The Frontiers of "Burma" Author(s): E. R. Leach Source: Comparative Studies in Society and History, Vol. 3, No. 1 (Oct.

, 1960), pp. 49-68 Published by: Cambridge University Press Stable URL: http://www.jstor.org/stable/177896 Accessed: 11/02/2010 20:54
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=cup. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access to Comparative Studies in Society and History.

http://www.jstor.org

THE FRONTIERS OF "BURMA"

The thesis underlying this essay may be summarized as follows: The modern but they are Europeanconceptsfrontier,state and nation are interdependent not necessarilyapplicableto all state-likepolitical organisations everywhere. In defaultof adequatedocumentary materialsmost historiansof South-East Asia have tendedto assumethat the stateswith whichthey have to deal were Nation-Statesoccupiedby named "Peoples"and separatedfrom each other by precise political frontiers. The inferencesthat have been made on the basis of these initial assumptionssometimesconflict with sociologicalcommon sense. It is not the anthropologist's task to write history,but if history is to be elaborated with the aid of inspiredguessesthen the specialknowledge of the anthropologist becomes relevantso as to point up the probabilities. What then do we mean by a frontier? In modernpolitical geographya frontieris a preciselydefinedline on the map (and on the ground)marking the exact divisionbetweentwo adjacentstates. Most such frontiers,as they exist today, are the outcome of arbitrary political decision or militaryaccident; very few correspondto any economicallysignificantfeature of the naturaltopography. Yet wars are fought to defend such frontiersand from such wars there has emergeda Europeanmyth which asserts,not only that every politicalstate must, ipso facto, have a definiteboundary,but also that the frontiersin questionought in some way to correspondwith differences of culture and language. This attitudeto frontiersties in with the dogma of sovereignty. In the ideology of modern international politics all states are sovereignand every of the earth's surface must, by logical necessity, be the rightfullegal piece possessionof one and only one such state. There are no longer any blank spaces on the map and, in theory at least, there can be no overlapbetween the territories of two adjacentstates. Whatever practicaldifficultiesthis may entail-as for examplein Antartica-, the principleis not in doubt;territorial sovereigntyis absolute and indivisible. The universality of this dogmais quite a recentdevelopment.In its present form it is a by-productof the clash of Europeanimperialistinterests. In Asia and Africa nearlyall the presentpoliticalfrontierswere first established

50

EDMUND LEACH

during the nineteenthor early twentiethcenturieseither as a compromise of EuropeanGreatPowersor else as an ad hoc betweenthe rival aspirations convenienceof some colonial invention designedto suit the administrative few in which in the cases Even genuinely non-European today agency. is often impossible. The determination of frontiers the still survive, regimes boundariesbetween Sa'udi Arabia and Trucial Oman are a case in point; the boundarybetween North-EastBurma and China is another. This last instanceis relevantto my theme. By the "Burma"of my title I wish to imply the whole of the wide impreciselydefinedfrontierregion lying between India and China and having modern political Burma at its core. In this region the indigenouspolitical systems which existed prior to the phase of Europeanpoliticalexpansionwere not separatedfrom one another by frontiersin the modernsense and they were not sovereignNation-States. is a frontierregion continuouslysubjectedto influThe whole of "Burma" ences from both India and China and so also the frontierswhich separated the petty political units within "Burma"were not clearly defined lines but zones of mutualinterest. The politicalentitiesin questionhad interpenetrating political systems, they were not separate countries inhabited by distinct populations. This concept of a frontier as a border zone through which in a dynamicmanneris not a new one' but it needs culturesinterpenetrate to be distinguishedclearly from the precise MacMahonlines of modern political geography. Existing histories of the Burma region do not interpretthe facts in this way. Insteadit is constantlyassumedthat frontiersof languagecorrespond to frontiers of culture and of political power. The populationis said to consist of a large numberof separate"peoples": Mons, Arakanese,Karens, Burmese,Kachins,Shans,Lisu and so on, each groupbeing assumedto have a separatehistory. Such peoples are never treatedas indigenousto Burma; each grouparrivedseparately from some remoteoriginalhomeby migration land. Such fables are like saying that the originalhome of Man was in the Garden of Eden. The theory that the Burmese came from Tibet is based on linguisticsimilarities betweenthe modernBurmeseand Tibetanlanguages. Similararguments that the originalhome of might be used to demonstrate the Englishwas in Italy or Persia or even Iceland. This myth of philologicalorigins, with its illusion of multiple discontinuities, has distractedthe historian'sattentionfrom those elements of the modernBurmesesocial scene which have been persistently presentthroughout the last 2000 years. In particular, the historianshave tended to neglect the continuing interactionbetween processes of political action and the permanentstructureof ecological relationships. In whatfollows I shall ignorethe problemsposed by languagedistributions
1Lattimore

(1940).

THE FRONTIERS OF "BURMA"

51

and I do so intentionallyfor I insist that very few valid inferencescan be constructedsolely on the basis of knowledgethus provided. Anyone who doubts this need only consider the relationsbetween history and the facts suggestedby a linguisticmap of contemporary Europe. Languagegroupings are of sociologicalratherthan historicalsignificance. Those who speak one mother-tongue necessarilyshare a certainsense of social solidaritywith one another,but this has no necessaryimplications for the historical antecedents of the individuals concerned. In present-day Ceylon most of the ruling elite speak English in their homes and have it would be completelyerroneousto supposethat any Portuguesesurnames; significantproportionof these people are of Europeandescent. The analysisand classification of the languagesand languagedistributions of "Burma" is an important scientificexercise,it cannotbe a contribution to history. As an alternativeI argue that the historically significant contrasts in are differencesof ecology and differencesof social orpresent-day"Burma" ganization. The two sets of differences nearlycoincide;roughlyspeakingthe Hill People are patrilinealand hierarchical, the Valley People have a nonunilineal kinship organisationlinked with charismaticdespotism. This coincidenceis not a necessarycoincidence;if we are to explain why it exists then we must seek an historicalexplanation. The explanationwhich I offer is that the Valley People took their social organisationand their politics from India while the Hill People took their social organizationalong with theirtrade and theirkinshipsystemfrom China. It is a possibleexplanation; I do not claim more than that. In place of the usual linguistic categories I would substituteecological categoriesand these I shall now specify.
"HILL PEOPLE" AND "VALLEY PEOPLE"

