Tribal absorption and Sanskritisation in Hindu society

http://cis.sagepub.com/Contributions to Indian Sociology

http://cis.sagepub.com/content/13/2/293The online version of this article can be found at:

DOI: 10.1177/006996677901300206

1979 13: 293Contributions to Indian SociologySurendra Munshi

Tribal absorption and Sanskritisation in Hindu society

Published by:

http://www.sagepublications.com

can be found at:Contributions to Indian SociologyAdditional services and information for

http://cis.sagepub.com/cgi/alertsEmail Alerts:

http://cis.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://cis.sagepub.com/content/13/2/293.refs.htmlCitations:

What is This?

- Jul 1, 1979Version of Record >>

at MICHIGAN STATE UNIV LIBRARIES on April 30, 2014cis.sagepub.comDownloaded from at MICHIGAN STATE UNIV LIBRARIES on April 30, 2014cis.sagepub.comDownloaded from

http://cis.sagepub.com/

http://cis.sagepub.com/content/13/2/293

http://www.sagepublications.com

http://cis.sagepub.com/cgi/alerts

http://cis.sagepub.com/subscriptions

http://www.sagepub.com/journalsReprints.nav

http://www.sagepub.com/journalsPermissions.nav

http://cis.sagepub.com/content/13/2/293.refs.html

http://cis.sagepub.com/content/13/2/293.full.pdf

http://online.sagepub.com/site/sphelp/vorhelp.xhtml



Tribal absorption and Sanskritisationin Hindu society1

SURENDRA MUNSHIIndian Institute of Management, Calcutta

The purpose of this article, as the title indicates, is to consider togethertwo of the most important processes that have been identified as operatingfor a long time in Hindu society. I have concentrated on Nirmal KumarBose for his formulations on tribal absorption and on M.N. Srinivas forhis elaboration of the process of Sanskritization in Hindu society. Theemphasis is on a systematic treatment of such contributions made by Boseand Srinivas as are relevant to the theme of this article. First of all, aneffort is made to reconstruct the expositions of their formulations, as faras possible in the language of the authors themselves. Next, Bose andSrinivas are subjected to a critical theoretical examination. Finally, draw-ing from the analysis, certain related substantive issues are raised. The

critique that emerges through this effort, thus, does not aspire to beexhaustive.2 I have tried to formulate an argument that has received rela-

tively little attention in Indian sociology. It is hoped that this article willbring out the relevance of such an analysis, particularly of treating Boseand Srinivas together.

1Grateful acknowledgements are made to my friends, Subas Kumar Biswas, IndianStatistical Institute, Calcutta, and Saila Ghosh, Indian Institue of Management,Calcutta, two of Prof. Bose’s direct students, who helped me through discussions inunderstanding the complex personality of Prof. Bose. My treatment of Prof. Bose wasconsiderably influenced by them. Tridip Chakrabarti provided considerable help inpreparing a bibliography for the work. Needless to add, they do not bear any res-ponsibility for the views presented in the paper.2For an earlier discussion of Bose, see Surajit Sinha (1972) and André Béteille (1975).

For comments on Srinivas and his contributions, there is a considerable literatureavailable: F.G. Bailey (1960), J.F. Staal (1963), Yogesh Atal (1968), T.K. Oommen(1970), David G. Mandelbaum (1972), Louis Dumont (1972), Ramkrishna Mukherjee(1977), Yogendra Singh (1977), and the review symposium in the Contributions toIndian sociology, January-June, 1978.

at MICHIGAN STATE UNIV LIBRARIES on April 30, 2014cis.sagepub.comDownloaded from


294

. NIRMAL KUMAR BOSE ON TRIBAL ABSORPTION

In a celebrated article, first published in 1941, Bose presented his viewson the Hindu method of tribal absorption. This article deserves close studyeven now. Basically, it is a work in the field of ’culture contact’, or

’acculturation’. It analyses the ’Brahminical way of acculturation’ - themethod of ’tribal acculturation, in Hindu society. The Hindu method oftribal absorption is found to be entirely different from that of Islam, whichis believed to follow the method of complete conversion (1967a: 222).

Bose points out that the Hindus generally ’exercised a policy of laissez-faire with regard to the social and religious practices of tribal peoples evenwhen they came within Hinduism’ (ibid.: 210). However, this ’catholicity’had three distinct features: (i) Although the policy was not to displace theoriginal social and religious culture of the tribes, but rather to preserveit (ibid.: 222), yet, something had to be done ’to bring the tribal culturesin line with Brahminism’. As a result, the ’Brahmins modified the oldculture when it went against the grain of their own ideas and left the rest

intact’ (ibid.: 212). (ii) ’Once a tribe came under the influence of theBrahminical people ..., a strong tendency was set up within it to remodelits culture more and more closely in conformity with the Brahminical wayof life’ (ibid.: 214, emphasis added). (iiij However, the tribes could not beallowed to come very close to their superiors, ’the Brahmins very often

put a stop to such progress ... towards higher and higher standards ofculture’ (ibid.: 214, emphasis added). Thus, many ’forms of culture’ cameinto existence.At the level of thought, Bose points out, a ’point of strength’ for

Hinduism was the belief that knowledge was available to a persona or

community only in ’fragments’. It was through the recognition ofthis ’fundamental truth’ that Hinduism became a ’federation of faiths’

(1972b: 79).Bose goes further than merely identifying the basic outline and features

of tribal acculturation in Hindu society. His treatment of the Brahminical

way of acculturation leads him to a general proposition regarding culturecontact. He emphasises the role played by ’economic matters’ even in

social and cultural spheres. He writes that, ‘culture ... seems to flow froman economically dominant group to a poorer one when the two are tied

together to form a larger productive organisation’ (1967a: 205). Politicaldominance has its own advantage for the ’dominant group’. In short, Boseconcludes: ‘Culture ... flows from a politically and economically dominantgroup to a subservient one. In social matters too, the former occupies ahigher status in contrast to the latter’ (ibid.: 214).

This is consistent with his earlier thinking on the subject. Thus, in awork first published in 1929, he admits that cultural contact is an event



295

of ’prime importance’ in cultural history (1961: 67) and identifies fourpoints that must be considered in studying culture contact: the central

ideas, the mental attitude, content of culture and the economic frame-work which sustains culture (ibid.: 92). Here, too, Bose emphasises thatnew material objects as well as new ideas associated with the ’culture ofthe conquerors’ naturally enjoy superior prestige and even economic valueamongst those belonging to the ’vanquished culture’ (ibid.: 92-3). In Thestructure of Hindu society, originally published in Bengali in 1949, Bosepointedly asks: ’Who does not crave to be honoured by those who areheld in honour’? He sees that in Hindu society ’there was a tendency ineach lower order caste to imitate the rites and customs of the higher ordercastes’ and comes to the conclusion that to be honoured ’the easiestdevice is clearly to imitate those held in honour’ (1975: 94).Three points are of importance in considering Bose on this issue. First

of all, it should be remembered that Bose is concerned with the ’soul ofculture’ which is defined as ’a constellation of ideas and emotionswhich forms the core of community’s culture’ (1961 : 36), as well as the’outward framework of culture’ which is seen as the ’institutional form’of the ’soul’ of culture (ibid.: 35). This emphasis on the institutionalform is related to his definition of culture as ’adaptive behaviour whichhas become institutionalised’ (ibid.:42), that is, as institutionalised adaptivebehaviour. Secondly, Bose realises that ’the value of culture lies in itsrelation to man’s life’ (ibid.: 67). In fact, he regards the ’human element’as an important ’operative factor’ in cultural evolution (ibid.:viii).The implication of these points will be taken up later. For the present,

let it be noted that, behind the multitude of cultural forms in Hindu

society, Bose identifies the ’soul’ that binds together various aspects ofculture by ’bonds of kinship’ (ibid.: 23), as the varna system. This is

brought out in chapter three of his book Cultural anthropology. He also seesin the varna syste:n the social ’design’ of Hindu society, which he descri-bes as a ’union of many communities’ (1975: 164). Thus, he sees in thecaste system an organising principle, an ’ideal’ that Hindu society hasevolved, for securing unity in diversity of many communities andcultures.

