'Indian Gift' Author(S): Jonathan Parry Source: Man, New Series, Vol. 21, No. 3 (Sep., 1986), Pp

Home , Generalized exchange, Raymond Firth, The Gift (essay)

The Gift, the Indian Gift and the 'Indian Gift' Author(s): Jonathan Parry Source: Man, New Series, Vol. 21, No. 3 (Sep., 1986), pp. 453-473 Published by: Royal Anthropological Institute of Great Britain and Ireland Stable URL: http://www.jstor.org/stable/2803096 Accessed: 08-04-2015 07:23 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp

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 [email protected].

Royal Anthropological Institute of Great Britain and Ireland is collaborating with JSTOR to digitize, preserve and extend access to Man. http://www.jstor.org

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions THE GIFT, THE INDIAN GIFT AND THE 'INDIAN GIFT'*

JONATHAN PARRY London School ofEconomics & Political Science

This articlecriticises much of the conventionalexegesis of Mauss's celebratedEssai surle don, and proposes a ratherdifferent reading of the textwhich stressesits evolutionaryaspects. The Hindu 'law ofthe gift' is shownto have a keyrole in thestructure of Mauss's argument,though in factit is quite inconsistentwith his centralthesis. In thisparticular instance he was rightwhere anthropologistshave generallythought him wrong, and wrong where anthropologistshave generallythought him right.In the Maori case, however,his interpretationhas much more to recommendit than has generallybeen recognised.Hindu and Maori ideologies of exchange representfundamentally opposed types;and it is suggestedthat we mightbegin to accountfor this kind of contrastin termsof broad differencesin politico-economy,and-more especially-in terms of the contrastbetween a World Religion and the kind of religion characteristicof small-scaletribal society. Following Mauss, an ideology of the 'pure' giftis shown to be inseparablefrom the ideology of thepurely interested individual pursuit of utility,and to emerge in parallelto it.

SirRaymond Firth recalls that on hisway to Tikopia he had to relyfor transport and hospitalityon theMelanesian Mission, and was forsome weeks theguest of theBishop on theMission yacht.

As we travelledtogether among the islands we discussedmany problems of human relationship in the islandcommunities. Malinowski had only recentlypublished his book Crimeand Customin SavageSociety in whichhe stressedthe importance of reciprocity as a forceof binding obligation in Melanesian social organization.. . . The Bishop borrowed the book fromme, read it, and stronglydisagreed. He arguedvehemently that Melanesians, like other people he said,performed manyacts forothers freely and withoutthought of return. . . and he deniedthe implication of self-interestedaction. . . . We arguedamicably about thisand otherthemes. . . . At lastthe time came forhim to land me on thebeach of Tikopia and leave me to myfate. He had shownme many kindnesses,which I could notrepay . . . he was retiringfrom the Mission after many years and we both knew thatit was unlikelythat we shouldever meet again. . . . As he said goodbye,leaving me alone in this remote communityhe shook me firmlyby the hand, said gruffly'No reciprocity!',turned his back and walkedoff down thebeach to theboat (FirthI973: 400-I).

This vignetteof the knightand the bishop, sailingover a distanttropical sea engaged in earnestdebate over the thesisof an expatriatePolish Professor, providesme withan aptintroduction to my centraltheme-for I wish to speak of ideologiesof reciprocity and non-reciprocity,of the 'Indian gift' and theIndian gift.

*The MalinowskiMemorial Lecture, I985

Man (N.S.) 21, 453-73

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions 454 JONATHAN PARRY

In Argonautsof the Western Pacific (I922: I76-9I), Malinowski rangestrans- actions along a continuumfrom 'pure gifts'to 'real barter'.The concept of reciprocityis central;but it has not hardenedinto the dogma it subsequently becamein Crimeand custom in savage society (I926), wherethe notion of'pure gift' is retracted.Taking a sufficientlylong-term view, we shallfind that even with regardto transactionsbetween a Trobriandman and his wifeand children'the mutual services balance' (I926: 4I) for 'keen self-interestand a watchful reckoning. . . runsright through' (p. 27). Ratherthan being slave to custom, the 'savage' is as cannyas the 'civilisedbusinessman' and has quite as sharpan eye forthe main chance. He cares more, it is true,for prestige than material pay-offs;and though he is certainlynot Economic Man, he is nonetheless MaximisingMan. Obligationsare kept because 'the chain of reciprocal gifts and counter-gifts. . . (benefit)both sides equally' (p. 40); and because the costsof reneguingon themare too greatin termsof 'self-interest,ambition and vanity' (p. 67). Supernaturalsanctions are eitherabsent altogether(p. 5I), or are relativelyeasily evaded by meansof counter-magic(p. 8o). The various elementsin this model-the tendencyto see exchanges as essentiallydyadic transactions between self-interested individuals, and as premissed on some kindof balance;the tendency to play down supernaturalsanctions, and the total contemptfor questions of origin-all theseconstituted an important legacy of Malinowski's teaching,and directlyor indirectlyexercised a major influenceover much of the subsequent literature.In Firth's writings,for example,we notonly find all Malinowski'smistrust of the sanctioning power of 'reconditebeliefs' (Firth I929: 4I 5), butalso a similaremphasis on theindividual choice-makingactor (e.g. FirthI 967). The same influencecan also be seen in Leach's criticismof Levi-Strauss's (I969) thesisthat ranked regimes of generalised exchange are inherently unstable since theyinhibit a closureof thematrimonial cycle. Although the wife-givers may not receive womenin exchange, says Leach, the rule of reciprocityis nonethelessmaintained since they are compensatedby all kinds of counter- prestationsof a differentnature; and itis forthis reason that the Kachin system is not in factsubject to the kind of instabilityenvisaged by Levi-Strauss(Leach I96I: 90). So while Levi-Strauss'smodel of generalisedexchange is based on a systemof indirectreciprocity and-relative to systemsof restrictedexchange -presupposes an expansionof trustand credit,Leach's alternativeinvites us to view thesituation in termsof an endlesssequence of dyadic exchanges which are in thelong termbalanced. Instead of a speculativeventure, exchange is a quidpro quobased on a certaintyof returns. Similarassumptions are builtinto Blau's (I967) discussionof exchangeand power, and Weiner's (I976, I980) criticismsof Sahlins. Both operatewith a similarpremiss of balanceso thatwhen-in Blau's case-the exchangeitself is notbalanced, the deficitis compensatedfor by an incrementin power to the creditorwhich restores the equilibrium. Both assumethat exchange takes place betweencalculating individuals, so thatWeiner can claim thatwith what she calls the'gift myth', the anthropologist is merely'perpetuating and creatingan imageof"the primitive" as a person,or "primitivesociety" as a way oflife, that has survivedon some fundamentalprinciple other than self-interest' (I976: 22I).

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions JONATHAN PARRY 455 It would be tedious, but not difficult,to multiplyexamples. The general messagewould be thesame. The giftis alwaysan 'Indiangift'-that is, one 'for whichan equivalentreturn is expected'1-and thenotion of a 'puregift' is mere ideologicalobfuscation which masks the supposedly non-ideological verity that nobodydoes anythingfor nothing. So it is thatanthropology often seems to be endlesslyrediscovering the moral of Mandeville's Fable of thebees. Publick Benefitderives from Private Vice. Societyis createdby, and itscohesion results from, an endless sequence of exchanges in which all pursues their own advantage(however conceived). All this may be obvious. But what is perhapsless so is thatthis habit of thoughthas distortedour reading of Mauss's essay on The gift.Though Malinowskiand Mauss are commonlytwinned as thejoint progenitorsof the anthropologicalunderstanding of exchange,it is as well to rememberthat the Durkheimianswere one of Malinowski's main polemicaltargets in Crimeand custom.Yet paradoxicallyMalinowskian premisses are only too oftenread into theMaussian text, which is unconsciouslyprocessed through a theoreticalfilter borrowedfrom his distinguishedcontemporary.

