This action might not be possible to undo. Are you sure you want to continue?
, Vol. 22, No. 1 (Feb., 1995), pp. 102-124 Published by: Blackwell Publishing on behalf of the American Anthropological Association Stable URL: http://www.jstor.org/stable/646048 Accessed: 09/04/2010 14:05
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=black. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact firstname.lastname@example.org.
Blackwell Publishing and American Anthropological Association are collaborating with JSTOR to digitize, preserve and extend access to American Ethnologist.
the spoken house: text, act, and object in eastern Indonesia
WEBB KEANE-University of Pennsylvania
Although it is part of the everyday work of anthropologists to share in local conversations about cultural things, they may listen to such reflexive discourse with some ambivalence. After all, much of what we know about our world remains tacit-it can go unsaid and often is difficult to put into words. Yet amidst the everyday flow of work and talk, moments occur in which people reflect on and even objectify their world. In particular, when they talk about their own culture, people often privilege descriptions and interpretations that have an almost textlike quality, as if simply awaiting the writer's pen. For Malinowski, such moments provide an unproblematic boon: of data of folk-lore,are not less valuableas ethnographic documents,and as illustrations the natives' of this attitude towardscustom.Incidently, feature nativemythologyshows thatthe taskof servingas is as informant not so foreignand difficult mightat firstappear.He is quite used to recite ethnographic in and one afterthe otherthevarious proceedings hisown narratives, he does itwith stagesof customary these qualities and an almostpedanticaccuracyand completeness, it is an easy taskforhimto transfer to the accountswhichhe is calleduponto makein the serviceof ethnography. 922):318] [1961(1 Indeed, the virtuosity in self-description that Malinowski describes seems to characterize entire societies. Barth (1990), for example, contrasts the high value Balinese place on decontextual explanation with the resistance to verbalization posed by the more taciturn Baktaman. Many observers have been struck by the way that regional patterns seem to have given rise to the "highly word-oriented approach to cultural analysis [that]has been advanced by ethnographers of Island Southeast Asia ... where ritualdiscourse and oratory abound and people engage one another in speculative exegesis, with or without the prodding of anthropologists" (Atkinson 1989:332, n.2). By this account, it is distinctive local ways of talking that shape the resulting ethnography. Sumbanese descriptions of the "traditionalhouse" as a microcosm and emblem of loca I identity are neither unproblematic expressions of a cultural totalitynorsimply objectifications imposed by ethnography or modernity. One way the house is able to serve as a discursive object reflects, in part, specifically Sumbanese models of action and beliefs about language as refracted in changing historical circumstances. In ritual,speakers seek to engage and elicit responses frompowerfulothers, whereas current religious and political developments reframe ritual words as means of describing a cultural world. Both sets of practices draw on the authority of "entextualized" language but interpret it in different ways. Emerging representations of cultural meaning are shaped by long-standing speech genres and by recent social and cultural transformations,mediated by shifting language ideologies. [culture theory, representation, discourse, ritual speech, language ideology, house, Indonesia]
American Ethnologist22(1):102-124. Copyright ? 1995, American Anthropological Association.
[T]he frequent, tedious repetitions and enumerations of customary sequences of events, interesting as
to On the otherhand, skepticsare reluctant take for grantedthe reflexiveease, schematic thatthese writersreport.Boas (1966) sounds an early distance, and verbaladroitness note of epistemologicalcaution, admonishingthe anthropologist againstthe operationsof rationalization." Morerecently,Asadassertsthat"[d]iscourse involvedin practice "secondary is not the same as that involved in speakingabout practice" (1983:243)-words must not be conflatedwith theirpurported objects.ForBourdieu(1977),talkabout practice,rules, and meanings tendsto arisein speech directedto outsiders. Suchtalkcan obscureor discredit tacit knowledge,helpingproducea pictureof cultureas a totalizable objectthat is cut offfrom the action. Historicizing epistemological Foucault (1972:44)observesthatthe question, very existence of somethingas an object of discourseonly arises as a functionof specific conditions,for "onecannot speak of anythingat any time."When the object of talk is "local culture,"add criticsof colonial knowledge,it may be in orderto serve externalpowersthat seek a positionfrom which to see and yet not be seen, so that local worlds may become and "picture-like legible... readablelikea book"(Mitchell1988:33;see Heidegger1977). Thussituated, even localwaysof talking aboutculture, "tedious especiallythe moreformulaic and enumerations," mustbe questioned:if somethinglooks like a book,we may be repetitions promptedto ask who the impliedreaderis. Nonetheless,I arguethatthe skepticsmust still "almost Suchexpansiveself-consciousness, contendwithMalinowski's interlocutor. pedantic" abundant in some places and rare in others, should neither be taken for granted as a straightforward expressionof autonomousculturalmeaningsnor dismissedout of hand as of the product "officializing reifica(Bourdieu 1977)or ethnographic merely strategies" tion. Thatreflexivediscoursecan providepowerfulinterpretive insightsor deceptive normative Boththe acceptingand the skepticalview, however, if taken as fully claims is indisputable. sufficientaccountsof the discourse in question,tend to reduce it to the service of a single function,usuallythatof talkingaboutthe world. But languageserves multiplefunctions,and even in face-to-faceinteraction liableto exceed any given context,purpose,and intention.' is Reflexive can drawon verbalresources are shapednot only by, say, an inherentneed talk that to interpret one's worldor to satisfyan ethnographer by otheraspectsof speech altogether. but The relationship betweendiscourseand action is not given a priori. extentto which talkis The "in" "about" or practiceis, in part,a functionof specificlanguageideologies-that is, speakers' assumptionsabout "the natureof language in the world" (Rumsey1990:346).2 Language in withsocial institutions, modesof action,andbeliefsaboutthe world, ideologiesarticulate turn all of which aresubjectto historical In transformation. thisarticleI lookat the shifting statusof talkas peoplereflect whatthey knowanddo, andsuggestthatthereis morethanone position on fromwhich to view a "picture-like" world.3 InAnakalang, exemplary one focus of reflexivetalkis the "traditional house,"which people commonlytreatas an interpretive to local cultureand identity.My interesthere is not in key the for reinterpreting house itselfbut ratherin seekingto accountfor its readyavailability talk aboutcultural as meanings.I do so by lookingat how it is constructed a verbalobject in ritual. I the Although do notclaimthatthis fullyexplains ways peopletalkaboutthe house, Iwill argue thatsuchtalkis influenced the demandsimposedby ritual action,whichareshapedby social by institutions cosmologicalbeliefsandby the semioticproblems and theypose. When institutions and cosmologieschange, so do the practices,problems,and languageideologiesthey entail. Atthe same time,as people drawon availablesymbolicresources, such as the wordsof ritual, them. InAnakalang, describeone resulting I they often recontextualize tendency,that many treat verbalizedhouseas a cultural the thatexistsindependently speakers of people easily entity and contexts, potentiallyable to totalize an entire social universewhose essence can be in summarized so manywords. Such representations work in partby appealto the authority and apparent concretenessof both the words and the thingsto which they refer.They select
the spoken house
and take advantageof alreadyexistingfeaturesof ritualspeech that make it susceptibleto treatment a sourceof stabletexts. as The resulting accountsof the traditional houseand itsmeanings,I suggest,can be misleading in two respects.Becausethey drawon genresof speech thatAnakalangese associatewith the that past,they can appearto be timelessartifacts expressa fullyreadableculturethatlies apart fromtheirown actions.Conversely, because representations the house have manyof the of schematizingand abstractfeaturesof talk addressedto outsiders,a skepticalanthropologist traditions" and Ranger (Hobsbawm mightassumethatthey are nothingmorethan "invented in 1983) or the effectof Westernor academicsortsof knowledge.To pose the matter termsof such simple alternatives threatensto restrictour view of people's verbal resourcesto an narrowrangeof possibilities. hintingat the richvarietyof ways thatpeople unrealistically By can formobjectsof discourse,I hopeto shed lighton the sourcesand development one sort of of culturalrepresentation its role in ongoing local effortsat self-definition. and
house structure as a paradigm of cultural order A common object of culturalrepresentation easternIndonesiais the house, renderedin in detailed homologies among architecture, society, and cosmos (Barnes1974; Cunningham when 1964; Ellen1986; Forth 1991; Kana1980; cf. Fox 1993; Waterson1990). Nonetheless, I startedresearchin Anakalang,an ethnolinguistic domain on the island of Sumba,I was at Thisbeganduring surprised how quicklyI foundmyselfthe recipientof similaraccounts.4 of was witha remarkable my firstweek on the island,when I, as a "student culture," presented culture"in Indonesian 1985). The nine-pagetypescriptaccountof "Sumbanese (Wohangara authorwas a Christian a retired and civil servant who lived in the largest town on the island-a market administrative and centerand home to a mix of ethnicitiesuncharacteristic mostof of Sumba, all of which may help explain the cosmopolitannatureof the document.Clearly of covers manyof the embodyingthe seriousreflections a speculativeintellect,the typescript of such topics favoredby otherIndonesian-language synoptictreatments cultureandtradition, treksand rulesformarriage for burial.5 as the ancestral and One striking feature thisworkis of the prominenceit gives to the house, to which, afteran initialhalf page definingthe word the 4 turns:"[The] centralpillars(kambaniru-lundungu) [are] marapu(ancestor spirit), author namedaccordingto theirrespective functionsas the FourPrinciples the Sumbanese of people TURAPARAINGU) form a basis for orderingthe way of life of the that (PATUKALARATU life."6 Sumbanesepeople throughout These pillars are identifiedwith "Godliness,""Marriage" (prescriptive marriagerules), and and of withthe fourcompasspoints,andwiththe (rites passage), "Prosperity," "Life Death" to entranceways the ancestorvillage. Here in a nutshellare the basic treatments: partsof the linkedto cultural houseare metonymically (see note 16), andthe whole is a diagram principles and world.Through such alignments authorportrays houseas the the of villagestructure larger and a schematicstructure microcosm,both a practicalgroundand conceptualoutlinefor an order. entiremoraland cultural of of thistreatment the house could be multiplied.In manycasualconversations, Examples told me aboutthe meaningof the house in preciselythe sortof "speculative exegesis" people describes.Thesimplest versiondividesthe distinctive thatAtkinson bamboo,wood, andthatch to see house(uma; Figure intothreelevels,corresponding the worldsof spirits atticunder (the 1) raisedon roofor "house'shump"[wogu uma]),humans(the livingplatform the high-peaked and underside" andanimals(corral stockpenbeneaththe floor,the "house's [lubuuma]). posts), Finerdiscriminations producefive levels-which, some tell me, parallelthe five-pointstate to (Pancasila--or more, in some cases corresponding the numberof clan ancestors, ideology
kadu uma (house horns)
wogu uma (house peak)
halikuru kalihungu (rafters;21)
1 /~l O
kabisu uma (house corner; 4-5) lubu uma (house underside; 1-2)
1. 2. 3. 4. 5. 6. 7.