°

With respect to the caste system, first of all, Bose draws a distinctionbetween varna system and several jatis (1967a: 227) that operate ’underspecific conditions’ of a village or region (ibid.: 2.t9). Secondly; he em-

phasises that ’the caste system is not exactly as immutable as it is gene-rally assumed to be’ (ibid.: 207). The varna system became increasinglymore complex due to the internal process of formation of new jatis aswell as by assimilation of external groups within its fold (1975: 88).Thirdly, he notes from the careful manner in which an attempt was madeto establish a close correspondence between jati and occupation in Hindu



296

society that the idea was to build up a ’social organisation on the basis ofhereditary, monopolistic, non-competitive guilds’ (1967a: 209). Fourthly,he points out that there was also an idea at the root of the caste systemthat ’man is subservient to society’. All the castes, high and low, maketheir living by serving the society in the prescribed manner. ’They attendto society and society attends to them. Rights and obligations are inextri-cably tied’ (1975: 168).

It was the idea of a social organisation based on the principle of ’heredi-tary, moz~.opolistic, non-competitive guilds’ that led to the process by whichonce a tribe came under the influence of ’the Brahminical people’ and wasconverted into a jati, it was left with hardly any economic freedom.

To each new jati, Hindu society assigned a particular occupation withina particular region (1967a: 209). Such a compulsory assignment withinthe graded varna system had mixed implications on the position of a tribethat had recently been converted into a jati. On the one hand, the mono-poly in a particular occupation that Hindu society guaranteed (or tried toguarantee a new jati gave it ’a certain minimum of economic security’(ibid.: 210), on the other hand, this process relegated most of thesegroups to an inferior position.An important assumption that Bose makes with respect to the tribal

population is that the Brahminical society was economically ’better off’ thanthe tribal society. Bose is also aware of the problem of the inadequacy ofof the tribal, ’productive system’. In most places in contenaporary India,Bose finds that the population pressure in tribal areas has crossed thecritical point affecting the land-man ratio on which the tribal productivesystem rests (1972b: 45). Thus, it was due to population pressure and the’attraction’ of the economically better off Hindu society that the Juangs,Oraons and other tribes were found ’gradually abandoning their indepen-dence and moving towards Brahminical society’ (1975: 114).

Bose recognises that the ’sad thing’ (1975: 166) or the ’unfortunatepart’ (1967a: 214) regarding the caste system has been that the idea of

non-competitive hereditary guild system was ’modified by a feeling ofracial superiority’ (ibid.: 221) and that the Brahmins ’never succeeded in

giving to the vanquished a place equal to their own’ (1975: 166). Thus,within the framework of the caste system, ’there was a broad class divi-sion into those who enjoyed privileges incommensurate with their servicesto society, and another who were deprived.... The former may have beenthe conquerors or the ruling class and the latter, the conquered’ (1967a:209). However, if the ’subjugated tribes’ did not rise in revolt againstthese deprivations, it was essentially due to a minimum security that theyenjoyed in the economic sphere and the freedom that they were allowedin the cultural sphere within the Hindu social organisation (ibid.: 210).No exposition of Bose’s views on tribal absorption in Hindu society can



297

be complete without reference to his policy formulations on the tribal

question in contemporary Indian society. As is well known, Bose occupiedthe influential post of Commissioner for Scheduled Castes and Tribes from1967 to 1970 and during this period he produced Annual Reports whichare of considerable significance (in parts reproduced in Bose 1972b). Evenearlier, he had presented his policy suggestions, most notably in his articleon ’Integration of tribes in Andhra Pradesh’ (1972b). Let me take thisarticle as the starting point of the brief resume presented below.

This article was written by way of a report on a seminar organised inDecember, 1963, in Hyderabad. In this article Bose recognises that at pre-sent the tribes are being exposed to the plains people on a massive scaleand at an accelerated pace and this is responsible for ’uncontrolled modi-

. fications in their life’ (ibid.:64). He also admits that the chief problemfacing the tribes is one of ’alienation of land’ (ibid.: 65). Yet, he arguesagainst the protectionist policy on ground of ’danger of a latent two-na-tion theory’ (ibid.:66,72). Instead, he suggests satyagraha, which will uniteall the under-privileged people, tribal as well as of the plains , in an efforttowards ’peaceful persuasion of the privileged class’. Basic to this proposalis Bose’s assertion that ’all plainsmen are not exploiters in their relation-ship to the tribesmen’ and that differences in terms of privileges are grow-ing up amongst the tribal people themselves (ibid.: 69-70).He repeats this point in his concluding address at the National Seminar

on the Hill People of North-East India, delivered at Calcutta in December1966. He opposes the demand of the delegates of the region for ’self-deter-mination’ even after they had clearly emphasised that this demand was nota demand of seperation from India. He asks them: ’Why can’t we viewone another sympathetically and come closer to one another in understand-ing ?’ (ibid.: 78). Reminding the hill delegates that the common economic

needs and demands of a common political system have brought ’us’ closerthan ever before, Bose refers to the ideal of Hindu civilisation as ’federa-tion of faiths’ and exhorts them to combine in the ’noblest of all advent-ures’ towards ’cultural federalism’ and swaraj (ibid.: 79-80).

SRINIVAS ON SANSKRITISATION IN HINDU SOCIETY

Basic to M.N. Srinivas’s views on Sanskritisation in Hindu society is thedistinction that he draws between varna and jati. This distinction is very im-portant to Srinivas for he sees varna as representing ’a model of the caste

system’, a ’literary or ideal model’ (1967: 75), which received a high degreeof elaboration at the hands of the Brahmin writers in the post-Vedic period(1966: 5). On the other hand, jati is an appropriate term for innumerable

’endogenous groups’ that exist in the country. Thus, whereas varna

gives us a Brahminical ’model’ of the caste system, jati represents caste at



298

the ’existential level’ (ibid.:4) or ’caste as it is in fact’ ( 1962: 63). As notedby Srinivas, the main features of the caste system embodied in the varnamodel are: (i ~ There is a single all-India ritual hierarchy without regionalvariations; (ii) there are only four varnas, excluding the Untouchables; (iii)the hierarchy is clear in terms of Brahmin (Brahmana), Kshatriya, Vaishyaand Shudra; and, (iv) it is immutable (1966: 3).On all these counts, Srinivas finds ’the system as it actually functions’

differing from or in conflict with tile ‘model’ of the system. The differenceis found to exist to such an extent that Srinivas finds it ’a matter for

wonder’ that the varna~model has continued to survive. His main contenta-tion against the varna model is that it obscures the ’actualities of caste’,particularly ’the dynamic features of caste’ (ibid.: 3), and that it is clearlyintended ’to support the theory of Brahminical supremacy’ (ibid.: 4). Hence,he advises ’the sociologist to free himself from the hold of the varna-modelif he wishes to understand the caste system’ (1962: 66).

In actual social life, real and effective units of caste, viz. jatis, are organi-sed in a ’ranked order of castes’, but there is considerable ambiguity regard-ing the actual position of castes, particularly in the middle region of theorder. This ambiguity regarding the precise position is important and is’an essential feature of caste as an ongoing system’. In fact, Srinivas,observes, when two castes claim equal status the situation is not accidentalbut a ’typical product of a dynamic system in which there is some pushingand jostling in the attempt to get ahead’ (1966: 4).

In other words, the traditional caste system is far from being a rigidsystem.Srinivas came to this conclusion in his Religion and society amongthe Coorgs of south India (t952). This is brought out in his introduction of.the concept of Sanskritisation. Srinivas writes:

Movement has always been possible, and especially so in the middle re-gions of the hierarchy. A low caste was able, in a generation or two, torise to a higher position in the hierarchy by adopting vegetarianism andteetotalism, and by Sanskritising its ritual and pantheon. In,short, it tookover, as far as possible, the customs, rites, and beliefs of the Brahmins,and the adoption of the Brahminic way of life by a low caste seems to

have been frequent, though theoretically forbidden. This process has beencalled ‘Sanskritisation’ in this book, in preference to ’Brahminisation’, ascertain Vedic rites are confined to Brahmins and the other ’twice-born’,castes (1952: 30).