Mauss'sgift Mauss's essay has acquiredfor anthropology many of the qualitiesof a sacred text.It is treatedwith reverential awe, thegreater part of its teaching is ignored, and it is claimedas thefonset origo of quitedivergent theoretical positions. It has been cited as a forerunnerof Barth's transactionalism(Kapferer I976: 3); as demonstratingan underlyingcontinuity between gift and commodityexchange (FirthI973: 370) andas demonstratingthe opposite (Gregory I983: i8 sq.). It has beenfound to containan implicitevolutionary model 'remarkablyparallel to Marx's argumentin Grundrisse'(Hart I983: 46), whileLevi-Strauss-modestly avoiding claims to the mantleofJoshua-likens Mauss to Moses leading his people into thePromised Land of Structuralism,though never quite makingit himself(Levi-Strauss I973: 37). Our undergraduatestudents, routinely ex- pectedto masterthis text in theirfirst year, might reasonably be forgivenfor wonderingjust how theyought to understandit. So ellipticalis his writingthat Mauss himselfdoes not always seem to be on theirside. Nor is the translation,which is both carelesslyinaccurate and an unconsciousmirror of the prejudices of its own period2.Mauss saysat theoutset thathe is concerned'with words and theirmeanings', and thathe has chosento concentrateonly on areas 'where we have access to the mindsof the societies throughdocumentation and philological research' (I966: 2-3; I973: I49). The translation,however, has littlepatience with these preoccupations. The textual notesare often elided or evensuppressed. But whatis morerelevant here is thatI believe that Cunnison's text both reflects,and has helped to perpetuate,a 'Malinowskian'reading of Mauss's original.Let me give youjust one example, taken from the very firstpage. Mauss writes of prestationsas having a 'voluntarycharacter, so to speak, apparentlyfree and without cost, and yet constrainedand interested. . . They are endowed nearlyalways withthe form

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions 456 JONATHAN PARRY of a present,of a gift generouslyoffered even whenin the gesturewhich accompanies the transactionthere is only a fiction,formalism and social deception,and whenthere is, at bottom,obligation and economicinterest'. In Cunnison's version what is voluntary'and yet constrainedand interested' becomes-in themanner of Malinowski-a disinterestedtheory contradicted by an interestedpractice; 'economic interest' becomes 'economicself-interest'; and 'even when' thegesture of generosityis onlya fictionis turnedinto an assertion 3 thatit is onlya fiction In fact, of course, Mauss repeatedlystresses a combinationof interest and disinterest,of freedomand constraint,in the gift.Nor could 'interest' possiblybe a matterof self-(in the sense of individual)interest. It is not individuals but groups or moralpersons who carryon exchanges. The individuals of modern society are endowed with interestsas against the world. The personswho enterinto the exchangeswhich centrallyconcern Mauss do so as incumbentsof statuspositions and do not act on theirown behalf(cf. Ekeh I974: 32). Nor do personsstand in oppositionto thethings exchanged. The giftcontains some partof thespiritual essence of thedonor, and thisconstrains the recipient to make a return-an argumentwhich has been thesource of some embarrass- ment to Mauss's admirers,who have tendedto dismissit as peripheral.The Maori notionof hau is generallytaken as the exemplary,or even as the only, instanceof this'spirit of thegift'. Sahlins (I972), forexample, suggests that in effectMauss has two differentanswers to the question:'why are giftsrecipro- cated?' The first-which is wrong-is the Maori hau 'raisedto the statusof a generalexplanation'. The second-which is right-is the Hobbesian State of Warre-the giftbeing theprimitive analogue of the social contract.Similarly Levi-Strausssees the discussionof the hau as a regrettableinstance of the anthropologistallowing himself to be mystifiedby thenative, whose culturally specificrationalisations cannot possibly explain a generalstructural principle. Happily,however, the argument only appears at thebeginning of the work, and is merelya pointof departuresuperseded by the end (Levi-StraussI973: xlvi). Firth(I929) comes to an oppositeconclusion. Mauss's argumentis not a Maori rationalisationbut a Frenchone-in supportof which we could point,I think,to itssimilarity to Levy-Bruhl'snotion of 'participations'(cf. MacCormack I982). I suspect,however-though I cannotprove-that thereal sourceof theidea is neitherMaori nor French,but Indian-for as we shall see the basic notionis clearlyarticulated in theHindu textson which Mauss had workedintensively long beforeLevy-Bruhl first published his ideas. As against Firth and Sahlins I will argue at a later stage that Mauss's interpretationof theMaori data has moreto recommendit thanthey allow. As againstSahlins and Levi-StraussI shouldlike at thisstage to suggestthat Mauss does not in factadvance a culturallyspecific ideology as a generalexplanation; noris hispoint peripheral and limitedto theopening sections of the essay. More or less the same argumentrecurs in relationto the pledge which must be exchangedbetween the parties to a contractin thelaw of theancient Germanic tribes.The indebtedparty is constrainedto make a returnsince he has handed over as pledgean objectwhich is imbuedwith his own personality,and which

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions JONATHAN PARRY 457

thereforeputs him quite literally in thehands of his creditor.By redeemingit he redeemshimself (I966: 6o-i; I973: 253-5). The generalprinciple-of whichthis and theMaori hauare only two amongst a whole batteryof illustrations-isthe absence of any absolute disjunction betweenpersons and things4.It is becausethe thing contains the person that the donorretains a lienon whathe has givenaway and we cannottherefore speak of an alienationof property;and it is because of thisparticipation of theperson in theobject that the gift creates an enduringbond betweenpersons. Sahlins thus misrepresentsMauss whenhe suggeststhat the argument about 'thespirit of the gift'is independentof the argumentabout the giftas social contract.The gift onlysucceeds in suppressingthe Warre of all againstall becauseit createsspiritual bonds betweenpersons by means of thingswhich embody persons.The two aspectsare inseparable;and if theywere not it would be hard to understand muchof Mauss's antipathyto themodern market. Whatis also strikingabout Sahlins'scommentary on theessay is thathe never actuallymentions its centralpurpose: to constructa kind of prehistoryof our modernkind of legal and economiccontract. Evolutionary speculations are at the heartof the enterprise,which displaysthat genetic concern for the originsof legal formswhich was the dominantcharacteristic of Durkheimianstudies of law (Vogt i983: 3 I). It is thisconcern with the originsof the moderncontract whichexplains why of the three obligations Mauss isolates,it is theobligation to make a returnwhich attractsthe greatestattention. Cases in which the giftis not reciprocatedare virtuallyexcluded fromMauss's purview by the way in which he has definedhis problemin termsof the archaeologyof contractual obligation. The broad outline of the evolution which Mauss traces is from 'total prestations'consisting of an exchangebetween groups in whichmaterial goods are only one item amongsta whole range of non-economictransfers, to gift exchange between persons as representativeof groups, to modern market exchangebetween individuals. The last of thesehas evolved fromthe first by a gradualprocess of attenuationor contraction.Exchanges between groups which had an aesthetic,religious, moral, legal and economicaspect have been stripped downto leave purelyeconomic exchanges between individuals5. (The objects of exchangethemselves undergo a parallelevolution. Ceremonial valuables of the kind representedby kula armshellsbecome detachedfrom the group and the person,and develop into the kind of depersonalisedmoney foundin modern economies[Mauss I966: 93-4; I973: I78-9]). Seen in thislight Sahlins's analogy with Hobbes looks problematic.Hobbes, who startswith the individual, was concernedwith the creation of a widerunity out ofan 'originally'atomised state of humanity; but Mauss, who startswith the group, has reversedthe sequence-from an originalholism, humanityand humaninstitutions have becomeatomised. What is morethe two essaysin Stone Age economicswhich immediately precede Sahlins's homage to Mauss ironically reveal a strikingdivergence-in a Hobbesian direction-fromMauss's argument. The giftrepeatedly stresses that thereis no such thing as a 'natural economy'where production isfor use and such exchangeas occursis of utilities. But Sahlins's Domestic Mode of Productionis surelyonly a variantof the

