Radio Timb Sharing and the Negotiation of Linguistic Pluralism in Zambia1

Pragmatics2:3.335 -35 4 InternationalPragmatics Association DOI: 10.1075/prag.2.3.12spi

RADIO TIMB SHARING AND THE NEGOTIATION OF LINGUISTIC PLURALISM IN ZAMBIA1

Debra Spitulnik

Theseven (7) mainvernacular languages on the Media representall seventythree (73) tribes foundin Tambia(Kapeya l98S: 3).

, Theproblem is, we havebeen recognized as a language,so why be mistreated?. . . It remains thatthe big two or the big four get mostof the time. . . But you can't raisethese issues because youmay be calleda rribalist(Anonymous).

L Introduction.

massmedia construct the communicativespace of the nation-state,all of a nation's dialects,and language varieties, and the speechcommunities associated with are automaticallydrawn into relations with one another. Some languagesand maybe dominantwith a greaterquantity and quality of media products,some berepresented only minimally,and still others may be completelyabsent, either or ignored.While there is a growing literature on how such institutional are bound up with the politics of ethnicity and national identity, the ing' of national communities,and the establishmentand maintenance of linguisticpractices2, it seemsto me that three basic issueshave remained understudied: (1) How do variousinstitutional practices differentially value languages (and variants),and which aspectsof languagein particular become important in institutionalconstruction of languagevalue? (2)Howdo peopletalk about (andjustify or contest)these institutional practices

I A muchshorter version of this articlewas presented at the AAA conference,November 23, 1991,. isdrawnfrom 1988-1990research at the TambiaNational BroadcastingCorporation, which was byFulbright-Hays and NSF fellowshipsand facilitatedby the Institutefor African Studiesar Univenityof Zambia. I am especiallygrateful to the numerousZNBC personnelwho generously I Oeirtime and encouragement to this project,and to SusanPhilips, Michael Silverstein,Edwin andtwo anonymous hagmatics reviewers for their insightfulcomments on earlier drafts. 2See,forexample, Anderson(1983), Collins (1988), Fishman (1989), Gal (1989),Silverstein (Lg87, andWoolard (1985, 1989). 336 DebraSpinlnik and valuations? (3) How are (1) and (2) related?How is discourseabout languagevalues related to the valuationsof languagein practice? This article attempts a preliminary exploration of these issuesthrough a case study of the politics of linguisticpluralism in Zambian radio. Echoing the introductory '73 quotes, we will examine how sevenradio languagesare made to "represent" ethnic groups, and how certain languages are accorded high status while others are "mistreated." Before turning specifically to the Zambian situation, however, we will elaborate the issuesraised above and sketch out their connectionsto the problem of language ideology as it has been variously defined in recent work. In some analysesthese types of commentsand practices- e.g. evaluating some languages as "better" than others or constructing a national scheme of linguistic pluralism - would be considered language ideologies, or evidence of language ideologies.Certainly thesephenomena are'ideological'in the sensethat they feed into and are fuelled by relations of power and interesr.They are also 'ideological' in the sense that they are tied up with the cultural conceptions(or modes of understanding 'ideological' and evaluation) that characterizeany given society.And finally, they are in the sense that languagescan representspeech communities, and thus become the focus of political strugglesover who counts and is counted in the national arena. As both Woolard (this volume) and Friedrich (1989) point out, such applications of the term "language ideology" traverse the complex (and often competing) formulations of ideolog5r in social theory, e.g. as shared beliefs, secondary rationalizations,and modes of legitimating power relations.These types can overlap to some degree and may also be differently defined in terms of where ideology residesand how, analytically,it can be distilled. Thus some definitions revolve around distinctions between "explicit" and "implicit" ideology, or the relationship of consciousnessand speech to ideologieswhich are embodied "in practice." But as the applications of the term "ideology" proliferate, it seemsto be less and less useful analytically. And if, as Friedrich suggests,"everything" is potentially ideological in the sensethat the cultural construction of meaning and value is bound up with relations of power and interest (1989: 300), then the term is only useful in a technical sense within a theoretical framework which specifies what kinds of data count as ideology (i.e. what if not "everything"is ideological), and what kinds of relations of determination exist between these levels of phenomena (e.g. between the ideologicaland the non-ideological,or between power and culture). I raise these difficult theoreticalissues not so much to take a stanceon them, but to clear some ground, and to return to some more basic questionsabout the kinds of evidence that we use (1) to identify and distill cultural conceptions (and understandings)of language,languages, their speakers,and languageuse, and (2) to investigate the role of powerful institutions in constructing and maintaining these conceptions(or valuationsor 'ideologies').Rather than looking directly for ideas,or who wins, it seems more productive at least initially to examine such questions of language ideology through the recurrent themes that emerge from a range of types of data (e.g. interviews,political documents,participant-observation, historical changes, Lingtisticpluralism in Zambia 337 etc.).Silverstein (1987) and Woolard (1989),for example,locate languagevaluations in therhetorical structuring, folk views,and common metaphors,analogies and tropes thatdominate certain languagepolicy debatesin the United States.While any or all of thesediscursive modes and moves might be called "languageideologies", neither of thetwo studiesuse this cover term, and instead frame the problem of the politics of languagevalue specifically in terms of the production of certain kinds of discourseand commentary. Followingthis approach here, I hope to introduce a further dimension to the problem,namely, how one might connect the ways that people talk about the values androles of languageswith broader institutional processes.Thus beyond identifying the commonthemes and tropes in the spoken battles over languagevalue in Zambia, I attemptto traceout their sourcesin and resonanceswith the national discourseon pluralism,the historical emergence of a scheme of language ranking, and the differentialvaluation of languagesin broadcastingpractice. And although this essayhas openedin a sensewith the spokenbattles, I begin with the more general processesto establishthe historicaland political momentum behind the various ways that people talk.