Srinivas acknowledges that he unduly emphasised the Brahminical ’model’of Sanskritisation in his earlier writings (1952,1962) and admits that other’models’ of Sanskritisation-Kshatriya, Vaishya and Shudra are possible(1966:7). In fact, he suggests that all these ’models’ of Sanskritisation are



299

mediated by the locally ‘dominant caste’, which is viewed as supplement-ing the Sanskritisation process (ibid.: 7-8). Srinivas’s revised description ofthe process of Sanskritisation is thus:

Sanskritisation is the process by which a ’low’ Hindu caste, or tribalor other group, changes its customs, ritual, ideology, and way of life inthe direction of a high, and frequently, ’twice-born’ caste. Generally such

. changes are followed by a claim to a higher position in the caste hierarchythan that traditionally conceded to the claimant caste by the local com-

munity. The claim is usually made over a period of time, in fact, a gene-ration or two, before the ’arrival’ is conceded. Occasionally, a caste

claims a position which its neighbours are not willing to concede

(1966: 6).

A few relevant points must be noted in order to appreciate this descrip-tion of Sanskritisation. Firstly, Srinivas asserts that it is not possible tounderstand Sanskritisation without reference to the ’structural frameworkin which it occurs’ (1962: 44). However, he points out that to describesocial change in India in terms of Sanskritisation is to describe changeprimarily in ’cultural and not structural terms’ (ibid.:55). The mobilityassociated with Sanskritisation results only in ’positional changes in thesystem and does not lead to any structural change’ (1966: 7). Secondly, henotes that ’caste mainly exists and functions as a regional system’ (ibid.:3). Thirdly, though he accepts, as noted above, that there are different

,

‘models’ of Sanskritisation, be emphasises that the Brahmins ‘may ... beregarded as ’better’ models of Sanskritisation than the others’ (ibid.:8).In addition, it is pointed out that the Brahmins had a special role in

legitimising the claims off a caste in the process of Sanskritisation.This role becomes particularly important in the contest of the followingtwo related points:-(i) Sanskritisation involves ’claimed’ and ’conceded’

status; and, (ii) such a status in not only a matter of opinion but refers tothe ’more important realm of institutionalised practice’ (ibid.:6).Fundamental to this approach is the assumption that ’structural basis

of Hindu society is caste’ (1962: 44). In view of this assumption, it is easierto understand the emphasis placed by Srinivas on group mobility (viz.,jati ) rather than on individual mobility. For, as seen by Srinivas, groupmobility is a characteristic feature of the caste system as distinct fromindividual or family mobility of the class system (ibid.: 58).

It will, however, be wrong to infer from the above discussion that theSanskritisation process covers only the Hindu castes. Srinivas has included’tribal or other groups’ in his revised description of the concept. His viewson the Sanskritisation process amongst the tribal groups deserve attention

particularly because he developed the concept in the course of his study of



300

the Coorgs who were originally outside the Hindu fold.In his study of the Coorgs, Srinivas clearly says that the Sanskritisation

process has been ’going on both inside and outside Hinduism’ (1952: 213).However, once a tribe is undergoing Sanskritisation it usually assumesthe characteristics of a caste and is brought within the fold of Hinduism(1966: 7). Caste guarantees ’autonomy’ and at the same time provides a

link with others ’all going to form a hierarchy’ (1952: 31). Thus, when theprocess is viewed on a continental scale and over a long period of history,it is possible to see how in the midst of regional diversity the ’Sanskritic

ideas and beliefs penetrated the remotest hill tribes in such a manner, asnot to do violence to their traditional beliefs’ (ibid.: 31).

Hinduism has not only absorbed other people but also several religiousideas and practices. Srinivas uses the concept of ’spread’ in his study of

the Coorgs and splits up Hinduism in the following categories: All-IndiaHinduism’, ’Peninsular Hinduism’, ’Regional Hinduism’ and purely ’LocalHinduism’. For the castes, there are two types of spread that are identified,’horizontal’ and ’vertical’ (ibid.: 214). Thus, for instance, Brahmins every-where have much Sanskritic ritual in common (horizontal spread) and, atthe same time, within a culturally homogenous area, they share some cul-

tural and ritual forms with all the castes, including the lowest, in that area(vertical spread) (ibid.: 215). There are two points to be noted in this res-

pect : (i) Srinivas formulates a general proposition, ’as the area of spreaddecreases, the number of ritual and cultural forms shared in common in-creases’ (ibid.: 215). Further, the higher castes have, more of the horizon-tal spread than the lower ones and conversely the lower castes share moreof the vertical spread than the upper castes (ibid.: 219 j. (ii) Srinivas seesSanskritisation as a ’two-way process’ though he also notes that ’the localcultures seem to have received more than they have given’ (1962: 59). Tovarying degrees, thus, local elements have been transformed in the courseof their ’ascent’ into Sanskritic Hinduism, and also, the components of all-India Hinduism (chiefly Sanskritic in character) have gone through ’paro-chialisation’ in the course of their ’descent’ into local religious culture, forwhich Srinivas’s preferred term is ’de-Sanskritisation’ (1967: 73).

In conclusion of this section I shall consider the concept of ’dominantcaste’ as developed by Srinivas. It has been noted above that Srinivas seesthis group as supplementing the Sanskritisation process. In one of the

early essays on his village Rampura, in Mysore, Srinivas describes a domi-nant caste as one that is numerically preponderant and one that wieldseconomic and political power in a region or locality. A large and powerfulcaste, he notes, can more easily be dominant if its position in the localcaste hierarchy is not too low (1955b: 18). Any caste may be a dominantcaste in an area, though Srinivas admits that he has not come across anyof the Untouchable castes being dominant anywhere. Occasionally, a



301

group originally outside the Hindu fold, such as the Coorgs or the RajGonds, may become dominant by virtue of its ’numbers, wealth, and mar-tial prowess’ (ibid.: 8).More recently, Srinivas has made his conception of ’dominant, landown-

ing castes’ more specific:

For a caste to be dominant, it should own a sizable amount of the arableland locally available, have strength of numbers, and occupy a high placein the local hierarchy.... New factors affecting dominance have emer-ged in the last eighty years or so. Western education, job in the

_

administration, and urban sources of income are all significant in contri-buting to the prestige and power of particular caste groups in the village(1966: 10-11).

A locally dominant caste serves as a traditional guardian of ’pluralisticculture and value system’ (ibid.: 15). Further, the power and prestige that

the ’land-owning castes’ command in relation to even those castes whichare ritually higher to them stimulates in the lower castes a desire to ’imi-tate’ their ’prestigious style of life’ (ibid.: 17). Thus, dominant castes ’setthe model for the majority of people living in rural areas including, occa-

sionally, Brahmins’ (ibid.: 21).

