CLIFFORD, J. and G. MARCUS.- Writing Culture. the Poetics and Politics of Ethnografy (Section) - Fix

7/22/2019 CLIFFORD, J. and G. MARCUS.- Writing Culture. the Poetics and Politics of Ethnografy (Section) - Fix

1/13

WritingCultureThe Poetics and Politicsof Ethnography

EDITED BY JAMES CLIFFORDAND GEORGE E. MARCUSA School of American ResearchAd vanced Seminar

Universily of California PressBEREELEY LO S ANGELES LONDON

University of California PressBerkeley an d Los Angeles, CaliforniaUniversity of California Press, Ltd,London, England(9 19BG byTh e Regellls of the University of CaliforniaPrinted il l lhe Uniled Slales of America3 4 5 G 7 8 9Library 01' Congress Datal\Iain entry ullder tille:\Vriting Culture,

Seminar held in Santa Fe, N.M. April 1984.Bibliography: p.Includes index,J:

2. I. (:lifford,JaI71cs,1945- II. rvlarc!1s, George E. Ill. School of AmericallResearch (Santa Fe, N,M.)GN307.7.W75 1980 30G ' '{)l'S 8S-14SGOlSBN 0-520-05729-.'l (alle paper)


2/13

JAMES CLIFFORD

On Ethnographic AllegoryI. a story in which people, things and hajJjJenings haveanother meaning, as in a fable or parable: allegtyries are usedFJr teaching or exjJlaining.2 . the presentation ufideas by means of such sturies . . . 1

In a recent essay on narrative Victor Turner argues that socialperformances enact pmverful stories-mythic and commonsensical-that provide th e social process "with a rhetoric, a rnode of emplot.-ment, an d a meaning" (1 q80: 153). In what follows I treat ethnogra-phy itself as a performance em plotted by powerful stories. Embodiedin written reports, these stories simultaneously describe real culturalevents an d make additional, moral, ideological, an d even cosmologicalstatements. Ethnographic \'/riting is allegorical at the level both of itscontent (what it. says about cultures an d their histories) an d of it.s form(what is implied by it.s mode of t.extualization).

An apparently simple example will introduce my approach. rvlar-jorie Shostak begins he r book Nisa: The and Words oj' (1 !KungWoman with a story of childbirth the! Kung wcty--outsidc the village,alone. Here are some excerpts:1 lay there and felt the pains as they carne, over an d again. Then I feltsomething wet. th e beg-inning of th e childbirt.h. I thought, "Eh hey, maybe it is

I, Webster's New Twentieth Century Dictionary, 2nd t:d. III literary studics definitionsof allegory have ranged from Angus Flcrcht:r's (J 961; 2) loose charactaiz

Th e slory bas great immediacy. Nisa's voice is unmistakable, theexperience sharplv evoked: "She Jay there moving her arms about

to suck he r 'fingers." Bu t as readers do m;Jre than register,:;l:ntque event.. Th e story's unfolding requires us, fIrst, to imagille a dif-ferent cultural norm (J Kung birth, alone in the bush) an d then to rec-ognize a common human experience (the quiet heroism of childbirth,feelings of postpartum wonder an d doubt). Th e story of an occur-

somewhere in the Kalahari Desert cannot remain just that. Itunpiles bot.h local cultural meanings an d a general story of birth.A difference is posited and transcended. IVloreover, Nisa's story tellsus (how could it not?) something basic about woman's experience.

life of a !Kung individual inevitably becomes an allegory of(female) humanity.

I ar?ue belc)'w 1ha t these kinds of transcendent meanings arc notabstractIOns or interpretations "added" to the original "simple" ac-count. Rather, they ar e the conditions of its rneaningfulness. Et.hno-

texts ar c inescapably allegorical, and a serious acceptance ofthIS 1a:1. changes the ways they can be written and read. Using Shostak'sexpenr.llent as a case study I examine a recent tendency to clisting).lishallegoncal levels as specific "voices" within the text. I argue, finally,that the very activity of ethnographic wri t ing-seen as inscription ortext.ualization-enacts a redemptive \Vestern allegory. This pervasivestructure needs to be perceived and weighed against other possibleemplotmenlS for the performance of ethnography.

Literary descrijJtion always oj)cns ontu another SCl'1le ,let, .\0 to speak,"behind" the this-worldly things it jJ1nj)orls to dejJ1:ct.MICHEL BEAliJOUR, "Some ,-,I'

Allegory (Gr. altos, "other," an d agot'euein, "to speak") usuallydenotes a practice in which a narrative fiction continuously refers toanother pattern of ideas or events. It is ;:t representation that "inter-prets" itself. I am using the term allewwy in the expanded sense re-


3/13

10 0 .lAMES CLIFFORD

claimed fo r it by recent critical discussions, notably those of AngusFletcher (19()'1) an d Paul De Man (1979)' Any story has a propensityto generate another story in the mind of its reader (or hearer), to re -peat an d displace some prior story. To focus on ethnographic allegoryin preference, say, to ethnographic "ideology"-alt.hough th e polit.icaldimensions ar e always present (Jameson 1981)--dnnvs alJention toaspects of cultural description that have unlil recelltly been mini-mized. A recognition of allegory emphasizes the fact tbat. realist.ic por-t.raits, to the extent that they are "convincing" or "ricb," ar e ext.endedmetaphors, patterns of associations that point. to coherent (t.heoreti-cal, esthetic, moral) additional meanings. Allegory (more stronglythan "int.erpretation") calls to mind the poetic, traditional, cosmologi-cal na t lire of such vniting processes.