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions 458 JONATHAN PARRY natural economy model6, and it locates the originsof exchange largely in the utilitarianneeds of the proportionof households which inevitablyface subsistencefailure7. To riska differentanalogy, there is-I would argue-a morethan superficial convergencebetween the evolutionaryschemes of Marx and Mauss (Cf. Hart I983); for much in Mauss's essay recallsthe progressiverupture Marx dis- coveredbetween man and the materialworld, man and his products,and man and man-this resultingin a fragmentedworld in which the relationsof the partsto thewhole can no longerbe discerned,and leavingthe person as a mere rump,an 'abstraction'(cf. Ollman I976: I33f). If theMaussian thesisis thatthe modern contract is theenduring remnant of archaicgift exchange, what then are moderngifts? Two readingsof the text seem possible.The conventionalone would stressa basic continuitybetween gifts in modernand pre-modernsociety. But what I understandto be the dominant proposition-which seemsto have been completelyoverlooked-is thatin our kindof society gifts come to representsomething entirely different. Gift-exchange -in which personsand things,interest and disinterestare merged-has been fractured,leaving giftsopposed to exchange, persons opposedto thingsand interestto disinterest.The ideology of a disinterestedgift emerges in parallel with an ideology of a purelyinterested exchange. So, for example, at the beginningof his chapteron archaicsocieties, Mauss writesexplicitly of a clear distinctionbetween obligatory prestations and (pure) giftsin our own cultural heritage8,and asks rhetoricallywhether such distinctions'are . . . not of relativelyrecent appearance in the codes of the greatcivilizations?' [I966: 46; I973: 229). Itis we,he sayselsewhere (I966: 46; I973: 229), whohave opposed 'theideas of thegift and disinterestedness'to 'thatof interestand theindividual pursuitof utility';and it is because the latterhave now become the guiding principlesof economiclife that Mauss wistfullylooks back on a primitivepast whereinterest and disinterestare combined. The whole ideologyof the gift,and converselythe whole idea of 'economic self-interest',are ourinvention; and the textexplicitly acknowledges the diffi- cultyofusing these terms for societies such as theTrobriands whereprestations-the word itselfmust have been chosenfor its connotations of constraint-area kind of hybridbetween gifts, loans and pledges. The Malinowskiof Argonautswas certainlyin errorto suggestthat what is givenby a fatherto hischildren is a 'pure gift'.But as thecontext makes entirely clear, Mauss's realpurpose here is not to suggestthat there is no suchthing as a puregift in anysociety, but rather to show thatfor many the issue simplycannot arise since they do not make thekinds of distinctionthat we make. So while Mauss is generallyrepresented as tellingus how infactthe gift is neverfree, what I thinkhe is reallytelling us is how wehave acquireda theorythat it shouldbe. The interestedexchange and thedisinterested gift thus emerge as two sidesof the same coin. Given a profounddislike of the first,mistrust of the second is onlylogical. The unreciprocatedgift debases the recipient, and thecharity of the 'rich almoner' is condemned (I966: 63; I973: 258)-presumably because it deniesobligation and replacesthe reciprocal interdependence on whichsociety is foundedwith an asymmetricaldependence. The remedyfor our modernills is

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions JONATHAN PARRY 459

a systemof social security founded on theold moralityof gift-exchange, to which we too are heirs. The beginningsof the ideological revolutionwhich destroyedthis ancient wisdom are located in the late Roman Empire with the legal separationof personsfrom things. But the main thrustof the discussionon Rome is that this distinction-which is centralto our concepts of propertyand market exchange-evolved out of earlierconcepts strictly comparable to those of the gifteconomies described for the 'primitive' world. The record,however, does notallow Mauss to establishthis as anythingmore than a 'likelyhypothesis'; and itis herethat India comes to hisaid as showingthat Indo-European law once had gift-exchangeinstitutions like those of the Pacificand America. In a manner which is thoroughlynineteenth century in spirit,and which like much in the essay is stronglyreminiscent of Maine9, India standsin-as Trautmann(n. d) notes-for Europe's missingpast. Though Maine and Mauss did not agreeon whetherits innovations stood formoral progress, for both it was Rome which carriedthe torch of history(as again it did in Mauss's essay on theperson) and Indiawhich revealed the fossil record of Indo-European law. It is notperhaps so hardto see why Mauss's theorieshave been subjectedto selectiveprofessional amnesia.

The Indiangift The centralthesis of the essay about the evolutionaryorigins of the modern contractthus hinges on showingthat gift-exchange in theHindu textsconforms to the model constructedfor pre-literatesocieties. Mauss focuses here on danadharma,the 'law of (religious)gifts', and in askinghow well his model in fact appliesI will followhim in this(without, of course,wishing to implythat this law covers all-or even a majority-of transactions).Like Mauss, then,my main concernis with those giftswhich rate as dana; and I shall referonly in passingto othercategories of 'gift', and in orderto contrastthem with dana. Lest the restrictionappear to confineme to the esoteric,let me remindyou of the enormouspolitico-economic significance of religious gift-giving in thesocieties which are heirsto thislaw: one has only to thinkof the huge landed estates donatedto thebig SouthIndian temples; of thefact that at thebeginning of the nineteenthcentury an estimatedI7 to 20 per cent.of thepopulation of Benares (thenone ofthe largest cities in India) wereBrahmans living off religious charity (Bayly I983: I26); or of estimatesof 25 to 40 per cent. of net disposable cash incomebeing given to monksin thetypical Upper Burmesevillage (Spiro I970: 459). Mauss's sourcesprovided him with many obvious illustrationsof theway in whichthe gift embodies the person. The donor of cattlesleeps in thebyre, eats barleyand cow-dung,and at themoment of transfer proclaims: 'what you are,I am; todayI am become youressence, and givingyou I give myself'(I966: 57; I973: 248). More specificallythe gift is held to embodythe sins of thedonor10, whom it rids of evil by transferringthe dangerousand demeaningburden of deathand impurityto therecipient (Heesterman I 964). Nor is itwithout peril to

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions 460 JONATHAN PARRY thedonor, for it bindshim dangerously close to one who mayprove unworthy. 'A Brahmanawho neitherperforms austerities nor studies the Veda, yetdelights in acccptinggifts, sinks with the (donorinto hell). .' (Manu 4: I90). The meritof thegift is thuscontingent on thatof a worthyrecipient. And who shouldthis be but theone who is mostunwilling to receiveit? The contemporarysignificance of thistheory has recentlybeen documented by severalethnographers. I have described,for example, how the Brahman priestsof Benaressee themselvesas endlesslyaccumulating the sin theyaccept with the giftsof the pilgrimsand mournerswho visitthe city,and how they liken themselvesto a sewer throughwhich the moral filthof theirpatrons is passed. Theoreticallythey should be able to 'digest' the sin by dintof various ritualprocedures of expiation,and by donatingthe gifts they receive to another Brahmanwith increment.But quite apartfrom the factthat this is plainlyan economic impossibility,thcy sadly admitignorance of the correctritual procedures.The sewerbecomes a cess-pit,with the result that the priest contracts leprosy and rots; he dies a terribleand prematuredeath and then faces the tormentsof hell. The donortoo is an endangeredbeing, for if the priest misuses his giftfor some evil purposehe sharesin the sin. He musttherefore give to a 'worthyvessel'; but the one who is preparedto accept his giftsis almost by definitionunworthy to receivethem (Parry I980; n.d.). Some of myIndianist colleagues will be as used to myrecital of thesedata as I to theirresponse. Such ideas, theyclaim, are ofpurely local distributionand do not reflectwidespread popular attitudesin ruralareas away fromthe major centresof Sanskriticculture, or existmuch outsidethe Hindi-speaking region (cf. Fuller I984: 67 sq.). In fact,however, thereis evidenceof very similar notions fromTravancore, Tamilnadu and Gujerat1";and thanksto Raheja's (I985; n.d.) excellentethnography there can now be littledoubt thatthey are pervasivein some ruralareas of thenorth. Raheja describestwo broad categoriesof prestationas ideologicallycentral in thevillage where she lived. The firstof theseis characterisedby reciprocityand an ideology of mutuality.By contrastwith this is the category of gifts genericallyknown as dana-these constitutingthe most importantfeature of most ritualsand festivals,being given almost daily, and utilisingenormous materialresources. Such gifts'send away' inauspiciousnessfrom the donor to therecipient, who maybe a Brahman,Barber, Sweeper or a wife-takingaffine, to whom thegift will bringmisfortune unless the correct ritual precautions are taken. Given thatthese data derivefrom a regionon which many anthropol- ogistshad previouslywritten, I stronglysuspect that such ideas aboutdana have a farwider distribution than has so farbeen recognised; and thatit is notbecause theyare absentthat they have not been more widely reportedin the village ethnography,but ratherbecause the fieldworkwas conducted with other preoccupationsin mind, and by a generationof ethnographersblinded by a deep-rootedprejudice that the spiritof the giftwas merelyMaussian meta- physicalmystification. In theHindu contextthis notion that the gift contains the person is associated withthe idea thatthe gift is a kindof sacrifice. It is in facta surrogatefor sacrifice appropriateto our degenerateage (Manu i: 86; Biardeau I976: 27). In termsof