BOSE AND SRINIVAS: TOWARDS A CRITICAL APPRECIATION

If one makes allowances for the considerable differences that exist betweenBose and Srinivas, the similarities in their approaches and substantive for-mulations, expressed with varying degrees of emphasis, appear quite strik-

ing. Thus, as Andre Beteille has pointed out in his excellent ’Introduction’to The Structure of Hindu society, Bose essentially views anthropology as afield science (Beteille 1975: 10). Similarly, Srinivas has clearly stated thatfield work experience is of critical importance in the career of an anthro-

pologist (Srinivas 1962: 121).As noted above, both Bose and Srinivas view the caste system as crucial

to Hindu society. In observing caste, Bose differentiate between the varnasystem and the local jatis, a distinction more insistently drawn by Srinivas.They have both emphasised the traditional complementary ties that the castesystem provided between different castes (Bose 1967a: 221; Srinivas 1955a:35, 1955b: 34-35). As early as 1941, Bose came to the conclusion that thecaste system was not as immutable as it was believed to be, a point madelater by Srinivas. Further, in his article, Bose emphasised the ’catholicity’of the Hindu method of tribal absorption, a point Srinivas later noted inhis Coorg study by suggesting that the process of penetration of Sanskriticideas and beliefs in the tribal groups was such that no ’violence to their

traditional beliefs’ was done. In 1949, in his Bengali work, Bose noted the



302

tendency amongst the lower castes and non-Hindu groups to imitate therites and customs of the higher order castes and, with the help of the Brah-min priests, to rise in status by making their ’rites and practices moreritually pure or Brahminical’ (1975: 87-88), a process fully elaborated bySrinivas under the name of ’Sanskritisation’. Moreover, the concept of

’dominant group’ that Bose formulates finds correspondence in Srinivas’sconcept of ’dominant caste’. Both Bose and Srinivas make use of the con-

cept’of ’dominance’ for the purpose of understanding the process of cul-

tural transmission.This is not to suggest that such similarities of views between Bose and

Srinivas exclude other observers of the caste system and tribal situation inIndia. Thus, for instance, in a work originally published in 1932, G.S.Ghurye clearly emphasised that to get a sociologically correct idea of the

institution, we should recognize the regionally confined local groups, thesub-castes, as real castes (Ghurye 1969: 20, 27). Further, tracing the originsof the caste system, Ghurye pointed out that the early term varna, mean-

ing colour, was used to specify the ’orders of society’. Later, the word jatiwas employed to denote caste. The word jati etymologically means a groupinto which one is born (ibid.: 176). Similarly, Ghurye has underlined the

fact that the acceptance of the caste system and the social and economicinter-dependence of different castes in a village contributed to ’harmony incivic life’ (ibid.: 28). On tribes, Ghurye noted in a book published in 1943,that the tribal groups in general had a Hinduised section, which soughtto improve its social position by asserting to be Hindus and then ’establish-ing a claim for a status higher than that of the lowest or even lowercastes’ (’.959: 19).To return to Bose and Srinivas, an important similarity in Bose’s formu-

lation of the Hindu method of tribal absorption and Srinivas’s concept ofSanskritisation is with respect to the emphasis put by them on the processof acculturation and cultural change. Bose’s study is admittedly in the fieldof acculturation. Srinivas clearly states that, ’Sanskritisation has been a

major process of cultural change in Indian history’ (1966: 23) and an im-portant tendency that Srinivas has observed with respect to the caste sys-tem is the ’imitation of the ways of higher castes’ (ibid.: 14). In this respect,an interesting point to note is that, whereas the ’cultural’ anthropologistBose accommodates the economic framework in his study of cultural con-tact, the ’social’ anthropologist Srinivas, trained in the tradition of British

anthropology, formulates a concept that indicates basically within culturalterms the direction as well as an explanation of clutural change in Indian

history and society.Bose emphasises ’an increase in the productivity of human labour’ (1961:

103) as a factor in cultural evolution. In fact, there is considerable evi-

dence in Bose’s work of his interest in the material culture of people. AQ



303

excellent example of such an interest is his study of the variety of techni-

ques and the implements used in oil extraction (1975: 73-86). However, itwill be wrong to ignore that for Bose the area of basic interest lies, as

pointed out by B4teille, in ’the ideals of social, economic and political lifeby which each society is governed’ (Beteille 1975: 19). Thus, with respectto the caste system, he is primarily concerned with identifying the Hindu’ideal’ behind the system or the ’idea’ at its root. Such an idealist view leadsBose to a description of the caste system that is formulated in terms ofcultural ’soul’ or social ’design’ of Hindu society.3

This idealism has certain implications. Thus, though Bose gives impor-tance to the criterion of ’productivity of human labour’ in studying cul-

tural evolution, he chooses value-oriented standards for the purpose of

comparison between two cultures, namely, ’the standard of human love’

(1961: 103) or of ’freedom of thought’ (ibid.:107). When the foregoing isviewed in connection with Bose’s emphasis on field work and on lettingthe facts speak for themselves, his approach appears to be an idealist viewlinked with an empiricist enterprise. Bose shares this approach with Srini-vas. This approach leads to certain difficulties.Thus, for instance, Bose is not consistent in his ideal standard of human

love or freedom of thought in concretely comparing the tribal with theBrahminical society. Here, he refers to the empirical criterion of the pro-ductive system or of economic organisation. For Srinivas, let us take thedistinction drawn by him between varna and jati, a distinction that he hasonce again drawn recently in terms of ’book-view’ and ’field-view’ (1978:146). As noted earlier, Srinivas has advised the sociologists to free themselvesfrom the hold of the idea of the varna model for an empirical study of thereal and effective units of jatis. To what extent has Srinivas himself beenable to free himself from this hold?

Srinivas has emphasised that the mobility associated with Sanskritisationresults only in positional changes in the system and does not lead to anystructural change. A caste may move up or down but all this takes placein ’an essentially stable hierarchical order’ (1966: 7). Now, which stablesystem or order does he refer to? He is obviously referring to the varnaorder. Thus, with respect to the ’rise and fall of particular varnas’, he givesthe example of Kshatriya and Vaishya claims in relation to Brahmins in

ancient India (ibid.:31 ). He illustrates the point regarding a degree of

3Such an approach takes Bose away from a concrete study of the relations of

production. Though Bose has studied the implements of production and has empha-sised the need to study caste in relation to occupation and division of labour, inthe absence of a concrete study of the relations of production, his approach is far fromthe study of ’mode of production’, the importance of which was not unknown to him.For example, he was not unfamiliar with the early works of Ramkrishna Mukherjee(1948, 1949, 1957, 1958).



304

topenness’ in the system by referring to the case of Kshatriyas who wererecruited in ancient times from several ethnic groups (ibid.: 32). That Sri-nivas is referring to the varna scheme comes out when he discusses the

possible ’models’ of Sanskritisation: Brahmin, Kshatriya, Vaishya andShudra. The concept of a local dominant caste is absorbed within this sche-me for it is thought to serve as a mediator for different ’models’ of Sans-kritisation (ibid.:l0).4 Further, Srinivas cautions against treating local

village system as completely independent from the wider ’all-India system’(ibid.: 15). After all, he is aware that to the average peasant the names ofcastes in other linguistic areas make sense only when they are fitted intothe varna frame (ibid.: 3), and also, that the mobility of a caste is frequentlystated in varna terms rather than in terms of the local caste situation

(ibid.: 69).Thus, we see that the ideal model of varna is at the core of Srinivas’s

empirical study of the caste system in Hindu society. Yet, he insists thatthe caste system does not fit with the frame of varna ’except at one ortow points’ (ibid.: 7). Having rejected and dissolved the idea of the varna

system from the empirical world of jati, Srinivas surreptitiously acceptsand restores in the end this idea in his empirical analysis. The act ofuncritical adherence to an idealist view and to empiricism reduces the

consistency of his argument.My more serious criticism againt Bose and Srinivas is that, lacking a

general sociological theory of society and social change within the frame-work of which empirical data are to be collected, interpreted and transc-ended, they end up with the transformation of the object of study into atheory that has conditioned the study itself. In other words, in their con-cern with the ideal sphere, they are compelled to accept the ruling ideasof the society, past and present, for providing them with the interpreta-tion of the corresponding empirical reality studied by them. In sum, theiranalysis is ideological.5 5

This is evident in Bose’s work. Bose accepts that in the caste systemdifferent castes attend to society and society attends to them. But,he is unable’ to see that it is intrinsic to the caste system, ratherthan its ’unfortunate part’, that the vanquished were never given an equalstatus in Hindu society. Further, he is unable to say definitely whether thecultural catholicity of the Hindus sprang from ’a genuine sense of humanlove’ or from the desire of the priestly caste to prevent the low castestom rising high in the social scale (1967a: 214). With the recognition that

4The empirical as well as theoretical limitation of the concept of ’dominant caste’has been pointed out by Dube (1968) and Oommen (1970).5That such an ideological position is not peculiar to Bose and Srinivas nor confined

to Hindu society is borne out by an interesting article on ideology and social structurein Indonesia (Kahn 1978: 103-22).