tolaku maneni (peak frame; 17-18) kerijialu (water jar base) leli (pillar disk; 23-24) walu toku (pillar carvings; 26-29) kabaringu padua (central pillars) karabuku (hearth) korungu marapu (spirit chamber; 12-15)
8. 9. 10. 11. 12. 13. 14. 15.
kabaringu padaku (short pillars; 19) halema (bench) pena bakul (main room) kabaringu uratu (divination pillar) tapu papalekaru (meeting mat; 9-10 ) hedi marapu (spirit shelf) panuangu (steps) pidu (door)
in refer linesof UmbuSebu'srecitation.) to 1. Figure Peakedhouse. (Numbers parentheses
with additionalhomologiessuch as the partsof the humanbody (cf. Barraud 1979:57; Forth 1981:29-30, 32). at ThatAnakalangese, least in speakingwith outsiders,may take this model as a key to an entirecultureis suggestedby two anecdoteswith which, as a searcheraftersuch knowledge The myself,I was often entertained(and,perhaps,admonished). firstconcernsa Sumbanese who, people say, was such a genius that he was sent to Javato study. There he was boy who asked,"Whatdoes your fatherdo?""He'sa herdsman." interviewedby the authorities, "Maizeandtubers." "What you eat?" do Basicmatters production consumption of and disposed the do "In the clincherwashisabilityto schematize house:"Where you live?" a three-leveled of, and the A house:thetopcontainsourfood,the middlethe humans, underneath, cattle."7 second in anecdoteconcernsa Dutchstudentwho had undertaken fieldwork Anakalang severalyears before I did, but who had never completedhis dissertation. One versionof his fate is that he his exam back in Holland,being unableto describethe house levels when askedto do failed so by his professors.
the spoken house
that Thehouse,as a structure embodiespubliclyknown,wide-ranging categories,formsone referent cultural for discoursein Sumba.Liketalkaboutgong tunes,stonetombs, paradigmatic marriageexchange, and stopping places on the ancestraltrek, this discourse often takes In schematicformslikethe list makingdescribedby Malinowski. focusingon speech, I do not meanto discountthe importance complexityof the house as a social concept, as a ritual and space, or as a criticalsite at which these two aspects articulatewith everydaypractice-a different longerproject(see discussionsinAdams1974; Forth and 1990). Rather, 1991; Kuipers my interestis provokedby the Sumbanesefacilitywith talkaboutcultural thingsand its links to speech practicesand languageideologies. Forall societies producedwellings, doubtless ladenwithmeaning-but noteveryonemakestalkaboutthe house as centraland as schematic For as do Anakalangese. example,Bourdieu's 979 )accountof the Kabyle housedraws (1 not on explicit verbalmappingbut on activities,collatedwith proverbs expressionsthat and are into speakers notthemselvesgivento assembling a synopticaccount(cf. Oliver1987:153). discourseis selective: it does not captureeverythingof interestor importance Furthermore, aboutthe house.8 Thisis to suggestthatthe treatment the house as an object of discourseis not simplyan of of its referent arises in partfromits place within speech practices.Indeed,to but expression focus exclusivelyon the referent be misleading: house and its meaningsmay not be can the sufficientto determinethe way they are spoken of. And conversely, I argue, even a fully "traditional" architectural structure not sufficient provideone with a cosmologicalhouse, is to in the absence of the ritualpractices,includingspeech. Thus,currentaccountsof the house that that builduponfoundations are laid in speech performances verballylay out the house as a formalstructure, and displayedpartby part.InAnakalang, these practicesare decomposed and As these challengesbecome less certaincommunicative pragmatic challenges. shapedby and and (withthe decline,forexample,of clan identity, spiritritual, the interlocutors important involve), the speech practices increasinglylend themselves to treatmentas agencies they of sourcesof textsthatreferto objects independent actions. ritual speech as cultural text witheverydayspeech. It Sumbaneseritualspeech is a highlymarkedregister contrasts that verbalmediumof ancestralrites(forexample,in prayer, and is the principal oratory, song)and of important interactionsamong humans, notablythe negotiationof marriageexchanges. ritualspeech, "patterned wordsof negotiation, alignedwordsof speech"(patali Anakalangese lola li panewi),comprisesa largebutfixed set of canonicalcouplets in semanticand kajiala, unitsand convenfor alongwith principles theircombinationinto larger syntacticparallelism, and tionsfortheirappropriate efficacioususe, all of which aresaidto havebeen createdby the often treatthe As ancestors.9 in many partsof Indonesia,people in contemporary Anakalang in artifacts theirown rightandas descriptors bothas treasured cultural of ritual speech couplets culture.Thisdouble role is implicitin the statement the Sumbanese of traditional compiler by of of a recentlexicon of ritualcoupletsthat his book could be called a "Dictionary Culture" 1987:9).10 (Kamus Kebudayaan) (Kapita ritualspeech addressesan other, 1981) of Anakalangese Fullperformance (Hymes humanor spirit,across some sort of social or ontologicaldifference.This differencecan be or termsas the distancepermitting defined,forexample,in kinship marriage, in politicalterms ancestral betweenfeudinglineagesegments, whenthe livingaddress asthe alienation or, spirits, betweenthe livingand the dead. The gap between interlocutors as the ontologicalseparation to efforts overcomethem.Two that specialcommunicative yet posesdifficulties require, permit, the of characteristics thisgap mark ritual speechthataimsto bridgeit. One is the reflexivitythat makesritualspeech a way of namingthe things,agents,and types of action involvedin the
and of actualevent in which it occurs.Theotheris the formality performance, poetic structure, canonical language that providesthis speech with aesthetic power, textlike qualities, and ancestral character." and arisesfromthe needfortalkaboutsocial identities cultural characteristic Thefirst objects. each otheracrosscultures,opens up within Thisneed, often apparentwhen speakersaddress and society-even in face-to-faceinteraction-at pointsof criticaldifference distance,motivatthe abstraction that characterizes discourse directed at "outsiders" (Bourdieu ing the ritualspeech oftenmediatesbetweengroupswho, through 1977:17).InAnakalang, formsof discoursethey use, definethemselvesas outsidersto one anotherand treattheir very of mutualcommunicationas uncertain.Theirspeech thus has many of the properties talk and constant with communicative betweenstrangers, uncertainty throughredundancy dealing to reference the relevantactors,events, and goals. ritualspeech help speakersidentifythemselvesto In addition,the formsof Anakalangese and Eventstructure performance ancestorsand claim some of theirauthority.12 style all help the situatedness speech acts of of mutethe particularities time and person,partially denying (Keane1991; Kuipers1990). These formswork in tandem with ritualpoetics that support of "entextualization," makingsalientformalfeatures the speech thatallow itto be treatedas an In or circumstances speakers.13 doing so, it stresses thatappearsto transcendparticular object of thoseproperties languagethatare leastdialogic,linkedto context,or to speakers'intentions. herebecauseit allows highlycontextualspeech events is interest Entextualization of particular to to createthe effectof texts,buildingon those aspectsof languagethatcontribute what have been treated,for quite differenttheoreticalends, as the "said"(Ricoeur1971), the "already 1973 ).Itthus that uttered" (Bakhtin 1981), or the writing "supplements" speech (Derrida articulatesin practice aspects of language often considered analyticallyto be in mutual opposition. underwritethe use of ritual speech as a source of cultural These two characteristics The reflexivedimensionof ritual speech providesit withformssuitablefortalk representation. and itsformality itstreatment a textthatexistsindependently as aboutcultural objects, supports events, persons, contexts, or practices.These of, and is highly portableamong, particular that however,derivefromfunctions arequitedistinctfromtheiruse as denotational properties, use ritualspeech in a rangeof contextsand purposesthat texts.Contemporary Anakalangese sets aboutthe nature languageand its effects.The different of bearcontrasting of implications differencesin languageideology,emphasizing roleof language the uses of ritual speech imply in problematic consequentialinteraction the one hand,and its functionin denotation and on on the other. Althoughin practicethe two aspects are often intertwined, they may pull in ritualspeech thatprovides way of speakingto powerfulinterlocutors a and different directions: can also be readas text abouta separateand self-contained world. inducingresponses associatedwith relatively Thesetwo emphasescan be seen in speech practicesrespectively and modernist of Anakalangese bothkinds conservative self-consciously parts society.Although evidence of historical of practicecoexisted in the 1980s, they mayprovide shifts(forwhichthe writtenrecord is scanty) in the languageideologythat informsthem. Forancestralritualis while Christianity, involvement withthe nation-state, and becomingincreasingly marginalized, and reflection the formsof cultural associated withthem,areclearlyascendant. issue At display herearespeech practices,since the actualwordsin questionneed notdiffer: the rather, crucial differences concernthe contextsin which they occur,theirpresuppositions, theirimplicaand tions. the To exemplifythe contrast, followingsectionpresents briefcanonicaltextdepictingthe a house."Itwas offeredto me, as a studentof culture,in the descriptivemode (and "traditional as a display knowledge). source,however,liesin ritual of Its addressed invisible to performances these two discursivepracticespresuppose distinctsocial fields of action, we spirits. Although
the spoken house
cannotunderstand former the withouttakingintoaccountthe social andcosmologicalconditions that shape the words in the latter.Theirtextualcharacteris in parta functionof the with distantinterlocutors, such as the dead, and of perceivedchallengesof communicating fromthemrecognition responses.14 character ritual and The of wordsinturnfacilitates obtaining their use in new contexts,supplyingan emergentfolkloricdiscoursewith materials that are of for primarily interest whattheydenote,or as a code awaiting exegesis. Itshouldbe no surprise to findthatthosespeechperformances mostlendthemselves entextualization favored that to are as lastingcultural objects.