2.An overviewof radio's role in the negotiation of linguistic pluralism.

Sinceits inception51 years ago, radio broadcastinghas been one of the most visible and contested arenas of language valuation and national language policy implementationin Zambia. Fundamentally, what underlies the political volatility of languageuse on radio is that Zambian broadcastingis the primary state institution for representingboth national unity and national diversity.Radio is restricted to use only ei$t of the country's 15 or so languages:English, Bemba, Kaonde, Lozi, Lunda, Luvale,Nyanja, and Tonga. At the sametime however,radio is chargedwith upholding a nationalphilosophy of ethnic egalitarianism(among the nation's"73 tribes") and the obviousaim of political encompassment,as encapsulatedin the national motto "One Zambia,One Nation." Collectively, then, the eight radio languagesare intended to rcpresentboth the nation as a whole and its internal diversity. In this processZambian radio operates with trvo competing visions of the nation'sdiversity. Radio's use of different languagesis structured by both a vision of ethnicegalitarianism in line with nationalistdiscourse, and a hierarchicalranking of the nation'slanguages which is linked to the political and demographicinequalities among theirspeech communities. As we will seein more detail below, Radio Zambia balances thesedifferent pluralisms in three basic ways: (1) through a differential allocation of radioairtime which both regulates language equality and gives languages different unequalvalues, (2) throughvarious modes of establishingevaluations of each of the sevenZambian radio languagesvis-a-vis the country's national language English, and CI)by accommodatingthe interestsof 73 ethnic groups within the programming constraintsof theseseven languages. Whilethese various valuations of language"in practice"do not directly dictate 338 DebraSpiulnik the ways that Zambians talk about the politics of language value, there are strong relations of determination.The discussionbelow illustrateshow certain institutional processesgive a fixity and legitimacyto (i.e. naturalize) certain languagevaluations, but are also at times contested,manipulated, or simplyineffective in producingconsensus. A different kind of determination arisesfrom the particular structuresand functions specificto an institution in the sensethat they may focuson some featuresof language (e.g. phonology, syntax, code choice) as more critical than others in constructing languagevalue. As radio is a medium definedand structuredby time, wavelengths,and sound, it is not surprising that much of the concern over radio language use in multilingual situationssuch as Zambia has been about the allocation of airtime and channelsto different languagesand the normativeuse of "standard"or "non-standard" phonology.3 Additionally, there are somemodes of evaluatinglanguages which while implicit in broadcastingpractice, actuallyhave their sources(or are just as solidly grounded) outside of radio, in national discourse,in labor markets, and in more general understandingsabout the relations between language,culture, and identity. For example, a major theme in the negotiation of linguistic pluralism in Zambia concerns the way in which "a language"is taken to emblematize a people, an ethnic group, or a political unit. Closely tied to this interpretation of languagesas representing collectivitiesare evaluationsof the communicativefunctions (or domainsor topics)that they canonicallyserye. Using Ferguson'sclassic formulation of diglossia,this would involve the "specialization of function" of different codes across different speech situations,contexts, and genres(1959: 328).But from the perspectiveof institutional practices,rather than speakerchoice or overall contextualdistribution of codes,this "specialization"depends on how the appropriatenessof differentlanguages for different functions and topics is legitimated by some regularity of practices.For example in Zambian broadcasting, some languages are constructed as more "intellectually equipped"and others as better suitedfor "cultural expression"through their exclusive use for certain program types. Significantly,these perceived qualities of languagesare entangled with particular assessmentsof their speakers,e.g. as rural people, urban consumers,"illiterate", "sophisticated", "insignificant", etc., and I would argue that the two modes of evaluationare not really separable.These evaluationsdo not emanate strictly from radio, however,but are more directly grounded in the overall political economy of languagesin the country, os linguisticcompetence (and membership in certainspeech communities) structures access to education,labor markets,and political power.

' Interestin normativelanguage use also focuses on issuesof "non-standard"vocabulary, idioms and syntax,as well as normsof languageinteraction (e.g. whether broadcasters exhibit proper turn-taking and deference). Linguisticpluralism in Zambia 339

3. Pluralismin national politics.

Ihe culturalization of ethnicity and the vices of tribalism.

Turning now to Zambia's national discourse on pluralism, there are 73 officially recognizedethnic groups which are constructedas equallydifferent, i.e. different in the sameway, in terms of their presumably harmless, apolitical 'cultural' differences. Amountingto a virtual culruralizationof ethniciyt,each ethnic group is said to have its ownunique traditions, dances,songs, and history, which it is encouragedto preserve and promote for the sake of the youth, tourism, and national identity. National discoursethus attempts to diffuse the political dimensionsof ethnicity at the same time thatit promotesan image of the state as tolerant of diversity.References to the more volatiledifferences between ethnic groups,particularly their inequalitiesin population, politicalpositions,and economicresources, are avoidedat all costs.Furthermore, there is no direct discussionof ethnic favoritism or conflict at the national level (except to denouncethem), although both significantlystructure social relations and politics, and arefrequent topics of everydayconversation. Instead, politicians constantly invoke the national motto "One Zambi4 One Nation"as a symbol of a national unity which is predicated on being able to hold togetherand tolerate ethnic difference. On this note, Zambia is rather exceptional amongmodern African states as it has experiencedalmost no violent ethnic conflict, andhas prevented any one ethnic group from monopolizing political power. But while claimingto have forged a national unity of."73 tribes united", the Zambian state has alwayssimultaneously attended to the politics of ethnic difference and special interest. Zambia'sfirst and long ruling presidentKenneth Kaunda carefully orchestrateda policy of "tribalbalancing" in political appointments,and early on articulated what became the generalconsensus in Zambia's political culture regarding the "deadly vice" of tribalism (Kaunda,L967: 11). Analogous to racism or sexism, tribalism is understood as discriminationor preference based on ethnic identity. Warnings against "falling victim t0 tdbalism" and denunciations of its divisive, retrogressive, and fundamentally un-Zambiannature pervade political oratory, and were continually raised during one-partyrule (1972-1990)in illustrationsof what would ensueif a multi-party system wereadopted. In a very curious and complex way, then, Zambia's rhetoric of encompassing nationalism("One Zambia, One Nation") and its philosophy of ethnic pluralism are mutuallyreinforcing, but are also in a constantstate of tension with the possibility of ideologicalimplosion. The national claim to build unity, fend off tribalism, and also encourageunique ethnic cultures,creates a cautiouspluralism within bounds, where diversityalways verges on divisiveness,and where attention to difference itself borders on subversion.As we will soon see, this, in a nutshell, describesRadio Zambia's positionconcerning language diversity as well.