305

the Hindu moneylenders from the plains were exploiting the tribals (1972b:70), Bose could not have exhorted the tribal people to come forward andunite in the adventure of making India ’the homeland of civilization oncemore where one faith does not dominate over another’ (ibid.:80) withoutaccepting as real the Hindu ideal of cultural federalism.The ideological character of Srinivas’s work is even more evident. He

confronts a problem relating to this issue quite early in his career. Thus, _writing on the social system of a Mysore village (1955b), Srinivas clearlyrecognises that the position claimed by a caste frequently differs from theposition conceded to it by others. The sociologist has either to accept oneof these claims or to construct his own picture of the hierarchy. However,he cannot claim complete objectivity for his construction involves theevaluation of the statements made by his informants. Further, he admitsthat there is at present ’a certain discrepancy between the hierarchy as itis conceptualised by the people and as it exists in behaviour. Discrepancyis due to the fact that, in conceptualising the hierarchy, ritual considera-tions are dominant, while in the day-to-day relationships between casteseconomic, political, and &dquo;Western&dquo; factors also play an important part’ibid. :26). Yet, the hierarchy that he constructs has admittedly ’ritual con-siderations as its basis’, that is, ’the castes are arranged in a particular ’

*

order on the basis of ideas regarding pollution, (ibid.:26). It may be put to

question whether the hierarchy conceptualised by different castes, highand low, is necessarily based on ritual considerations, particularly as Sri-nivas himself notes that there are different ’axes of power’, of which ritualis only a part, and that all these ’axes may be said to be implicit in any

single act of contact&dquo; (ibid.:2&).s Even if one does not raise this question,Srinivas does not provide the justification of accepting a particular con-ceptualised hierarchy in preference to the hierarchy expressed throughsocial relations. This is important to note in the context of Srinivas’sown emphasis that society should be studied as it is in reality. It may be

recalled that this forms the ’basis of his explicit though not actual rejec-tion of the varna model in preference for ’caste as it is in fact’.

Such uncritical acceptance of the ritual hierarchy becomes ideologicalin the context of Srinivas’s recognition that ‘Brahminical life was domina-ted by ritual’ (1962: 11), Brahmins had a special role and interest in legiti-mising the claims of a caste in the process of Sanskritisation, and thatthe role of Brahmins was particularly crucial in spreading all-India

Hinduism, chiefly Sanskritic in character. Above all, Srinivas has noted

6For a view of caste ’from the point of view of those at the bottom’, Joan P. Men-cher has presented interesting empirical data (see Mencher 1974). It is not necessarilytrue that ritual considerations are dominant in the conceptualisation of hierarchyby different castes. In this respect, there might be significant differences between ’low’caste and ’high’ caste population (Berreman 1975: 23).



306

that Brahmins were the creators of the varna model of caste, as well asof much of the sacred literature of the Hindus, and had the privilege ofreserving a position of ritual supremacy for themselves.. It strengthenedtheir position to ensure that ’Sanskritic ideology ... emerged as the mostprestigeful of indigenous ideologies’ (1967: 79).

This twin defect of uncritical ideological perspective following from an- equally uncritical approach of idealism coupled with an empiricist enter-

prise sets fetters on the interpretations attempted by Bose and Srinivas.

Thus, Bose is unable to examine ’why and how the set of dominant thou-ghts changes in a civilization’ (1961: 23). Srinivas similarly leaves manypoints unanswered or inadequately grasped. Let me take a few examples.Srinivas wonders how people living in villages were made to obey thecaste rules or punished for violating them (1966: 5). Further, Srinivas hasclaimed that the Brahmin varna may be regarded as a ’better’ model of Sans- ,

kritisation. Why is it so? He avoids the question by offering the followingobservation: ’It would be fascinating to trace the gradual emergence over

the centuries of puritanical style of life as a dominant feature of Hinduism,and the association of that style of life with Brahmins ... , but that is notmy task here’ (ibid.: 25). The unity of the Brahmin varna cannot be an

answer for that would raise the question, even if such a unity is acceptedfor the sake of an argument, as to why there has been greater unity in thatvarna. Srinivas is forced to turn to ritual considerations for an explana-tion and he offers the suggestion that the Brahmin varna provides a

’better’ ’model’ because the Brahmins are ’most particular about the perfor-mance of these (Vedic) rites’ (ibid.: 8)The problem involved in Srinivas’s position is obvi ous. He prefers the

abstraction of hierarchy in ritual terms against the hierarchy expressedthrough.social relations, notes the empirical fact of closer association ofBrahmins with rituals, and argues that the Brahmin varna provides’ abetter model of Sanskritisation because the Brahmins are most particularabout these rituals. In this stand of Srinivas, the sociological explana-tion of the empirical position of the Brahmins does not go beyond theideological justification that this group itself has created for itself its

privileged status in the society.’

7That one need not be a Brahmin himself for having adopted the Brahminical ideo-logy in one’s work on Hindu society is best illustrated by Louis Dumont (1972). For acriticism of Dumont for his Brahminical view of caste see Berreman (1975). Srinivashas been more recently criticised for his Brahminical ’odyssey’ (Parvathamma 1978).Ravindra K. Jain has pointed out that Srinivas’s approach is particularly suited for

providing a view of the community from the top, that is, from the ’vantage point of thepowerholders’. Further, Jain notes that, Srinivas has tried very hard to keep his cog-nitive process sharply differentiated from those of the villagers studied by him. Thus,he seems to have emerged unsullied and resplendent as does ’the lotus from the sur-



307

Such specifically Brahminical bias, however, does not exist in Bose.In understanding Bose it is good to remember the complexity of his

personality. His commitment to nationalism and the national freedom

struggle, his adherence to Gandhism with personal devotion to MahatmaGandhi, even though he had reservations about him, and Bose’s, involve-ment in the cause of the national unity must be noted.’ It is not widelyknown that Bose made personal organised efforts to save the Muslims

during the worst days of communal riots in Calcutta. These values pro-vided Bose with a broader view of Hinduism and its role in society.He was opposed to the ’egocentricism’ of capitalism and he also rejectedwhat he called the ’sociocentricism’ of socialism, which he regarded as a’society of ants’ (1975: 170). He was looking for an alternative and this wassuggested to him in an approach that would penetrate the smoke that wasthere in traditional society in India. He wanted to use the ’burning flame’beneath it for the happiness of mankind. This was the purpose of his

knowledge as far as he himself defined it.

SOME SUBSTANTIVE ISSUES ABOUT THE COORGS

In a recent answer to his critics Srinivas has emphasised that he arrived atan idea of the traditional Indian society being far more flexible and permis-sive of mobility than was generally accepted when he started his work.This he has done, claims Srinivas, starting with Radcliffe-Brown’s func-

tionalism which is supposed by its critics to be ’particularly unsuitable forthe analysis of change’ (1978: 147). This point made by Srinivas deservesa particular note because it raises an issue that is of a broader theoreticalsignificance as well as of relevance for understanding the process of Sans-kritisation as elaborated by Srinivas.

I propose to consider Srinivas’s functionalism in relation to his work onthe Coorgs, a study which is deservedly acknowledged as his best work sofar (Madan 1978: 3). Besides, this work was most directly influenced atthe writing stage by the master himself. What does Srinivas achieve in hisstudy of the religion and society among the Coorgs? As Radcliffe-Brownsees it, Srinivas has studied ’Hinduism as a system of ritual and belief’

rounding mud’ (1978: 50-51). In relation to Berreman, while I agree with what he haswritten about Dumont, I do not think that the view of caste that Dumont holds wouldhave been changed had he been more attentive to the empirical literature on caste andvillage India. For, should my critique developed so far be acceptable, it may be seenthat one may convey a view of caste that is ’idealized’ even with considerable empiricalfocus.8Sinha has noted that Bose insisted on pointing out that a major source of his learn-

ing was through involvement in the ’problems of regeneration of a tired ancient societygroaning under the shackles of colonial rule’ (Sinha 1972: 4).