the house in words
I spentmanyafternoons with an elderlyritual sitting specialist,whom Iwill call UmbuSebu, An conversations. memberof a powerfuland ambivalently engaged in rathermeandering modernist but his family,he had recentlyconvertedto Christianity, he enjoyeddisplaying rich historical memoryand facilitywith poetic speech. Likemanypeople, he could easily launch into longrecitations, trek. itemizing exchangevaluablesor the stoppingplaceson the ancestral Once, for example, he asked whether I knew about the house, and presentedit to me as
yili pakowa yili pataukaru pakabuangu kamumu pabongu wai baha deta ta kabisu uma katikupenang tomaja ta tapu papalekaru yeka da nula pakahorungu liya ta koru mamutu na bola mamutu katikuoli nula karasaoli tapu lisa da tolaku maneni aharu lagapa lisaka nau deta ta kabaringu padaku halikuru kalihungu purungu ta leli mangu isi ta kajanga ma wua liya na walu toku na walu lara na dakut nibu na koba wai dug out site excavated site discarded remnants thrown out bath water goes up to house corner head of the main floor arrives at unrolled mat and the offered pillow stops by there at warm chamber the warm basket head's companion pillow flank's companion mat stops by the attic uprights crossbeam of lagapa stop by up there at short pillars encircling rafters descend to the full disk to the fruited branch stops by there the eight pokings the eight flutings the sharp spear the water cup 25 20 15 10 5
As Umbu Sebu spoke, he gestured,as if to map out a series of points leadingthe spirit addresseeupwardand intothe house fromthe villageplaza (although house in which we the and The sat lackedthe "traditional" peak, altars,centralhearth, pillars). firstcouplet, referring to the foundation holes in which the house pillarsareplaced,denotesthe housesite withinthe of that village,a permanent possessionof the clan and independent any physicalstructure may stand there. The succeeding couplets name the pen housing pigs and horses momentarily beneaththe livingplatform (lines 3-4); the frontcorner(lines 6-7); the benches in the most indicated (lines9-10); publicroomof the house, metonymically throughimagesof hospitality the interior space in which the seniorcouple sleep (lines12-15); the roofbeams(lines 17-18); the postsand beamsthatreinforce centralstructure the shelves (lines20-21); the disk-shaped attachedto the fourcentralpillars(lines23-24); and the pillarcarvingsalong which commuwiththe spiritsare saidto travel(lines26-29). Bythe time of thisconversation,I was nications no longersurprised such displaysof schematicknowledge, people hadstrongideasabout at for what I was-and should be-seeking. But was this nothing more than a product of the I are encounter? will arguethatmatters not thatsimple. ethnographic ways of looking at a house Textssuch as the precedingone do not come from nowhere-nor, if summonedup and refittedfor my benefit, were they merely invented for "the service of ethnography," in Malinowski's phrase.Neithereverysociety noreveryspeakerreadilyprovidessuch texts-nor do they providetexts on any possible topic. If the partsof the house are common, everyday knowledge,what social conditions favor their codificationas schematic objects of verbal Neitherexplicitpedagogy(largely communication? absentin traditional nor Anakalang) interThis providethe only occasionsforformulaic lopers(suchas ethnographers) self-presentation. to was coupletsquoted manytimes and had fragment familiar me, as I had heardindividual recordedlongerversions,often extendingthe enumerations partsof the village, in several to who come to negotiatemarriage kindsof ritual events. Visitors exchangesare verballyled up intothe houseto the matson whichthey will sit andfacetheirhosts-even when the encounter shelter.When mortuary takesplace in a temporary directthe deceasedto the villageof prayers the dead, these words lead them into houses that lie along the way. Other rites lay out the in progression an effortto drawthe good fortuneof the deceased back intothe house to bless After the bereaved.16 a burned a enumeration pointssummons of villagehasbeen rebuilt, similar the spiritsback fromthe forestswhere they have fled. The orderof the partscan be reversed. sits Whena priest(ratu) on the floor,at the base of the divination pillar,with offeringsin front of him(speaking arenormally the attention the spirits the atticmustbe directed of in men), priests downward receivethem. Ritual to (heku)afterincestexpels the wrong stepby step purification 1988:106). by a further listingof parts(cf. Kuipers The currentprominenceof the house is doubtlessdue in partto its nationwiderhetorical for differences exemplary an tokenof safeethnic importance, the statehasfoundin architectural difference.This is most evident in TamanMini Indonesian, nationaltheme parkoutside the whereeach provinceis represented an oversized"traditional" house (see Anderson Jakarta, by and 1990). As with well-known instancesin Sumatra Sulawesi,the house providesa visualemblemof cultural for the on distinctiveness; instance, firstlocalpublications Sumbanese a culturefeature pictureof the house on theircovers(Kapita 1976a, 1976b, the latter devoting to two chapters listingitsparts), the shoulderpatchesof some local officialsin EastSumba and a portray roofpeak. But in Sumbathis contemporary role intersectswith and takes advantageof preexisting In significances. additionto the threelevels notedabove,the house'sinterior spaces,physically indicated benches, or mats,are in practiceassociatedwith distinctions only by low partitions,
the spoken house
of genderand clan affiliation. veranda(Iaga) shouldface the centralpartof the village, The locationof the ceremonialplazasand the tombsthatserve as foci of clan identity. Thisis the outermost of the houseandthefreestof access, forentrance the doorrequires part through front invitation recognizedpurpose. placeforcasual interaction, is where peopleexchange and A it theirfirstbetelchew beforefurther moreformaltalk,andwatchthe everydayandceremonial or events in the plaza;when largepartiesstay duringmarriage and negotiations funerals,it can also serveas a sleepingplace. Behind above it is a low wall on which manyhousesdisplay and the buffalohornsremaining fromformer sacrificesor feasts.To the right, one facesthiswall, as up a steportwo is a low doorintothe mainroom(penabakul),the mainindoorstageforformal events.17In such events, the visitorssit along the outer benches facing inward,while those associatedwith the house sit across from them, their backs to the heartharea. This inner elements of the structure, four centralpillars(kabaringu quadrantis framedby the sturdiest and padua).Divination(uratu) prayers (nyaba)usuallytake place at the foot of the divination which isclosesttothe door.18 of uratu), pillar(kabaringu Onlyintimates the inhabitants normally inward. venturefurther containsthree There,at the centerof the house, the hearth(karaluku) stones(tularu), identified maleandtwo as female,of whichonlythe latter one as maybe moved when adjusting cookingpot. The hearthrestsin a slightlyrecessedsand box andformsboth a the and literally figuratively warmcenterof the household,a comfortable placeto loungewhile chattingand pokingat the fire.Aroundit sleep the youngermembersof the household,while the couple who head the householdtake a separatechamber(korung), which may also hold inalienable the "ancestors' valuables, portion" (tagumarapu). Overnight guestsremainon the Diametrical acrossfromthe front benchesof the frontroomortheveranda. dooris an informal ly entranceand the "baseof the waterjar" the areareserved the washingand food for (kerijialu), activities women and young men. of preparation In fact, however,only some existinghouses actuallyapproachthe canonicaldescription. of numbers people live in cementor stonehouseswithcorrugated roofs. zinc Today,increasing houses but in simpler Moreover,even in the past many people lived not in "traditional" the even thecentralpillars elevatedplatform.19 or dwellings,lacking peakedroofandsometimes Whethera house took the fullycanonicalformwas a functionof the wealth and authority of the owners, of its ritualstatuswithinthe clan, and of whetherit was located in an ancestral or village (paraingu) in a gardenhamlet(kalebugalu).20 Evenin the ancestralvillage, houses often fall shortof the ideal-in manycases, all that is to be found is a vacanthouse site (yili of uma).Conceptually, however,these sites, which remainthe inalienableproperty the clan, "traditional" structure is with houses,and in practice,a full-fledged continueto be identified natureof ritual,a temporary shelter for not always required the rites.Giventhe performative and conditionsare met.Conversely, can serve, if the properofferings otherperformance many zinc otherwise"modern" houses,churches,and government buildingsare given high-peaked betweentheseextremeslie villagesof basically"traditional" roofs.Somewhere peakedhouses where the hearthhas been removedto a cook shed and the altarslie in disuse because the ownershave convertedto Christianity. social action and semiotic difficulty InAnakalang, in other"housesocieties"(Levi-Strauss as 1982; Macdonald1987),the heavy for to attached the physicalhouse is yet further loadof significance overdetermined, it is closely of withthe sociological"house" identified (uma),the basic constituent the clan. InAnakalang, in and house andclan identity,membership, possessionsmustbe continuallynourished rituals to Theserituals forancestral togetherdispersedmembers the villageswheretheir bring spirits.21 as with the ancestors active theiridentification houses and house sites are locatedto manifest collectivities.Ritualspeech is pivotalto politicaland affinalties among the livingas well as
with the spiritworld,the latterstillconsideredby manypeople to be crucialto their relations of the social and politicalefficacyand standing the group. the The livingperiodically gatherto face theirancestors,speaking through mediumof ritual to they speakers.Bymeansof speech performances, seek to compel the spirits recognizetheir ancestrallygroundedsocial identity and to acknowledge the obligations between them, to or debts.Theyattempt inducethe ancestors grantthemhealth to fulfilling, renewing creating, andsafety,fertility, good crops,andsuccess inexchange.Clansmayvarygreatlyintheirsuccess in stagingrituals, whichcan be logistically politically and complexandexpensive,anddemand the cooperationof often rivalrous affines,and skillful-and at times highlyrecalcitrantkin, bothmanifest compoundthe strengthening weakeningstate and ritualists. Successandfailure or theirsocial honor,and theireconomic and physicalwell-being. of theirperceivedcharisma, and People knowthatgood ritesdemandstrategy economic calculation,but they oftensee these as manifesting to Effective sources,foundinthe qualityof theirrelations ancestors. deeper with is communication spirits neverassured.Forone thing,the verypresenceofinvisiblebeings it that is notcertain,and,once they are in attendance, is notguaranteed theywill recognizethe astheirproper descendants. Evenifallthatisgiven,ritual actionsareinherently difficu It speakers characterare the semantic and their outcomes uncertain.Threeaspects of this problematic that and and ambiguities opacitiesof ritualspeech, the restrictions constrainits performance, the social constructionof the interactionitself. Ambiguity,allusiveness,and opacity are All to that of commonfeatures ritual speech andcontribute the generalperception it is difficult. fromthe ancestors are supposedto have been transmitted house-basedlinks couplets through to the present.Manyare semanticallyobscure because of esoteric vocabularyand allusive the reference" (see metaphors, latterknown in east Sumbaas "disguised (hangindingu ngara) Adams1974:331;cf. Forth1988:135), and these featuresare oftenexacerbatedby truncated in of can The and syntax. specificimport anexpression anygiveninstance be ambiguous, people of are veryconcernedwith the possibilities misconstrued intentions. that constrainactual performance, These possibilitiesare underlinedby the restrictions risk of reminding participants the ever-present of failureand of the wrathof offendedaddresses. concernthe actualchoice of couplets,forconventionsrestrict Some restrictions some couplets to specific types of event, lest one "wronglylifta sharpenedstake,wronglytake up a stone" kahala dekiwatu) therebyinvitesanctions. and tebahoraku, Performance demands (kahala style limitwho may speak and where. Most highlyformalbody postureand prosody.Restrictions to communications ancestors shouldoccuronly inthe ritual "cool" of important (maringu) night. When people face the spirits,minoroffenses,such as forgetting mentiona certainancestor to or omittinga stage in a sequence of prayers,may draw seriousconsequences,such as theft, disease, accidents, conflagrations,infertility, drought, and, if certain clans are involved, strikes.22 Indeed,priestshave been knownto trembleor faintfromanxiety.One ritual lightning I witnessedcame to a haltbecause a priestlost his nerveand refusedto perform. sudden The in deathof anotherpriest's to he daughter 1993 is commonlyattributed mistakes had made in ritualspeech only a few days before. The formaland performative characteristics ritualspeech in Anakalang of take ultimately shape in referenceto the social actions that they mediate.Whetherused by contemporary or ritualspeech marapuritualists betweengroupsof livingpeople, as in marriage negotiation, is preeminently wayof addressing listener a a acrossa significant power-laden and socialdivide. The kind of ritualdialogue in which formalspeech occurs presupposes-and helps make The between participants imaginable-differenceand a distinctsortof interlocutor. separation in the speech event is represented physical.Thus,for instance,to prayto the spiritsis to be as "faceto face"(pahagangu) them, requiring with directthe spirits passageacrossa gap. Prayers to "descend" to (or, (purung) the offerings for largesacrifices,to tokens,which sit in frontof the priest).In other situations,words are directedupwardsand outwards.Forexample, oratory