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions JONATHAN PARRY 46I theirsymbolism and structurethere are many obvious parallelsbetween the two procedures12.But what is mostrelevant here is thatthe identification between the sacrifierand the victim-which in the classical theoryis explicitlya substitutefor his own person-is carriedover into the theoryof dana as an identificationbetween the donor and his gift.As thevictim is a surrogatefor the sacrifier,so thegift is a surrogatefor the donor. It stands,moreover for what he mustexpiate. Now it is when the objectiveof the sacrificeis to eliminatebad sacrednessthat Hubert and Mauss (I964: 55) predictthat the identification with the victim will be closest beforethe immolation,and that theirsubsequent separationwill be as finalas possible. Consistentwith this, the identification of the donor with his giftis oftenhighly elaborated before the transfer(as for example when he is weighed against a valuable substancewhich he then donates),and his separationfrom it afterwardsis absolute. Thereis no question,then, of thegift being a loan or pledge. It is alienatedin an absolute way, and the very definitionof the giftis that it involves the completeextinction of the donor'sproprietary rights in favourof the recipient (AiyarI94I: 77; Law I926: i). The giftthreatens to cementthe two togetherin a dangerousinterdependence; but everyattempt is made to severtheir bond by insistingon thecomplete alienation of thething. Under no circumstances,and on painof terrible supernatural penalties, is thegift resumed. Its evil 'spirit'must notcome back. While Mauss originallyintroduced this notion of 'spirit' to explainthe inalienability of the object and thenecessity of making a return,what it in factexplains in thiscontext is why the giftmust be alienated,should never return,and shouldendlessly be handedon. The obligationto make a returnis nottherefore encoded in the danadharma. Mauss (I966: I23; I973: 243) himselfwas uneasyhere and concededin a footnote that'on the obligationto make returngifts-our main subject-there are few facts except perhaps Manu VIII, 2I3. The clearestrule consists in a rule forbiddingthe return of gifts'.In trutheven theverse cited demands the most wilfullycock-eyed reading to make it say anythingabout reciprocity.'The clearestrule' is in realityunqualified. A pure asymmetrymust obtain. The donorshould seek out thereluctant recipient and givefreely, for the genuine gift is neversolicited. No returnof any earthlykind is countenancedand even an incrementto theprestige of thedonor weakens the gift, which should therefore be made in secret.It is as if-to paraphraseTrautmann (I98I: 28I-2)-the ancientPandits had arrivedat themodern theory of reciprocity, didn't like what theyfound, and smartlyturned heel. The patternof affinal relations amongst the high castes in contemporarynorth India clearlyreflects this ideology. Kanyadana, the 'giftof a virgin'along with her dowry, is merelythe beginningof an endless series of giftswhich flow unilaterallyfrom wife-givers to wife-receivers.Not even a glass of watermay be acceptedin a village to which one of the daughtersof the lineage has been givenin marriage;and suchprohibitions may even extendto thosewho rateas wife-takersto one's own wife-takers(Ibbetson quoted in Lewis I958: I88-9; ParryI979: 304-5). The chainis conceptuallynever closed-as is illustratedby Vatuk's (I969) analysisof the Hindi kinshipterminology. Levi-Strauss (I969: 398-9), of course, argues that this theory of marriageby gift is merely

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions 462 JONATHAN PARRY generalisedexchange in an illusoryguise. In the hypergamousvariant of the system,however, female infanticide was in thepast widespread, and thehighest rankinglineages receivedbrides but gave to none"3. The very existenceof hypergamy,of course, immediately suggests a reciprocalexchange in whichthe wifc-giversin factgain an incrementin status.In thccase of manynorth Indian castes,however, there is no hypergamyin thesense of a systematicranking of descentlines, and any inferiorityon thepart of thebride's family is createdby themarriage itself (e.g. Vatuk I 975). Clearlyin such circumstancesthe endless streamof giftscannot be a quidpro quo foran elevationin rankand thereis no questionof a dyadicexchange. A rangeof quite differentexamples of thissame denialthat the gift sets up an obligationto makea returncould be given.But whatthey would all show is that thetheory is-as Trautmann(I98I: 279) putsit-a 'soteriology,not a sociology of reciprocity . '14. The giftdoes indeedreturn to thedonor, but it does so as the fruitsof karma.It is this 'unseen fruit'(adrstaphala) which witherson the branchif any returnis accruedin thehere and now (cf.Aiyar I94I). The return is deferred(in all likelihoodto anotherexistence); its mechanismhas become entirelyimpersonal, and therecipient is merelya 'vessel' (patra)or conduitfor theflow of merit and is himselfin no way constrainedby thegift or bound to the donor. Even a spiritualaccounting is sometimeslooked upon with suspicion, and so thebest gifts are givenmerely from a detachedsense of dutyand without thinkingof them as giftsat all (Kane I974: 8: 42; Mahabharata I3: 49: 3). Whetherwe emphasisethe impersonality of thereturn, or theideology which deniesthat a 'true'gift is made 'withdesire' for any kind of reward, it seems clear thatwe aredealing with a transactionaltheory quite unlike Mauss's Melanesian, Polynesianand Americanexamples. The Hindu 'Law ofthe gift' does notcreate societyby institutingthat constant give-and-take which Malinowski described forthe Trobriands. The Trobriandgift may be an 'Indian gift',but theHindu giftis not. In passing we mightnote that much the same themerecurs in Theravada Buddhism,where indeed we findthe gift without a recipientat all. Offeringsare regularlymade to theBuddha, buthe has attainednirvana, and no longerexists. It is the gods who governthe pragmaticaffairs of the world, and with them significantlythe ideology is one of reciprocity-offeringsfor boons. The reciprocatedgift belongs to theprofane world; the unreciprocated gift to a quest forsalvation from it (Ames I966). Giftsmade to thosewho carrythe soterio- logical message of the Buddha-as theyat any rateinsist-are neverrecipro- cated. Out of compassionthe monk merelyprovides a 'fieldof merit'for the laity;but he is not the donor of the meritacquired throughthe offeringshe receives(Strenski I983). Nor is an obligation to receive the giftentirely clear. According to the well-knowntextual formula it is in anycase onlythe Brahmans who have sucha duty-and this is paradoxicallyevaded by the best of them. The Hindu ascetic-unlike theBuddhist monk-is certainlyunder no suchobligation. Nor is a willingnessto receivegifts entirely consistent with the honour of themartial Kshatriya,who must never be a supplicant.Hence forhim-as Hara (I974) shows-the mostappropriate form of marriagewhich the texts can envisageis

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions JONATHAN PARRY 463 marriageby captureand the open exerciseof violence (cf. Trautman I98i: 283). The king'sduty is to makegifts and he shouldfund his generosity-as the sacrifice(Heesterman 1959)-through forceand valour. Hence the king is in manysymbolic ways oftenidentified with the bandit (Shulman I980; cf. Dirks I982). Danadharmathus poses a numberof difficultiesfor our generaltheories of exchange.It consciouslyrepudiates Gouldner's (I960) universal'moral norm of reciprocity'.Nor is it clearthat the unreciprocated gift produces the differentia- tionin power predictedby Blau (I967)-for in northIndia wife-givingaffines are commonlyrequired to put up with the most peremptoryand disdainful treatmentat thehands of thoseto whom theyact as perpetualdonors. Withthe hypergamousvariant of this systemit seems thatHindu ideology has even succeededin periodicallyexcluding segments of north Indian society from what Levi-Strauss(I969: 143) calls 'the universalformof marriage'-one based on reciprocity.Nor does Sahlins's(1972: I85 sq.) typologyof exchange,in which the 'solidaryextreme' of generalisedreciprocity is seen as coincidingwith the closest social relations,seem wholly applicable. Here the most unbalanced exchangesare representedby the giftsmade to thewandering ascetic or to the priestof a farawaypilgrimage centre; while transactions between, for example, a fatherand son areoften talked about in theidiom of a quidproquo in whichthe son finallysettles his debts throughthe performanceof his father'smortuary rites(Parry I985). As for Mauss, I thinkthat with regard to the kinds of giftscovered by danadharma,that part of his thesiswhich anthropologists have generallyfound most problematic(the spiritof the gift)is-with the qualificationsI have registered-in fact the most acceptable;while that part of his thesiswhich anthropologistshave acceptedmost readily (the obligations to receiveand make a return)is actuallythe most problematic.Now it is, of course, true that danadharmadoes not embracethe whole rangeof transactionswhich we would rateas gifts.As was indicatedearlier in connexionwith Raheja's material,there are manyother kinds of prestation-mostof which are explicitly reciprocal (cf. Parryn. d.). Here, however,there is littleindication that the giftcontains the donor-except in the loose sense thatit may be interpretedas an objective manifestationof his subjectivedispositions (or perhapsas materialtestimony to theskills with which his casteis innatelyendowed). Wherewe have the'spirit', reciprocityis denied;where there is reciprocitythere is not much evidenceof 'spirit'.The two aspectsof themodel do not hangtogether.

