Pragmatics2:3. 445 -449 InternationalPragmatics Association

MULTIPLICTTY AND CONTENTION AMONG IDEOLOGIES:A COMMENTARY

SusanGal

It is a pleasureand a challengeto comment on this fine set of papers becausethey bringtogether a great diversityof ethnographicmaterials and a range of approaches towards"linguistic ideology."They are all trying to define and use this conceptual categoryin new ways, while sensing its familiarity. Woolard urged uS, in her introduction,to start not with strict definitionof termsbut from the broad premisethat "linguisticideology is a mediatinglink betweensocial structures and forms of talk." I hopethat the term can provide a reframingor Nekker cube effect:the realizationthat manyof the things we have been studyingunder different labels - common sense notions,metalinguistics, status/solidarity, rhetoric, languageattitudes, worldviews in language- can be brought together,revealing family resemblances,and inspiringnovel analyticconnections. Therefore, my aim here is simply to highlightthe most important themes that emergefrom thesepapers despite the followingimportant differencesin subjectmatter: Thesocial groups discussedrange from face-to-facecommunities of hunter-gatherers in Venezuela,to Quebecoisindustrial and clericalworkers, Haitian farmers, Catalan students,US, Haitian and Indonesianscholars and academics.The linguisticmaterials discussedare similarly diverse, including interactional genres in classrooms,in mourhing,in shamanisticsong, as well aswriting systems,details of linguisticelicitation sessionsand, most generally,what people explicitlysay - or have historicallywritten -- abouttheir languagesand languageuse. One kind of evidencenotably missingis full scaleformal or variational analysisof linguisticstructure. But common themes are more interesting.It may point to a welcome blurring of the well-known divisionsof scaleand geographywithin our sub-disciplinethat the scholarsgathered here includeseveral who havewritten aboutsocial units of more than one scale e.g. national languagepolicy as well as face-to-faceinteraction (Collins, Errington,Heller, Schieffelin),and have conductedresearch in severaldifferent geo- politicallocations (Briggs, Collins, Schieffelin). The themesand questionsI would like to emphasizecan best be grouped around three points.

1.Idea, practice, power

Ideologyis conceptualized- implicitly or explicitly - not only as systematicideas, culturalconstructions, commonsense notions, and representations,but a/so as the 446 SusanGal everydaypractices in which such notionsare enacted;the structuredand experienced socialrelations through which humansact upon the world. Thus the papershere discuss not only explicit ideas about language.They also infer ideasabout languagefrom the way that people use it. This combinationrejects the implicit linguisticideology of much socialtheory and of structuralistlinguistics, which strictlyseparates the ideationalfrom the material, words from actions. And the combination draws on a number of traditions,some more "local" to linguisticanthropology than others.First, the analysis of systematic ideas - here ideas about some aspect of language - continues ethnoscientific attempts to get at worldview by specifyingthe semantic relations between native terms. It continues,as well, symbolicanthropology's intent to treat ideologiesas cultural systemsand to deciphertheir logic.Joined to thesetraditions are contemporaryliterary and philosophicalapproaches that haverediscovered rhetoric and the question of how texts make meaning. Second, in their attention to practices,these papers rely not only on the contextual approach of the ethnographyof speaking,but on (Continental) traditions which, though differing from each other in important ways, neverthelessshare the emphasis on ideology as embedded in everydaypractices. These include structural Manrist "subjectivities,"Bourdieu's "symbolic domination," Gramsci's alwayscontested "cultural hegemony" and evenFoucault's discussion of the "disciplines"of everydaylife . Thus, Collins infers a set of beliefs about languagefrom Tolowa speakers' unsoliciteddiscussions of "favoritewords," DiGiacomo interpretsclassroom question- sequencesbetween Catalan studentsand teachersas enactinga set of assumptions about what a languageshould be like. Briggs infers ideas about intentionality from the recipient designand participantstructure of Warao ritual wailing.Errington alludesto the stratification system encoded in the use Javanesevariants. All buttress their analysesof practiceswith explicit statementsthat speakersmake about beliefs and ideas.And it is clear that explicitstatements need not be congruentwith what practices imply. Indeed, Heller focuseson the recurrent contradictionsbetween the beliefs of bilingual women in Canada, and the bilingual practicesthat they adopt. Similarly DiGiacomo showsthat the classroominteraction contradicts what the teacherexplicitly claims about Catalan. But the papersin this sessionare inspirednot only by the emphasison practice, but also by social theorists' various attempts to redefine the relationship between languageand power. In addition to the familiar idea that linguisticpractices provide access to material resources (see Heller) and further provide the means for participation in decisionmaking ("a voice"),there is the important notion that ideas about languageand particular linguisticpractices provide representationsof reality.If such representationsare authorizedby societalinstitutions, they seryethe interestsof some groups better than others,and are thus sourcesof socialpower. In one way or another the papersin this sessionexplore this "realitymaking" aspect of linguisticideas and practices,and its relation to socialpower. They reject a view of linguisticideology as neutral cultural constructions,but equallyreject the view that ideasabout language and its role in sociallife are simply distortionsof a separatelyknowable reality. They employ the fruitful strategyof focussingon how actors conceptualize,formulate and Multiplicity and contention among ideologies 447 expresstheir disagreementsand conflicts,especially concerning particular linguistic practices,e.g. arguing and disagreeing(Briggs), writing systems(Schieffelin, Collins) or grammars(DiGiacomo, Errington).