;. * I 340 Debra Spitulnik

4. Democratic vs. hierarchical linguistic pluralism.

A convergenceof language history, po@, and demographics.

Alongside the national discourse of apolitical ethnic egalitarianism, Zambia has developed a national language policy which also stresses pluralism but ignores difference. This has been accomplishedmainly through sanctioningan extraordinarily high number of "official languages"- seven - which are all subordinated to the ethnically neutral "national language",English. The seven official Zambian languages are Bemba, Kaonde, Lazi, Lunda, Luvale, Nyanja, and Tonga, and they are the only Zambian languages used in broadcasting, schools, and government publications. English, the former colonial language,is the languageof government,higher education, and international communication.The choice of.whichseven, out of a total of between 15 and 20 distinct languagesand perhaps as many as 50 different dialectsa,was fairly well determined before independenceby the convergenceof severalprocesses: colonial language policy (especiallyin broadcastingand education),missionary work, and labor migration patterns which establishedBemba and Nyanja as the colony's two urban lingua francas.s Crucially, English, Bemba, and Nyanja were the first languagesof Northern Rhodesian broadcasting,which started in 1947as a World War II information service. Following the war, Lnzi and Tonga were added as the serviceexpanded. According to Harry Franklin, the Northern Rhodesian Director of Information, nearly everyone understood these four "main 'root' languages of the country", and thus it was unnecessaryto add more languagesdespite several requeststo do so (1950: 16). Significantly, Franklin saw radio as having a key role in helping to standardize languagesand promote the dominance of some over others: Africa'sbabel of tonguesis oneof themain causes of its people'sbackwardness, and we hope eventuallyto reducelanguages broadcast rather than to increasethern, and so to playa partin the eventualdevelopment of oneAfrican tongue . . . andstill moreremotely, in the universal useof English(1950: 16). Over time, the selection and dominance of the four languages became mutually reinforcing, and this arrangement was extremely convenient administratively, as their geographic distribution virtually aligned with the four compasspoints North (Bemba), South (Tonga), East (Nyanja), and West (I-nzi), which defined the major provinces of

a The mostcomprehensive studies suggest thar thereare 15to 20 distinct(non-mutually intelligible) languagesin Zambia, but many of the 73 ethnic groupsclaim to speak their own unique language (Kashoki,1978). Behind this is the complex,political issue of defining"a language"vs. "a dialect",which is further compoundedby the limited comparativedata on languagevariation in the country. Cf. Kashoki and Mann (1978)and lrhmann (1978).

5 Historical data is drawn mainly from Kashoki(1978), Roberts (1976), and Mytton (1978). Linpisticpluralism in Zambia 34I

the colony.6In 1954,however, the two Northwestern Province languagesLunda and Luvalewere added to broadcasting (no single language was prominent in the province),and at Zambia's independencein 1964,an additional Northwestern Province language,Kaonde, was included. The contemporarylinguistic situation inZambia is in large part a legacyof these trendsin colonialadministration, and the related trends in ethnic politics. In effect, the useof numerousZambian languagesin early radio broadcastingdid not help to reduce "theAfrican Babel", but helped to permanently establishsome of the basic terms in whichethnolinguistic diversity is structured and experienced in Zambian society. Differentlanguages became associatedwith different prestige values and domains of usein the nationalcon_text, primarily through the "uneven development" of the ethnic goupsspeaking them/ and their concurrentunequal valuation in broadcasting.The highstatus and widespread use of Bemba and Nyanja, "the big two", grew as their value in theurban labor marketsbecame established, and this was accompaniedby higher proportionsof radio airtime for them. These two, along with Lozi and Tonga came to beknown as "the big four" as their first languagespeakers emerged as the major ethnic and regional power blocks in national politics (cf. Molteno 1974). For most of broadcastinghistor!, Lozi and Tonga have had slightly lower proportions of airtime thanBemba and Nyanja,but more than the three Northwestern languageswhich are associatedwith much smaller populationsand ethnic power bases.The use of these threein contemporaryZambia is fairly limited to their own mother-tonguepopulations (e.g.l%oof all Zambiansspeak Kaonde, while 37oof the countryis ethnicallyKaonde). By contrast,the Bemba and Nyanja languagesfar outreach their ethnic populations (l9%oand l6Vo) and are spoken by 56Vo and 42% of all Zambians, respectively (Kashoki,1978). What obtainsthen in Zambia is not a situation of democrarrclinguistic pluralismin linewith the nationaldiscourse of ethnic egalitarianism.All languages(or evenjust the seven chosen languages)are not equally valued, rather, they are hierarchicafuranked (Figure 1). This languageranking is both a historicalproduct of colonialadministration and ethnic politics,and a partial diagram of the demographic andstatus inequalities among speech communities.

,l I ii I ii

6 The.ont.mporarydiscourse of regionalism,which flowsdirectly from theserelations of power establishedduring the colonialperiod, is one of the major tropesof ethnic and languagepolitics in Zambia,and deserves much more treatmentthan is possiblehere (cf. Mitchell 1956:28-30 and van Binsbergen1985: 203).