the spoken house
by (taunguii)in the villageplaza mustbe conveyedto the spirits a singerwhose voice must"go In (li alongthe gong rack,go alongthe drumbridge" ta ladatala,li ta lediuedu).23 eithercase, hardto achieve, and discourseis achieved acrossa space thatis felt to makecommunication which it is the functionof the elaboratelistingof places to define. One is never sure thatthe spiritsare present,that they are listening,and that they recognize their descendants.The to invisiblelistenersmustattendand be prepared engage: the whenI awaken sleeper tamangadu baku bangu letthemnotbe startled halawada abudakabataku the whenIarouse dozers kokida majiuda ta baku letthemnotmishear abudakadadangu rajiaku rangu what is being askedof them: Theymustalso understand
abu mu urungbamurangu don'thearpoorly
elu bamu abumuyabaru
don't unclearly see
to Having summonedthe spirits,the priestmightthen call out, referring both spatial and nutu?) Are (Na!Jara "There! you horsesnow complete?" ganakukadimika temporalpresence: whetherthey have arrived to are accompaniedby carefulattention any sign indicating Prayers in the appointedplace and have heardwhat was said, althoughthe clinchingevidence may to are occuronly long after,when subsequentmisfortunes attributed retrospectively pastritual
to crucialforthe discursiverelationship culture. encountersthusexhibittwo features Ritual One is thatthey requirethatpeople speakto othersacrossa sociological,physical,and even and is ontologicaldivide.The second featureis thatthiscommunication displayedas difficult is of Thiscombination semioticandpragmatic uncertainty evoked in the way subjectto failure. for one man interpreted me the couplet, "Soundof the hawk, rustleof the duck"(li hangula for The kuala,li hamowiradi). hawkis too highoverhead usto see, butit alwaysgivesa warning beforesnatchinga chick. Inthe darkof the night,we can hearthe wings of otherwiseinvisible to duckspassingoverhead.This, he explained,refers the obscurityof the knowledgethatthe livinghave receivedfromthe ancestors.What littlewe know of the correctrites involvesus with powersthatwe sense are there, but that remainelusive. Althoughthey are out of sight, they bearthe threatof dangeras well. On the one hand,this is clearlya way of talkingabout betweenthe livingandthe dead, and the crypticnatureof the media relations the problematic that linkthem.Butit mayalso be one way of acknowledging interaction that escapes ultimately of intentions any given participant. the speaking to ancestors like ancestors
The words that Umbu Sebu uses to denote a culturalartifacthave their source in actions Theirpurpose,when used, forexample,to summona directedto ends otherthandescription. intothe house, explainsthe orderin which the partsare named,fromoutsideto inside, spirit the bottomto top, iconically reproducing relevantaction. The role of predicationhere is a of function the task,of the sense thatitcouldfail,andof the delicacyof the etiquetteof directing a deceasedelder.Thedesiredmovementsof the spiritarespelledout in step-by-step detail,for about the action being undertakencan be a minimal degree of shared information only presupposed,and no agreementon intentionscan be taken for granted.Only the one-line directives(lines5, 8, 11, 16, 19, 22, 25), such as "goesup to," indicatethe highlyperformative do these directives notformcouplets,and so arenotembodied of nature thiscitation.Strikingly, in fully canonical form. Thus, when people cite couplets in conversation,quote them in textsto the ethnographer, they oftenomit subsequentritualspeech, or offerthem as cultural the unpairedlines-here, the directives-as well as other contextualizingmarkers (Kuipers
1990:60). That is, they foreground as more essential those parts of the text that are least contextual. A significant verbal difference between Umbu Sebu's performance and that which addresses spirits is that the former omits the words of the ritual respondent who sits facing the speaker. A characteristic of full Anakalangese ritual speech performance, the respondent's role structures the performance as a stylized dialogue. Every few couplets, the speaker cries out the ritual address name (kadehang) of this house, to which a second person replies with the cry "Go ahead!" (Malo!). This omission in the cited text is part of a general deemphasizing of the interactive-and problematic-character of ritual speech. It also affects its practical authority and efficacy. When the living address the dead, they speak as their legitimate descendants. Knowledge of couplets provides indexical evidence of the speakers' relationship to ancestral powers, something also denoted by the kadehang, the ritualavoidance names of the ancestors. These structure speech as a dialogue in which verbal recognition of the speaker in terms of the ancestor is continually asserted, and, with it, the legitimacy of their claims on the ancestors. In addition, the gap across which speech is transmitted intervenes not only between groups but also within them. Forexample, a preliminaryspeaker makes an offeringspeech (palaikungu) to identify the offerings, their recipients, and their beneficiaries for the priest who will then reformulate this information in prayers (for examples from other genres, see Hoskins 1988; Keane 1991). In most cases given as instructions to intermediaries ("Youwill say 'such-andsuch' "), they situate the moment of speaking amid both of a chain of speakers and a temporal sequence of interdependent speech events. Given the delegated structure of performances, principals cannot assume that even their own speakers know what is going on-any relevant actions, agents, instruments,and goals must be put into words.
words and consequences
The social structure of speech events helps explain why much of the referential content of ritual speech consists of metalanguage that announces the purpose, context, and participants in the event. Prayer, for example, insists on its purposefulness and indicates the spirit couple that it addresses, invisible but still forming a distinct spatially localized other to be faced: in orderthathe hearwell kanapekukolungukanapekurangu ubuna ma rarana matana red-eyedlord that'swhy I put it beforethe horse jiayadukutauyata hagajara in orderthatshe hearfeelingly kanapekupadangkanaita i rabuna ma miangna kuruna blushingchested lady that'swhy I put it at dog'ssnout jiayadukutauyata oraahu These lines refer to the message, but the offerings must also be specified, and the ancestral warrant for the action noted in terms of foundational promises now being fulfilled: na kawadaku hawala na manuhangiwu tu kapawolu kawungaya tu kaparawi madainguya the one metalsliver the one chicken as established the firstpeople by as createdby the ancientpeople
Physically placed in front of the speaker, they are directed at the invisible interlocutor, with an explicit demand for a response: kanakayiwena lima kanahimawena aba in orderto receivewiththe hand in orderto respondwiththe mouth
The structure of performance, its uncertainties, and its consequential nature underwrite and mobilize its powers of reference.
the spoken house
As a result,the construction ritual of encountersupports certaindiscursive a objectification. it a First, presupposes sociologicalandeven ontologicaldivideacrosswhichpeople mustspeak about who they and the interlocutor and how they are to act. In contrastto everyday, are face-to-faceinteraction, ritualspeakers cannotassumethattheirinterlocutors awareof the are relevant elementsof the context,can interpret messagesbeingconveyed,or are even able the to hearone another.Second,theircommunication displayedas difficult, is highlyconsequenwhich they communicateis best fitted tial, and subjectto failure.Third,the mediumthrough to this purpose by virtueof its ability to transcendparticular contexts throughthe formal of it properties entextualization: is ancestralspeech. The presenceof ancestors,embodied in the actionsof speakers,is evoked by foregrounding absence of those speakersfromtheir the identities speakers the colloquial.Fourth, as of understand semioticproblemin the participants termsof pragmatic outcomes.Any given performance authorized the commitments is that by linkit to previousand futureperformances instance,to give thanksor to make up forthe (for of shortcomings the presentrite)and lookstowardoutcome(thefame,wealth,and healththat shouldfollow good performances, the misfortunes may motivateor resultfromerrors). or that Reference predictionin marapuritual,then, are functionsof how speakersunderstand and theirwordsto act and haveeffects,andof the kindsof relations betweenspeakers addresses and Central the strategic interpretive to and discussions surround event that the theycan presuppose. is the bearing any given stretch of speech may have on the situation at hand and its of aim (cf. consequencesforthe listeners Rosaldo1982). Thewordsandstructure performance at engagingan interlocutor orderto imposeor respondto obligations. in The risksare misfire, and error. practical incapacity, procedural
speech without obligation I have arguedthatthe referential characterof ritualspeech operatesin relationboth to the distancethatis assumedto lie between interlocutors to speakers' and beliefs aboutthe words that can most effectivelycross that distance. In contrast,a quite differentrelationship was enacted when Umbu Sebu produceda formaldescription the house, offeringa nuggetof of cultural In his knowledgeand displayinghis own authority.25 the contextof our conversation, wordswereto be understood as referential function,servingto pointout and name in primarily the partsof the house. I was not the spiritaddressee,nor was this partof a speech event to responding an obligationor demandingrecognitionand action. In fact, he spoke "outof in context" a breachof performance restrictions a non-Christian that wouldconsiderdangerous. Aselsewherein Indonesia restrictions loosenedif one's speech are (Fox1988:20),performance is thoughtto concern "culture" ratherthan "religion" (Indonesiankebudayaan) (Indonesian agama). To use ritual of speechwithoutriskis madefeasibleby changesin the nature practical power under increasingly coercive state rule and the ontologicaltransfigurations broughtabout by in in Church stateintervene local institutions practices interlocking and and Christianity.26 ways. Forone, bothchallengethe authority ritualspeech, or at leastresituate as the languageof of it, The trade,and, to a largeextent,the church,is Indonesian. state,havingalready government, the mainpowersof clans, continuesto downplaytheir identitiesand discourage superseded theiractions.27 and Throughregulations, pedagogy,exhortations, events such as interdistrict of them with allegiancessituatedwithina nested hierarchy sportsevents, it seeks to supplant effortshave begun administrative-territorial Althoughnot fullysuccessful,governmental units. to undermine clan as an effectivepoliticalprotagonist ancestralties as a presumed the and the sourceof well-beingand, in the process,have obliterated contextin which muchof ritual speech operates,the differencesit mediates,and the consequencesit entails.