The 'Indiangift' I have shown, then,that the Hindu law of thegift does not displayconvincing evidence of continuitywith the exchange systems Mauss described for Melanesiaand Polynesia.But how well do thesecases themselves conform to his model?Better, I think,than is sometimesclaimed. It was Mauss who originallynoted theproblem with Malinowski's classifi- cation of the presentsmade withinthe Trobriand domestic group as 'pure

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions 464 JONATHAN PARRY gifts'. The Trobriandersthemselves describe them as mapula,a termwhich Malinowski translatedas 'payment'or 'equivalent',and theyclearly state that thepresents which a man makesto his wifeand herchildren are a returnfor her sexual and domesticservices. I am aware thatWeiner (I980) has deniedthat this is the exchange,and has questionedMalinowski's gloss formapula. What she shows neverthelessis thatsuch transfersare visualisedas partof a long-term cycle of reciprocaland ultimatelybalanced exchanges. Gifts of descent-group propertyto outsiderswill eventuallybe reclaimed,and therebyreturn-as Mauss said of theMaori gift-to theirancestral hearth. In view ofJohansen's(1954: II8) expertconclusion that 'a certainuncertainty' precludes'actual certainty'on thematter of theMaori hau,and of Gathercole's warning(I978) thatall we reallyhave to go on is 'thedetritus of an (anachronis- tic) discoursebetween various Victoriangentlemen', one might thinkthat Ranapiri's famous partingshot should be taken seriously-'enough on that subject'.I thereforecomment on theMaori case withall due trepidation. The crucialevidence is containedin Ranapiri'sexplanation of a sacrificial offeringof birds to theforest. This is elucidatedby analogywith a giftexchange in whichA givesa valuableto B who passesit on to C. WhenC makesa returnto B, he mustgive it to A becauseit is thehau of thefirst gift, and he will become sick or die ifhe retainsit (Mauss I966: 8-9; 1973: 158-9; Sahlins1972: 152). To Mauss it seemed thatthe giftitself is animatedwith the spiritof its original homelandand donor,to whom it strivesto return. This interpretationwas severelycriticised by Firth,who confirmedthe Maori preoccupationwith reciprocity, but arguedthat the real sanctionsbehind it are the threatof witchcraftand the economicand social costs of defaultingon the exchange. The hau is not a 'purposiveentity'; nor can the hau of thingsbe identifiedwith that of persons(1929: 413). These criticismshave been generally accepted (e.g. Johansen1954: II7; Forge 1972: 529; Parkin 1976: 171)15. In contestingthem I pay tributeto Firth'sown ethnography,and acknowledgethe centralimportance of thesanctions he outlined. In thePolynesian context Mauss makesit clearthat the kinds of thingswhich embody personsbelong to the categoryof valuables known as taonga,which constitutethe sacra of the family.The supreme example of taongaare the 'treasureitems', or 'heirloomvaluables', of the kinshipgroup-with whose land and historythey are identified,and into whose genealogiesthey enter. Johansen(1954: 104) reportsthat one is relatedto themas a kinsman,and that theyare greetedand honouredlike chiefs.Such valuableswere exchangedas giftsbetween groups, and were used to concludea peace treaty,for the mana inherentwithin them has thecapacity to createa strongbond betweenpeople. These heirloomsare closelyassociated with rights to land (Firth1929: 348), and the land is conceived to be the source of a Maori's spiritualwell-being and identity(Hanson & Hanson I983: 65). It thereforeseems clear that in giftingsuch valuablesa Maori was in an importantsense gifting an aspectof his personhood. This identificationbetween persons and thingsis again illustratedby thebelief thatthe possessions of a man ofrank-or indeedanything he calledby hisname or referredto as a partof his body-were permeatedby his sacredpower, and thereforedangerous to others(Firth 1929: 336). Given all thisI cannotsee how

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions JONATHAN PARRY 465 therecan be any clear-cutdistinction between the hau of personsand things. Perhaps theirblending together reflects not so much the metaphysicalpre- occupationsof theCollege de France,as theirseparation reflects the pragmatic rationalismof theLSE. At leastin thecase of heirloomvaluables the idea of the thingitself striving to returnto its homelandseems perfectlyconsistent with Maori representationsof theseobjects. But thereis also, I think,a more importantkind of connexion between reciprocityand thehau. At thehighest level of abstraction,Sahlins glosses the hau as a 'generalprinciple of productiveness',which in specificcontexts might be renderedas 'yield', 'returnon' or even 'profit'.Thus thehau of a good is its 'yield', and when Ranapirisays thatB must returnthe second valuable to A because it is the hau of the firstvaluable, what he is reallyenunciating is the preceptthat 'one man'sgift should not be anotherman's capital'. 'We have', says Sahlins(I972: I62), 'to deal witha societyin whichfreedom to gainat another's expenseis notenvisioned by therelations and formsof exchange'.The hauthus representsa kindof pre-emptiveideological strike against market principles. It is herethat I findSahlins's argument weakest. Ranapiri said nothingwhatever about an equivalencein exchange.On thecontrary, a 'profit'is clearlyimplied by the veryinstance he was tryingto explain. The huntersreturn some of the birdsto theforest, but keep therest. As Sahlinssees it, the great advantage of his analysis over previous ones is that it enables him to explainwhy Ranapiriinvoked a three-partyexchange-the thirdactor being logically necessary to illustratethe principle of a yieldor profit on the originalgift. Sahlins analyses accounts of threedifferent exchanges in which thehau figures,and claimsthat all reveala similartriadic structure. But whata close readingof theseexamples actually reveals is thatit is onlyby sleight of handthat two of themcan be representedin thisway. The onlyone whichin factdisplays this structure is thatcontained in Ranapiri'stext, and thisis easily accountedfor in otherterms'6. What was beingexplained after all was a sacrifice whichrequired the mediationof priests. It was, in other words, a three-party transactionbetween hunters,priests and forest;and the economic analogy thereforehad to be triadic. What is,in my view, valuablein Sahlins'sanalysis is his stresson thehau as a generalprinciple of productiveness;and thisinsight can-I think-be taken further.What Ranapiriis actuallytrying to tell us, I would argue, is thatthe sourceof well-beingand productivenessis reciprocalexchange-and it is this principlethat fecundity and increasestem from reciprocity which he emphasises by introducingthe hau into his discussionat the stage of the return.That exchange itselfis fertileand promotesincrease must have appeared as a self- evidenttruth to theMaori, sincethe gift normally attracts an increment,and as it circulatesit grows (Firth1929: 416; Hanson & Hanson I983: i io). By returning A's gift,B 'nourishesthe hau' and therebyensures-contra Sahlins-a future 'profit'or 'yield'. By failingto do so, by 'avertingthe hau', B would destroythe sourceof his own productivenessand vitality,and would thereforesuccumb to witchcraftor otherwisesicken and die. Again experienceproves the theory, for one who defaultson his exchangeswill cease to be an acceptablepartner, and will be excludedfrom this apparently magical source of growthand productiv-

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions 466 JONATHAN PARRY

ity. The giftof a Maori valuable, I conclude, does embody the person; and what Firth dismisses as 'recondite beliefs' about the hau do sanction a return. By contrast with the 'spirit' of the Hindu giftwhich brings destitutionand death to one who fails to pass it on along a chain which is conceptually never closed, the 'spirit' of the Maori giftentails like consequences for one who fails to return it to the original donor.

The ideologyof the 'puregift' It is in any case obvious that both Trobriand and Maori exchange reveal a preoccupation with reciprocityas a norm of social conduct which contrastswith its denial in the Hindu law. In what kinds of social system, then, do we find values of this latterkind, and why? Since the notion of reciprocityhas been used so uncriticallythat it is oftenunclear whether what is being described is a matter of empirical fact, indigenous theory or anthropological assumption about the nature of human behaviour (MacCormack 1976), an answer to these questions can only be tentative. In somecontext reciprocityis surely a normative expectation in every society; and I thinkit probable thatthe vast majority also make some place forthe notion of a free gift. A return, argued Simmel (1950: 392), is always ethically con- strained; but the very firstgift which initiates a relationship has (or better, is often seen as having17) a voluntary and spontaneous character which no subsequent gift can possess, and for this reason it can never be entirely reciprocated. Schwimmer (1973) describes a Melanesian society in which every social relationship is ideologically premissed on exchange. Yet how could exchange ever begin in the firstplace? Only, the Orokaiva myths tell us, by an original freegift of the primal ancestors18.Empirically, then, it is not a question of either an ideology of reciprocity or of its repudiation, but rather of a significantdifference in the extent to which these possibilities are elaborated. The premium placed on reciprocityin Melanesian societies is so strikingthat Levi-Strauss (1973: 33), commenting on the convergence between Mauss and Malinowski, was led to wonder whether it was not the Melanesians themselves who were the true authors of the theory. It was the societies of Melanesia and Polynesia which firstattracted explanation in these terms, and the principle does not have the same prominence in the Africanistliterature (MacCormack 1976). For 'traditional' African societies, however, I can find little evidence of any elaborated ideology of the 'pure gift'. It is surely rather the ancient literate civilisations of Europe and Asia which have stressed this notion. Mauss-as I have shown-provides some preliminary hints as to why this might be. Those who make freeand unconstrained contractsin the market also make freeand unconstrained giftsoutside it. But these giftsare defined as what market relations are not-altruistic, moral and loaded with emotion. As the economy becomes progressively disembedded from society, as economic relations become increasinglydifferentiated from other types of social relationship, the transactions appropriate to each become ever more polarised in terms of theirsymbolism and ideology. We might thereforeargue thatan ideology of the