7 Thehistorical connections between colonial language policy and the constructionof ethnicityin Zrnbiaremain ro be fully investigated.But seeMitchell (1956)and van Binsbergen(1985) on how dbpritiesin missionization,rural development,and acressto educationand labor markets were imtrumentalin the articulationof ethnicdifference and affiliation during the colonial period. 342 DebraSpirulnik

Figure 1. Hierarchicallinguistic pluralism rn Zambia. colonial language English "nationallanguage' international language government,higher education, I TV, radio, daily newspapers urban lingua francas Bemba,Nyanja I languages of dominant Bemba, Nyanja, Lnzi, Tonga ethnic groups \ "official languages' schools, radio, Northwestern Province - monthlynewspapers r r languages Kaonde,Lunda, LUVale I other languagesand dialects Luunda, Mambwe, Nkoya, Soli, Tumbuka, etc.

5. The politics of languagevalue in contemporary radio.

Partitioning the linguktic universe.

The politics of languagevalue, and especiallythe grounding of this schemeof hierarchical linguistic pluralism, occur most prominently in four domains in which the Zambian linguisticuniverse is essentiallypartitioned through broadcasting practice: the radio channel,airtime, administrative resources, and programcontent. In the following, we will examine how the very way that the medium of radio is carved up establishes basic structural contrastswhich can become focal points for the construction (and contestation)of languagevalue. Since the early daysof broadcastingin Zambia the airwaveshave been carefully divided up among different languages,and concurrentlydistinct administrativeunits have been assignedresponsibility for the output in each of the different languages. Most pronouncedhas been the segregationof Englishvs. Zambian languagesin terms of both airwaves and program types - and one of the key issuesof attention and contestationhas been which Zambian languagesare treatedmore (or less)like English, which clearly occupies the privileged position in the language hierarchy. Table 1 illustratesthe specialstatus of English in terms of ZNBC's radio channels.8Not only are there currently more than twice as many hours for English as for all Zambian languages combined, but the two English channels also monopolize the more

8 While technically"channeln denotes a broadcastfrequency, it ZNBC's term for a distinct programmingschedule. ZNBC runsthe country's only broadcasting operations: one televisionand four radio channels.Radio 3 is an externalservice for SouthernAfrica. The three domesticradio channels are all centrallyrun from the capitalLusaka and havesimultaneous national transmission on their variousfrequencies. Linguisticpluralisnr in Zambia 343 prcstigiousFM frequencies.Furthermore, up until 1991,Zambian televisionbroadcast exclusivelyin English,and presentlyit carriesonly one hour of Zambian languagesper day.The names "Radio 1" and "Radio 2" were introducedin 1989with the creation of "Radio4", and were precededby five other pairs of labels. The various alterations in boththe channels'names and languagesreveal a constantattention to the politics of languagevalue, and illustratethe ongoingpractices which are necessaryto sustainand justiffcertain valuations; we considerthem in detail below.e

Trble1. ZNBC radio operations,1990.

Hours Frequencies languages7o of totalradio broadcast time RadioI 04.50-24.05 SW,MW 7 Zambianlgs 3lVo Radio2 04.50-24.05 SW,MW, FM English 3l Vo Radio 4 Contin. 24 hrs FM stereo Enplish 38Vo

From the late 1940s up to Zambian independence in 1964, broadcasting operationsfor Africansand Europeanswere separate.The "African Service"emanated fromLusaka (the Northern Rhodesiancapital) and the "EnglishSeryice" came from Salisbury(the Southern Rhodesian capital). The African Servicecarried a few programs in Englishfor Africans, but most English language programs were produced in the SouthernRhodesian studios. Zambianbroadcasting inherited this operational imbalance andlanguage segregation at independence,and out of this situation the "Vernacular Scwice"(with 7 Zambian languages)and the "National Service" (with English only) werecreated in 1964.Racial overtones were thus eliminatedwith both the rejection of thelabel "African" and the revaluing of English as something "National", i.e. rightfully belongingto the newlyindependent nation-state and not any one race. But even with thesenew labels, the comparativescopes of the two channelsand their implicationsfor languagevaluations remained fundamentallythe same: Zambian languageswere still segfegatedfrom English,which stood apart as the languageof wider, national, and moremodern communication. Oppositionsto these distinctions erupted during the two years following independenceas the channels' names were changed three more times, and controversiescontinued as the allocationof languagesto each channelwas altered as well.ln1965, for example,the National Servicewas relabeled"the English Sewice",in anattempt to dispelany implicationthat the "vernacular"languages of its counterpart werenot also'national'. The implicationendured, however, and in 1966the Vernacular and EnglishServices were renamed "Home" and "General" thus even further 'not-general', accentuatingdifferences in scope. The Home Service,as something caniedconnotations of being specific, local, and even narrow. Resonating with

9 Sour..s, Handbookto the Federationof Rhodesiaand Nyasaland,1960 (Salisbury: Government Printe4p.552. A Handbookto theRepublic of Zambia,c. 1965(Lusaka: Government Printer) p.105-7. l*hilo,nos. 183-236 (January 29,1965 - February10, 1967). 344 Debra Spitulnik

Zambian idioms of ethnic origin and rootedness,the word "home" also signalled that the "vernacular" broadcastswere for rural and ethnic-basedaudiences, and tapped into their essentialidentities, e.g.