The terrainof Burmais very mountainous but not uniformlyso. There are parts of the area where the valleys between the mountainridges are narrow gorges where no human habitationis possible except for those who are preparedto scarp a livelihoodfrom the steep mountainface. But elsewhere the valleys form flat well-wateredalluvial basins perfectly adapted to the needs of the rice farmer. My terms "Hill People" and "ValleyPeople" are intendedto denotethe diametrically opposedmodes of subsistenceassociated with these two types of terrain. These two modes of subsistencehave been historicaltimes2and any hypothesisconcerning presentin the areathroughout historicalprocess must take this into account.
2

E.g., Pelliot (1951), pp. 20-25. This is a translation of the only first hand account of Angkor at the height of its splendour. The original author was a Chinese who

52

EDMUND LEACH

the people so describeddo in The term "Hill People" is unambiguous; fact live in steep hill country. In the main they enjoy a somewhatmeagre standardof living sustainedthroughthe aid of shifting cultivation,though will be consideredlater. Among certain exceptions to this generalization the "Hill People" there is a great range of variety both in language and tribal organisation. The indigenousreligion of most groups comes within Tylor's category of animism; it usually involves some form of ancestor worship. Over the past century the Christianmissions hav_ made many converts,but true "Hill People"are never Buddhists. The term "Valley People" is not quite so straightforward; it is not the equivalentof "Lowlander".The major populationsof the lowland plains of Burma, Thailand and Assam are "Valley People", but so also are the Yunnan dominantelementsin the populationof the Shan States,South-West and Laos-all of which are upland districts. The characteristicalluvial terrainwhich makes wet rice cultivationa profitableenterpriseoften occurs at high altitudes. Some settlementsof "Valley People" are located nearly 6000 feet above sea level. My term "ValleyPeople" also covers other ambiguities.The greaterpart of "Burma" is a region of high annualrainfall,in which every level stretch of ground can readily be developed into a rice field. But "Burma"also includes certain Dry Zones in which the characteristic rice farming techof the are in association with large scale only possible niques Valley People in irrigationengineering. Consequently, these Dry Zone valleys, the population is divided between two distinct sociological categories. On the one hand there are the prosperousrice farmers who are concentratedaround the areas of artificialirrigation;on the other there are the people of the parched outlands, whose living standardsare at an altogetherlower level. In this essay I shall ignorethis distinction.My Valley People are all assumed to be wet rice cultivatorsliving in conditionshighly favourableto wet rice cultivation. The languagesof the Valley People are diversebut much less so than is the case with the Hill People. The majorityof Valley People speak Khmer (Cambodian),Tai and Burmese dialects;3 these languages are not spoken is common. by any Hill People as a mother-tongue, thoughbilingualism The most distinctiveculturalcharacteristic of the Valley People-apart from the practice of wet rice farming-is their adherence to Hinayana Buddhism. The Valley People think of themselvesas the civilisedsector of
travelled with an embassy from the Mongol Emperor in the year 1296. He distinguishes three elements in the population, the Cambodians proper of the city area, the savages who sell themselves as bond slaves to Cambodian masters and work for them in the city, and the brigands of the mountains who form a race apart. The first two categories are dependent on intensive rice agriculture. 3 This ignores the highly complex language pattern among the Valley People of Eastern Assam and Manipur.

THE FRONTIERS OF "BURMA"

53

the total "Burma" population,and in that contextBuddhismand civilisation are synonymous. In Burma proper, the Valley People are mostly either Burmese or Shan; they express their contemptfor their hill neighboursby using the epithetKha ("slave","savage").Nevertheless,a Kha who becomes a Buddhist is thereby civilised, he has "become a Shan", and within a may be forgotten. origin of his descendants generationor two the barbarian has been going on for centuries. What is recorded This type of assimilation of Cambodiain the 13th centuryis strictlyin accordwith what we know of North Burmain the 19th century.4 My generalisationthat Hill People are never Buddhist needs further there are certainexceptional conversions qualification.Apartfrom individual in which whole groupsof Hill People have become economicircumstances cally sophisticatedand have adopted the religion and manners of theil Valley neighbours. For example, the Palaunginhabitantsof Tawngpengin the Burma Shan States who are prosperouscultivatorsof tea have become Buddhistsand have organisedtheir TawnpengState in exact imitation of the political model providedby their Shan neighbours-who are typicalrice growingValley People.5 In generalhoweverit is only the true Valley People who can affordto be civilisedand Buddhist. The fully documented goes back only for a few centuhistoryof "Burma" ries and is very largely concerned with the relations between European colonialistsand native rulers. For periods more remote than the fifteenth century we have only a kind of proto-history,a mixture of legend and to those historiesof Troy which ingeniously inspiredguessworkcomparable Iliad to the latest findings of Turkish to fit the stories of the manage If is such archaeology. proto-history to be convincingit must be sociologically probable;it must not neglect the fixity of ecologicalfacts and it must not postulatesharpculturaland politicalboundariesin a regionwhere none exist even to this day. Besidesthe two internalcontinuities-the ecologicalcategoriesHill People and Valley People-there have been two externalcontinuities, the persisting influenceof India and China. Whatis their nature?

INDIA AND CHINA

Throughoutrecorded history there have been two main foci of cultural developmentin EasternAsia; one in India and the other in China. Every society in South-EastAsia of which we have knowledgewhich has possessed even a modest degree of culturalsophistication has been quite emphatically to Indian or Chinese to influence;usually both. subject
4 Pelliot (1951), p. 19; Leach (1954), p. 293. M Milne (1924).