2. Multiplicity of ideologiesand expert knowledge

Thefocus on conflictshighlights the most important aim of thesepapers: to document themultiplicity of languageideologies in eachof the socialorders discussed. This goes directlyagainst two widespreadtendencies. In traditional ethnographicapproaches, including the ethnography of speaking, we often attributed a single, patterned worldviewto a socialgroup. Briggssuggests this in his critique of Rosaldo'swell-known descriptionof Ilongot speech acts. Even in variationist studies, which otherwise celebrateheterogeneity, the speechcommunity was defined as the locus of shared evaluationsand attitudestowards varieties. This excludedvariation from the realm of whatwe here would call ideology. More recently,as linguisticpractices and ideashave been discussedin terms of symbolicdomination, there has been a tendency to identify a single, rnonolithic "dominant"ideology. Similarly, although cultural hegemony implies multiplicity and the likelihoodof resistanceor opposition, analysesbased on this notion have usually juxtaposedone dominant practiceor ideologywith one form of resistance.In contrast, thepapers in this sessionshow that evenin smallface-to-face communities ideas about languageare various and can be contradictory.Briggs'examples about the precedence of the "thinking-speaking"model of discourseover other modelsamong the Warao also suggeststhat even in such face-to-facecommunities, some models are more socially powerfulthan others. In larger social systemsit is clear that there are discursivebattles amongelites for which ideasare to be institutionalizedin law, education,industry. And withinalready established institutions the debatesamong dominant ideologies continue. Theseessays are strong in demonstratingthe diversityof ideas about language,how theychallenge each other, formulate different social realities, and thus support different configurationsof power. Further work might suggestthat such ideas are not easily resolvedinto dominant and alternativeversions. An important, related theme is the role of experts- philologists,sociolinguists, grammarians,teachers - and expert knowledge,in producing conflicting linguistic ideologies.The frequent and fertile slippagebetween analytic and everydayuses of terms,the usageof this slippagein creatingcultural capital, the naturalizationand later de-naturalizationof linguistic facts, and the common strategy of appealing to the authorityof science,are issuesthat all deservefurther investigation.Several papers attemptto critique our own academicideologies of languageby putting them in the sameanalytic frame as those of the people studied(Briggs, Collins). Drawing on Foucault,Collins notes how ideasabout language(e.g. that it is an autonomoussystem that can be studiedin isolationfrom context)provide self-definition and self-justificationfor expert groups such as linguists in the US academy, and managersin educationalbureaucracy. He describes"multi-party conflicts" waged with 448 Susan Gal such ideas, in which the stakesare jobs, and other material advantages.Collins and Heller are particularly precisein describingdebates among different groups who are in a position to define a broad socialreality, as well as the responsesof thosewho must collude and calculate with or resist that reality. Schieffelin and Doucet show that debates about Haitian orthography not only divide the experts, but create/define constituenciesamong different fractionsof the Haitian elite. Literary people, school teachers and parents bent on upward mobility for their children defend the pro- etymology position, and are opposedby thosein governmentagencies working for adult literacy. This study, along with Errington's discussionof debates among Javanese intellectualsand politiciansabout the refigurationof the Javaneselanguage, bear on my next point.