Protestantism influences also and speech practice linguistic ideology.Contemporary marapu ritualists themselvesfromChristians callingthemselves"peopleof prayer" (tau distinguish by When ritualspeech is directedto marapuit is doublysituatedin temporal relations of nya'ba). cause andeffect.Theconsequentiality performance protected strong of is taboosthatprohibit by out as Christians, membersof a global institution, performance of context.Incontrast, praying should not aim to createor fulfillspecificobligations with individuated, potentially responsive ancestral exhorttheircongregations agents.Nor do they aim to speaklikeancestors.Ministers to pray directly from the heartor "liver"(ati). Telling them to shut their eyes (a sign of and-in contrast Catholicpractice-the lack of prayerbooks),theyexplicitly to introspection ritual: use of poeticallystructured the oppose two core featuresof marapu speech,which they see as not spontaneous; the delegationof speakingroles,which they see as irresponsible. and to are Enjoined heartfelt speech in religiousfunctions,Christians leftto see ritualspeech as Churchmembers sometimesambivalent different. are about"tradition" (Indonesian something is tradisi). Christianity identifiedwith modernityand the West, and church servicesclosely forms.Yetgovernment follow European policy,thoughdemandingadherenceto one of thefive world religions,also encouragesthe preservation local culture.28 of As legally recognized elsewhere in Indonesia,the crucial issue in Anakalang become that of how to separate has as (whichareto be eschewed)fromculture(whichis to be promoted furnishing paganpractices emblemsof local identity) and Rodgers Volkman1990). 1987; (Kipp of ritualspeech can arisein at leasttwo basicforms: as Non-marapu appropriations marapu or as a culturaltext in citationor exegesis. Common contexts for display include display inductionceremoniesfor local officials,culturalshows held duringinterdistrict events or for of tourists, oratory and the raising stone tombs.The firsttwo contextsare part of parties during of effortsto link culturaldisplay with territorial identitiesand often presume a nonlocal audience.The lastof these contextsinvolvescompetitivestatusclaims:although sharing many featuresof speech directedto spirits,it does not usuallypresuppose same kindof efficacy. the This is evident in the fact thatChristians expresslyforbiddento makeor even to consume are a to ritualists offerings, restriction which marapu objectas an evasionof the chainsof obligation in which the event is situated(Keane1994, 1995; cf. Volkman1990). When Christians speak but omit the material of sign, this suggests,likeotheralterations speech events (performances in daylight,inappropriate and eitherthat participants, otherovertbreachesof ritualpropriety), an interlocutor thattheydo so withoutanycommitment or theydo notface andaddress invisible to, and fearof, the results. Ifperformance withoutdangerrevealsa shiftin linguisticideology,the second use of ritual speech, as culturaltext, bringsout the focus on the denoted object. This can be seen in and I Wohangara's typescript is also evidentin a conversation had in 1993 withAmani Delu, a middle-aged school teacher.Proudof his evangelicalefforts,he showed me how elementary he used the textualhouse to verypersuasive effect:
Who is at the "housecorner,headof the mainfloor"? What's there? tellyou, it's I'll Why is it calledthat? Satan[Indonesian who causedhumanity fall. Becauseif you prayto marapu, to look iblis],the tempter at where you go. When you "stopby the headof the main floor,"it nevermentionsthe door-it goes straight fromthe plaza intothe room.Now who travelssecretlylikethat,withoutgoingby the front up door?Itmustbe the tempter. Andthat'swhy he's below Nuku[a termforthe spiritin the house peak], below "thecreator humans" wolutau].Now, ifwe'reto believe in Jesus,we've got to go in by the of [ma door.Thedoor is Jesus:"Iam thedoorof truth."29
to Christianity. long afterward,overheard usingit againwith a marapu I him (Not As priest.) in Wohangara's exegesis quotedabove, he buildsan analysison the physicalhouseas a site and sourceof broader UmbuSebu,he does thisby drawing the ritual treating on text, meanings.Like its pragmatic sequencingin termsof a spatiallogic:the absence of a door in the textexpresses the somethingabout the world it denotes. In this treatment, effects of speakingthe text are
Ama ni Delu went on to tell me that once after he made this argument, a whole valley converted
the spoken house
and to of subordinate the objectsthat it denotes,which exist independently the performance itseffects.30 referentiality and what counts as culture directives talk of Ihave mentioned number waysAnakalangese aboutthe house,fromritual a with and casualconversations cultural to written (withethnographers, representatives synopses I of the nationandstate,andwithone another). have emphasizedUmbuSebu'sspeech, in part actionand cultural text. the becauseof the way in which it straddles differencebetweenritual of When UmbuSebuspoke, he stillcould assumethe authority one who is used to facingand this exhibitsa lack of In typescript, recitation engagingthe spirits.31 contrastto Wohangara's to forcethatcannotsimplybe attributed the operations and performative contextual specificity in the to however,UmbuSebu'swordswere of, say, literacy opposition orality.Like typescript, directednot at an addresseefromwhom he demandscountering actions,but at an audience. who listeners Such an audiencepotentiallyincludesa world beyond Anakalang, introducing are not able-or, at least, lack the authority-to speak by ancestralmeans and respondto ancestral obligations. his was clearlydistinctfroman addressto a spirit: wordswere not UmbuSebu'srecitation with which it should be linked,they were not embedded in the otherspeech performances and nameof an interlocutor, we did in predicated earlierpromises,he did not cryout the ritual evidence not havethe appropriate offeringin frontof us. Eachof these absenceswas sufficient he was expected.Whatever thatno response authority mighthave invokedas one who knows the texts, he was not for the momentclaimingto speak either as or to an ancestor.Forhis audience,this use of coupletsposed problemsless of responseand outcome thanof interpreis In tationandtranslation. thiscontext,his recitation to be evaluatedprimarily itstruthand by than its consequences-and it is to such an evaluationthat completenessof referencerather the Ama ni Delu appealsin identifying door altarwith Satan. the Inthe contextof his recitation, text that Umbu Sebuofferedme bore two implications. ladentraditional the As an informative itdenotesa referent, symbolically house,something text, or of the performance of the demandfor response.As a display and existing independently that as of it student culture, offers referent anexemplary to transmittal knowledge a note-taking of In culturalartifact. both respects,howeversimilarthe wordsmightbe, it departsfromspeech has to that is addressed a spirit-speech in which the criticalissue is whethercommunication been achieved,the spiritinducedto move, and itsblessingssecured.Insuch speech, reference to the partsof the house is a functionof its role in directingand defininga place for actions: as likeotherformalized lists,it emergesfromefficaciousperformances, tracesof the moves in of The which layingoutthose moves in speech playsa crucialpart. elaboration thispart-by-part the encounter,the the accountreflects need forprecisionin the face of the problems posed by that cautiousetiquette,and the consequentiality these entail. I suspect that as the powers that Anakalangeseritualspeech was meantto mobilize lose and interlocutor the purchase,the emphasison denotationmaygrow.Whenthe presupposed as are both lost, what is left remainsthat much more availablefor treatment text. difficulty where the denotativefunctionbecomes central,new aspectsof its object may Furthermore, come to the fore.One is the alteredtemporaldimension:becausethereis no event,there is no beforeand no after.The step-by-step descriptionof the house (likeother rituallists,such as a trek on the ancestral or componentsof exchange)does not construct motion stoppingpoints is that indicatesa structure potentially fullypresent.Thesecond acrossspace in timebut rather follows fromthe first:as a picture, it neitherevokes nor respondsto possible failure.With and uncertainty dangerless at issue, the picturemightbe incomplete,but the actionwill not for At be infelicitous. most,the description mightsimplylacka target ostensivereference-the
house in question might be absent. The fact that Umbu Sebu did not speak to me in a "traditional" housewould nothave affectedthe efficacyof the wordshadthey been spokenin the rightframe,with appropriate speakers,antecedents,and offerings.It did, however,affect the referential felicityof the text:likethe wordsof a book, it did not denotethe place wherewe ideal house.Withsuch a shiftin emphasis,this house becomes spoke, but some hypothetical, less a site for encountersof consequence than an object of talk and of potentiallytotalizing interpretation.
I began this articleby arraying "almostpedantic" againstMalinowski's speakera series of to suggestthatwhat makesit possibleto talkaboutsomethingis in parta functionof skeptics what makesit necessaryto do so. Inthe exampleof the spokenhouse, I have focusedon ritual and speech because of the way it foregrounds respondsto that necessity.The formsof ritual are affectedby the actions it mediatesand the assumptions about language,and the speech thatinhabit world,thatunderliethem.Atthe same time,the verytextualdimension the beings in of ritualspeech leaves it open to appropriation new contextsand to new purposes.Insome or Yet contexts,itcan give riseto apparently objectifying totalizingcultural representations. this is only one partof the story,for ritualencountersdemandnot only entextualizedspeech but also interactive, whatcan happenwhen one of these consequential performance. suggesting By dimensionscomes to be emphasizedover another,I have attempted tracea few of the links to among ways people think about theirwords, their capacities to act, and how they portray themselves.A speakerin Anakalang mightemphasizethatritual speech is a powerfulmeansof with powerfulothers across social and communicativedistance, or that it is the interacting bearerof timelesstextsand codes. Inpractice, these treatments intertwine oscillate:indeed, or ritualspeech requiresboth possibilities operate,for the power of ritual to fully performative terms(1986:74),as eitherutteranceortext, but that it speech is not that it works,in Bakhtin's is both. To askwhatmakescertainkindsof talkplausible, even necessary,is to placeepistemology or in socialand historical context.As Rosaldo (1973)pointsout, languageideologiesoftenchange as functionsof transformations institutions, in ontologies,and the sourcesof power.Evenwhen the words, such as Umbu Sebu's text, are not new, these transformations supportnew can of what it means to speak them. Neitherthe words nor the objects that are understandings there fully determinewhat will be made of them. In dwellingon formalgenresof "already" I haveemphasizedtheircontribution self-representations playan important to that role speech, in the interactions the "local"and "outside," of where "traditional" are reconfigured practices and culturalidentitydebated. Culturalmetalanguagesmay give strengthto ongoing local self-definition struggle(Comaroff Comaroff or and 1991:212).Alternatively, they may add to the perceptionthatcultureis somethingthatexists primarily a self-contained, as stableobject of discourse(Handler art fromeconomic 1984) or as an expressive existingin a sphereseparate or politicalaffairs (Bowen1991:127;Rodgers 1979).Theycan providepeople withnew Siregar sourcesof self-description recognition-or tell themthatthe heartof theiridentitylies out and of reachin the past,perhapsin anotherland,or inthe handsof thosefew people who are"best" able to represent To follow these developments another it.32 is task.I havetriedto suggesthere that, in listeningto reflexiveor objectifying discourses,we shouldconsiderthe rangeof their sources and possibilities,and their mediationby speakers'understandings their circumof stancesand of whattheirwordscan do as they respondto the shifting termsof engagement.