54

EDMUND LEACH

of such influenceare very diverseand I only propose The manifestations to considercertainmajor aspects. Withinthese limits I shall propoundthe theses that the influenceof Chinahas been mainlyin the fields of trade and communicationand has affected the Hill People rather than the Valley People and that, in contrast,the influenceof India has been felt particularly in the fields of politics and religionand has affectedthe Valley People rather than the Hill People. There is no mystery about this-the Chinesehave never been interested in "Burma"as a potential dominion-they have believed it to be too unhealthy. But they have had a persistentinterestin overlandroutes to India mountainsand forests. The and also in the naturalresourcesof "Burma's" of the records tell us about the Chinese nothing political organisation early that the such facts as and also but record detailed itineraries6 region they land of the b'uok7 tribes living South-Westof Yung Chang producedrhinoceros, elephant, tortoise-shell,jade, amber, cowries, gold, silver, salt, cinnamon and cotton, hilly-paddyand panicled millet, a catalogue which, to this day. apartfrom the cinnamon,is accurateand comprehensive the in 13th under the Even Mongols was adopting centurywhen China blatantly imperialisticpolicies her ambassadorremarksof Cambodiathat "thiscountryhas long had commercial relationswith us". He does not claim ancient political suzerainty.8 any So also in recent centuries when Northern Burma had been the main sourceof jade for all China,the jade mines were owned and workedby Hill People and Chinese interests remained basically commercial rather than political. The Valley People of the jade mines areabenefitedonly indirectly. between I shall returnagain to this matterof the economicinterrelations us the Chinese and the northernHill Peoples, but first let consider some featuresof the political structure. B.C., the Chinesehad Alreadyin early Han times, in the first millennium an the idea of the Nation State to developed conceptof imperium comparable whichthe Romans developedin Europaa few centurieslater. This ideology which is the ultimatepolitical authorityfor postulatesa centralgovernment the whole of a largeterritorial area delimitedby frontiers.The administration of this empire is in the hands of office-holders,an Emperorwith an administrativestaff of bureaucrats.9 The authorityof the central government is maintainedby militaryforce, exercised by garrisontroops permanently the countryand at appropriate dispersedthroughout positionson the frontier. Administrationis financed by taxation which is levied in a systematic
6

Pelliot (1904).

8 Pelliot (1951), p. 10.

See Luce and Pe Maung Tin (1939), p. 267. Chinese sources date back to the 4th century A.D. Certain of the Kachin groups of North Burma are still referred to as p'ok by their Shan neighbours. See Leach (1954), p. 248f.

9 I use Weber's terminology: cf. Weber (1947); (1951).

THE FRONTIERS OF "BURMA"

55

mannerand not accordingto the arbitrary whim of local war"legitimate" lords. No doubt the practical applicationof such theories often deviated very far from the ideal, yet the basic structure of both the Chineseand the Roman systems possessed an extraordinary degree of stability. In both cases the Empire was able to survive long phases of catastrophicincompetenceand of administrative was almost corruptionat the centre;the structure authority to the of effects revolutions and impervious palace dynasticchange. The Indian political model is very different. Here the ideal ruler is not an office-holder,the Emperor,but an individual,Asoka; the patternis one of charismatic leadershipratherthan bureaucratic continuity. Now it is an establishedfact that all the early historicalstates of the Burmaregionwhich achieved any internationalrenown were of an Indian style. Coedes calls them les etats hindouises,and Hall, elaboratingthis, says that their organisationalpatternalways had four common elementswhich he lists as: (i) a conceptionof royaltycharacterised by Hindu or Buddhistcults; (ii) literaryexpressionby means of the Sanskritlanguage; (iii) a mythologytaken from the Epics, the Puranas,and other Sanskrit texts containinga nucleus of royal traditionand the traditionalgenealogies of royal families of the Gangesregion; the sacred laws of Hinduism (iv) the observanceof the Dharmashatras, and in particular that versionknown as the Laws of Manu.10 It is quite outside my field to discuss just how this Indian colonisation came aboutbut what I must emphasizeis the pervasiveness and wide extent of the politicalinfluencein question. In the second centuryB.C. the Westernborder of China lay along the Salweenbut after 342 A.D. the official frontierwas withdrawn much further to the North-East. This was a consequenceof the development of Nanchao as an independent politicalentity centrednear modernTali. Now Nanchao, a state of Indianratherthan despiteits remoteposition,was unquestionably Chinese type. It had no bureaucraticstability and its fortunes fluctuated of successiverulers. violently accordingto the individualaggressiveness Nanchao provides an excellent example of the confusion which arises when such states are thoughtof as nation states of moderntype. Nanchao was inhabitedby people of Tai speech; it ceased to exist as an independent politicalentityin 1253, followingconquestby KublaiKhan. In the centuries which follow, monarchswith Tai soundingnames are recordedas the rulers of petty principalities all over "Burma". This has been interpretedas evidence that, followingthe destruction of Nanchao,therewas a mass migration of Tai-speaking peoples to the South-West.1 This in turn is linked with the
Hall (1955), p. 13. When Hinayana Buddhism replaced Saivite Hinduism, Pall replaced Sanscrit. 11 Hall (1955), pp. 144-146.
10

56

EDMUND LEACH

more general thesis that since Tai is a language of Chinese type it must have "originated" somewhere in Central China. Yet in fact there is no evidence at all of any migration of Tai-speaking peoples into "Burma" from the North-East, and recent trends in linguistic research seem to indicate that Tai speech has no close affinities with Chinese. Its closest links appear to be with languages further south such as Mon and Indonesian.12 Moreover Nanchao should not be thought of as a state with borders but as a capital city with a wide and variable sphere of influence. The inhabitants of Nanchao had no specific identification with the state, there was no Nanchao nation which would be dispersed by the elimination of Nanchao as a separate political entity. Indeed Kublai Khan's occupation of the capital-which was notably peaceful-need have had no effect on the population whatsoever. The common-sense assumption is that there must have been Valley People in "Burma" in the fifth century just as there were Valley People in "Burma" in the fifteenth century and that the Valley People of the two periods spoke much the same sort of language or set of languages. The migration hypothesis of the historians is both improbable and unnecessary.
CHARISMATIC KINGSHIP