3. Linguistic ideologiesand other conceptual systems

Ideologiesof languageare important for socialanalysis because they are not only about language. They envision and enact connections between linguistic and social phenomena. Briggs comments explicitly on this, but all the papers offer interesting examples.As Ochs, Borker and others pointed out years ago, cultural constructionsof gender are often linked to ideasabout waysof speakingin diversesocial groups. Many of the papers I am discussinghere focus on another modern category of identity: 'nation.'By contrast,a somewhatdifferent body of literaturein linguisticanthropology shows language ideology to be about conceptionswe might label aesthetic,moral or epistemological. These concern intentionality, sincerity, emotion, truth and evidence, sourcesof thought and responsibility,clarity, simplicity. If at first glance these do not appear to be related to group identity, a closer look reveals that they too are often deeply enmeshedwith conceptionsabout the nation. 'nation' 'language' Both and are terms for cultural categoriesthat are used for analysis as well as in everyday life. In the course of the 19th century, European philosophy,social thought, and political practicedid the ideologicalwork of making the connectionbetween them appear a necessary,natural, and self-evidentone. Although later generationsof linguistsand anthropologistswould work similarly hard to de- naturalize the link between language, race and social group, the connection was borrowed and recontextualized all over the globe. There are, however, at least two 'nation,' modern sensesof as Eric Hobsbawm has recently reminded us, both conceivedas having a'state'as its ideal home, One is a cultural relation betweenstate and subject basedon political participation and known as citizenship;the other is based on some imagined, shared cultural unity. Their relation to ideas about languageare interestingly different. In the first case, a uniformity of language throughout the nation-state is understood as necessaryfor free communication,an informed and mobile citizenrywho can exercisepolitical rights. Notice that referential adequacy,simplicity, clarity, and individual intentionality in languageand thought are aestheticand moral notions often invoked in such discourse.In the secondcase language is understoodas an expression Multiplicity and contention among ideologies 449 of communal spirit and the uniformity of languageis important not for efficient communicationand broad participationbut as proof that the speakingsubject is an authenticmember of the nation, linking speaker and languageto the past and its (invented)traditions. Different moral and aestheticissues are involved in each case. In both of thesewell known conceptionsthere is a logic of boundednessthat invitesus to imagine - againstthe ubiquitous evidenceof variation in languageand society- the happyfusion of a circumscribedand internallyhomogeneous language with a similarlyconfigured nation. There are certainlyother culturalrelations between state andsubjects, for instanceone typical of state-socialistsocieties. Any suchrelation, like "nation,"and in common with other categoriesof political theory and practice,such as "the people,""the masses,""the public" includesimplicit assumptionsabout language thatdeserve exploration. Notice too, that both ideasof the nation can be evokedin the samecountry, can be usedby minorities,indigenous populations or competingelites to constitutethemselves and argue with each other. Claims to the national self are madenot only to internal audiencesbut to other nations,and as Errington showsfor Indonesia,are involved in handlingcomplex relations to internationalcapital. The implicit logics I have outlined form the backgroundto the more specific discursivebattles describedin these papers. What part of the population is really Haitian and what standardizedwritten or spokenlanguage will representthem? One versionof Catalan is to be the authenticone, but which current speakerswill be able to understandit? Why make Javanese- with its levelsof socialdifferentiation - into an internallyhomogeneous ethnic language? In the courseof highlightingthe multiplicityof linguisticideologies - as ideasand practice- these papers begin two further and related tasks: 1. to understand the semioticprocesses by which chunks of linguisticmaterial (".g. orthographic systems, archaicvs. new forms) gain significanceas representationsof particular parts of populations;2. to unravel the semioticand rhetoricalmeans by which our own expert theoriesas well as everydayand political argumentslink together such apparently diverseculturalcategories as language, spelling, nation, gender, simplicity, intentionality, authenticity,development, tradition.