the spoken house
Acknowledgments. Fieldwork in Anakalang (January 1986 to November 1987) was funded by a Department of Education Fulbright-HaysDissertation Fellowship, the JointCommittee on Southeast Asia of the Social Science Research Council, and the American Council of Learned Societies with funds from the National Endowment for the Humanities and the Ford Foundation and the Wenner-Gren Foundation for Anthropological Research, under the sponsorship of the Lembaga Ilmu Pengetahuan Indonesia and the Universitas Nusa Cendana. Fieldwork in 1993 was funded by the Southeast Asia Council of the Association for Asian Studies and the Luce Foundation Small Grantsfor Isolated Scholars. Further support was provided versions of this article were presented at the Department by the CharlotteW. Newcombe Foundation. Earlier of Anthropology, Universityof Chicago (1991), the Center for South and Southeast Asian Studies, University of Michigan (1992), the annual meetings of the American Anthropological Association (1992), and the Department of Folklore and Folklife, University of Pennsylvania (1993). I have been helped by the discussions on those occasions, especially by RichardBauman, Charles Briggs,JeanComaroff, and Michael Silverstein. I am grateful for the comments of Donald Brenneis, Joseph Errington,Marilyn Ivy, John Pemberton, Adela Pinch, Rafael Sanchez, PatriciaSpyer, and four anonymous reviewers. I am most deeply indebted to all who worked with me in Anakalang. This article is dedicated to the memory of Umbu Pada Buli Yora. 1. Bakhtin 1986 and/or Voloshinov 1973 emphasize(s) the interactive and dialogic nature of speech as an activity. The field of pragmatics draws particularattention to the role of context (for example, Duranti and Goodwin 1992; Hanks 1990). At the same time, as I discuss below, an emphasis on situated speech can be complemented by attention to the role of decontextualizing features of language in actual use (Bauman and Briggs 1990). Although some ritual forms, such as divination, "actively subvert the expression of intention"(Du Bois 1993:48), even the performatives that inspired speech act theory depend on conventional codes that cannot be fully grounded in the context of use and the intentions of speakers (Derrida 1982). 2. Differences in language (or linguistic) ideology, such as privileging reference and predication over performativity,can affect the ways people actually use language (Silverstein 1979) and embody historical and cultural differences in concepts of person and agency (Rosaldo 1973, 1982; see the useful review in Woolard 1992). Even people who speak "the same" language may operate from differentassumptions about its functions, and these assumptions may produce important distortions of cross-cultural conversations (Briggs1 984). 3. Speech "about" practice is itself "in" a form of practice. This means also that any instance of metalanguage is itself a form of language and thus, potentially, the object of furtherreflexivity(Lucy 1993; Silverstein 1979; Urban 1991). Bateson (197211955]) argues that the problems of framing, recursiveness, and vulnerability to reinterpretationare not just linguistic issues but a fundamental aspect of sociality more generally. 4. Anakalang (population 16,000) is a one of the dozen or so related societies of Sumba. As in east Sumba (Adams 1974; Forth 1981; Kapita 1976a, 1976b; Onvlee 1973, 1977), to which it is linguistically closest, asymmetric marriagealliance, mediated by multigenerationalexchanges of valuables, continues to structure a large part of social interaction. But Anakalang also resembles its geographically nearest neighbors to the west (Geirnaert-Martin 1992; Hoskins 1988, 1993; Kuipers1990), for until the 1920s, although cooperating in large rituals, "clans" (kal?isu) did not acknowledge a central authority, and their mutual relations were realized as much through warfare and competitive feasting as through marriage. 5. D. H. Wohangara was introduced to me during my initial week-long visit to the eastern part of the island in 1985, when he gave me this document. While the paper was typed for me and also addressed to me, the care with which it was thought out suggests that the basic text had been prepared beforehand, though for what audience I cannot say; he apparently also wrote a paper on marriage rules (Wohangara 1963), probably for a conference on customary law convened by the Regency government. I therefore offer this as no more than a particularlyapposite illustrationof how easily one encounters schematic descriptions of Sumbanese culture, and how central the house is in them. On several other occasions people offered me similar texts, some produced for themselves, others for cultural affairsofficials. 6. The expression "FourPrinciples"(CaturSila) is formal, possibly meant to invoke the "Five Principles" (Pancasila) of the official Indonesian state ideology. "Sumbanese People" (Suku Sumba) is rendered with a term usually denoting an ethnic group. Although this degree of generality is common in colonial and Indonesian language writing, Sumbanese rarely identify themselves in terms of the entire island. The capitalized words in parentheses are a couplet from east Sumbanese ritual speech that might be rendered "four rules, establish village" (Onvlee 1980). 7. The attic serves as storage bin for staple grains (rice and maize), including small portions set aside as offerings to the spirits who watch over them. Whether the speaker emphasizes one function or the other may well reflect the extent of the listener's perceived sympathy to "paganism." 8. Formal descriptions do not, for example, tell you how to build a house, nor how to live in one, and may even neglect visually prominent features such as the "house horns"(fora criticism of totalizing models of the house, see Ellen 1986). In one colloquial discussion about rebuilding a long missing clan house, I heard a man say, "Ifwe didn't have a house, where would we gather? Where would we speak? Where would we confer? Where would we dwell, huh?"Note his emphasis on the house as a site for certain socially
marked actions. But this is not simply a logistical matter;after all, the speakers were not prevented by lack of a "traditionalhouse" from conferring. Itwas their public identity and status as a clan that were at stake. In contrast, talk about cultural preservation, usually by people identified with development, tends to emphasize architectural form over practices, for example, proposing that pockets of tradition be required to retain thatched rather than zinc roofs. When the Department of Education and Culture sponsored rebuilding in LaiTarung, a nearly abandoned ritual center expected to attracttourists, it was done without rites. 9. A rough idea of the size of the corpus is given by the 3,200 EastSumba couplets published by Kapita (1987). Formore on Anakalangese ritualspeech, see Keane 1991; for other Sumbanese societies, see Forth 1988, Hoskins 1988, and Kuipers1990. Parallelism and ritualspeech are discussed more generally in Fox 1971, 1988; Jakobson 1960. 10. By Indonesian religious policy, possession of a sacred book is an important part of the definition of a legally recognized religion (Kipp and Rodgers 1987). Sumbanese marapu followers, however, have not sought to gain state recognition by producing a sacred book based on ritual speech. Many Anakalangese claim that the ancestors once possessed a full book of ritualprocedures. Some insist that the ancestors had their own book but left it behind in Bali during their trekto Sumba. This, they say, explains the phenomenal success of the Balinese in achieving legitimacy in the national scene. The privileged status of sacred writing surely also contributes to a linguistic ideology that privileges fixed texts and denotation. 11. In addition, fully authoritative performance of ritual speech must be accompanied by either prestations or offerings, which make words binding and contribute to the temporal dimensions of ritual. Usually a dish containing betel and either the prestation or some token to represent it lie in front of the speaker, which helps define the space of speech performance. Prestationsto the living and tokens left from sacrifices to the dead remain as evidence of former speech acts. The current willingness of some Christians to perform without such materials both transformsthe political economy of speech practices and indicates importantchanges in language ideology (Keane 1994). 12. Ritual speech is usually performed in public, and most Anakalangese adults have some knowledge of it, as they often show by citing couplets in everyday conversation. Performances usually do not consist of fixed texts, but are improvised. Because speakers must draw on the canonical set of couplets, follow performance conventions, and are sometimes restricted by ancestral prerogatives, they are subject to challenge by listeners and disapproval by spirits. This, combined with the difficulties created by esoteric vocabulary and indirection, means that few people have the skill or self-confidence to perform. Most importantnegotiations and rituals are delegated to specialists, normally men. Their skills are respected but do not contribute much to their own social standing. 13. Bauman and Briggs (1990) show how a wide range of linguistic and performance properties, such as reduced deixis, highly stylized prosody, and increased focus on poetic structure,contribute to "entextualization." For a thorough analysis of the full range of linguistic resources that promote entextualization in .the ritualspeech of a West Sumbanese society, see Kuipers1990. Kuipers(1990:1 75-1 78) points out that formalized speech has many of the features that supposedly distinguish "literacy" from "orality." The salience of such features is especially marked in places such as Anakalang, where the indigenous language is rarelywritten, and literacy in the national language, Indonesian, is largely restrictedto the young. 14. Marapu, a term for the most distant and powerful ancestors, is commonly used in Anakalang as a cover term for ancestral ritual and its adherents. By "Christians" mean the dominant Calvinists. Although I Christiansand non-Christiansdo not formdistinct sociological or economic groups, Christianityis associated with education and participation in the cash economy, and is virtuallya legal requirement for civil service employment. In the 1986 census, greater Anakalang was 42.4 percent Protestant, 6.48 percent Catholic, 50.6 percent marapu (KantorStatistik1987). 15. This is transcribedfrom dictation. Canonical parallel couplets are indicated by adjacency or parallel indentations. Consonants marked [. ] are implosive. 16. There is great interclan variety in mortuary ritual, but in general the deceased is told the way to the village of the dead at the time of burial. Three days later, a rite "receives the spirit"(kayi dewa) back into the house. Afterthree years, wealthy families can "place in the shelter of the shade" (pahangerang ta mawu) to secure furtherthe presence of the deceased's good fortune(kanyuru)forthe survivors.The relations among these events hinge in part on distinctions among differentaspects of spirit (dewa and hamangu or hamawu), which are raiher esoteric; as Forthobserves: "ndewa... is probably the most elusive term in the eastern Sumbanese metaphysical vocabulary" (1981:77; cf. 439). 17. Formal speech at these events, such as marriage negotiations, divination, and mortuary and other rituals, is-with the exception of keening for the dead-dominated by men. In some houses, this room is forbidden to in-marryingwomen, who must enter the house by the back door. Just inside the front door, in the corner where the roof meets the wall, is the "spiritshelf" (hedi marapu), a small, rough-hewn platform. Betel offerings are placed there for a guardian spirit, sometimes denoted by its location, "house corner." Betel is also placed on the disk (leli) of the divination pillar, and, in some houses, at an altar above the door spirit shelf ("egg spirit, chicken spirit,"marapu tilu, marapu manu) and at the rearof the main room ("horse spirit, dog fate," dewa jara, ura ahu). 18. Much house-building ritual centers on obtaining and erecting the central pillars, especially the "divinationpillar."The pillars bear a varietyof names, not all of which are common knowledge. One version, given to me by a Christian minister whose father and brother were marapu priests, is coordinate with Wohangara's four principles above. The terms derive from activities associated with the quadrants of the
the spoken house