What then are the empirical characteristics of les etats hindouises? Most of them have been small, most of them have been shortlived; the continuity of the state depends upon the personality of the monarch; every monarch has a successor, but every succession is an issue of dispute; the state dies with the King, the successor must create a new state from his own personal endeavours. There was continuity of a sort, for the states were in every case built up around a heartland of irrigated rice cultivation and, whatever the vicissitudes of politics, the rice-land stayed in one place. But the state had no fixed frontier, no permanent administrative staff. Scott's comment on the Shans is applicable to all the Valley Peoples of "Burma": Shan history more than that of any other race, seems to have depended on the characterand personal energy of the Sawbwa (Prince). An ambitiousruler seems always to have attempted, and often to have effected, the subjugation of his neighbours. When there were two or more such there was perpetualwar; when there was none there were a numberof practicallyindependentchieftainsdwelling in their own valleys. Hence the astoundingnumber of huge ruined cities which are found all over Indo-China.'3
12

13 Scott and Hardiman (1900/01), Part II, Vol. II, p. 333.

Benedict (1942); Taylor (1956).

THE FRONTIERS OF "BURMA"

57

In this respect the historical kingdoms of Arakan, Pagan, Pegu, Thaton, Ava, Ayut'ia, Manipur,and Assam (as well as some hundredsof smaller locatedwithinthe same generalarea) all had much in common. principalities HinayanaBuddhismwas everywherethe state religionmixed, as in Ceylon, with many explicit elements of Saivite Brahmanism. Everywhere royal polygyny was an exaggeratedfeature of the royal prerogative. The King was regardedas a Chakravartin-a "UniversalEmperor"-or else as an incipientBuddha.14Ritual and myth both implied that he was playing the role of the secular Gaudamaprior to his enlightenment.Every feature of the system implied that governmentwas regardedas personal rule by a divinely inspiredmonarchconsideredas an individual. Let me elaboratethese sweeping generalisations so as to bring out the differencebetweenthe Indian model and the Chinese. In Chinathe successionwas governedby law; each Emperorhad a single legitimateheir specifiedby rules of descent. If the heir was a minor at the time of his successionhe still became Emperoreven though a close relative might act as regent. Usurpationwas relativelyrare and occurredonly with a change of dynastyor in times of political chaos. Day to day government was in the hands of the literati,personswhose status as bureaucrats was, in based on merit favour or aristocratic and not on theory, personal royal blood. In practicethe literaticonstituteda largelyhereditary class but they were not close relatives of the Emperor. The Princes of the blood royal held highly privileged positions but this did not give them office as administrators. In the "Indian"states of "Burma"any one of a King's very numerous succeed him and palace murderswere the offspring might "legitimately" norm. The first act of any successfulclaimantwas to carryout a holocaust of his most immediate rivals-that is to say, his half-brothersand stepmothers. He then apportionedout his realm in fiefs to those of his close relativeswho had survivedand were consideredtrustworthy; that meant, in the main, the King'sown wives and sons. The natureof this fiefdomis well indicatedby the Burmeseterm for a fief holder-myosa-"the eater of the township". Since the myosa's tenure of office was notoriouslyshortlived, he made the most of his opportunities. It is true that in additionto these licensed royal plunderers the structure includeda hierarchyof commonerofficialswith fanciful and of government elaboratetitles-the Burmesewun, Shan amat, Siamese brahya. But these offices too were directlyin the King'spersonalgift. There was no criterion of achievedqualification as in the case of the Chineseliterati. In thirteenth
14 Quaritch Wales (1931), Chapter IV; cf. Cady (1950), Chapter I; cf. Hall (1955), pp. 93-94. Pelliot (1951), p. 16, credits the 13th century Cambodian monarch with 5000 concubines.

58

EDMUND LEACH

century Cambodia the two recognised ways of obtaining administrative office were (a) to make onseself the client of a royal prince or (b) to donate a daughter to the royal harem.15 First hand observers of Thai and Burmese monarchs during the 19th century were all unanimous in emphasising the complete absolutism of the monarch's authority16 and the arbitrariness of the resulting administration. The position is thus summed up by Scott: The coolie of today may be the ministerof tomorrow;and a month hence he may be spread-eagled in the court of the palace with a vertical sun beating down upon him and huge stones piled on his chest and stomach... When King Tharrawaddy ministerswork as slaves on the roads for a time, succeeded,he made Ba-gyee-daw's and when this exercise had quite worn them out, charitablyput them to death... When an official displeasedthe king (MindohnMin) in some way, he said emphatically "I don't want to see that man any more"... A day or two afterwardshis majestywould ask where so-and-sowas. "Alas Sire,"was the answer, "he died of chagrin shortly after the lord of the earth and ocean cast eyes of displeasure
on him."l7

Absolute tyranny was tempered only by the fact that the King, though also head of the Buddhist church, had relatively limited power to manipulate clerical offices. A hierarchy of relatively permanent Church officials operated in parallel to the secular hierarchy of royal appointees and seems to have introduced at least a few elements of stability and mercy into a governmental system ordinarily controlled by arbitrary whim. The typical "Burma" state consisted of a small fully administered territorial nucleus having the capital at the centre. Round about, stretching indefinitely in all directions, was a region over which the King claimed suzerainty and from the inhabitants of which he extracted tribute by threat of military force. These marginal zones all had the status of conquered provinces, and their populations were normally hostile to the central government. Insurrections were endemic and the political alignments of local leaders possessed the maximum uncertainty. Practically every substantial township in "Burma" claims a history of having been at one time or another the capital of a "kingdom", the elleged frontiers of which are at once both grandiose and improbable. It is consistent with this general pattern that those who are now remembered as great Kings were practitioners of banditry on a grand scale whose fame rests solely on their short-term success in carrying fire and slaughter into the territory of their more prosperous neighbours. The "Just Ruler", that archetypal figure upon whom the Confucian ethic lays much stress, had no
15 Pelliot (1951), p. 14. 16 For summarised evidence see in particular Quaritch Wales (1931); Graham (1924);

Scott and Hardiman (1900/01).