308

and has helped us to see ’the social function of these rituals as expressingthe solidarity and continuity of the structural system’ (Radcliffe-Brown1952:x, viii). Srinivas has described various features of the cult of okka

(patrilineal and patrilocal joint family) as well as the cults of the largersocial units, particularly of the village.

This is consistent with the method to which Radcliffe-Brown contributed

significantly. At the time when Radcliffe-Brown was working, he was

basically responding to the ’conjectural history’ approach of ethnology.In an article, characteristically entitled, ’Historical and functional inter-

pretations of culture in relation to the practical applications of anthropo-logy to the control of native peoples’, the ’founder’ of the scientific studyof social anthropology makes a strong case against the speculative histori-cal method. His argument against such a method rests on the followingpoints: (i) The historical method is not applicable for ’uncivilized peoples’without historical documentary records; (ii) the validity of the hypotheticalreconstructions ’depends on that of the assumptions (generally implicit)on which they are based’; and, (iii) the ’greatest weakness’ of the methodis that it is entirely devoid of any practical value (Radcliffe-Brown1958: 39-41).

Against this, the functional method of analysis is of immense practicalvalue, more particularly for those who are concerned with the governmentof the ’backward peoples’ (ibid.: 37). Besides, the functional method of

social anthropology relies solely on facts and on well-authenticated obser-vations of the facts (bid.: 26).The fundamental assumption that the functional method makes is that

’a culture is an integrated system. In the life of a given community eachelement of the culture plays a specific part, has a specific function. The

discovery of those functions is the task of a science...’ (ibid.: 40). It is notmy task here to examine the weaknesses of this position. Suffice it tomention here that the functional method of Radcliffe-Brown suffers fromthe same limitation as identified by him for the historical method of

ethnology, namely, the functional interpretation is as much dependent, if

not more, on the assumption that it makes. Besides, contrary to its in-

ductive emphasis, a priori general reasoning is necessarily involved in itsformulations.

Consider the following significant sentence from Radcliffe-Brown:

if it is valid generalisation to say that the chief function of ritual or

ceremonial is to express and thereby maintain in existence sentimentsthat are necessary for the social cohesion, we can &dquo;explain&dquo; any givenritual or ceremony by showing what are the sentiments expressed in itand how these sentiments are related to the cohesion of the society(ibid.: 41 ).



309

If the assumption of the cohesion of society and that each componentof society has a specific function is made anyway, then, there is little togain, except richness of an ethnographic description, by way of a detailedstudy. For, even if it were possible inductively to arrive at a formulationafter considerable and painstaking comparative studies of the followingkind, ’the chief function of ritual is to contribute towards the cohesion ofan integrated system’, we would still not know more than what we startedwith within a functionalist enterprise.9 Moreover within this approach,there would generally be no possibility of arriving at facts that would

’speak’ differently. Thus, for instance, the functionalist Srinivas was con-cerned with showing the vertical unity in the village of Rampura, and hehonestly admits himself, that it had not occurred to him during his fieldwork how important it was to examine how far the unity of the villagereally included such polar groups as the Brahmins and the Untouchables,and also a peripheral group like the Muslims (1955b: 33).

Turning now to the Coorg study, Srinivas has noted the prevalence ofthe pre-Sanskritic religious forms amongst the Coorgs, expressed in theform of ancestor-propitiations by animal sacrifices, offering of meat and

liquor with participation of the oracles (1952: 159-67), offering of meatand liquor to the ‘village deity’ (ibid.: 171), and so on. He has observedthe unorthodox diet of the Coorgs which includes pork and liquor andthat the Coorgs do not perform any of the Vedic rituals and mantras are

not recited when a Coorg is given a name, nor on the occasion of his

marriage, or death (ibid.: 33). Unlike the Amma Coorgs, who numbered666 persons only according to the 1941 Census, the rest of the Coorgs donot wear the sacred thread (ibid.: 34). All this suggests, at the least, im-perfect Sanskritisation. This raises an important question: Why is it thatthe bulk of the Coorgs are imperfectly Sanskritised?An adequate answer to this question cannot be given in terms of their

objectionable dietary habits of the Coorgs which they are unable to giveup because more than these preferences the pre-Sanskritic religious formsand other practices are involved. We cannot look for an answer in therefusal of the Brahmins to afford acceptance to the Coorgs for, as Srinivashas vividly described, the Brahmin priests have gone a long way toaccommodate the pre-Sanskritic religious forms of their aflluent clients

(ibid.: 194-97). Nor can it be acceptable in terms of the values of the

qoorgs themselves that the Amma Coorgs are decades, if not centuries, in

9In this respect Levi-Strauss has a relevant point to make: ’What interests the an-thropologist is not the universality of the function—which is far from definitely establi-shed, and which cannot be asserted without a careful study of all the customs of this

type and their historical development—but, rather, the fact that the customs are sovaried’ (Levi-Strauss 1977: 14).



310

advance of the broader group of Coorgs, as Srinivas seems to believe

(ibid.: 167). These Brahminised Coorgs do not appear to have much statuseven in ritual terms for, as Srinivas records, the Brahmins of SouthCanara are priests at very many temples in Coorg and the Coorgs had(and continue to have) considerable respect for them (ibid.: 183).

Srinivas gives us an excellent ethnographic description of the ritualsand beliefs of the Coorgs, the social function of these rituals, and the in-formation that the Coorgs were greatly imitated by people around themwho accepted the Coorg’s dominance. But, yet, we do not know why theCoorgs themselves did not Sanskritise their practices more thoroughly.

Is it possible that the Coorgs did not choose to Sanskritise their way oflife more thoroughly not because of their being ’conservative’ and beingbehind the small group of Amma Coorgs, as Srinivas seems to believe, butbecause of very different reasons? Two of the Coorg proverbs may be re-

called here, ’Him who holds sway we must obey’ and ’He who can spendmoney is accounted an honourable man’. These spirited hill people who inthe British eyes resembled more the Scottish Highlanders than the Hindu

castes of the plains of India by and large made their peace with theirBritish masters after the annexation of Coorg by the British in 1834, per-

haps more particularly so the affluent~ section of the Coorgs.l° Their new

10This is not to deny the participation of the Coorgs later in the national movement.The influence of Mahatma Gandhi was felt there too and the Coorgs took some partin his ’non-violent war’. Mahatma Gandhi visited Coorg in February; 1934; and headdressed large public meetings there (I.M. Muthanna 1953: 136). With respect to theBritish annexation; it is important to remember that the annexation was pre-plannedand the statement made by the British of the ’unanimous wish’ of the inhabitants to

be ruled by the British untrue. As it has been noted by a Coorg scholar, ’the Britishgot fullest advantage from the Dewans of the day who betrayed the people in order toachieve their own ends and in which they were evidently successful’ (ibid.: 70-71).Even in 1834; there was considerable economic differentiation amongst the Coorgsthemselves. Thus, according to an estimate, the privileged jamma land tenure was heldin the following manner:

Soyrce: I.M. Muthanna (1953: 154).



311

masters found them ’a race more loyal and devoted to the British Govern-ment than ... any in India’ (Cole 1897: iii) and, in turn, the Coorgs sang,

,

’By the grace of our Queen/All the people of this country/Suf~er neitherwant nor hardship’ (Richter 1887: 48). It has been admitted that, underBritish rule, ’an indulgent Government and enthusiastic local administra-tors did their part to help them into the saddle’ (ibid.: 22). Under thecategory of ‘the landlord or warrior caste’ only the Coorgs were so classi-

fied, who were allowed to continue as lords of the soil (Richter 1870).,Under these conditions, is it possible that Hinduism did not any longerrepresent the ruling dominant ideology, which it did under the rule of the

Lingayat Rajas,’-1 from whom too the Coorgs had received considerablefavours? Could it be that the Coorgs adjusted to the changed reality anddid not find it necessary for the maintenance of their privileged positionto further adopt the Sanskritic ideology? This question is relevant even

though there was perhaps no reversal, instead, the Coorgs tried to maketheir belief system more consistent, which was easier with the spread ofeducation amongst them. 12

.