house. Moving counterclockwise fromthe divination pillar,they are the pillarsof "theway of life and death" (li luri, li mati), "rice and corn" (uhu, watar), "way of marriage"(li lalawi, li mangoma), from the place for the corpse during a wake, the food preparation area, and the main sleeping chamber. 19. The peaked ("able, good") house (uma piaku) is distinguished from huts (kawarung) or minor buildings (uma an) along various crosscutting, but similarly hierarchical, dimensions. Nonpeaked houses can be categorized by roof shape-four eaves ("square,"uma kabalolu) or two eaves (long like a "ship," uma tena)-or by whether they are elevated (uma deta) or stand on the ground (uma tana). Indonesian tax categories look at materials and structure:in 1984, Umbu Pabal township (Indonesian desa) reported two stone houses, 27 impermanent or rough houses (Indonesian rumah darurat),and 136 platform houses; Anakalang township listed 27 partly stone, 47 permanent stone, 115 board, and 121 bamboo. As many houses form a subset of the last category platformand bamboo houses do not have peaks, fully "traditional" (even if the question of each list;thus, well under half of the houses in Anakalang township are "traditional" of roofing material, noted above, is disregarded). 20. Ritual,politics, and economics all tend to support the exemplary status of the peaked house. To build one requiresthe ability to mobilize and feed workers. Although laborand most materialsmaybe contributed, they flow more readily to possessors of position or wealth. Of particularimportance is access to large trees forthe central pillars;on this largelydeforested island, sources of wood maybe distant and require payments to other clans. In contrast to garden houses, the building of those in the ancestral village, which serve as ritual centers, is the responsibility of larger social units, often consisting of one or more clans. Residence and status are therefore not isomorphic: people of low rankmay occupy large peaked houses that manifest the status claims of clans whose more powerful members dwell elsewhere. 21. Houses are usually spoken of as offspring of the shared clan ancestor or in terms of their special contributions to clanwide rituals.As the boundaries between house and clan are not always clear in actual practice, and since the basic principles discussed in this article remain the same for both, Iwill not distinguish between them here. Forthe complex intersection of marriagestrategies, exchange, ritual, gender, and rank in other eastern Indonesian "house societies," see Barraud 1979; Fox, ed. 1980; Lewis 1988; McKinnon 1991; and Traube 1986. 22. In one ritual to summon the village spirits back after a fire, a priest arose in the middle of the night and went out alone. The next day he told me he had forgotten a step and had gone to make an offering and prayer. He feared that the success of the rite would be jeopardized and that he himself was at risk. The presence of the interlocutor poses temporal problems that parallel the spatial ones. A great deal of effort goes toward telling the spirits not just where to show up, but when. Many large rituals are preceded by smaller rites to announce the forthcoming event. 23. For reasons of space, I mention only lexicalized examples of the space constructed by ritualspeech. It is importantto bear in mind, however, that other elements contribute to this sense of facing an other, such as performance structureand linguistic features such as deixis (Duranti and Goodwin 1992; Hanks 1990). Thus, the directives in the quotation above point to parts of the house as in some sense present to both speaker and addressee (cf. Kuipers1990). 24. Speech to ancestors is normally accompanied by a sacrifice, which provides entrails or a liver for divinatory reading (Hoskins 1993; Kuipers 1990:99-105). In addition, however, people are alert to all possible indicators: I have been present at worried discussions concerning how to interpreta stumble, a coughing fit, and a spilled cup of coffee. 25. Presumably, this performance, like other ways of displaying knowledge, can arise from many motivations at once, including pleasure, and it may, in this rivalrous social world, make an implicit claim to a superior command of tradition. Our interaction is also pedagogic, but I should note that ritual speech is not normally taught. Most specialists claim that they never learned ritualspeech but that it just comes to them-which is a common way of experiencing the process of acquiring practical knowledge (Lave and Wenger 1991). They usually attributetheir abilities to intervention by spirits. Some will also acknowledge having acquired it through long experience on the sidelines. The point here is that explanation and purposeful memorization play a very small role. 26. Under colonial rule, beginning early in this century, the chief mission on Sumba was Calvinist (van den End1987; see Keane 1995). By the time the church came under local control in 1946, missionaryefforts in Anakalang had resulted in only 180 baptisms in a regional population of over 22,000 (Luijendijk1946). Large-scaleconversion began afterthe violent end of the Sukarno regime in 1965-66, when nonadherence to a monotheistic religion was taken by many to be synonymous with communism. Government pressure on nonconverts through regulation of identity cards, requirements for civil service positions, classes in the schools, and direct exhortation has gradually increased. By the late 1980s, it was clear that Christianswould soon be in the majority in Anakalang (Keane, in press). 27. Inthe firsttwo decades of the century, the Dutch consolidated a loose system of indirect rule through petty kings of their own creation. One of the chief consequences was the removal of most legitimate force from local hands. Nonetheless, until several administrative restructuringsduring the 1950s and 1960s gradually implemented Indonesian state control and produced the present hierarchy of territoriallydefined units, clans continued to hold much of the access to farmland,forests, water, and other resources. This clan control has now been abolished by state decree, and the clans' ritual powers over the fertilityof land and women have been eroded by religious conversion. Clan elders still retain considerable control over the valuables necessary for marriage and, within the township structure,adjudicate legal matters such as theft,
boundary disputes, and other quarrels. Their authority is implicated, albeit to a diminished extent, in the continued coherence of clan identity. 28. In fact, a major sponsor of the publication of cultural materials in Sumba is the church (Kapita1976a, 1976b, 1987, n.d.). 29. Presumably an allusion to John 10:1, 9. 30. Of course, he does indeed speak to achieve specific effect, as he points out himself. But this is the persuasive effect of an argument constructed around the object portrayedin the ritualtext-it does not draw on indexical links to ancestral addressees, on conventional performativity,or on poetic structure. 31. People in Umbu Sebu's position can play self-consciously with these different roles, each with its distinctive forms of authority. See Hymes' (1981) account of a man who is willing to tell a myth but resists the ethnographer's desire to have him "perform"it, preferringthe role of linguist to that of "native." Forthe role of citation and displaced agency in "traditionalization,"see Bauman 1992. 32. A brief note on the large topic of emerging hierarchies is called for here. The reader may already have observed that my examples of authoritativewords are all spoken by men-although not all men speak with authority, and the ability to speak formally does not map directly onto social standing (Keane 1991; cf. Lederman 1980). In addition, today it is sometimes people who are distant from ritualpractices and can draw on new sources of status, such as formal education, who seem most articulateabout "tradition." They may undermine the authorityof those who in fact act but cannot explain (for unintended consequences of see "traditionalism," Hill 1985). Relatedto this is the experience of nostalgia and inauthenticity:when people focus on denotation, they may find a mismatch between the referentand the text; for example, actual houses frequently lack many of the features mentioned in the text. As noted above, to the extent that the text is performative, this may not necessarily be surprising or significant. When the text is treated as meaningful chiefly as evidence of a world to which it refers, however, the mismatch between text and objective correlative may support a common perception that "authentic culture" either is imperfectly realized in the given instance or has been perfectly realized once, but on ly in the past. This may help explain my experience that many of the people most committed to what they see as modernity are also the most vocal about the loss of tradition:they measure their vision of tradition against the imperfections of existing houses and the apparent inarticulatenessof actual practitioners.
Adams, Marie Jeanne 1974 Symbols of the Organized Community in EastSumba, Indonesia. Bijdragentot de Taal-, Land-en Volkenkunde 130:324-347. Anderson, Benedict 1990 Cartoons and Monuments: The Evolution of Political Communication under the New Order. In Language and Power: Exploring Political Cultures in Indonesia. Pp. 152-193. Ithaca, NY: Cornell University Press. Asad, Talal 1983 Anthropological Conceptions of Religion: Reflections on Geertz. Man 18:237-59. Atkinson, Jane Monnig 1989 The Art and Politics of Wana Shamanism. Berkeley: Universityof California Press. Bakhtin, M. M. 1981 The Dialogic Imagination: Four Essays. Michael Holquist, ed. Caryl Emerson and Michael Holquist, trans. Austin: University of Texas Press. 1986 Speech Genres and Other Late Essays. Vern W. McGee, trans. Austin: University of Texas Press. Barnes, R. H. 1974 K6dang:A Study of the Collective Thought of an EasternIndonesian People. Oxford: Clarendon Press. Barraud,Cecile 1979 Tanebar-Evav:Une societe de maisons tournee vers le large. Cambridge: Cambridge University Press. Barth, Fredrik 1990 The Guru and the Conjurer:Transactions in Knowledge and the Shaping of Culture in Southeast Asia and Melanesia. Man (n.s.) 25:640-653. Bateson, Gregory 1972  A Theory of Play and Fantasy. In Steps to an Ecology of the Mind. Pp. 177-193. New York: Ballantine Books. Bauman, Richard 1992 Contextualization, Tradition, and the Dialogue of Genres: Icelandic Legends of the Krataskild. In Rethinking Context: Language as an Interactive Phenomenon. Alessandro Duranti and Charles Goodwin, eds. Cambridge: Cambridge University Press. Bauman, Richard,and Charles L. Briggs 1990 Poetics and Performance as Critical Perspectives on Languageand Social Life.Annual Review of Anthropology 19:59-88.
the spoken house
Boas, Franz 1966(19111 Introductionto the Handbook of American Indian Languages, FranzBoas, Indian Linguistic Families of America, J. W. Powell. Preston Holder, ed. Pp. 1-79. Lincoln:University of Nebraska Press. Bourdieu, Pierre 1977[19721 Outline of a Theory of Practice. Richard Nice, trans. Cambridge: Cambridge University Press. 1979 The Kabyle House or the World Reversed. In Algeria 1960. Richard Nice, trans. Pp. 133-153. Cambridge: Cambridge University Press. Bowen, John R. 1991 Sumtran Politics and Poetics: Gayo History, 1900-1989. New Haven, CT: Yale University Press. Briggs,Charles L. 1984 Learning How to Ask: Native Metacommunicative Competence and the Incompetence of Fieldworkers. Language in Society 13:1-28. Comaroff, Jean, and John Comaroff 1991 Of Revelation and Revolution: Christianity,Colonialism, and Consciousness in South Africa, 1. Chicago: University of Chicago Press. Cunningham, Clark E. 1964 Order in the Atoni House. Bijdragentot de Taal-, Land-en Volkenkunde 120:34-68. Derrida, Jacques 1973 Speech and Phenomena: Introductionto the Problem of Signs in Husserl's Phenomenology. In Speech and Phenomena and Other Essayson Husserl's Theory of Signs. David B. Allison, trans. Pp. 3-104. Evanston, IL:Northwestern University Press. 1982 Signature Event Context. In Margins of Philosophy. Alan Bass, trans. Pp. 307-330. Chicago: University of Chicago Press. Du Bois, John W. 1993 Meaning without Intention: Lessons from Divination. In Responsibility and Evidence in Oral Discourse. Jane H. Hill and Judith T. Irvine, eds. Pp. 48-71. Cambridge: University of Cambridge Press. Duranti, Alessandro, and Charles Goodwin, eds. 1992 RethinkingContext: Language as an InteractivePhenomenon. Cambridge: Cambridge University Press. Ellen, Roy 1986 Microcosm, Macrocosm and the Nuaulu House: Concerning the Reductionist Fallacy as Applied to Metaphorical Levels. Bijdragentot de Taal-, Land-en Volkenkunde 142:1-30. Forth,Gregory 1981 Rindi: An Ethnographic Study of a Traditional Domain in Eastern Sumba. The Hague, the Netherlands: Martinus Nijhoff. 1988 Fashioned Speech, Full Communication: Aspects of EasternSumbanese Ritual Language. In To Speak in Pairs: Essays on the Ritual Languages of EasternIndonesia. James J. Fox, ed. Pp. 129-160. Cambridge: Cambridge University Press. 1991 Place and Space in EasternIndonesia. Canterbury, UK: University of Kent at Canterbury:Centre of South-EastAsian Studies Occasional Paper No. 16. Foucault, Michel 1972 The Archaeology of Knowledge and the Discourse on Language. A. M. Sheridan Smith, trans. New York:Pantheon Books. Fox, James J. 1971 Semantic Parallelism in Rotinese Ritual Language. Bijdragen tot de Taal-, Land-en Volkenkunde 127:21 5-255. 1988 Introduction. In To Speak in Pairs:Essays on the Ritual Languages of EasternIndonesia. James J. Fox, ed. Pp. 1-28. Cambridge: Cambridge University Press. Fox, James J., ed. 1980 The Flow of Life:Fssays on EasternIndonesia. Cambridge: HarvardUniversity Press. 1988 To Speak in Pairs: Essays on the Ritual Languages of EasternIndonesia. Cambridge: Cambridge University Press. 1993 Inside Austronesian Houses: Perspectives on Domestic Designs for Living. Canberra, Australia: Research School of Pacific Studies, AustralianNational University. Geirnaert-Martin,Danielle 1992 The Woven Landof Laboya: Socio-Cosmic Ideas and Values in West Sumba, EasternIndonesia. Lieden, the Netherlands: Centre of Non-Western Studies. Handler, Richard 1984 On Sociocultural Discontinuity: Nationalism and Cultural Objectification in Quebec. Current Anthropology 25:55-71. Hanks, William F. 1990 ReferentialPractice: Language and LivedSpace among the Maya. Chicago: University of Chicago Press.