17 Scott (1896), pp. 484-5.

THE FRONTIERS OF "BURMA"

59

place in the value system of "Burma" kingship. The kings of Ava, Arakan, Pegu and Ayut'ia were forever pillaging each other's capitals, but conquest by the sword was never followed up by any serious attempt to establish a permanent political hegemony. Military success was simply a manifestation of the monarch's personal power, it did not serve to establish authority and it did not alter political frontiers. But in what sense did these explosive, ephemeral, yet recurrent states really possess "frontiers" at all?

POLITICAL INTERDEPENDENCE OF HILLS AND VALLEYS

Let us go back and resume our consideration of the ecological as distinct from the politico-historical factors in the situation. The political states which we have been discussing have always included elements of both my main population categories, Hill People as well as Valley People. The heartland of the state, with the King's capital, was always a rice-growing valley inhabited by Valley People but the outlying parts of the state normally included Hill sectors as well as Valley sectors. The pattern of development was as follows. The King would first establish authority over his own home valley -ideally by succession, but more frequently by usurpation. He would then spread his authority to a neighbouring valley. This might be achieved by conquest or by marriage treaty or sometimes simply by colonization. Finally the King would claim sovereignty over all the hill country separating the two valley sectors of his total domain. Thus most Hill People were, at least in theory, the subjects of a Valley Prince. But the control which the Valley Princes were able to exercise over the Hill subjects was seldom more than marginal, and the Hill People were quite indiscriminate in their favours. If it suited his convenience a Hill chieftain would readily avow loyalty to several different Valley Princes simultaneously. There were two recognised methods by which the Valley Prince might assert his authority; he could organise a punitive expedition and levy tribute, or he could pay protection money to the hill tribesmen as a reward for their loyalty.18 Some form of the latter procedure seems to have been the most common. What I must emphasise is that the nominal overlordship of a Valley
In Assam there is a special term posa for this type of payment. Even the British authorities with their overwhelming military superiority found it convenient to make posa payments to the hill tribes throughout most of the nineteenth century. Another method of appeasement was for the Prince formally to grant his Hill Chieftains the right to levy toll on trade caravans passing through the mountains. This practice likewise was kept up by the British colonial authorities for many years.
18

60

EDMUND LEACH

Prince over a tract of Hill countrydid not entail the mergingof the Valley People with the Hill People in any cultural sense. Whateverthe overall political structurethe two categoriesremaineddistinctin language,religion and ecological adaptation. It is true that the mannersand customs of the Valley People provided,in certain respects, a model of politenesseven for the hill barbarians. In Burma proper the Hill Chieftainswhom the first European travellersencounteredwere often dressed in Chinese, Shan, or Burmesestyle and took pride in listing the honorifictitles which had been bestowedupon them by their elegantValley overlords,but at the same time they themselvesclaimedto be lords in their own right,subjectto no outside authority. the Valley People and the to represent But it would be equallymisleading Hill People as permanently in ranged implacable hostility. The two categories of populationare symbioticon one another;they interpenetrate territorially and politically as well as culturallyfor, in the course of centuries,"civilisation", as represented by the cultureof the Valley People, has fanned out and the river infiltratedupwardsinto isolatedpockets rightin along valleys the heartof the hill country. Some of these small pockets of uplandValley People may have originated as militarygarrisonsguardinga strategicroute, othersmay have been started by private colonists seeking to escape the burdensof war and tyranny,but the fact that they have survivedand still managedto retaintheircharacteristic Hill Valley Culture shows that the professed hostility of the surrounding People is seldom carriedto extremes. The high degree of political interconnectedness between adjacentgroups of Valley People and Hill People may best be demonstrated from an; example. The far North-West of Burmais dottedwithtiny Shansettlements surrounded by vast areas of mountain country inhabited only by Kachins. Chinese documentsshow that some of these Shansettlements were alreadyin existence in the 8th centuryA.D. Thoughwidelyscattered,these variousShanstatelets claim a culturalunity; they are Hkamti Shans and formerlyfell within the domain of the Prince of Mogaung. The Kachins of the surrounding hill country admit no kinshipwith the Shans nor do they admit that they were ever the subjectsof the Prince of Mogaung. They point out that the jade and ambermines which were the main source of Mogaungprosperitylie in Kachin and not in Shan territory. The Kachins, they say, were the allies but not the servantsof Mogaung.It is a matterof historythat the independent of Mogaungwas extinguished principality by militaryforce in 1765, but the ancient ideology persists. The present day Kachin chieftainwho owns the to himself the title of Hkamti Prince (Kansi jade mines has appropriated Duwa). The Shan stateletsof the formerMogaungrealmwere widely scatteredas the followingtable shows:

THE FRONTIERS OF "BURMA"

61

Modern Map Names

Distance (miles) and Direction from Mogaung

Mogaung Kamaing Mohnyin Mainghkwan Mong Kong (Maing Kaing) SingkalingHkamti Taro Hkamti (HeadwatersIrrawaddy) Hkamti (In Assam)

22 50 73 93 90 90 140 160

N.W. S.W. N.N.W. W.S.W. N.W. N.N.W. N.N.E. N.N.W.