This suggests a possibility that remains unexplored. However, it is clearthat, under the British rule, the Coorgs could continue with their existingway of life because the British rulers were careful not to disturb the sourceof their traditional dominance and material relations; in fact, the favouredjamma tenure was confirmed by the British government.13 The introduc-tion of coffee plantation, however, had consequences along expectedlines.

Even though Srinivas takes a brief note of the presence of the British inCoorg in his introductory chapter of the Coorg study, the implications ofcolonial rule are not woven in his functional interpretation of religion andsociety among the Coorgs. Such a line of inquiry could have been relevant.Thus, for instance, Srinivas treats the Coorgs in general as a homogeneousgroup. However, we know that it was the conscious policy of the Britishrulers to form ’a kind of otficial aristocracy’ amongst the privileged Coorgs.The principal offices in Coorg had been monopolised by a few families who

11The rule of the Lingayat Rajas was, according to a Coorg opinion, imposed by’deceit’ and the rule of the dynasty was despotic, carried out with ’hated tyranny’(Pandanda Muthanna 1931: 31).12One of such efforts was the exposure of the Brahminical fabrication of the Kav&emacr;ri

Pur&amacr;na as an ’eternally glorifying monument to their perversive ingenuity’ (PandandaMuthanna 1931: 69). The importance of this view is to be assessed in the context of thefact that Muthanna’s work was originally published in Kanarese as early as 1927 andit had a good circulation. Muthanna holds the view that the Coorgs are Aryans andthat they are derived from the ’Chalukya lineage’.13The number of jamma tenure holders increased to 3,275 in 1866 (see I.M. Muthan-

na 1953: 156).



312

made good use of their opportunities to acquire land, wealth and influence(Richter 1887: 22). It would have been of interest to know the role of theseespecially privileged families. In religious terms, did they constitute a moreSanskritised section among the Coorgs? What was their involvement in thepre-Sanskritic religious forms? Was it from these families that the more

~enterprising Coorg individuals emerged who broke ofI’ from the traditionaljoint family and established their own family units? Srinivas does not askthese questions. Yet, these questions are relevant because Srinivas has him-self acknowledged that some Coorg families were more Sanskritised thantheir neighbours (1952: 38).

Srinivas’s approach does not explain inconsistency and variation insocial life. This is another limitation of the functional method.&dquo; The varia-tions from the abstracted regularities are ignored, there is little interestin the problem of conflict, including conflict resulting from the introductionof a new religion or a new government. Thus, Srinivas not only finds his

assumption of consistency, homogeneity and relative stability unshakenby the introduction of the British rule in Coorg but also the spread ofSanskritic Hinduism is seen as contributing to such stability even thoughstrong pre-Sanskritic religions forms existed and continued to exist. Myown study of the prevalence of different religious forms amongst the

Tamangs of Darjeeling suggests several variations which. have serious

implications on the social life and ethnic identity of the group (Munshiand Lama 1978).

Leaving these issues aside, let me now consider the concept of Sanskriti-s-ation itself. Srinivas has suggested that the concept emerged through anempirical study of the social and religious life of the Coorgs (1962: 42).However, a closer examination of his book on the Coorgs suggests the .following sequence in the context of presentation of the concept. The firstreference to Sanskritisation that Srinivas makes is in the context of a

general account of the caste system in India. The second reference, which

refers to the Amma Coorgs, also relates to ’a tendency which has alwaysbeen present in the caste system’ (1952: 35), and so on. Thus, Srinivasintroduces a general notion, developed a priori, in his effort to- empiricallyexamine a given reality of the Coorgs. However, the basis of this formu-

lation of a prior general notion remains unclear. In fact, to appear con-sistent with his empirical approach, Srinivas has later presented the notionto appear as an outcome of his specific empirical study. Thus, it is not

true, as Srinivas seems to believe, that he was forced out of his microshellin the course of the Tagore Lecture for 1963 (1962: 2). He has been mak-ing useful statements claiming to hold good for Hindu society and Hindu-ism right from the days of his Coorg study.

14For a criticism of this kind of field work, see Van Velsen (1967: 133-37).



313

The discrepancy between Srinivas’s general notion of Sanskritisationand his study of the Coorgs is made obvious when we consider that the

description of Sanskritisation that Srinivas gives in the book, as quotedearlier in this article, refers to ’vegetarianism’ and ’teetotalism’; in short,to the Brahminical ’model’ of Sanskritisation. However, the main body ofthe Coorgs, if anything, manife~t a natural affinity with the Kshatriya wayof life rather than suggest a cultural imitation of the Brahmins

Thus, if it is true that the concept was developed a priori by Srinivas,then, we need to look for the sources of this general notion. Srinivas hashimself referred to Suniti Kumar Chatterjee who actually used the word’Sanskritisation’ in a work published in 1950 (Srinivas 1967: 67). The con-cept was not unknown when Srinivas was doing his held work. It has beennoted earlier that Bose had talked of the soread of Brahminical influencebefore Srinivas did so. It has been pointed out elsewhere that the Britishadministrators like Lyall and Risley had anticipated the concept beforethe close of the last century (Mukherjee 1977: 49) Moreover, it is notknown that two of the British observers of the Coorgs, to whom Srinivas

frequently refers in his study, had already developed understanding of the‘Brabmanisation of Coorg tradition’ (Cole 1897: V) and the process, ini-tiated by the Brahmins, of ’Brahmanising the worship of the river Kaveri

(Richter 1887: 42) much before Srinivas took up his field work.The work of Brahmins in incorporating the Coorgs in the Brahminical

fold was noted as ’a matter of necessity’ by Cole. He wrote:

The Brahmanisation of Coorg tradition presented no small difficulty, onaccount of the tough materials of the wild world of Coorg. These illi-terate and untameable hunters seem to have ever had an instinctive

antipathy to, and thorough contempt for, the sanctities and pretensionsof the smooth and crafty Brahmin ... It was ... rather a difficult task todress up a history of Coorg in Brahminical fashion. But the attempt wasa matter of necessity. The story of the country from which the holyKaveri descends into the eastern plains, could not be left to the Coorgthemselves. It had to take its place in the Kaveri Purana and to be inharmony, as much as possible, both with the realities and traditions of

Coorg (Cole 1897: V).

Similarly, Richter has observed, ’the Brahmins who are domiciled in

Coorg have succeeded in introducing Mahadeoa and Subrahmanya, in en-tirely Brahmanizing the worship of the river Kaveri, in having templeserected and idols set up, in spreading puranic tales and in usurping to

some extent the puja at the places of the worship’ (Richter 1887: 42). ThatRichter was aware of the broader cultural process at work is borne out bythe following observation, ’superior Brahmanical priestcraft found a way



314

either to adapt itself to existing superstitions or to envelop them in pura-nic lore and thus disguised a new religious system with deities from theBrahmanical pantheon, found gradual acceptance in addition to and mix-ed up with the old popular belief’ (ibid.: 3).Among other points, a crucial difference between the early British ob-

servers of the Coorgs and Srinivas lies in the fact that, whereas Srinivas

emphasises the internal process on the part of the Coorgs to Sanskritisethemselves, as, for instance, with respect to Amma Coorgs who broke offfrom a larger whole of which they were a part and Sanskritised their cus-toms and rituals (1952: 35), or in stressing that the cultural and ritualforms found among Coorgs should not be in the majority of cases beattributed to diffusion from outside (ibid.: 183), the British observers havenoted in the process of Brahminisation an external imposition by theBrahmins. As far as the substitution of the term ’Brahmanisation’ by theterm ’Sanskritisation’ is concerned, it has not been a very happy choicefor more than one reason. As J.F. Staal, a scholar of Sanskrit and classi-cal Indology, has pointed out, ’the term itself seems to be misleading,since its relationship to the term Sanskrit is extremely complicated’ (Staal1963: 261). Further, Staal has pointed out that Sanskritisation coverscases where the influence of Sanskrit and the amount of Sanskrit materialhave decreased (ibid.: 275).