Allegory drav,rs special attention to the narrative character of cul-tural representations, to the stories built into t.he representationalprocess itself. It also breaks down the seamless quaEt y of cultural "de-scription hy adding a temporal aspect to t.he process of reading. On elevel of meaning in a text will always generate ot.her levels. Thus therhetoric of presence that has prevailed in much post.-romanticI.ure (and il l much "symbolic anthropology") is interrupted. De Man'scritique of th e valorization of symbols over allegory in romantic es-thetics also questions th e project of realism (De ?vIall 1969). The claimthat nonallcgorical description was possible--a position underlyinghoth positivist literalisrn an d realist synecdoche (the organic, func-tional, or "typical" relationship of parts to wholes)-was closely alliedto th e romantic search for unmediated meaning in the event.. P ositiv-ism, realism, an d romanticism-nineteenth-century ingredients oftwentieth-century anthropology--all njectcd the "false" artifice ofrhetol'ic along- with allegory's supposed

.. _!lS t)ot 11 of em .. ei ty (Ong197 1 : deductive, too much an open imposition ofmeaning on sensible evidence. The recent "revival" of rhetoric bya diverse group of literary an d cultural theorists (Roland Barthes,Kenneth Burke, Gerard GeneHe, Iv'lichel de Certeau, Hayden 'White,Paul De Man, an d Michel among others) has thrO\vn seriousdoubt on the positivist-romantic-realist consensus. In ethnographythe current. turn to rhetoric coincides with a period of political an depistemological reevaluation in 'which the cOllstructed, imposed na-ture of representational authority has become unusually visible an dcontesled. Allegory prompts us to say of any cultural description......-- ." "this is a (nlorallylhat.."2 ..

2. An "allegorical anthropology" is suggested fairly explicitly in recent works byBoon (1977, 1982). Crapanzano (1980), Taussig (1981), an d Tyler (1

On Ethnographic Allegory >01The specifIC accounts contained in ethnographies can never be

lirnited to a of scientihc description so long as th e guiding taskof th e work is to make the (often strange) behavior of a different wayof life humanly To say that exotic behavior an d sym-bols make sense either in "human" or "cultural" terlllS is to supply t.hesame sorts of allegorical added meanings that appeal' in older narra-tives that saw actions as "spiritually" signifIcant. Culluralist an d hu -manist allegories stand behind th e controlled fictions of differencean d similitude that we call ethnographic accounts. What is maintainedin t.hese texts is a double attention to the descriptive surbee an d tomore abstract, comparative, an d explanatory levels of meaning. Thistwofold structure is se t ou t by Coleridge in a classic definition.\Ve ma y then safely define allegorical writing as the employment or on e sel ofagents and images with actions and accompaniments correspondent, so as toconvey, while in disguise, either moral qualities or conceptions of the mindthat are not in themselves of the senses, or other images, agents, for-t.unes, ;-llld cin:umstances so that the difference is everywhere present ed to theeye or imagination, while the likeness is suggested to the mind; and this con .nectedly, so that t.he parts combine to form a consistent whole. (1 \J3(): 3C)\-\'hat on e sees in a coherent et.hnographic account, the imagedstruct of the ot.her, is connected in a continuous double struct.ure withwhat on e understands. At times, th e structure is too blatant: "Duringthe ceramic manuLteturing process, \vomen converse gently, quietly,always wilhout conflict, about ecosystem dynamics ." (vVhitlCll

847). Usually it is less obvious and thus more realistic. AdaptingColeridge's formula, \vItat appears descriptively to the senses (and pri-marily, as he suggests, to the observing eye) seelllS to be "other," whilewhat is suggested by the coherent series of perceptions is anlying similitude. Strange behavior is portrayed as meaningful within acommon network of symbols-a common ground of understandableactivity valid for both observer an d observed, an d by irnplication forall human grou ps. Thus ethnography'S narraYive of specific differencespresupposes, an d always refers to, all abstract plane of similarity.

It is worth noting, though I canIlot pursue the theme here, t.hatbefore the emergence of secular anthropology as a science of humanan d cu.ltu.ral phenomena, ethnographic accounts were connected todifferent allegorical referents. Father Lafitau's famous comparison

of Native American customs with those of the ancienl Hebrev,'san d Egyptians exempli ties an earlier tendency 1.0 map descriptions ofth e other onto conceptions of the "jJremiers temps." More or less explicitbiblical or classical allegories abound ill t.he early descriptions of theNew vVorld. For as Johannes Fahian (1983) argues, there has been apervasive tendency to prebgure others ill a temporally distinct, but. 10-


4/13


5/13

JAMES CLIFFORDlogue). Nisa is the pseudonym of a fifty-year-old woman who llas livedmost of he r life in semi-nomadic conditions. Maljoric Shostak belongsto a Harvard-based research group that has studied the !Kung Sanhunter-gatherers since the 1950s. Th e complex truths that emergefrom this "life an d "vords" ar e not limited to all individual or to he rsurrounding cultural world.