Heidegger, Martin 1977 The Age of the World Picture. In The Question Concerning Technology and Other Essays.William Lovitt,trans. Pp. 115-154. New York: Harper and Row. Hill, Jane 1985 The Grammar of Consciousness and the Consciousness of Grammar. American Ethnologist 12:725-737. Hobsbawn, Eric,and Terence Ranger, eds. 1983 The Invention of Tradition. Cambridge: Cambridge University Press. Hoskins, Janet 1988 Etiquettein Kodi SpiritCommunication: The Lips Told to Pronounce, the Mouths Told to Speak. In To Speak in Pairs:Essayson the Ritual Languages of EasternIndonesia. JamesJ. Fox, ed. Pp. 29-63. Cambridge: Cambridge University Press. 1993 Violence, Sacrifice, and Divination: Giving and Taking Life in Eastern Indonesia. American Ethnologist 20:159-1 78. Hymes, Dell 1981  Breakthroughinto Performance. In "InVain I Tried to Tell You":Essays in Native American Ethnopoetics. Pp. 79-141. Philadelphia: University of Pennsylvania Press. Jakobson, Roman 1960 Closing Statement: Linguisticsand Poetics. In Style in Language. T. A. Sebeok, ed. Pp. 350-377. Cambridge: MITPress. Kana, N. L. 1980 The Order and Significance of the Savunese House. In The Flow of Life: Essays on Eastern Indonesia. James J. Fox., ed. Pp. 221-230. Cambridge: Harvard University Press. KantorStatistik(KabupatenSumba Barat)(Office of Statistics [Regency of West Sumba]) 1987 Sumba Baratdalam angka 1986. (West Sumba in Figures 1986). Kapita, Umbu [Oemboe] H[ina] 1976a MasyarakatSumba dan adat istiadatnya. (Sumbanese Society and Its Customs and Traditions.) Waingapu, Indonesia: Naskah-naskah Kebudayaan Daerah Sumba Dewan Penata Layanan Gereja KristenSumba. 1976b Sumba di dalam jangkauan jaman. (Sumba through the Ages.) Waingapu, Indonesia: Naskahnaskah Kebudayaan Daerah Sumba Dewan Penata Layanan Gereja KristenSumba. 1987 Lawiti luluku humba/pola peribahasa Sumba. (The System of Sumbanese Proverbs.) Lembaga Penyelidikan KebudayaanSelatana Tenri. n.d. Pamangu ndewa/perjamuan dewa. (Communion of the Spirits).Ende, Indonesia: Arnoldus. Keane, Webb 1991 Delegated Voice: Ritual Speech, Risk, and the Making of Marriage Alliances in Anakalang. American Ethnologist18:311-330. 1994 The Value of Words and the Meaning of Things in Anakalangese Exchange.Man (n.s.) 29:605-629. 1995 Materialism, Missionaries, and Modern Subjects in Colonial Indonesia. In Conversion to Modernity. Peter Van der Veer, ed. New York: Routledge, in press. Inpress Religious Change and Historical Reflection in Anakalang, West Sumba, Indonesia. In Protestents and Tradition in Southeast Asia. Cornelia Ann Kammerer,ed. Singapore: Journalof Southeast Asian Studies Monograph Series. Kipp, Rita Smith, and Susan Rodgers, eds. 1987 Indonesian Religions in Transition. Tucson: University of Arizona Press. Kuipers,Joel C. 1988 The Patternof Prayerin Weyewa. In To Speak in Pairs: Essays in the Ritual Languages of Eastern Indonesia. James J. Fox, ed. Pp. 104-128. Cambridge: Cambridge University Press. 1990 The Power of Performance: The Creation of Textual Authority in Weyewa Ritual Speech. Philadelphia: Universityof Pennsylvania Press. Lave, Jeane, and EtienneWenger 1991 Situated Learning:LegitimatePeripheral Participation. Cambridge: Cambridge University Press. Lederman, Rena 1980 Who Speaks Here? Formalityand the Politics of Gender in Mendi, Highland Papua New Guinea. Journalof the Polynesian Society 89:479-498. Levi-Strauss,Claude 1982 The Way of the Masks. Sylvia Modelski, trans. Seattle: University of Washington Press. Lewis, E. Douglas 1988 People of the Source: The Social and Ceremonial Order of Tana Wai Bramaon Flores. Dordrecht, the Netherlands: Foris Publications. Lucy, John A., ed. 1993 Reflexive Language:ReportedSpeech and Metapragmatics. Cambridge:CambridgeUniversityPress. Luijendijk,P. J. 1946 Zeven jaar zendingswerk op Soemba (1939-1946). (Seven Year Missionwork on Sumba ). Groningen, the Netherlands: J. Niemeijer. Macdonald, Charles 1987 De la hutte au palais: Societes "a maison" en Asie du sud-est insulaire. Paris:CNRS Press.
the spoken house
Malinowski, Bronislaw 1961  Argonauts of the Western Pacific. New York:E. P. Dutton. McKinnon, Susan 1991 From a Shattered Sun: Hierarchy, Gender, and Alliance in the Tanimbar Islands. Madison: University of Wisconsin Press. Mitchell, Timothy 1988 Colonising Egypt.Berkeley: University of CaliforniaPress. Oliver, Paul 1987 Dwellings: The House Across the World. Austin: Universityof Texas Press. Onvlee, L. 1973 Cultuur als Antwoord. The Hague, the Netherlands: Martinus Nijhoff. 1977 The Construction of the Mangili Dam: Notes on the Social Organization of Eastern Sumba. In StructuralAnthropology in the Netherlands. P. E.de Josselin de Jong, ed. The Hague, the Netherlands: Martinus Nijhoff. 1980 Kamberaas (Oost-Soembaas)-Nederlands woordenboek met Nederlands-Kamberaas register. (Kambera[East-Sumbanese]-Dutch Dictionary with Dutch-KamberaRegister.)Dordrecht, the Netherlands: Foris. Ricoeur, Paul 1971 The Model of the Text: Meaningful Action Considered as a Text. Social Research 38:529-562. Rodgers Siregar,Susan 1979 Advice to the Newlyweds: Sipirok Batak Wedding Speeches-Adat or Art? In Art, Ritual, and Society in Indonesia. EdwardM. Brunerand JudithO. Becker,eds. Pp. 30-61. Athens: Ohio University Center for InternationalStudies. Rosaldo, Michelle Z. 1973 I Have Nothing to Hide: The Language of llongot Oratory. Language in Society 2:193-223. 1982 The Things We Do with Words: llongot Speech Acts and Speech Act Theory in Philosophy. Language in Society 11:203-237. Rumsey, Alan 1990 Wording, Meaning and Linguistic Ideology. American Anthropologist 92:346-361. Silverstein, Michael 1979 Language Structureand Linguistic Ideology. In The Elements:A Parasession on Linguistic Units and Levels. Paul R. Clyne, William F. Hanks, and Carol L. Hofbauer, eds. Pp. 193-247. Chicago: Chicago LinguisticSociety. Traube, Elizabeth 1986 Cosmology and Social Life:RitualExchangeamong the Mambai of EastTimor.Chicago: University of Chicago Press. Urban, Greg 1991 A Discourse-Centered Approach to Culture: Native South American Myths and Rituals. Austin: University of Texas Press. van den End,Thomas 1987 Gereformeerde Zending op Sumba, 1859-1972. (Calvinist Mission in Sumba, 1859-1972). Alphen aan den Rijn, the Netherlands: Raad voor de Zending der Ned. Herv. Kerk,de Zending der Gereformeerde Kerkenin Nederland en de Gereformeerde Zendingsbond in de Ned. Herv. Kerk. Volkman, Toby Alice 1990 Visions and Revisions: Toraja Culture and the Tourist Gaze. American Ethnologist 17:91-110. Voloshinov, V. N. 1973 Marxism and the Philosophy of Language. Ladislav Matejka and I. R. Titunik, trans. New York:Seminar Press. Waterson, Roxana 1990 The Living House: An Anthropology of Architecture in South-East Asia. Singapore: Oxford University Press. Wohangara, D. H. 1963 Kawin mawin menurut adat istiadat suku Sumba Timur.(MarriageAccording to the Customs and Tradition of the East Sumbanese People.) Unpublished manuscript cited in Taro Goh 1991. Sumba Bibliography, Canberra: Occasional Paper of the Department of Anthropology, Research School of Pacific Studies, The Australian National University. Unpublished manuscript. 1985 Sejarah singkat kebaktian dan kebudayaan Sumba berdasarkan keyakinan marapu (Animisme). (A Brief History of the Religious Ceremonies and Culture of Sumba Based on the Marapu Belief [Animism].)Unpublished manuscript. In possession of the author. Woolard, KathrynA. 1992 Language Ideology: Issues and Approaches. Pragmatics2:235-249.
submitted August 3, 1992 revised version submitted January27, 1993 revised version submitted May 7, 1993 accepted August 5, 1993
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue listening from where you left off, or restart the preview.