Thus, if Srinivas has arrived at an idea of the traditional Indian societybeing relatively flexible and providing for some mobility, it is not derivedfrom Radcliff-Brown’s functionalism. In fact, as I have tried to show, hisfunctionalism makes it difficult for him to adequately explain the incom-

plete Sanskritisation of the Coorgs.

- CONCLUDING REMARKS

My criticism presented above should not detract from the significant con-tributions made by Bose as well as Srinivas in our effort to understandIndian society. They are deservedly acknowledged as the doyens in thefield. My attempt here has been to show the limitation of their method.The limitation of their method that has been highlighted should warn

us against sharing the ’conception’ and ’illusion’ of a particular peopleand epoch (Marx and Engels 1976: 63). What Marx and Engels wroteduring 1845-46 is still of great relevance to students of sociology inIndia:

,

The ideas of the ruling class are in every epoch the ruling ideas: i.e., theclass which is the ruling material force of society is at the same time its

ruling intellectual force. The class which has the means of material pro-duction at its disposal, consequently also controls the means of mental



315

production, so that the ideas of those who lack the means of mental. production are on the whole subject to it. The ruling ideas are nothing

more than the ideal expression of the dominant material relations, thedominant material relations grasped as ideas; hence of the relationswhich make the one class the ruling one, therefore, the ideas of its domi-nance (ibid.: 67).

We need to study ’the dominant material relations’ more closely. Workin this area has yet to be done. Further, we need to recognise that obser-

vations always imply interpretations within the framework of previoustheories (Habermas 1976: 202). This makes it necessary to adopt and deve-lop a consciously formulated sociological theory of society and socialchange, a need that has remained largely unfulfilled here in India.15 Forsuch a purpose, a functionalist approach derived from Radcliffe-Brownis no answer.

References

ATAL, Y. 1968. The changing frontiers of caste. Delhi: National Publishing House.BAILEY, F.G. 1960. Tribe, caste, and nation: a study of political activity and political

change in highland Orissa. Bombay: Oxford University Press.BERREMAN, G.D. 1975. The Brahmanical view of caste. Contributions to Indian socio-

logy (n.s.)BETEILLE, ANDRE. 1975. Introduction. In N.K. Bose, The structure of Hindu society.

New Delhi: Orient Longman.BOSE, N.K. 1961. Cultural anthropology. Bombay: Asia Publishing House.———. 1967a. Culture and society in India. Bombay: Asia Publishing House.———. 1967b. Problems of national integration. Simla: Indian Institute of Advanced

Study.———. 1971. Tribal life in India. New Delhi: National Book Trust.———. 1972a. Some Indian tribes. New Delhi: National Book Trust.———. 1972b. Anthropology and some Indian problems. Calcutta: Institute of Social

Research and Applied Anthropology.———. 1975. The structure of Hindu society. New Delhi: Orient Longman.COLE, R. 1897. A manual of Coorg civil law. Coorg: Mercara Jail Press (reprint).DUBE, S.C. 1968. Caste dominance and factionalism. Contributions to Indian sociology

(n.s.) 2:58-81.DUMONT, LOUIS. 1972. Homo hierarchicus: the caste system and its implications.

London: Paladin (paperback).GHURYE, G.S. 1959. The scheduled tribes. Bombay: Popular Book Dept.———. 1969. Caste and race in India. Bombay: Popular Prakashan.HABERMAS, J. 1976. A positivistically bisected rationalism. In Theodor W. Adorno

et al., The positivist dispute in German sociology. London: Heinemann EducationalBooks.

15For a recent reminder of this limitation, see Yogendra Singh (1977).



316

JAIN, R.K. 1978. Lotus in the mud pond? Contributions to Indian sociology (n.s.) 12,1: 49-55.

KAHN, J.S. 1978. Ideology and social structure in Indonesia. Comparative studies in

society and history 20, 1: 103-122.LEVI-STRAUSS, C. 1977. Structural anthropology. Middlesex: Penguin Books.MADAN, T.N. 1978. M-N. Srinivas’s earlier work and The remembered village: an

introduction. Contributions to Indian sociology (n.s.) 12, 1: 1-14.MANDELBAUM, D.G. 1972. Society in India. Bombay: Popular Prakashan.MARX, KARL and F. ENGELS. 1976. The German ideology. Moscow: Progress Pub-

lishers.

MENCHER, JOAN P. 1974. T. e caste system upside down, or the not-so-mysteriouseast. Current anthropology 15, 4: 469-78.

MUKHERJEE, R. 1948. Economic structure in rural Bengal. American sociologicalreview. December, 14, 2.

———. 1949. The economic structure and social life in six villages of Bengal. Americansociological review 14, 3: 415-25.

———. 1957. The dynamics of a rural society. Berlin: Akademic-Verlag.———. 1958. Six villages of Bengal. Calcutta: Asiatic Society of Bengal.———. 1977. Trends in Indian sociology. Current sociology 25, 3.MUNSHI, S. and U. LAMA. 1978. The Tamangs of Darjeeling (in two parts). Journal

of the Indian anthropological society [to be published soon].MUTHANNA, I.M. 1953. A tiny model state of south India. Pollibetta, Coorg.MUTHANNA, PANDANDA. 1931. Coorg and the Coorgs. Siddapur, Coorg: C.M.

Ponnappa.OOMMEN, T.K. 1970. The concept of dominant caste: some queries. Contributions to

Indian sociology (n.s.) 4: 73-83.PARVATHAMMA, C. 1978. The remembered village: a Brahmanical odyssey, Contribu-

tions to Indian sociology (n.s.) 12, 1: 91-96.

RADCLIFFE-BROWN, A.R. 1952. Foreward. In M.N. Srinivas, Religion and societyamong the Coorgs in south India. Bombay: Asia Publishing House.

———. 1958. Method in social anthropology. Chicago: The University of ChicagoPress.

RICHTER, G. 1870. Manual of Coorg: a gazetteer of the natural features of the countryand the social and political conditions of its inhabitants. Mangalore: Basil MissionBook Depository.

--

——. 1887. Ethnographic compendium on the castes and tribes found in the province ofCoorg, with a short description of those peculiar to Coorg. Bangalore: MorningStar Press.

STAAL, J.F. 1963. Sanskrit and Sanskritisation. The journal of Asian studies 22, 3:261-275.

SINGH, Y. 1977. Social stratification and change in India. New Delhi: Manohar BookService

SINHA, S. 1972. Anthropology of Nirmal Kumar Bose. In S. Sinha, ed., Aspects ofIndian culture and society: essays in felicitation of Professor Nirmal Kumar Bose.

Calcutta: The Indian Anthropological Society.SRINIVAS, M.N. 1952. Religion and society among the Coorgs of south India. Bombay:

Asia Publishing House.———. ed., 1955a. India’s villages. Bombay: Asia Publishing House.———. 1955b. The social system of a Mysore village. In McKim Marriott, ed., Village

India: studies in the little community. Chicago: The University of Chicago Press.———. 1962. Caste in modern India and other essays. Bombay: Asia Publishing House,



317

———. 1966. Social change in modern India. Bombay: Allied Publishers.———. 1967. The cohesive role of Sanskritisation. In Philip Mason, ed., India and

Ceylon: unity and diversity. London: Oxford University Press.———. 1976. The remembered village. Delhi: Oxford University Press.

———. S. SESHAIAH and V.S. PARTHASARTHY, eds., 1977. Dimensions of social changein India. New Delhi: Ailied Publishers.

———. 1978. The remembered village: reply to criticisms. Contributions to Indian

sociology (n.s.) 12, 1: 127-52.VAN VELSEN, J. 1967. The extended-case method and situational analysis. In A.L.

Epstein, ed., The craft of social anthropology. London: Tavistock PublishersLimited.



Tribal absorption and Sanskritisation in Hindu society

Documents