The book's lhree registers ar e in crucial respects discrepallt. First,th e autobiography, cross-checked against other 1Kung women's lives, isinserted within an ongoing cultural interpretation (to which it adds"depth"). Second, this shaped experience SOOIl becomes a story of"women's" existence, a story that rhymes closely with maIlY of the ex-periences an d issues highlighted in recent feminist thought. Third,Niwz narrates an intercultural encounter in \vhich two individuals col-laborate to produce a specific domain of truth. The ethnographic en -counter itself becomes, here, the of the book, a fable of com-municalion, rapport, and, finally, a kind of fictional, bu t potent,kinship. NLw, is thus manifestly an allegory of scientific compre-hension, operating at the levels both of cultural description an d of asearch fo r human origins. (Alollg with other students of gatherer-hunters, the Harvard project-Shostak i l lcluded-tend to see in t.hislongest stage of human cultural development a baseline fo r humannat ure.) j\hsa is a Western feminist allegory, part of th e reinvention ofthe general categoq..' "woman" in the 1 an d 80S . j\ / isa is an allegoryof ethnography, of contact an d comprehension.

A hraided narrative, the book moves constalltly, at times awk-wardly, be!:\veen its three meaningful registers. Nisa is like manyworks that portray common human experiences, conflicts, joys, work,an d so on. Bu t t.he text Shostak has made is original in the way it re -fuses to blend its three registers into a seamless, "full" representa-tion. They remaiu separate, in dramatic tension. This polyvocality isappropriate to the book's predicament, that of lIlallY self-consciousethnographic writers who kind it difIicult to speak of well-defined"others" [rom a stable, distanced position. Difference invades the text;it can no IOllger be represented; it must be enacted.Nisa's first register, that of cultural science, holds its in hr mrelation to a social world. I t explains Nisa's personality in terms oflKung ways, an d it uses he r experiellce to nuance an d correct gener-alizations about he r group. If Ni.\a reveals mechanismsin unusual depth, its polyvocal constructioll shows, laO, that the tran-sition to scientific knovvledge is no t smooth. Th e personal does no tyield to the general without loss. Shostak's research was based on sys-tematic interviews with more than a score of !Kung women. Fromthese conversations she amassed a body of data large enough to reveal

OI l Ethnographic AHegorytypical attitudes, activities, and experiences. Bu! Shostak was dissatis-Jleel by the lack of depth in he r interviews, an d this led he r to seek ou tan informant able to provide a detailed personal narrative. Nisa "vasquite unusual in he r ability to recall and explain her life; moreoverthere developed a strong resonance between he r stories and Shostak'spersonal concerns. This posed a problem for the expectations of ageneralizing social science.

At t.he en d of he r first sojourn in th e field, Shostak was troubledby a suspicion that he r interlocutor might be too idiosyncnHic. Nis;lha d known severe pain; he r life as she recalled it \vas often violent.Most previous accounts of the! Kung, like Elizabeth Marshall Thomas'sThe llarrnless People (1959), had shown them to be peace-loving. "Did Ireally want to be the one to balance the picture?" (:350). On a returntrip to the Kalahari, Shost.ak found reassurance. Though Nisa stillexerted a special fascination, she now appeared less unusual. And theethnographer became "more sure than cver that ou r work togethercould an d should move forward. Th e int.erviews I was conductingwith other women werc proving to me that Nisa \vas fundamentallysimilar to those around her. She \vas unusually articulate, an d she ha dsulfered greater than average loss, but. in most other important re -spects she was a typical !Kung ,voman" (358).Roland Barthes (1981) has written poignantly of an impossiblescience of t.he individuaL An insistent tug toward the general is feltthroughout Nisfl, an d it is no t without paill that we find Nisa geller-alized, tied to "a n interpretation of lKung life" (350). The book's sci-entific discourse, tirelessly contextual, typifying, is braided throughthe other two voices, introducing each of the fIfteen thematic sectionsof the life with a few pages of background. ("Once a marriage has sur-vived a fev,' years beyond the young wife's first menstruation, th e rela-tionship between th e spouses becomes more equal" l1691. An d soforth.) Indeed, on e sometimes feels that the scientiIlc discourse func-tions ill th e text as a kind of brake on the book's other voices, vvhosemeanings ar e excessively personal and intersubjective. There is a realdiscrepancy. For at the same time that Nisa's story contributes to bettergeneralizations about the !Kung, its very specificity, an d the particularcircumstances of its making, create meanings that. ar e resistant to thedemands of a t.ypifying science.