Each of these localities is a small rice plain inhabited by a Tai-speaking Buddhist population ranging in numbers from a hundred to a few thousand. In addition the Mogaung Prince claimed suzerainty over all the hill country lying in between, that is an area of some 10,000 square miles. For that matter, he also claimed overlordship over all Assam. In turn the Prince himself offered ambivalent allegiance to both the Emperor of China and the King of Ava, a circumstance which proved disastrous when, in the latter part of the eighteenth century, the King of Ava went to war with China over the control of the jade trade. My point is this. The seventeenth-eighteenth century realm of Mogaung may be regarded as a typical "Burma" state. It had from certain points of view a very real existence; there was a Mogaung Prince and his kingdom had a name (it appears in the records as Nora, Pong, etc.); yet in another sense the kingdom was a fiction. Its Shan inhabitants were widely scattered and by no means numerous. The Prince could only undertake effective military or political action with the aid and consent of the Hill "subjects", who were not subjects at all. His claims regarding territorial suzerainty were optimistic in the extreme. This Mogaung example is in no way an extreme or a typical instance nor have the conditions which prevailed in the 18th Century altered substantially in recent times. The authority exercised by the central government of the Independent Sovereign State of Burma over its outlying regions in the year 1959 is of a very similar kind. I believe that nearly all the Indian style states of "Burma" history have been of this general type. My main purpose in citing this example was to indicate the kind of relationship which existed between the civilised, nominally dominant Valley People on the one hand and their barbarous Hill neighbours on the other. In the Mogaung case, Kachins fought in the Shan armies and they traded in the Shan markets and they admitted the lordship of the Shan Prince. But the Shan Prince exercised no administrative authority and levied no tribute. Shans and Kachins did not intermarry and the Kachins had no truck with the Buddhist priesthood. Yet assimilation could and did take place. The labourers on the Shan ricefields were mostly settled Kachins living in volun-

62

EDMUND LEACH

tary serfdom. We have actual historical evidence that such groups, by adoptingthe manners, and dress and language of their masters tended to The Valley mergewith them completelyin the courseof a few generations.19 descendants of an be of as the should not of immigrant thought People today alien race, they are simply descendantsof Hill People who have settled in the Valleys and adoptedcivilisedcustomsalong with the practiceof Valleytype wet rice agriculture.But in makingthis culturaltransferthey have cut themselvesoff completelyfrom their former associatesin the Hills. In this part of the world a Buddhistcannotbe a kinsmanof a non-Buddhist. from ecology deserveattention. The Certainother aspectsof the argument Valley People because of their wet rice farminglive in locally dense aggrefactor in their culturalcohesion; gates of populationand this is an important in contrast,most of the Hill People, being shiftingcultivators,live in small widely scattered settlements. I have no means of computingthe precise figures but, very roughly,in Burma proper, the Hill country takes up ten times as much space as the Valley countrybut there are ten times as many Valley People as there are Hill People. some of them The Hill People are not in every case shiftingcultivators; terraces. But the groupswhich do this resortto fixed cultivationon irrigated are not a distinct categoryin any linguisticor culturalsense and their existence does not affect the general argument. Hill farming of any kind requiresa very high labour effort in relation to yield and consequentlyit can very seldomprovideany economicsurplusover and above the immediate subsistence needs of the local population. It is the existenceof such a surplus in the Valley economy which permits the Valley People to maintaintheir more elegantstyle of life. the whole Hill region wherever The converseis likewise true; throughout its membersshow a has become a particular exceptionally prosperous group tendencyto adopt a Shan (Tai) or Burmese style of living and to become convertedto Buddhism. It follows that the contrastsof cultureand language which have led to the conventionalclassificationof "tribesand peoples of can startas a member Burma" have no intrinsicpermanence.Any individual of one categoryand end up in another. Although the greater part of the "Burma"hill country has been for centuries under the (nominal) political suzeraintyof Valley princes, the Valley influencehas not been evenly distributed.Valley culturehas spread only to areas where there is suitable alluvialfarmingland so that in some of the more remote areas an indigenousHill populationhas been allowed to develop on its own without political interferencefrom would-be Valley rulers. If then we want to considerthe natureof Hill Society as an ideal type so
19 Cf. supra, note 4.

THEFRONTIERS OF "BURMA"

63

as to contrast it with Valley Society as an ideal type then it is here in the more remote hill areas that we can observe it.

POLITICAL STRUCTURE OF HILL SOCIETY

On this basis Hill Society can be said to possess the following general characteristics: (1) The Valley pattern of a semi-divine Prince, surrounded by a harem, and ruling by divine right in his personal capacity, is wholly absent. (2) Two contrasted patterns of authority structure stand out and are nearly always juxtaposed in immediate association... These are: (a) an ideology of rule by aristocratic chiefs. The chief is not endowed with personal charisma but holds his office by hereditary right as senior member of a royal lineage; (b) an ideology of "democratic" rule by a council of elders. Each elder acts as representative of a particular lineage but no one lineage is intrinsically superior to any other. The elder may achieve his office either by seniority or as a consequence of passing some test of merit.20 In either case offices of authority are representative offices and are derived from status at birth. The granting of office is never linked with political patronage as in the Valley society. It is very remarkable that both types of ideology, the aristocratic and the democratic, are regularly found to coexist side by side throughout the whole of the northern and western parts of the "Burma" hill country. I will list only a few examples:
Hill Group Kachin21 Konyak Naga22 Central Chin24 Aristocratic gumsa thendu Democratic gumlao thenkoh

SouthernNaga23 WesternManipurHills25

Sema
Zahau

Chakrima(Angami)
Zanniat

New Kuki (Thado)

Old Kuki

For the more easterly hill tracts of the Southern Shan States, Karenni, Northern Thailand and Laos, the ethnographic descriptions are too defective to permit confident generalisation, but here too the same two contrasted types of political ideology appear to coexist.26 Elsewhere I have argued at
20 21 22 23 24 25 26

Stevenson (1943). Leach (1954).

Von Fiirer-Haimendorf(1941).
Hutton (1921a); (1921b). Stevenson (1943). Shakespear (1912).

Scott and Hardiman (1900/01), Part I, Vol. I.