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions JONATHAN PARRY 467

'pure gift' is most likely to arise in highly differentiatedsocieties with an advanced division of labour-such an ideology being a logical end-product of the kind of sequence Levi-Strauss (I969) traces from restrictedto generalised to complex systems of reciprocity,where each step implies a greaterindirectness of returns and an expansion of the social universe. Again, in an economy with a sizeable market sector gift-exchange does not have the material significance it has for the many tribal societies in which it provides the only access to crucial scarce resources. Gifts can therefore be given with the sole objective of cementing social relations and without any insistence on an equivalent return (cf. Schwimmer 1973: 49). Moreover if-as Mauss argues-gifts are the primitive analogue of the social contract, then they clearly carry a social load which in centralised politics is assumed by the state. In other words, gifts can assume a much more voluntaristic character as their political functions are progressively taken over by state institutions. I am suggesting, then, that an elaborated ideology of the 'pure' giftis most likely to develop in state societies with an advanced division of labour and a significant commercial sector. But what is also in my view essential to its articulationis a specifictype of beliefsystem, as is suggested by the factthat in all of the major world religions great stress is laid on the merit of giftsand alms, ideally given in secrecy and without expectation of any worldly return. There are, as Obeyesekere (I968; I980) has argued, certain fundamental differencesbetween these historical world religions and the religions character- istic of small-scale tribal society. Though the idea of salvation-defined as 'a state or condition from which sufferinghas been eliminated'-is found in both contexts, the majority of tribalreligions are not salvation religions; and in those which are, 'compensation for suffering is meted out in the other world irrespectiveof the actor's behaviour in this world. Ethical considerations do not influencethe topography of the other world. . . . The kingdom of heaven is for saint and sinner alike' (I968: 12, 14). In such societies social behaviour tends to be sanctioned largely by secular rather than religious morality; 'there is no systematic attempt to incorporate the secular moral code into a religious one' and hence no thoroughgoing 'ethicization' (I980: 154). Where religious norms are violated, supernatural sanctions tend to be immediate ratherthan saved up for the after-life'9.In the world religions, by contrast, social behaviour is sys- tematically ethicised. This 'implies the religious evaluation of moral action, actions thatare morally good or bad are . . . also religiously good or bad' (I980: 147). The consequence is an elaboration of the concepts of sin (in the sense of 'a violation of the religious ethics of morality' [I968: 14]) and religious merit. These determine the individual's ultimate destiny on the principle of contin- gency of supernatural reward, and entail the bifurcation of the other world- hell forsinners and heaven forsaints. In the South Asian context, Obeyesekere's contrast is clearly illustrated by the comparisons which Furer-Haimendorf (I967; 1974) and Bailey (I98I) have drawn between tribal religions and the Indian world religions. An ethicised salvation religion, in which rewards are contingent on conduct, is clearly likely to have the effectof orienting the ideal goals of social action towards a futureexistence. Those whose horizons are limited by the rewards of

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions 468 JONATHAN PARRY this world will not gatherthe 'unseen fruits'of the next; and the ethic of intentionrequires that the expectation of a returnin thehere and now shouldbe denied. Moreover,the notion of salvationitself devalues this profane world of suffering.The unreciprocatedgift becomes a liberationfrom bondage to it, a denialof theprofane self, an atonementfor sin, and hencea meansto salvation. The more radicalthe oppositionbetween this world and a world freefrom sufferingto come, themore inevitable is thedevelopment of a contemptusmundi whichculminates in theinstitution of renunciation,but of whichthe charitable gift-as a kind of lay exercisein asceticism-is also oftenan expression.In abandoningits millenarianexpectations in favourof an eschatologyof heaven and hell,the early Christian Church widened the chasm between this world and the Kingdom of God, and therebyenormously boosted both the spiritof asceticismand a preoccupationwith charitable'good works' (Troeltsch193 I: 113). It is of coursethe case thatthe common fate of such 'free'gifts is to become a purchaseprice of salvation,resulting in the kind of actuarialcalculation representedby the Merit Books of the pious Buddhist(e.g. Spiro 1970), or the arithmeticalrelationship between alms-offeringsand the eliminationof sin establishedby Cyprian(Westermark I906: 555). But thissame ideology may react againstsuch reckoningin the name of an ideal of purelydisinterested action. I do not, of course, deny the importantdifferences between the World Religions. I have spoken of an other-worldlyorientation, but in the case of certainProtestant sects (and of some brandsof Islam) it mighthave been more accurateto speakof an emphasison creatingan imageof thatother world in this one. The effectis nonethelessto directaction towards a transcendentalideal, and to devalue theworld whichactually exists along withthe returns which can be expectedwithin it. Though thestress which Hinduism and Buddhismplace on thenotion that the spiritual worth of the gift is contingenton thatof its recipient has many parallelsin medieval Christianity(Lawrence I984), again thereis certainlya differencein emphasis.To a fargreater extent than in the Indian religions,Christianity-with its notion that all menare fashioned equally in the image of God-has developeda universalisticconception of purelydisinterested giving. But it was also, of course,the Christian world which developed the theory of pure utility,and that-as Mauss indicated-is perhapsno accident.Since the thingsof thisworld are seen as antitheticalto theperson's true self, his soul, an ethicisedsalvation religion is I thinklikely to encourage that separationof personsfrom things which is an ideologicalprecondition of marketexchange, and whichsignificantly was firsteffected in theWest by thelaws of a Christian Emperor. It is surely significantthat the pagan practice which the early missionarymonks who Christianisedthe Germanicpeoples were most con- cernedto extirpatewas the buryingof treasurewith the dead (Little1978: 5). 'The Christiansoul', as Kiernan(1978: 374) notes,'was purifiedin heavenfrom all taintof ownership',and not even theToriest theologian has thoughtto ask 'whethereach new arrivalwill be assignedhis own personalharp'. More importantlya universalisticethic of disinterestedgiving can surelyonly

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions JONATHAN PARRY 469 encouragethe creation of a separatesphere which is immunefrom the require- mentsof such a demandingprecept. The ideologyof the pure gift may thus itself promoteand entrenchthe ideological elaborationof a domain in which self- interestrules supreme. It is notI thinkcoincidental that the ideology of the 'pure gift' is accorded such prominenceamong groups-such as the Jews and Jains-which have a particularlyclose historicalassociation with market trade, for the two spheresdefine each other-sometimes less, but with us today as sharplyas ever.With renewed ideological stress on theautonomy of themarket go renewedpleas for philanthropy to assumethe responsibilities it denies.It was possiblythrough such speculations that Mauss arrivedat hisnow perhapsnot so quaint-soundingmoral conclusion-that the combinationof interestand dis- interestin exchangeis preferableto theirseparation. Even beforethe publication of Crimeand custom, Mauss had shown thatboth positionsin thatlong-running argument sparked off by Malinowski's book, adjournedon thebeach at Tikopia and endlesslyresumed in theanthropological literature,are entirelygiven by this ideological separationand belong to a discoursepeculiar to a certainkind of society.