Th e hook's second and third registers ar e sharply distinct fromthe first. Their structure is dialogical, an d ,if times each seems to existprimarily in response to the other. Nisa's life has its own textual auton-omy, as a distinct narrative spoken in characteristic, believable tones.Bu t it is manifestly the product of a collaboration. This is particularlytrue of its overall shape, a full lifespan-hfteen chapters including


6/13

1 06 JAMES CUFFORD

"Earliest Memories," "Family Life," "Discovering Sex," "Trialriages," "Marriage," "Motherhood and Loss," "Women an d Men,""Taking Lovers," "A Healing Ritual," "Growing Older." Although atthe start of th e interviews Nisa had mapped out her life, sketching t.hemain areas to be covered, th e thematic roster appears to be Shostak's.Indeed, by casting Kisa's discourse ill the shape of a "life," Shostak ad-dresses two rat.her different audiences. On on e side, this intenselypersonal collection of memories is made suitable fo r scient.ific typifica-tion as a "life-history" or "life-cycle." On th e other, Nisa's life bringsinto playa pOlent an d pervasive mechanism for t.he production ofmeaning in th e West- the exemplary, coherent self (or rat.her, th e selfpulling itself together in autobiography). There is nothing universalor natural about the fictional processes of biography an d autobiogra-ph y (Gusdorf 1956; Olney Lejeune 1975). Living does no teasily organize itself into a continuous narrative. When Nisa says, asshe often does, "We lived in that place, eating things. Then we left an dwent somewhere else," or simply, "we lived an d lived" (69), th e hu m ofunmarked, impersonal existence can be heard. From this blurredbackground, a narrative shape emerges in th e occasion of speaking,simultaneously to oneself an d another. Nisa tells he r life, a processtext.ually dramatized in Shostak's book.

As alter ego, provoker, an d editor of the discourse, Shostak makesa number of significant interventions. A good deal of cutting an d re -arranging t.ransforms overlapping stories into "a life" that does notrepeat itself unduly and thal develops by recognizable steps an d pas-sages. Nisa's distinct. voice emerges. But Shost.ak has syst.ematically re -moved he r ow n int.erventions (though they can often be sensed inNisa's response). She has also taken ou t a variet.y of narrative markers:he r friend's habit.ual comment at the en d of a story, "t.he wind hastaken that away," or at. the start, "I will break open the story an d tellyou what's t.here"; or in the middle, "\Vhat am I trying to do ? Here Iam sitting, talking about one story, an d another runs right into myhead and into my thoughtsl" (40). Shostak has dearly thought. care-fully about th e framing of he r transcripts, and one cannot have every-th ing- the performance with all its divagations, an d also an easily un -derstandable st.ory. IfNisa's words were to be widely read, concessionsha d t.o be made to th e requirements of biographical allegory, to areadership practiced in the ethical interpretation of selves. By theseformal means the book's second discourse, Nisa's spoken life, is broughtdose to its readers, becoming a narration that makes eloquent "hu-tnan" sense.

The book's third distinct register is Shostak's personal account offieldwork. "Teach me what it is to be a lKung woman" was the question

011 Ethnographic Allegory 10 7she asked of her informants (319). If Nisa responded vvith peculiaraptness, he r words also seemed to answer another question, "What isit to be a woman?" Shost.ak told he r informants "that I wanted to learn\-vhat it meant to be a vvoman in their culture so I could better under-sland what it meant in my own." With Nisa, tbe relationship became,in 1Kung terms, that of an aun\. talking to a young niece, to "a girl-woman, recently married, struggling '"vith the issues of love, marriage,sexuality, work an d identity" (4.). Th e younger \VOnlall ("n iece, " some-times "daughter") is instructed by an experienced elder in the arts an dpains of womanhood. Th e transforming relationship ends with anequality in affection an d respect, an d wit.h a final word, pot.ent infemillist meaning: "sister" (371). Nisa speaks, throughout, no t as aneutral witness bu t as a person giving speciflc kinds of advice to some-on e of a part.icular age with manifest questions an d desires. Sh e is notan "informant" speaking cultuml truths, as if to everyone and no one,providing information rather dlan circumstantial responses.

In he r account, Shostak describes a search for personal knowl-edge, for something going beyond the usual ethnographic rapport.She hopes that intimacy with a woman will, somehow, enlargeor deepen he r sense of being a modern \Vestern woman. With-out. drawing explicit lessons {"rom Nisa's experience, she dramatizest.hrough he r own quest. the way a narrated life makes sense, allegori-cally,f()Yanother. Nisa's story is revealed as ajoint production, the out-come of an encount.er that. cannot be rewritten as a di-chotomy. Something more t.han explaining or representing the lifean d words of another is going more open-ended. Th ebook is part of a new interest in revaluing subject.ive (more accurat.ely,intcrsubjcct .ive) aspect.s of research. I t emerges from a crucial momentof feminist politics an d epistemology: consciousness raising and thesharing of experiences by women. A commonality is produced that,by bringing separate lives together, empovvers personal action, recog-nizes a common estate. This moment of recent feminist. consciollsnessis allegorized in Nisa's fable of its own relationality. (I n other eth-nographies, traditionally masculine stories of initiation an d penetra-tion differently stage the productive encounter of sciI' an d other.)iiShostak's explicit. feminist allegory thus reflect.s a specific moment inwhich t.he construction of "woman's" experience is given center stage.I t is a moment of continuing importance; bu t it has been challengedby recent counLercurrents within feminist theory. The assertion ofcommon female qualities mel oppressions) across racial, et.hnic, an dclass lines is newly problematic. An d in some quarters "woman" is

