic Press.

Silventein, M. ( l99Za) T h e Indeterminacy of Contextiialization: When is Enough Enough." in Auer and Luzio eds. nit Gntarualization ufkrnguage, Amsterdam: John Benjamins.

Silventein, M. ( 19920) 'The uses and utiiity of ideology: Some reflections," in P. Kroskrity. B. Schieffeh. and K. Woolard eds. Pragntatics 2(3):3 1 1-24.

Silven tein. M. ( 1 993) 'Metapragmatic Discourse and Metapragmatic Function." in J. Lucy ed. Rejkiw Language: Reporfed Speech and Mrtapragnatics, Cambridge: Cambndgeuniversity Press,

Silven tein, M. ( 1 996) " Indexical order and the dialectics of sociolinguistic Me. " in R Ide et al. eds. SALSA III: Pmceedingr 4th thrird Annual Sy?lposium about Language and Society - Austin. T m Linguistic Fomm 36.

Sïngler, j.V. ed. ( 1990) Pidgin and Cmole Tme-Moud-Aspect Sysumc. Amsterdam: John Benjamins.

Smith, RT. ( 1988) Kimhip und C l p s in the Wert Indin. Cambridge: Cambridge University Press.

Smith. RT. (1995) " ' W g in the gun mouth': Race, dass, and political violence in Guyana." Nm West Indies Guide, vol. 69 no. 3&4.

Smith, Raymond T. and Chandra Jayawdena ( 1 959) 'Marriage and family amongst East Indians in British Guiana." Social and Ecomrnik Studies. Vo1.8, no. 4.

Thorne, B. ( 1990) "Children and Gender: Constructions of Difference." in D. Rhode ed. Theontical Pmpecti~ on Smral DiJimice. New Haven: Yale University Press.

Trouillot. ( 1992) The Caribbean Region: An Open Frontier in Anthropological Theory." Annual Revim ofAnfhmpology 2 1 : 19-42.

Trudgill, P. ( 19 72a) 'Sex, covert prestige, and linguistic change in the urban British Engiish of Nomich." Lanpage in Soc* 1:179-195.

Trudgill. P. ( 1972b) The M u 1 Dzfmntiaiation ofEnglirh in Nomich. Cambridge: Cambridge University Press.

Van Valin, R and R LaPolla (n.d.) "Semantic Representation 1: Verbs and Arguments." chapter 3 of the manusaipt Syi- Sm<chrn, Meaning and Functiun. to appear in Cambridge Textbooks in Linguistics Series.

Voiosinov, V.N. ( 192911973]) Marxisrn and the Phihophg ofLanguage, Cambridge: Harvard University Press.

Vygo tsky. L. ( 1 9 78) Mind in Society. Cambridge: Haward Universisr Press. Washabaugh, W. ( L 975) 'On the development of complementizers in aeolization." Working

Papm on Lnnguage Univemals #17, Stanford University. Washabaugh. W. ( 1977) 'Consuaining Variation in Deaeolization." Languuge 53(2): 329-

52. Watson-Gegeo and Gegeo ( 1986) 'CalLing out and repeating routines in Kwara'ae duldren's

Language socialization. " In Language Sucialirntion A m s Cultuns , eds. B. Schieffelin and E. Ochs. Cambridge: Cambridge University Press.

Wertsch, J . ( L 985) Vygotiky and the Social Formation of Mind, Cambridge, Mass.: Harvard University Press

Wertsch, j. ed ( 1985) Cultirn, Communication and Cognition, Cambridge: C.U.P.. Wek, J . ( 1982) Amts of English, vol. III Bqond the Britirh I s h . Cambridge: Cambridge

University Press. Williams, Brakette (1984) "Ef me naa bin corne me naa been know': sociai control and the

Afro-Guyanese wake, 1900-1948." Caribbean Quattmiy, vo1.30. nos. 3-4. Williams, Brakette ( 1986) "Humor, linguistic ambiguity, and disputing in a Guyanese

community." Intemational /oumal of the Sociobgy of Langwigc 65, pp.79-94. Wiliiams. BDkette ( 199 1) Stains on My Name, War in My Vcins: Guyana and the Politia of

Cultural Stnrglerc, Durham: DukeUMversity Press. Wion. P. J . ( 1969) " Reputation and Respectability: A Suggestion for Caribbean Ethnology. "

Man (ns.) 4:70-84. Wilson. Peter . J. ( L995[ 19731) C d Antics: A Canbbeun Cuse Study of the ConJict betwem

Reptation and Respect. Iiiinois: Waveland Press.

Winford, D. ( 1 9 72) A sociolinguiztic desnptim ofmmuniticr in Tdidad. D.PM niesis, Universitv of York

Winford, D. ( l985a) "The concept of diglossia in Caribbean mole situations." Language in Society L4(3):345-56.

Winford. D. ( l98Sb) 'The syntax offi complements in Caribbean English Creole." L a n p g e 6 1(3), 588-624.

Winford, D. ( 1988) 'The aeole continuum and the speech cornmuni~y as the locus of language" in The Intemational Journal of the Sociology of Language, 7 1.

Winford, D. ( 1990) "Copula variabüity, accountabiiïty and the concept of 'polylectal' grammars." Journal ofpidgin and C d Lunguages 5223.252.

Winford, D. ( 199 la) 'The Caribbean." in J. Cheshire ed. Engiirh h u n d the Wod: S~n'olinguiitic Pmpectivcs. Cambridge: Cambridge University Press.

Winford, D. ( 199 lb) 'The passive in Caribbean English Creole." in W. Edwards and D. Winford eds. Vn6 Phracc Pattmrr in Black Engiish and Cnolc. Detroit Wayne State University Press.

Winford, D. ( 1992) 'Another look at the copula in Bhck English and Caribbean Creoles. " Ammimn Speech 67.1:2 1-60.

Winford. D. ( 1993) Predication in Caribbean Engiish Creoles. Amsterdam: John Benjamins. Woolard, K. ( 1985) "Language Variation and Cultural Hegemony: Toward an Integration of

Soaolinguistic and Social Theory." Amdan Ethnulagbt 12: 438-448. Woolard, K ( 1992) "Language Ideology: Issues and Approaches." in Kroskrity, P. and B.

Schieffelin. K. Woolard eds., Pragmatia 2(3):3 1 1-24. Young, Iris Marion ( 1990) Iustice and the Politics ojfDflmrce. Princeton: Princeton Universitv

Press.

IMAGE EVALUATION TEST TARGET (QA-3)

APPLIED IMAGE. lnc - = 1653 East Main Street - -- - , Rochester. NY 14609 USA -- -- - - Phone: 71 6/482a300 -- -- - - Fax: 716/2865989