64

EDMUND LEACH

length that these two types of political organisation represent different aspects of a single "cyclic" type of system viewed at different phases of its growth.27 (3) In the aristocratic type of regime a single Chief usually claims dominion over a number of scattered villages. Each of these villages has a headman who holds office by hereditary right. The headman's lineage and the chief's lineage are usually linked by ties of affinity. In contrast, in the democratic regime, each village is on its own. Democratic villages are not necessarily weaker politically than aristocratic chiefdoms, for some democratic villages are relatively very large. (4) The great majority of the Hill People are organised in exogamous unilineal descent groups of lineage type. The Karens may be an exception to this rule but the available information is inadequate and inconsistent. In contrast, among the Valley People unilineal descent groups are not a normal feature of the social structure. Where such descent groups occur, as sometimes among the aristocracy, they are not exogamous. (5) Hill Society attains its highest elaboration in areas which are remote from the contaminating influence of Buddhist civilisation. This proposition is not self-contradictory. From the viewpoint of the external observer it is legitimate to regard the Buddhist Valley People as the civilised element in the total population in contrast to the Animist Hill People, who are, by comparison, barbarians. But Hill Society has its own scale of values and these show up best when they are unadulterated. The following remarks by Scott concerning the Wa headhunters of the Eastern Shan States might be applied to almost any of the Hill Peoples: Materialprosperityseems to exist in inverseratio to the degree of civilisation. The HeadhuntingWa have the most substantialvillages and houses, the broadestfields, the greatestnumberof buffaloes, pigs, dogs and fowls. They also have the greatest conceit of themselves,the most ornamentsand the least clothes. The Intermediate Wa fall some way behind in material possessions. The Tame Wa with their civilisation,find their houses dwindleto hovels, their fields shrinkto plots... and beyond this there are Wa who put up no heads at all; some of them claim to be Buddhists,othersmake no claim to anything,not even the pity of their neighbours.28 There is a genuine paradox here. The Valley People as a whole are vastly more prosperous and sophisticated than the Hill People as a whole, yet in the context of a hill ecology the trimmings of civilisation are disadvantageous. Certain features of this pattern deserve attention. In the first place, it is very clear that the process of interaction between the Hill People and the Valley People has not been one of simple cultural diffusion. Hills and Valleys stand in radical opposition and there is evidently a certain level at which Hill culture and Valley culture are totally inconsistent with one another just as one might say of early mediaeval Europe that
27 28

Leach (1954).

Scott and Hardiman (1900/01), Part I, Vol. I, p. 511.

THE FRONTIERS OF "BURMA"

65

Christianity and Paganism were inconsistent. There are cultural elements which are common to both groups but these similarities are remarkably few. Culturally there is far more in common between the Lakher in Assam and the Lamet in Laos29 than there is between either group and their nearest Valley neighbours. The same is true of any of the Hill peoples throughout the area. Yet the pattern of political relations which I have previously described might have led us to expect something different. After all, the Hill People and the Valley People are racially the same and languages are very easily changed, so what is it that keeps the two groups apart? I do not think that the anthropologist or anyone else can say why the distinction exists but I think it may be illuminating to point out some of the associated correlations.

CHINA AND THE HILL PEOPLE

Earlier in this essay I remarked that although the Hill People of "Burma" have for centuries come under the spasmodic political influence of Indian style states, their most direct economic contacts have been with the Chinese. In some cases this is true even at the present time. I myself have first-hand acquaintance of a number of North Burma localities which were "unadministered territory" throughout the period of British colonial rule; all of them were regularly visited by Chinese traders, but never by Burmese. It is relevant here to remember that Chinese society, like that of the Hill People, is structured into a system of unilineal descent groups and also that the "animism" of the Hill People is fundamentally a cult of dead ancestors which has many Confucianist parallels. These similarities make it possible for the Chinese and the "Burma" Hill People to communicate with one another and to establish permanent social relationships in a way which is impossible for the Hill People and the Valley People. Chinese villagers actually settle in the hill country and then live much like ordinary hill folk. They will even intermarry with their "barbarian"neighbours; but the Valley People will never do either of these things. The contrast in the pattern of marriage seems to me particularly significant. In the Valley culture the population of each local rice plain tends to be endogamous. The Prince, who has many wives, may take daughters from neighbouring Princes but he also takes women from his own immediate followers. He receives the latter women as tribute. Thus, in terms of kinship, Valley society as a whole forms a closed system; the Valley People do not give their women away to strangers. Furthermore, each marriage is an individual affair between a particular man and a particular woman; it
29

Parry (1932); Izikowitz (1951). These two "tribes" are about 500 miles apart.

66

EDMUND LEACH

does not establishan alliancebetweenkin groups. All this is consistentwith the fact that elementsof Hindu caste ideology have all along been present in the Valley culture. Valley People repudiatemarriagewith the barbarians even when they are willingto accept their economicservices.30 In contrast,in the Hill culture,marriageis closely mixed up with trade. Girls are marriedagainst a bride-priceand the objects involved in brideprice transactionsare the same sort of objects as are met with in dealings with a Chinese trader.31Thus the ties of affinal kinship ramify widely, following trade routes and jumping across languagefrontiers and political boundaries. In 1942 a Gauri acquaintanceof mine from East of Bhamo found himselfin a Singfovillagein Assam 250 miles from home, but it took him only a day or so to persuadehis hosts that he was one of their relatives. I am not arguingthat a single kinshipnetworkramifiesover the whole of Hill countrybut everywherein the hills a very high valuation the "Burma" is placed on extendedkinship relationsand also upon the permanenceand of stability of such relations. Individualsare regardedas representatives and are and classed on that account lineages particular they places particular as friend or foe. Women who are given in marriageserve to establish a which is likely to be repeated between lineages-a relationship relationship or furthertrade. This is the antithesis laterin furthermarriages of the Valley Culture theory which treats women either as separate individualsor as chattel slaves. A Valley Prince receives women as tribute;a Hill chieftain gives them out as pledges of economic cooperation. Hill cultureis not a direct imitationfrom the Chinesebut it parallelsthe Chinese system in a way that the Valley culture does not. Just as we can say that the Valley culture has an Indian style without implyingthat the Shans are Hindus or that the Burmesehave a fully developedcaste system, I thinkwe can say that the Hill culturehas a Chinesestyle withoutimplying that Naga tribesmenare devout adherentsof the Confucianethic. The really crucialdistinctionhere is that betweencharismatic (individual) the one hand office the on and traditional other. In on the Valley authority all and and is for that individual reason system authority temporary tyrannical. The Valley tyrant does not display his merit by justice but by acts of
self-glorification.