G. Oil ethnography as an allegory of conquest and initiation, see Clifford


7/13


8/13

1 1 0 JAMES CLIFFORDmask, or direct, mult.iple allegorical processes. For may not every ex -tended description, stylistic turn, story, or metaphor be read 1.0 meansomething else? (Need we accept the t.hree explicit. levels or allegory ina book like Ni.sa? vVhat about it.s photographs, which tell their ownstory?) Ar c not readings themselves undecidable? Critics like Dc Ma n(1.979) rigorously adopt such a position, arguing that. the choice of adominant. rhetoric, -igure, or narrative mode ill a t.ext is ahvays all im-perfect attempt to impose a reading or range of readings 011 an intcr-pret.ive process that is open-ended, a series of displaced "meanings"with no full stop. But. whereas the free play of readings may in theorybe infinite, there arc, at any hist.orical moment, a limit.ed range of ca-nonical an d emergent allegories available to t.he competent reader(t.he reader whose interpretat.ion will be deemed plausible by a specificcommunity). These structures of meaning arc historically boundedan d coercive. There is, in pract.ice, no "free play."

Within this historical predicament, the critique of stories an d pat-terns t.hat persistently inform cross-cultural accounts remains an im-portant. political as well as scient.ific task. In the remainder of this essayI explore a broad, orienting allegory (or more accurately, a pattern ofpossible allegories) t.hat has recently emerged as a contestedstructure of retrospect.ion that may be called "ethnographic pastoral."Shostak's book an d t.he Harvard hunter-gatherer studies, to t.he extentthat they engage in a search fo r fundamental, desirable human traits,are enmeshed in this structure.

In a t.renchant article, "The ese an d Abuse of Anthropology: Re-flections on Feminism an d Cross-Cultural Understanding," l'vIichelieRosaldo 11(1.1 questioned a persistent tendency to appropriate ethno-graphic clata in the form of a search for origins. Analyses of social"givens" such as gender and sexuality show an almost reflexiveneed fo r anthropological just-sa-stories. Beginning with Simone deBeauvoir's founding question, "What is woman?" scholarly discussionsmove to a diagnosis of contemporary subordination and from

t.hen on to th e queries '''Vere things always as they are today?' an d tllen'vVhen did "it" stan?'" (1980:391). Enter examples drawn front e111-llography, In a practice no t essentially different from that of HerbertSpencer, Henry Maine, Durkheim, Engels, or Freud, it is assumedthat evidence from "simple" societies \vill illuminate the originsan d Slructure of contemporary cultural patterns. Rosaldo notes thatmost scientific anthropologists have, since the early t.wentieth century,abandoned the evolutionary search for origins, but her essay suggestst.hat the reflex is pervasive and enduring. ]\''1orcovcr, even scientifIcethnographers GUHlOl fully control the meanings--readings-pro-yoked by their accounts. 'rhis is especially true or representations t.hat

OIl Ethllographic Allegory I I Ihave not historicized their portraying exo-tic societies In an"ethnographic present" (which is always, in fact, a past). This syn-chronic suspension effectively textualizes t.he other, and gives thesense of a reality !lol in temporal (lux, nol. in th e same ambiguous,moving hi,


9/13

11 2 JAMES

an d attractive anarchy or Nuer society ar e distant, irretrievable. Theyar c lost qualities, textually recovered.

This ironic appeal belongs to a broad ideological pattern that ha soriented much, perhaps l1lost:, twentieth cent.ury cross-cultural repre-sern


10/13

JA:V1F.S eLIHORDWilliams does not dismiss this structure as sil'nply nostalgic, which

it manifestly is; but rather follows ou t a very complex set of temporal,spatial, an d moral positions. He noles that pastoral frequently in-volves a critical nostalgia, a \vay (as Diamond [19741 argues for acept of the primitive) to break ,,,,ith the hegemonic, corrupt present byasserting the reality of a radical alternative. Edward Sapir's "Culture,Genuine and Spurious" (1966) recapitulates these critical pastoralvalues. And indeed every imagined authenticity presupposes, an dis produced by, a present circumstance of felt inauthenticity. Bu tWilliams's treatment suggests that such projections need not be consis-tently located in the past.; or, what amounts to the same thing, that the"genuine" clements of cultural life need not be repet.itiously encodedas fragile, threatened, an d transient. This sense of pervasive socialfragmentation, of a constant disruption of "natural" relations, is char-acteristic of a subjectivity vVilliams loosely connects \\'lth city life an d\\'ith romanticism. Th e self, cut loose from viable collective ties, is anidentity in search of wholeness, having internalized loss an d em-barked on an endless search fo r authenticity. Wholeness by definitionbecomes a thing of the past (tural, primitive, childlike) accessible onlyas a f-iction, grasped from a stance of incomplete involvement. GeorgeEliot's novels epitomize this situation of participant-observation in a"common condition. . a knowable community, belong[-ing] ideally inthe pasL" khddlemarch, for example, is a generat.ion backfrom th e time of its \\Titing to 1 R30. An d this is approximately th etemporal distance that many conventional ethnographies assumewhen they describe a passing reality, "traditional" life, in the presenttense. The fiction of a knowable community "can be recreated therefo r a ,videly ranging moral action. But the real step that has beelltaken is withdrawal from any full response to an existing society. Valueis in th e past, as a general retrospective condition, an d is in the pre-sent only as a particular an d private sensibility, th e individual moralaction" (180).