"[ZNBCrealizes] the importanceof a vernacularlanguage as being the [sic]most andproper communicatorand a trueidentifier of man'sbelonging. . . (Kapeya1988: 2).' Even after the Home Service was relabelled "Radio 1" in 1989, ZNBC broadcasters continued to draw upon the themes of home and belonging in promoting the channel, e.g.in Bemba sloganssuch as "Kumwesuhualiwama",'At home (our place), it's nice'. The General Service,on the other hand,as something'not-home', did notbelong to anyone in particular. Associated with wider (more general) spheres of communication, and seeminglydissociated from locality, ethnicity, and special interest, the General Serviceused the one languagewhich symbolizesthe transcendenceof such boundaries, English. This situation did not remain for long, however, and soon several controversieserupted over which languagescould rightfully share time with English on the General Service. In early 1966 "the big four" (Bemba, Nyanja, I-ozi, Tonga) were added with the rationale that their audienceswere larger and therefore they needed more hours. But a year later, I-ozi and Tonga were removed, leaving "the big two", which remained on the General Service with English for nearly 20 years, thus acknowledgingthe wider scopeof the linguafrancas and the possibilitythat they did not necessarilybelong to any one people. The problem, was however, that while Bemba and Nyanja are widely used by people of all ethnicities, their links to specific ethnic constituenciesremains a continually invocable reason for limiting their use at the national level. Bemba and Nyanja are simply not 'general' like English, and in selectingthem over others one is open to being "called a tribalist", as this complaint to the Ministry of Information and Broadcastingimplies: nSome languages have more hours than others [and] only nvo of the seven languages are on the nOne General Service . . . [This is] unfair as far as the national motto of Zambia, One Nation" is concerned. (Lrtter from Northwestern Province, received 13 August 1981, ZBS|101/1/5)." Since the national motto "One Zambia, One Nation" is about uniting an ethnically plural society,to these letter writers unity was violated by separatingout two Zambian languagesto share a channel with English. As the pressuresincreased for the General Service to "include everybody",the other five languageswere added in the mid 1980's but equal time was not granted. The situation was then dramatically reversed in 1988. All Zambian languageswere removed from the General Service, and it resumed its much earlier status as English only. Significantly,this 'all or nothing' decision evoked both praise and criticism, essentiallydrawing from two competing visions of pluralism. On the one hand there was the danger of having too much pluralism, i.e. a pluralism verging on divisiveness and confusion ("the African Babel"), with the possibility of contamination especiallywhen it came to mixing on a com-mon channel with English. One listener,for example,lauded the removal of Zambian languagesand expressedthe Lingubticpluralism in Zambia 345

"sincerehope that these changes will helpbring back to life the [sic]turnished image of the generalservice . . .^(Times of Zambia,October 16, 1988, p.4). Othersfelt that the presenceof Zambian languageson a channelwith English was too disruptive,or that only English was necessaryfor General Service programs such as health,farming, and educationalbroadcasts. The colonial bifurcation of English vs. all other languagesthus remained with lingering connotations of English as the sole languageof prestigeand progress. On the other hand, there was the view of pluralism as positive and necessaryfor nationalcommunication and the promotion of Zambian culture. Several listeners stronglyopposed the removal of all Zambian languagesfrom the General Service, complainingfor examplethat "English has the lion's share" while Zambian languages were "deprived of adequate air time" (Times of Zambia October 1988). One commentedsarcastically, "I suppose that is the penalty they get for being Zambian languages"(Times of Zambia October 11, 1988p.2). The problemof singlingout two languagessurfaced again in 1989when ZNBC introducedRadio 4, a 24 hour FM stereo channel designedto carry a mixture of pop musicprogramming and commercial advertising.toInitially, some ZNBC staff thought Radio4 woulduse all eight languages,but from the outset it was English only. This was justifiedby claimsthat advertisersall preferred English, and that listenerswho spoke only"the home languages"had no buying power and would therefore not be good audiencesfor a channelthat was to be supported by advertisingrevenue. Other staff notedthat since FM radio covered only urban areas, there was no need for the vernacularswhich wete for villagers. But despite all of these seemingly objective appealsto externalfactors, none of these claims were supported by market research, andin fact they ran contrary to the well known fact at ZNBC that urban people do listento Bembaand Nyanja programs. Sowhy weren't the two linguafrancas added to the new urban-basedchannel? Overtwo-thirds of all Zambiansspeak at leastone of theselanguages, and nearly half of all Zambianslive in urban areas.Certainly the listenerswere there. And regarding businessinterest, one senior executiveacknowledged that "Bembaand Nyanja are the only two languages that bring any money [on Radio 1]. That'swhy theyare slotted at lunchtime, and they have mmmercial programs . . . Thesponsors shun some ib of theseother sjI6 flanguages]." F ZNBCexecutives knew that the two linguafrancas could have been successfulon Radio ;. 4, but the choicewas political. In terms of ethnolinguisticimpartiality, the issue paralleledthat of the embattled General Service languages:either have all Zambian languageson the new channel,or none at all. Even if Bemba and Nyanja reach wider

10 Rrdio Zambiahas had commercialadvertising since the 1960s,but Radio 4 was createdto increaseadvertising income. As ZNBC is a parastatalbody with corporatestatus, it can generateand investits own revenues, but it is substantiallycontrolled by the Ministry of Information.Radio 4 is thus lol a "commercial"(privately owned) radio channelin the Westernsense, and outside commercial interestshave had only minimalimpact in shapingprogramming decisions. 346 DebraSpirulnik audiences,their connectionsto specificethnic groups lurks in the backgroundof any decisionover their use. With the additionalaim of stylingRadio 4 after Western FM pop radio, the exclusiveuse of Englishnot only dispelledcontroversies over "tribalism", but was a 'natural' consequenceof its statusas the modern, internationallanguage.

The linguistic division of labor.