In the Hill system, as in the Chinese, all offices are vested either in particularlineages (which are conceptuallyimmortal)or else are reserved for individualswho have achieveda particularsocial status (e.g. by passing examinationsor by working through a graded series of sacrificialfeasts). The acts of the ruler are themselvesgovernedby rules, everythinghe does carriesthe sanction of legitimatecustom. My suggestion is that this contrast of ideologies about the nature of
30 31

Cf. Pelliot (1951), p. 19; Milne (1910), p. 50. Cf. Leach (1954), references to hpaga.

THE FRONTIERS OF "BURMA"

67

authorityilluminates,even if it does not explain why 2000 years of Indian rule has not eliminated the radical separation between Hill and Valley society.
SUMMARY

"Burma" is a regionlying betweentwo greatcentresof civilisation. Throughout history it has been influenced simultaneouslyfrom India and from in China, not only at the trivial level of Court politics but fundamentally terms of the culturalsystem as a whole. But this influence has not been an indiscriminatediffusion of ideas. Politics, ecology, kinship and economics provide in some degree separate and separableframesof reference,and I have thereforeinvitedthe historian of Burmato look upon the presentas part of a continuingprocess of interaction between two kinds of political structure,two kinds of ecology, two distinctpatternsof kinshiporganisation, two sets of economicinterests. There are other frontier regions where a very similar style of analysis might apply and it is on that accountthat I feel justifiedin offeringthis as a contribution with potentialcomparative value.
E. R. LEACH

Cambridge University

REFERENCES

No attempt has been made here to support my more general statements with detailed references. For Burma proper and the regions to the East the most useful select bibliography for English language readers is that appended to Hall (1955). Ethnographic sources for the whole region are well covered by Embree and Dotson (1950). For the Naga Hills area Hutton (1926) covers 19th century sources very thoroughly. I know of no general bibliography of sources for Assamese history but Mackenzie (1884) and Michell (1883) and Selection of Papers (1873) give summaries of many of the key documents for the 19th century period. These source books have been supplementedby Reid (1942). Leach (1954) mentions a number of items relating to the North Burma/ Assam region which do not appear in other bibliographiclists. P. K. BENEDICT, 1942
CADY, JOHN F.

"Thai, Kadai and Indonesian: a New Alignment in

Southeastern Asia", American Anthropologist, 44, pp. 578-601. A History of Modern Burma (Ithaca). Bibliography of the Peoples and Cultures of Mainland Southeast Asia (New Haven). "Dass Gemeinschaftsleben der Konyak-Naga von Assam", Mitt. d. Anthrop. Gesellsch. in Wien, LXXI.

1958
L. 0. EMBREE, J. F. & DOTSON, 1950
FiFRER-HAIMENDORF, C. VON

1941

J. S. FURNIVALL,
1939 GRAHAM,W. A.

"The Fashioning of Leviathan: The Beginnings of

British Rule in Burma", Journal of the Burma Research Society, XXIX, Pt. I. Siam, 2 vols. (London).

1924

68
HALL, D. G. E.

LEACH EDMUND A History of South-East Asia (London). (a) The Angami Nagas (London). (b) The Sema Nagas (London).

1955
J. H. HUTTON,

1921 1921 1926

K. G IZIKOWITZ, "A Bibliography of the Naga Hills with some Adjacent Districts". Appendix VI to J. P. Mills, The Ao Nagas (London).

Lamet: Hill Peasants in French Indochina (Goteborg). Inner Asian Frontiers of China (New York). Political Systems of Highland Burma (London). The Tribes of Burma. Ethnological Survey of India: Burma, No. 4 (Rangoon).
"Economic Life of the Early Burman", Journal of the

1951
O. LATTIMORE,

1940
LEACH,E. R.

1954
LowIs, C. C.

1919
LUCE, G. H.

1940
TIN LUCE, G. H. & PE MAUNG

Burma Research Society, XXX.

"Burma down to the Fall of Pagan, Part I", Journal

1939
A. MACKENZIE,

1884
MICHELL, ST. J. F.

of the Burma Research Society, XXIX (Part II was never published). History of the Relations of the Government with the Hill Tribes of the Northeastern Frontier of Bengal
(Calcutta).

Report (Topographical,Political and Military) on the

North-East Frontier of India. Confidential (Calcutta).

1883
MILNE, L.

Shans at Home (London). The Lakhers (London).

"Deux itineraires de Chine en Inde a la fin du VIIIeme

1910
PARRY,N. E.

1932
PELLIOT,P.

1904 1951
QUARITCH WALES, H. G.

siecle", B.E.F.E.O. (Hanoi). Memoires sur les Coutumes du Cambodge de Tcheou Ta-Kouan. Oeuvres posthumes de Paul Pelliot, No. 3 (Paris). Siamese State Ceremonies (London). History of the Frontier Areas Bordering Assam from
1883-1941 (Shillong).

1931
REID, R.

1942
YOE) SCOTT,J. G. (SHWAY

The Burman, His Life and Notions (London). Gazetteer of Upper Burma and the Shan States. Part
I: 2 vols; Part II: 3 vols. (Rangoon).

1896
J. P. J. G. & HARDIMAN, SCOTT,

1900/01 Selection of Papers 1873

J. SHAKESPEAR,

Selection of Papers regarding the Hill Tracts between Assam and Burma and on the Upper Brahmaputra
(Calcutta).

The Lushei Kuki Clans (London). The Economics of the Central Chin Tribes (Bombay).
"General Structure of Languages Spoken in Burma",

1912
H. N. C. STEVENSON,

n.d. (1943)
L. F. TAYLOR,

1956
WEBER, MAX

1947 1951

Journal of the Burma Research Society, XXXIX. The Theory of Social and Economic Organisation (London). The Religion of China (Glencoe).