In George Eliot we can see the development of a style of sociologi-cal writing that will describe whole cultures (knowable worlds) from aspecific temporal distance an d with a presumption of their transience.This \vill be accomplished from a loving, detailed, hut ultimately dis-engaged, standpoint.. Historical worlds will be salvaged as textual fab-rications disconnectcd from ongoing lived milieux an d suitable fo rmoral, allegorical appropriation hy individual readers. In properlyethnographic pastoral this lextualizing struCI ur e is generalized beyondthe dissociations of nineteenth-century England to a wider capitalisttopography ofWestern/l1on-Western, city/country oppositions. "Pr im-itive," nonlitcrate, underdeveloped, tribal societies are constantly yield-

011 Ethnographic Allegoryin g to progress, "losing" their traditions. "In th e name of science, weanthropologists compose requiems," writes Robert i\'1urphyBut the mosl problematic, and politically charged, aspect of this "pas-toral" encodatioll is its relentless placement of others in a present-becoming-past. V\'hat "wulrl it require, fo r example, consistently toassociate the inventive, resilient, enormously varied societies of:[lvlelant'sia \vith the cultural Iuture of tlie planet.? How might ethno-graphies be differently conceived if this standpoint could be seriouslyadopted? Pastoral allegories of cultural loss and textual rescue would,in any event, have to IJe transformed.')

Pervasive assumptions about ethnography as writing would alsohave to be altered. For allegories of salvage are implied by th e verypractice of textualization that is generally assumed to be at th e core ofcultural description. vVhatever else an ethnography does, it translatesexperience into text. There are various ways of effecting this transla-tion, ways that have significant ethical an d political consequences. Onecan "write up " the results of an individual experience of research.This may generate a realistic account of the unwritten experience ofanother group or person. On e can present. this I.extualization as th eout.come of observation, of interpretation, of dialogue. On e can con-struct an ethnography composed of dialogues. On e can feature mul-tiple voices, or a single voice. On e can portray the other as a st.able,essential whole, or on e call show it to be th e product of a narrat.ive ofdiscovery, in specific historical circumstances. I have discussed someof these choices elsewhere (1983


11/13

11 6 JAMES CLIFFORDthentic), writing endures (as supplementarity an d artifice). Th e textembalms the event as it extends its "meaning." Since Socrates' refusalto write, itself po"verfully "vriaen by Plato, a profound ambivalenceto\van:l t.he passage from oral to literate has characterized Westernthinking. An d mucll of the pcnver and pathos of ethnography deri\'esfrom th e fact. that it has situated its practice "vithin this crucial transi-tion. The fieldwOI-ker presides over, an d controls in some degree, themaking of a text ou t of life. His or bel' descriptions an d interpret.a-tions become part. of the "consultable record of what man has said"(Geertz 1973: 30). Th e text is a record of something enunciat.ed,in a jJast. The structure, if not the thematic content, of pastoral isrepeated.A small parable may give a sense of why this allegory of eth-nographic l'escue an d loss has recently become less self-evident. I t is atrue parable.l() A student of African ethno-history is conducting fieldresearch in Gabon. He is concerned with th e M pong\.ve, a coastalgroup who, in the nineteent.h century, were active in contacts with Eu-ropean traders and colonist.s. T"he "tribe" still exists, in the regionof Libreville, an d t.he et.hno-historian has arranged to interview thecurrent chief about traditional life, religious rit.ual, an dso on . In preparation fo r his interview the researcher consult.s a COlTl-pendium of local cust.om compiled in the carly twentiet.h cent.uryby a Gabonese Christian an d pioneering ethnographer, theRaponda-Walker. Before meeting with the r"Ipongwe chief t.he eth-nographer copies ou t a list. of religious terms, inst.itutions an dcept:;, recorded and defined by \V"alker. Th e interview willfollow this list, checking whether the customs persist, an d if so, withwhat. innovat.i ons. At first things go smoothly, wit.h the I\'lpongwethoritY'providing descriptions an d interpret.ations of the terms sug-gested, or else noting that a practice has been abandoned. After atime, however, when the researcher asks about a particular word, th echief seems uncertain, knits his brows. 'Jusr a moment," he says

an d disappears into his house to return wit.h a copy of Ra-ponda-\V"alker's compendium. For the rest of the interview t.he booklies open 011 his lap.