NOTES This articlerepresents a slightlyrevised version of theoriginal lecture, during the preparation of whichMaurice Bloch, ChrisFuller, Liz Nissan andJockStirrat gave me some sound advice,many usefulsuggestions and muchmoral support. I am also particularlyindebted to Tom Trautmannfor his expertcomments on the text of the lecture,permission to cite his unpublishedarticle, the stimulusof his workand thegenerosity of his encouragement.Responsibility for the deficiencies of theend-product remains, of course,entirely my own. 1 This definitionis given both by the SOED and, in I764, by Hutchinsonin The historyof the colonyofMassachusets-Bay (cited in Hyde I979: 3). One or two nativeinformants have toldme thatin theirunderstanding the expression refers to a giftwhich is reclaimedafter it has beenmade; but here I followthe dictionary usage. 2 The deficienciesof thetranslation have been commentedon by Leach I955; SchwimmerI973: IO; van Baal I975: io; Trautmann I98I: 279 and n.d., and Fuller (in a letterto Man which substantiatedthe charge of inaccuracyin some detail,but whichwas not published). ' Mauss (I973: I47) wroteof 'le caract&revolontaire, pour ainsi dire,apparemment libre et gratuit,et cependantconstraint et interess6de ces prestations.Elles ontrevetu presque toujours la formedu pr6sent,du cadeauoffert g6nereusementmeme quand, dans ce geste qui accompagnela transaction,il n'y a que fiction, formalismeet mensongesocial, et quand il y a, au fond,obligation et inter&t6conomique'. Cunnisonby comparison,speaks of 'prestationswhich are in theoryvoluntary, disinterested and spontaneous,but are in fact obligatoryand interested.The formusually taken is thatof the giftgenerously offered; but the accompanyingbehaviour is formalpretence and social deception,while the transactionitself is based on obligationand economicself-interest' (Mauss I96I). The firstsentence quoted above was one of Fuller'sexamples of theproblems with the translation. 4 Testimonyto theimportance which Mauss attachedto thisinseparability of persons and things in primitiveand archaicsocieties is again foundin his muchlater essay on theperson (Mauss I979). But in view of theenormous significance for our modernconcepts of property and exchangewhich Thegift attributes to theireventual separation, it is perhapssurprising that in thesubsequent essay he did notreverse the perspective to explorethe consequences of thishistorical break for the concept of theperson. 5 In thecontext of sacrificeMauss (I973: I67) observesthat one of thefirst groups of beingswith

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions 470 JONATHAN PARRY whom menhad to contractwere the gods and thespirits of thedead. This mightperhaps be takento suggest thatjust as Durkheim and Mauss (1973) had discoveredthe originsof the theoretical classificationsof modernscience in religion,so the giftessay tentativelylocated the originsof modernsecular forms of contractand exchangein sacrifice. 6 It is, he says, 'cousin to Marx's "simple circulationof commodities",thus to the celebrated formulaC-I M-* C' . . . primitivepeoples remainconstant in theirpursuit of use values, related always to exchange with an interestin consumption,so to productionwith an interestin pro- visioning'(Sahlins I972: 83). 7 'Almost everyfamily living solely by its own means sooner or laterdiscovers it has not the means to live'. Unless, therefore,'the domesticeconomy is forcedbeyond itself the entire society does not survive'(Sahlins 1972: IOI, 86). 8 It is truethat the adjective'pure' is Cunnison's interpolation,but it seemsjustified in thatit drawsattention to thestrong distinction Mauss emphasisesbetween 'l'obligation et la prestationnon gratuite,d'une part,et le don, de l'autre'. ' More specifically,I have in mindhere their common interestin theevolution of the contract; and the ideas thatancient law knew 'next to nothingof individuals'(Maine I960: 152), thatthe conflationof personsand thingsis characteristicof societiesbased on status,and thattheir legal separationconstituted a majorhistorical watershed (I960: 164). 10 Mauss(1973: 249-50; I966: 125-6) acknowledgedthis hastily in a footnote,only to dismiss it as an 'absurdtheological interpretation'. Since what-as we shallsee-the notionactually explains is whythe gift should not be acceptedin thefirst place, and cannotin thesecond place be reciprocated, Mauss clearlyhad difficultyin comingto termswith it. 11 Fuller(I984: 67 sq., 196) refersto an arrayof sourceswhich clearly document the extremely wide distributionof such notions,though paradoxically his own discussionis premissedon their lack of Pan-Indian significance.On the prevalenceof these ideas in the South Indian literary tradition,see ShulmanI985. 12 At themost obvious level,both require a consecration;both transform the religious state of the donor/sacrifier,and bothconstitute a meansof communicationwith the divine via an intermediary. Given these parallelsit is indeed curious that,as Fuller (I984: I96) notes, Mauss 'failed to tie explicitlyhis analysisof giftsto his earlierwork on sacrifice'.In a paragraphwhich strikingly presagesThegift, Hubert and Mauss (I964: IOO) describedsacrifice in termsof 'disinterestedness. . . mingledwith self-interest.That is why it has frequentlybeen conceived as a formof contract'. Sacrifice,they continue, presupposes an intermediarywhich (like the gift) simultaneously unites and separatestwo opposed partieswho 'draw close to each otherwithout giving themselves entirely'. They windup, whatis more,by endorsingTylor's (I904) accountof an evolutionfrom the sacrifice madein theexpectation of a return,to an ideologyin whichit becomes an actof self-abnegation-on myreading precisely the development which Mauss tracedfor the gift (or to be moreprecise, for one aspectof thearchaic prestation). 13 It is truethat such systems are prone to thekind of instability which L6vi-Strauss predicted; but it is also the case thatthey have a tendencyto re-establishthemselves-along with the patternof matrimonialnon-reciprocity they institute (Parry 1979: 247 sq.). 14 In thearea of politicalvalues the same ideologicaldenial of reciprocityhas been documentedin Mayer's (i98i) finediscussion of theconcept of seva;and it is such ideas whichunderlie the Bhudan (or 'land gift')movement of Vinoba Bhave-whose objectivewas quite as much to provide the landedwith a routeto salvationthrough disinterested giving as it was to provideland forthe landless (Oommen I972: 35f;Gonda I965: 228). 15 Two recentdissenting voices are MacCormack (I982) and Weiner(I985). Unfortunatelythe latterhad not been publishedat thetime of writing,for it containsmuch to supportthe view that Ranapiriwas talkingabout taonga-valuables;that these valuables do indeedembody the person and thatMauss was essentiallycorrect in suggestingthat the giftis not inert.In certainother respects, however,Weiner's interesting interpretation differs from my own. 16 The othertwo exampleswere those of the cape whichBest had orderedfrom a weaver(Sahlins 1972: i6i); and the'payment' made to his teacherby the'sorcerer's apprentice' (I972: 103-5). In the firstcase theweaver refused to deal withthe trooper who triedto buy thecape, and therebypreserved theproper dyadic transaction; while the second case can onlybe made triadicby treatingthe victim as a partyto theexchange (rather than as the'thing' exchanged).

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions JONATHAN PARRY 47I

17 The qualificationis necessary,for it is surelyalso the case thatrelationships are sometimes thoughtto be initiatedout of self-interest. 18 It might,I recognise,be possibleto arguethat we shouldinterpret these myths not so muchas a statementabout theimpossibility of initiatingexchange without priming the pump witha freegift, butmore as a statementthat the only beings capable of making such a giftwere the original ancestors. Eitherway, mycentral point would stand:even in thiscase we findsome ideologicalspace (however minimal)for the notion of a freeand unconstrainedgift. 19 We are dealing,of course,with an ideal type,and Obeyesekere(i980: 153) makesit clearthat empiricallythere is no religionentirely devoid of ethicalimplications.