The "penalty" that Zambian languages"get for being Zambian languages"thus extends beyond the allocationof airtime and channelsinto a virtual institutionalizationof what countsas their appropriate program content.English has long been establishedas the languageof internationalnews, current affairsdiscussions, and the majority of scientific and formal education programs, while Zambian languagescarry mostly cultural and development programs.lr Ethnolinguisticpluralism is therefore contained not only within one channel,but alsoin termsof certaintopics, and specificallythrough an equal subordinationunder English in this linguisticdivision of labor. Most prominent is the constructionof language equality through linguistic uniformity. To take one example,the sameEnglish script is used to prepare the daily Zambian language newscasts,so the news in each of the seven languagesis identical in format and content. The master text is extractedfrom the script of the preceding English language"Main News",but it is half the length,with the international items usually omitted. In this equal subordinationunder English, Zambian languagesare often equated with the rural folk, even though nearly half of all Zambians and over three-quartersof all radio listenerslive in urban areas.Thus similar to the Radio 4 case above, news staff claim to be just following the norms "out there" when they explain that the "sophistication"of the Radio 1 audiencediffers from the English language audience,and that a five minute excerptfrom the ten minute Englishnews is adequate for them. As one senior news editor elaborated,"We wouldn't give an item on the SALT talks in Geneva. If they could understandthat, they could listen in English." Another said, "You don't pick somethingin Saigonor China,because you maybe reachingilliterates . . . There are complicated issues which wouldn't even make sense to them.' There was no audienceresearch, however, to substantiatethese views, and in a sense, the ability of newscastingto widen people's horizons was short-circuited by the assumptionthat their intelligencewas unalterablylimited. Many Zambianswho know little or no English do have an interest in world events,and would like more international news in their own languages. In apparentcontrast to the newseditors'statements, ZNBC administratorspoint to the useof sevenlanguages as examplesof broadcasting'sdedication to the national

ll The functional segregationof languagesis even more severein the education system. English is the medium of instruction after Grade 3, and Zambian languages are taught as subjects, with traditional culture as typical content. policyof "takingthings to the people." But what,, J;';'":::;:: :;:: people?Most informational programs in Zambian languagesare on national politics anddevelopment topics such as farming and health,and I would argue their audiences arein a sense"the imagined others" for the highly cosmopolitanEnglish speaking broadcastingelite. The audiencesthought of as "illiterates"are basciallythose who havenot reacheda certain level of formal education,in English, and to some media elitethis precludes an ability to comprehendor sharethe sameworld views. Not onlydoes radio's institutionalization of limited programcontent for Zambian languagesperpetuate this perceiveddivide, but the use of English master scriptsfor theseinformational programs further regulateslinguistic uniformity (or egalitarianism), andconstructs English's high value as the scientificand factual language. By contrast, mostindependently scripted programs in Zambian languagesadhere to the politically acceptableexpressions of ethnic diversity - that is, traditional music, storytelling, and culturaltopics. Through these broadcastingpractices, then, the sevenvernaculars are valuedas equally transparent vehicles for communicating the same information to ethnicallydiverse people, and as equally particularisticcodes for gaining accessto differentethnic cultures. 12

Tnu sharingon Radio I: Regulating language samenessand difrerence.

fu indicatedabove, the allocation of radio airtime itself has emerged as one of the key measurementsof linguistic equity: not all languageshave the sameamounts (and types) of air-time,and thusthe promotion of a simple linguisticegalitarianism as one would expectfrom the national philosophy is far from clear. Instead, Radio Zambia has developeda complexschedule which simultaneouslyregulates language equality and giveslanguages different, unequal values.With such a dynamic of languagevaluation, radioboth upholdsa vision of balanced pluralism and operates with its seeming contradiction,a scheme of hierarchicallinguistic pluralism. Significantly,the partitioning of radio airtime among seven Zambian languages andtheir ability to represent73 ethnic groupsis limited to one radio channel. Figure 2 illustrateshow Radio 1's schedule carefully staggers the seven Zambian radio languagesacross the different days and times of the week, and this establishesa basic imageof one national pie that is getting divided up. As people talk about this radio timesharing, various discourses of ethnolinguisticdemocracy are invoked which play intoand seize upon thesestructural contrasts.

12 If Zt*birn languagesare cultural reservoirsthen Englishis an economicreservoir, or at least theavenue for tappinginto economicresources. This issueneeds more extensivetreatment in terms of the overallpolitical economyof languagesand classdynamics in the country, but briefly its significancein radio's negotiation of linguisticpluralism lies in English'sexclusive use as the neutral languageof translationand encompassment.The materialhere suggeststhat radio's (and the media elita')role in constructingthis seeminglynatural neutralityand transcendenceof English cannot be undentated. 348 Debra Spirulnik

Figure 2. Time allocation of Radio 1 languages (1990).13

Time Mon Tues Wed Thurs Fri Sat Sun 04.50-08.00 B T LnlLv Lz K N B 08.00-10.00 Lz K T Lv T Ln B 10.00-12.00 T Lv Lz T Ln Lz T 12.00-14.00 N B N B N B N 14.00-16.00 T LnlLv Lz Lz K N Lz 16.00-18.00 Lz K T Ln Lz B N 18:00-20.00 T Lz Ln K Lv Lv T 20.00-24.05 Lz N B N B T B

[B Bemba; K Kaonde; Ln Lunda; Lv Luvale;Lzl.ozi; N Nyanja; T Tonga]

Below, we see that the national motto "One Zambia, One Nation" becomes a dictum of democracy, i.e. equal representation for all, which then feeds into various modes of measuring language "equality" and "inequality." The middle term in this equation is the essentialistnotion that a languageis a "true identifier"of a people,thus: if languages represent people, then equal representation for everyone in radio will occur through equal distribution of radio's resources (airtime, programs, staff, etc.) across these different languages.One problem which emerges however, is that if languagesdo emblematize people, how are 73 different peoples represented by only sevenlanguages? This sectioncloses with an illustrationof how ZNBC handlessuch difficult cases. In one sense the schedulein Figure 2 presents a picture of distributional equality, where no particular language seems privileged over others. Each day a different languageopens the station, no languagedominates any large time blocks, and acrossthe weekly cycle, every time slot is shared by two or more languages.A similar construction of distributional equality occurs in the Radio 1 news schedule,where a sequenceof back-to-backnewscasts in each of the sevenlanguages is broadcastthree times daily, starting at 6.00, 12.20,and 17.20.Within each of these three news periods the newscastsare identicalin format and content,but the orderingof languagesdiffers at different times of the day, thus in effect no single languagealways gets first place at news time.