Versions of this story, in increasing numbers, are to be heard int.he folklore of ethnography. Suddenly cull ural data cease to movesmoothly from oral performance into descriptive writing. Now dataalso move from text to text, inscription becomes transcription. Uothinformant. and researcher are readers an d re-writers of a cultural in-vention. This is no t to say, as some might., that the interview lias ended

to . My thanks LO lIenry Bucher for (his (rut; sLory. I have told it. as a parable, bothbecause it is one, an d because I suspect he would tell it somewhat dilferendy, havingbeen there

On Ethnographic Allegory 11 7in a st.erile short circuit. No r need one, like Socrates in the Phaedrll.s,lament t.he erosion of memory by literacy. Th e interview has Hot, sud-denly, become "inauthentic," the data merely imposed. Rather, wItaton e must. reckon with are new conditions of ethnograpllic prociuctiOlI.First, it is no longer possible to act as if the outside researcher is t.hesole, or primary, bringer of the culture into writing. This has, in fact,seldom heen t.he case. Ho\-vever, there has been a consistent tendencyamong ficlclworkers to hide, discredit, or marginalize prior \vritten ac-counts (by missionaries, travelers, administrators, local authorities,evell ot.her ethnographers). Th e field worker, typically, starts fromscratch, from a research eXj)(;rience, rather than from reading or tran-scribing. Th e held is no t conceived o/" as already hlled witll texts. Yetthis intert.extual predicament. is more and rnore the case (l.arcom1983). Second, "informants" increasingly read an d write. They inter-pret prior versions of their cult.ure, as well as those being byethnographic scholars. \Vork with process of inscription,rewriting, an d so no longer (if it. ever was) t.he exclusive do-main of outside authorities. "Nonliterate" cultures are already tex-tualized; there are fevv, if any, "virgin" lifeways to be violated an d pre-served by writing. Third, a very widespread, empowering distinctionhas been eroded: the division of the globe into literate an d nonlit.eratepeoples. This distinction is no longer "widely accurate, as l1on-vVestern,"t.ribal" peoples become increasingly literate. But furthermore, onceon e begins t.o doubt the ethnographer's monopoly on the power t.o in-scribe, on e begins to see the "v.niting" activities that have always beenpursued by native collaborators-from an Amhrym islander's sketch(in a famous gesture) of an intricat.e kinship system In the sand forA. B. Deacon to the Sioux George Sword's book-length cultural de-scription found in the papers of James \;\Talker. (See the Introductionof this volume, p. 15.)

Bu t the most subversive challenge to the allegory of textualizationI have been discussing here is found in the work of Derrida (1974),Perhaps the most. enduring cifect of his revival of "grammatology"has been t.o expand what \-vas conventionally thought of as writing. AL-jJitaoelic writing, he argues, is a rest.rictive definition that. tics the broadrange of marks, spatial articulat.ions, gest.ures, anel other inscriptionsat. work in human cultures too closely to the represent.ation of speech,the orallaural word. In opposing logocentric representation to icri-lu.re, he radically extends the definition of the "writ.ten," in effect.smudging its clear distinction from the "spoken." There is no needhere to pursue in detail a disorienting t.hat is LJY now wellknown. What matters for ethnography is the claim that aLl humangroups they articulate, classify, possess an "oral-literature,"or inscribe t.heir world in ritual acts. They repeat.edly "tcxtualize"


12/13

CUFFORD

mcamngs. Thus, in Dcrrida's epistemology, the writing of ethnogra-phy cannot be seen as a drastically new form of cultural inscription,as an exterior imposition on a "pure," unwritten oral/aural uni-versc. The logos is no t primary and the gm'mme its mererepresentation.

Seen il l this light, t.he processes of ethnographic ""Tiling appearmore complex. If, as Derrida "vould say, the cultures studied by an-thropologists are always already \\Titing themselves, the special statusof the ficld\vorker-scllOlar who "brings the culture into writ.ing" isundercut. \Vho, in fact, writes a myth that is recited into a t.ape re -corder, or copied down to become part of held not.es? vVho writes (ina sense going beyond t.ranscription) an interpret at.ion of cuslom pro-duced through iruense conversations with kll(ndedgeable native col-laborators? I have argued that such questions can, and should, gener-ale a rethinking of ethnographic authority (Clifford 198:1'1)' In thepresent. context 1 want merely to underline the pervasive challenge,both historical an d theoretical in origin, that presently confronts th eallegory of ethnographic practice as textualization.

It is important to keep th e allegorical dimensions in rnind. For int.he \V'est the passage from oral to literate is a potent. recurring story-or power, corrupt.ion, an d loss. It replicates (and to an extent pro-duces) the st.ruct.ure of pastoral that has been pervasive in Iwenrieth-century ethnography. Logocentric writing is conventionally conceived10 be a rejJresentation of authentic speech. Pre-literate (the phrase con-tains a story) societies are oral societies; \vril.ing comes to them from"outside," an intrusion from a wider world. 'Whether brought by lllis-sinnary, trader, or ethnographer, writillg is both empowering (a nec-essary, eiTect.ive way of sloring and manipulating knowledge) an dcorrupting (a loss of' immediacy, of t.he face-to-face communicationSocrat.es cherished, of [he presence an d intimacy of speech). A com-plex an d fertile recent debate has circled around the valorization, his-torical signihcance, an d epistemological stal us of writing. I vVhatevermayor may no t have been settled in the debate, there is no doubt ofwhat has become unsettled: the sharp distillction of the world's cul-tures into literate an d pre-literat.e; t.he not.ion t.heH et.hnographic tex-tualizatioll is a process t.hat enacts a fUlld,lmental transition from oralexperience {O written representation; the assumption that somethingessential is lost when a culture becomes "etllllognlphic"; the strangelyambivalent. authority of a pract.ice that salvages as text a cultural lifebecoming past.

component.s of what I have called ethnographic pastoral no11. Th e "dcbate" (cuters Oil Lile confrontation of Ollg (1967,19'77. 1982) an d

Derrida (1973, ]9(74). Tyler (1978, 1981b) tries to work past the opposition. Goody(1977) an d EisenMein (19'79) haH' made importallt recent contribulions.