REFERENCES

Aiyar, K. V. Rangaswami 1941. Introductionto Krtyakalpataruof BhattaLaksmidhara, vol. 5, Danakanda.Baroda: Gaekwad's OrientalSeries 9I. Ames, M. i966. Ritualprestations and the structureof the Sinhalesepantheon. In Anthropological studiesin TheravadaBuddhism (ed.) M. Nash (SE Asia Stud. I3). New Haven: Yale Univ. Press. Bailey,F. I 98 I. Spiritualmerit and morality.In Cultureand morality: essays in honour of Christoph von Fiirer-Haimendorf(ed.)A. C. Mayer. Delhi: OxfordUniv. Press. Bayly, C. I983. Rulers,townsmen and bazaars:north Indian society in theage ofexpansion, 1770-1870. Cambridge:Univ. Press. Biardeau,M. I976. Le sacrificedans l'hindouisme.In M. Biardeau& C. Malamoud: Le sacrificedans l'Indeancienne. Paris: Presses Universitaires de France(Bibl. Ecol. hautesEtud., Sci. relig.79). Blau, P. M. I967. Exchangeand power in social life. New York:John Wiley. Dirks, N. I982. The pastsof a Palaiyakarar:the ethnohistory of a South Indianlittle king. J. Asian. Stud.6i, 655-83. Durkheim,E. & M. Mauss I963. Primitiveclassification (trans.) R. Needham. London: Cohen & West. Ekeh, P. 1974. Socialexchange theory: the two traditions. London: Heinemann. Firth,R. 1929. Primitiveeconomics of the New ZealandMaori. London: Routledge. I967. Themes in economic anthropology:a general comment. In Themesin economic anthropology(ed.) R. Firth.London: Tavistock. 1973. Symbols:public and private. London: George Allen& Unwin. Forge,A. 1972. Thegolden fleece. Man (N.S.) 7, 527-40. Fuller,C.J. I984. Servantsof the goddess: the priests of a SouthIndian temple. Cambridge: Univ. Press. Fuirer-Haimendorf,C. I 967. Moralsand merit: a studyof values and social controls in South Asian societies. London: Weidenfeldand Nicholson. 1974. The senseof sin in cross-culturalperspective. Man (N.S.) 9, 539-56. Gathercole,P. 1978. Hau,Mauri and Utu: a re-examination.Mankind II, 334-40. Gonda,J. I965. Changeand continuity in Indian religion. The Hague: Mouton. Gouldner,A. W. I960. Thenorm of reciprocity. Am. sociol. Rev. 25, I6I-78. Gregory,C. A. I983. Giftsand commodities. London: AcademicPress. Hanson, F. & L. Hanson I983. CounterpointinMaori culture. London: Routledge& Kegan Paul. Hara, M. 1974. A noteon theraksasa form of marriage.J. Am. orient.Soc. 94, 296-306. Hart,K. I983. On commoditization.In Fromcraft to industry (ed.) EstherGoody. Cambridge:Univ. Press. Heesterman,J. C. 1959. Reflectionson thesignificance of theDaksina. Indo-IranJ.3, 241-58. I964. Brahmin,ritual and renouncer.Wien. Z. KundeSiid-Ostasiens 8, 1-3 I. Hubert,H. & M. Mauss I964. Sacrifice:its nature andfunction (trans.) W. D. Halls. London: Cohen & West. Hyde, L. I979. Thegift: imagination and the erotic life ofproperty. New York: Random House. Johansen,J. P. 1954. The Maori and his religionin its non-ritualisticaspects. Copenhagen: Ejnar Munksgaard. Kane, P. V. 1974. Historyof Dharmasastra, vol. 2. Poona: BhandarkarOriental Research Institute. Kapferer,B. 1976. Introduction:transactional models reconsidered. In Transactionand meaning (ed.) B. Kapferer.Philadelphia: Institute for the Study of Human Issues.

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions 472 JONATHAN PARRY

Kiernan,V. G. I978. Privateproperty in history.In Familyand inheritance:rural society in western Europe1200-1 800 (eds)J. Goody,J. Thirsk& E. P. Thompson. Cambridge:Univ. Press. Law, B. C. I926. The law ofgift in British India. Calcutta: Eastern Law House. Lawrence,C. H. I984. Medievalmonasticism:forms ofreligious life in WesternEurope in the Middle Ages. London: Longman. Leach, E. R. I955. Review of M. Mauss Thegift(trans. I. Cunnison).Man 55, 30. I 96 I. Rethinkinganthropology. London: Athlone. Levi-Strauss,C. I969. The elementarystructures of kinship(trans.) J. Bell, J. von Sturmerand R. Needham. London: Eyre& Spottiswoode. I973. Introductiona l'oeuvre de Marcel Mauss. In M. Mauss: Sociologieet anthropologie. Paris:Presses Universitaires de France. Lewis, 0. I958. Villagelife in northern India: studies in a Delhi village.New York: Random House. Little,L. I978. Religiouspoverty and the profit economy in Medieval Europe. London: Paul Elek. MacCormack, G. I976. Reciprocity.Man (N.S.) II, 89-I03. I982. Mauss and the'spirit' of thegift. Oceania 52, 286-93. MahabharataI905. A proseEnglish translation by ManmathaNath Dutt. AnushasanaParva, vol. 13. Calcutta. Maine, H. S. I960. Ancientlaw. London:J.M. Dent. Malinowski,B. I922. Argonautsof the western Pacific: an accountof native enterprise and adventure in the archipelagoesofMelanesian New Guinea.London: Routledge& Kegan Paul. I926. Crimeand custom in savage society. London: Kegan Paul, Trench,Trubner. Manu, Thelaws of, I 886; translatedfrom Sanskrit into English by GeorgBuhler (Sacred Books ofthe East 25). Oxford:Clarendon Press. Mauss, M. I966. The gift:forms and functions of exchangein archaicsocieties (trans.) I. Cunnison. London: Cohen & West. I973. Essai sur le don:forme et raisonde 1'echangedans les societesarchaiques. In M. Mauss, Sociologieet anthropologie. Paris: Presses Universitaires de France. I 979. A categoryof the human mind, the notion of person,the notion of 'self'. In M. Mauss: Sociologyand psychology (trans.) B. Brewster.London: Routledge& Kegan Paul. Mayer, A. C. I98I. Public serviceand individualmerit in a town of centralIndia. In Cultureand morality:essays in honour of Christoph von Fiirer-Haimendorf (ed.) A. Mayer. Delhi: OxfordUniv. Press. Obeyesekere,G. I968. Theodicy, sin and salvationin a sociology of Buddhism. In Dialecticin practicalreligion (ed.) E. R. Leach. Cambridge:Univ. Press. 1980. The rebirtheschatology and its transformations:a contribution to the sociology of early Buddhism. In Karma and rebirthin classicalIndian traditions (ed.) W. D. O'Flaherty. Berkeley:Univ. of CaliforniaPress. Ollman, B. 1976. Alienation:Marx's conception ofman in capitalist society. Cambridge: Univ. Press. Oommen, T. K. 1972. Charisma,stability and change: an analysisof the Bhoodan- Gramdan movement in India.New Delhi: Thomson Press. Parkin,D. 1976. Exchangingwords. In Transactionand meaning(ed.) B. Kapferer.Philadelphia: ISHI. Parry,J. P. 1979. Casteand kinship in Kangra. London: Routledge& Kegan Paul. 1980. Ghosts,greed and sin: the occupationalidentity of the Benaresfuneral priests. Man

(N. S.) I5, 88-II I- 1985. Death and digestion:the symbolism of food and eatingin northIndian mortuary rites. Man (N.S.) 20, 612-30. n.d. The moralperils of exchangein a Hindu pilgrimagecity. In Thesymbolism ofmoney and themorality of exchange (eds) M. Bloch &J. Parry. Raheja, G. I985. Kinship,caste, and auspiciousnessin Pahansu. Thesis, Universityof Chicago, Illinois. n.d. Auspiciousnessin northIndian social life.In Beyondpurity and pollution: studies in Indic categoriesand everyday experience (ed.) M. Moffat. Sahlins,M. 1972. Stoneage economics. Chicago: Aldine. Schwimmer,E. 1973. Exchangein the social structure ofthe Orokaiva: traditional and emergent ideologies in theNorthern District of Papua. London: C. Hurst.

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions JONATHAN PARRY 473

Shulman,D. I980. On SouthIndian bandits and kings.Ind. econ.social Hist. Rev. 17, 283-306. I985. Kingshipand prestationin SouthIndian myth and epic. AsianAfr. Stud. 19, 93-117. Simmel,G. 1950. Thesociology of Georg Simmel (ed.) K. H. Wolff.Glencoe, Ill.: FreePress. Spiro, M. 1970. Buddhismand society: a greattradition and itsBurmese vicissitudes. Berkeley: Univ. of CaliforniaPress. Strenski,I. I983. On generalizedexchange and the domesticationof the sangha.Man (N.S.) i8, 463-77. Trautmann,T. I981. Dravidiankinship. Cambridge: Univ. Press. n.d. Thegift in India. Paper deliveredbefore the eleventh Annual Conference on SouthAsia, Universityof Wisconsin,November I982. Troeltsch,E. 193 I. The socialteachings of the Christian churches, vol. I. London: George Allen & Unwin. Tylor,E. B. 1904. Primitiveculture, vol. 2. London:John Murray. van Baal, J. 1975. Reciprocityand the position of women: anthropological papers. Assen: Van Gorcum. Vatuk,S. I 969. A structuralanalysis of the Hindi kinshipterminology. Contrib. Ind. Sociol (N. S.) 3, 94-115. 1975. Giftsand affinesin northIndia. Contrib.Ind. Sociol.(N.S.) 9, 155-96. Vogt, W. P. I983. Obligation and right:the Durkheimiansand the sociology of law. In The sociologicaldomain: the Durkheimiansand thefounding of Frenchsociology (ed.) P. Besnard. Cambridge:Univ. Press. Weiner,A. 1976. Womenof value, men of renown: new perspectives in Trobriandexchange. Austin: Univ. of Texas Press. I980. Reproduction:a replacementfor reciprocity. Am. Ethnol.7, 7I-85. I985. Inalienablewealth. Am. Ethnol.12, 210-27. Westermark,E. I906. The originand development ofmoral ideas, vol. i. London: MacMillan.

This content downloaded from 130.223.251.15 on Wed, 08 Apr 2015 07:23:22 UTC All use subject to JSTOR Terms and Conditions