13 Based on Zambia National Broadcasting Corporation Radio One hogramme Schedule, First Quarter 1990. Linguisticpluralism in Zambia 349

At the sametime, however,there are great disparitiesin the quantity and qualigof airtime allocatedto each language,and this is summarizedin Table 2.

Table2. Radio time sharingon Radio 1.14

% Vo Days Weekend l-ate Lunch Zambians Radio 1 broadcast hours nights slots speaking airtime language PEAK TIMES Bemba 56% 19Vo daily TT WFSu TuThSa Nyanja 42 15 daily 7 TuTh MWFSu Tonga 23 19 daily 8 Sa Lozi I7 18 daily 4 M Luvale 8 9 Tu-Sa 2 Kaonde 7 8 TuThF 0 Lunda 5 8 Tu-Sa 2

'The big four" have roughly the same amount of airtime, while the three Northwestern Provincelanguages have much less,and do not occur every day, except in newscasts. In termsof measurablehours then there are two tiers - Bemba, Nyanja, Tonga, and Lozion one level,ranked above Luvale, Kaonde, and Lunda. The high valuation and demographicdominance of Bemba and Nyanja over the other five languages are reflectedmore in the allocation of peak times, and other forms of institutional favoritism.The rwo linguafrancas have the most late night slots, numerous weekend programs,and the dailylunchtime slots, all of which attracthigh advertisingvolumeand alsocorrespond with the typical urban laborer's off-work schedule. The most pervasivetheme in both listeners' and broadcastingpersonnel's discourseabout these language valuations is that radio languages represent homogeneousand relatively monolingual speech communities, i.e. ethnic groups. Radio'selaborate partitioning of its universeof addresseesat the most general level, asseparate but equalspeech communities based on language,draws from and feeds intothis essentialist view. Thus, while managementstatements say that "the seven(7) mainvernacular languages on the Media representall seventythree (73) tribes found in Zambia"(Kapeya, 1988: 3), somepeople simply do not feel represented;they want theirpiece of the nationalpie. Many suchcomplaints, as we have alreadyseen, allude to thenational motto "One Zambia, One Nation" in making demands for a linguistic

14 Colurnn1 is basedon Kashoki1978: 39. Other tabulationsare basedon Fisure 2. 350 DebraSpitulnik democracy.While this motto signifiesa vision of equal inclusion,the closelyrelated nationalistslogan "One Man, One Vote" hasbecome a major trope for measuringequal representation,as in this letter to Radio Zambia: 'lf we are all Zambians, then we should all get the same privileges. Kaonde should be on the radio for the same amount of time as the other languages.Equal hours for each tribe." (Mytton, 1978: 217,) Building on the view that one languageis "a true identifier" of one people, a whole series of one-to-one correspondencesor distributionaltropes have thus emerged as assessmentsof ethnolinguisticequity: one language,one people, one culture; equal hours for every language,etc. In this ongoingdebate over the three Northwesternlanguages, egalitarianism is also measuredin terms of anotherkind of one-to-onecorrespondence, that of regional distribution. Some broadcastersand listenerssay for example that the Northwestern Province is unduly favored with three radio languages,while the five remaining languagesare made to cover the eight other provinces.According to this view, some provincesare actuallyshortchanged by havingno unique radio languageof their own, or by having their other languagesgo unrepresented.But Northwesternerssee it differently.To them the criterion of impartialityis not one languageper province,but equal hours per language,as this broadcasterexplained in 1989: "Theproblem is, we have been recognized as a language,so why be mistreated? . . . It remains thatthe big two or thebig four get most of thetime. . . Butyou can't raise these issues because youmay be calleda tribalist." During 1989,ZNBC's concernover suchgrowing sentiments was considerable. Originally Radio t had been divided into five different departments:Bemba Section, I-ozi Section, Nyanja Section, Tonga Section, and Northwestern Section. This bureaucraticdivision in itself clearlymarked the statusof "the big four" as againstthe others,and ZNBC's responsewas to administrativelyrepartition this linguisticuniverse. The Northwestern Section was split into three distinct languageunits (the Kaonde Section,the Lunda Section,and the Luvale Section),and eachwas allocatedthe same number of employeesas the other four sections.An administrativeanalog of linguistic egalitarianismwas thus created:one language,one section.Radio time was not changed however. Other voicesof contestationcome from Zambianswhose first languagesare not any of the seven national languages(Mytton I974, 1978). While national statistics suggestthat suchpeople compriseroughly 40Vo of. the total population,multilingualism is extremely high and in actuality lessthan 5Voof all Zambians do not understand any of the sevennational languages(Kashoki, 1978: 19-21,38-9). Still, periodic demandsfor recognition are common, as this letter from a Nkoya listenerillustrates:

"L-ozi isn't representative of Nkoya, so we should have our own broadcasts. . . . We won't accept a situati