On Ethnographic Allegory 11 9longer appear as common sense. Reading an d writing af C general-ized. If the ethnographer reads culture over the native's shoulder, :h enative also reads over the ethnographer'S shoulder as be or she wntescach cultural description. Fieldworkers arc increasingly constrainedin what they publish by the reactions of those previously classified asnonlilerale. Novels by a Samoan (Alfred vVendt) can challenge theportrait of his people by a distinguished anthropologist. Th e notionlhat writing is a corruption, that something irret.rievably pure is lostV\dlen a cultural world is textualized is, after Derrida, seen to be a per-vasive, contestable, vVestern allegory. Walter Gug and others haveshown that something is, indeed, lost with the generalization of writ-ing. But authentic culture is not that something-to be gat.hered up inits fragile, final truth by an ethnographer or by anyone else.

rvlodern allegory, vValter Benjamin (1977) tells us, is based on asense of the world as transient an d fragment.ary. "History" is graspedas a process, no t of inventive life, but of "irresistihle decay." Th ematerial analogue of allegory is thus the "ruin" (178), an always-disappearing structure that invites imagintltive reconstruction. Ben-jamin oGserves that "appreciation of the transience of things, an d theconcern to redeem them for eternity, is one of the strongest impulsesin allegory" (quoted by \Volin 1982: 71). rvIy account of ethnographicpastoral suggests that. this "impulse" is to be resisted, not by abandoll-in g allegory-an impossible hy opening ourselves toerH hist.ories.

Allegories are secuIed .. . by teaching /leo/Jie to read in ceJ"tain ways.TAI.AL ASAD (comment on this essay al th e Santa Fe seminar)

I have explored some importallt allegorical forms that ex-press "cosmological" patterns of order and disorder, fables of per-sonal (gendered) identity, and politicized models of temporality. Th efut.ure of t.hese forms is uncertain; they ar e iJeing rewritten an d criti-cized in current practice. A few conclusions, or at least assertions, maybe drawn from t.his exploration.

There is no way definitely, surgically, to separate the betual fromthe allegorical in cultural accounts. Th e data ofmake sense only within patterned arrangements an d narratives,an d these ar e conventional, political, an d meaningful in a morethan referential sense. Cultural facts arc not true an d cultural al-legories false. [n the human sciences the relation of fact to alle-gory is a domain of struggle and institutional discipline.


13/13

12 0 JAMES CLIFFORD

Th e meanings of an ethnographic account ar e uncontrollable.Neither an author's intention, nor disciplillary t.raining, no r therules of genre can limit the readings of a text that \,vill emerge\vith new historical, scientific, or political But jf ethnog-raphies arc susceptible to multiple interpretations, these ar e no!at any given moment inf-lnite, or merely (in th e pe-jorative sense). Reading is indetermillate only to the extent thathistory itself is open-ended. If there is a common resistance to therecognition of allegory, a fcar that it le::tds to a nihilism of reading,this is no t a realistic kar . It. confuses contest.s fo r meaning v./ithdisorder. An d often it reflects a 'wish to preserve anrhetoric, refusing to locate its o\vn mode of production wit.hin in-ventive cultun: an d historical change.A recognition of allegory inescapably poses the political an d ethi-cal dimensions of ethnographic writing. It suggests that these bemanifest.ed, not hidden. In this light, the open allegorizing of aMead or a BCllcdict enacts a certain probity-properly exposingitself to th e accusatioll of havillg used u'ibal societ.ies /(:)r pedagogi-cal purposes. (Let t.hose free of such purposes cast th e first stone!)One need not, of course, purvey heavy-handed "messages," ort\vist cultural facts (as presently kllmvn) to a political purpose. [would suggest as a model of allegorical tact Marcel Mauss's TheGift. No on e would dell)' its scientific importance or scholarly com-mitment. Yet from the outset, an d especially in its concludingchapter, the work's aim is patent: "to draw conclusions of a moralnature about some of th e problems confronting us in our presenteconomic crisis" (1967: 2). TIlt' book was written in response toth e breakdown of European reciprocity in 'World vVal' L The trou-bling proximity it shmvs betvl'Tcn exchange an d warfare, th e im-age of the round table evoked al. I.he end, these an d other urgentresonances mark the work as a socialist-hulllanist allegorydressed to th e political world ofthe twenties... This is not the work'soIlly "cOllLcnt." Th e many rereadings The Gift has generated tes-(ify to its productivity as a text. I t call even be cert.aingraduate seminars'-

CLIFFORD, J. and G. MARCUS.- Writing Culture. the Poetics and Politics of Ethnografy (Section) - Fix

Documents

Transcript of CLIFFORD, J. and G. MARCUS.- Writing Culture. the Poetics and Politics of Ethnografy (Section) - Fix