6. Conclusions.

ln conclusion,I would like to briefly consider how extensive and effective Radio Zambiais in negotiatingtwo kinds of linguistic pluralism, and then return to the initial questionsof how to characterize"language ideology." Zambian radio has been faced withthe immensely complex task of reconcilinga nationalphilosophy of egalitarianism witha fairlyrigid linguistichierarchy grounded in the demographicsof its audiencesand the long history of differential language valuation in the country. Since the two pluralismsare fundamentallyopposed and never fully resolvable, as we have seen, 352 DebraSpiulnik apparentinconsistencies in broadcastingpractice can be seizedupon as focal points for contestation,especially by thosewhose languages are excludedor usedonly minimally. Even so, it is to Radio Zambia's credit that it has supported such a high number of languageswithin the economicand physicalconstraints of one broadcastingoperation. The only drawback,it seems,is that in its responseto the dominant interpretationsof pluralism in national politics - where diversity is seento verge on divisivenessand acts of recognizingdifference are potentially "tribalistic" - Zambian broadcasting'sextensive promotion of English has weakened its ability to adequately provide for all of its listeners in the seven Zambian languages.In this sense, pluralism is not really negotiatedat all, it is muted. It is tempting to relate this handlingof linguisticpluralism to the wider issues of national integration and the diffusion of ethnic conflict in the country, but these processesfar outreach the modesof languagevaluation discussed here, and are more dependent on factors such as economic stability and political representation.More narrowly however, Radio Zambia's role and effectiveness in establishing and reproducing languagevalues is extensiveand beginsduring the colonial period, as we have seen. In particular, radio's differential allocation of resources(e.g. channels, airtime, staff, and program types) among the eight radio languagesis very closelylinked to the construction of socialinequalities among the speakersof theselanguages (as well as others who do not feel "represented"by radio languages),and this deservesmore detailed study. The approach here, however,has not been so much to assessthe role of a powerful institution in producing a net linguistic effect, but rather to trace out how radio's negotiation of linguisticpluralism is structured(and understood)in terms of various historical processes,political discourses,and cultural understandings of the relation between languagesand peoples. If these processesare mutually determining and complexlyentangled over a 51 year history, on which side of the equation does one locate "languageideology"? One might call the languagevaluations that emergethrough the dynamics of broadcasting "languageideologies which are embodied in practice", while the political discoursesand cultural understandings(as evidencedin verbal data) might be termed "explicit languageideologies." But basicallythis implicit/explicit dMde is just a shorthand for the different kinds of data used to distill "ideologies." As we have seen. both schedules and talk about schedules establish and construct interconnected languagevaluations. Invariably theseare tied up with relations of power and interest,in terms of both the nationalpolitics of representationand broadcasters' perpetuation of their elite distinctions.It has been suggestedhere that to call all of these valuations, interests,discourse modes, and cultural understandings"language ideologies" conflates several social, cultural, and linguistic processesunder one cover term, and hopefully this case study has sketched out some avenues for investigating their complexity. Linguisticpluralism in Zambia 353 I References

I I Anderson,Benedict (1983) Imagined communities: Reflections on the oigin and spread of nationalism. london: Verso.

Binsbergen,Wim van (1985)"From tribe to ethnicity in Western Tambia:The unit of study as an ideologicalproblem." In W. van Binsbergenand P. Geschiere(eds.), Old modesof productionand capitalistencroachment. [,ondon: KPI, p.18 I -233.

I Collins,James (1988) "Hegemonic practice: Literacy and standard language in publiceducation.' Working ' papersand proceedings of the Centerfor PrychosocialSmdies, no. 21. Chicago:Center for Psychosocial Srudies.

Ferguson,Charles A. (1959)"Diglossia.n Word 15:325-40.

Fishman,Joshua A (1989)Langtage and ethnicityin minoity sociolinguisticperspective. Philadelphia: MultilingualMarrers.

'the Franklin,Harry (1950)Repon on saucepanspeciat': Thepoor ntan's radio for wal populations. Lusaka:Government Printer.

I Friedrich,Paul ( 1989)"L-anguage, ideology, and political economy." Ameican anthropologist9l:. 295-312.

Gal,Susan (1989) "l-anguage and political economy."Annual reviewof anthropologt18:345-67.

Kapyr, Mwansa(1988) "The use of vernacularlanguages in broadcasting."Paper presentedat a broadcastingseminar in Maputo,Mozambique.

Kashoki,Mubanga E. (1978)"The language situation in Zambia."In S. Ohannessianand M. E. Kashoki (eds.),p.9-46.

Kashoki,Mubanga E. andMichael Mann (1978)"A generalsketch of the Bantu languagesof Zambia." ln S.Ohannessian and M. E. Kashoki(eds.), p.47-100.

Kaunda,Kenneth D. (1967)Humanism in Zanzbiaand a gtide to its implementdtion,part L Lusaka: GovernmentPrinter.

l.ehmann,D. A (1978)"l.anguages of the Kafuebasin: Introductory notes.n In S. Ohannessianand M. ' E.Kashoki (eds.), p.101-20.

Mitchell,J. Clyde( 1956)The kalela dance. Rhodes-Livingstone Papers, no, 27 . Manchester:Manchester UnivenityPress.

Molteno,Robert (1974) "Cleavage and conflict in Zambianpolitics: A study in sectionalism."In W. Tordoff(ed.), Politics in Zantbia.Berkeley: University of CaliforniaPress, p.62-106.

I Mytton,Graham (1974) Listening looking and leaming: Report on a national mass media audience survey inZambia(197A-B). Lusaka: Institute for African Studies. Mytton,Graham (1978) "language and the media in Zambia." In S. Ohannessian and M. E. Kashoki ' (eds.),p.207-27. & 354 Debra Spiutnik

Ohannessian,Sirarpi and Mubanga E. Kashoki (eds.) (1978) Langtage in Zambia. [,ondon: InternationalAfrican Institute.

Roberts,Andrew (1976)A hbtory of Zambia.New York: Africana.

Silverstein,Michael (1987) 'Monoglot 'standard'in America."Working papers and proceedingsof the Centerfor PsychosocialSrudies, no. 13. Chicago:Center for PsychosocialStudies.

Silverstein,Michael (1990)'The skin of our teeth:Registers, poetics, and the First Amendment."Paper presentedat the American AnthropologicalAssociation annual meeting, New Orleans.

Woolard, Kathryn A. (1985) "l-anguagevariation and cultural hegemony:Toward an integration of sociolinguisticand socialtheory." Ameican ethnologist12:738-48.

Woolard, Kathryn A. (1989) "Sentencesin the languageprison: The rhetorical structuring of an American languagepolicy debate."Ameican ethnologist16:28-78.