The Indian Journal of Social Work, Vol. XXV, No. 2 (July 1964).

THE PROBLEM OF JOHN V FERREIRA MARATHA TOTEMISM

In the Marathi-speaking areas of Western India, there are several castes and tribes which exhibit the phenomenon of the devak. Many interpreters of the phenomenon, which is found at its strongest among the Marathas and the occupational castes related to or influenced by them, and less strongly among the primitive tribes of the. region, have characterized it as a form of totemism. A few interpreters, however, have called the totemic" character of the devak into question. The object of this article is, therefore, to re-examine the factual evidence and the more prominent of its interpretations in order to arrive 'at the Origin, the true nature " and the significance of the phenomenon. - - "Mr. Ferreira is Reader in Sociology, University of Bombay.

An early observer, James Campbell, noted The most detailed accounts of the that the Marathas of the Bombay Presidency Maratha devaks, however, have been given were divided into families, each with its to us by R. E. Enthoven. Referring to the devak or sacred symbol The devaks were Marathas. proper, to the Maratha Kunbis, patrilineally inherited, and worshipped at and to the Maratha occupational castes (the marriages and other important occassions. Bhandari, the Chitrakathi, the Gavandi,, the Persons with the same devak were not Kumbhar, the Lohar, the Mali, the Nhavi, permitted to marry. Of the devak listed, the Parit, the Sutar, the Taru., the Teli and 18 are trees and their products and 9 are so on), he says that their exogamous groups inanimate objects. One refers to a Brahminic have devaks or marriage guardians which in rishi or ancestral sage, and the meaning of their origin seem to have . been the, exact another is unknown. One of the devaks is equivalents of the Kanarese, balls or totems. the panchpalvi or leaves of five trees—the The devaks of the Marathas and related Ficus indica, the Ficus religiosa, the castes are not as highly revered as the balis; Cynodon dactylon, the Bauhinia racemosa but similarity of devaks is a bar to marriage, and the' Eugenia jambolana. Carhpbel and some of the Maratha kuls are named classed the Maratha devaks and the Kanarese after devaks (More after mor, Salunke after balls together, and regarded them as guardiari salunki). The devak is usually some spirits and ancestors and heads of house common tree like the bel (Aegle marmelos); (Fxazer 1910: II: 276-278; Enthoven the pipal (Ficus religiosa), the Vad (banyan); 1911 : 65). or the shami (Prosopis spicegera). In its commonest form it is the leaves of five trees, So far as the Marathas in the Central of which one is held specially sacred and Provinces (old provincial names have been was the original devak of the section. The retained here for the sake of convenience devak is worshipped at marriages, at the time are concerned, Russell and Hiralal (1916: of entering a new house, when the IV: l98-204) report that they have several threshing floor is prepared at harvest time; exogamous clans with Rajput names and during the thread ceremony. In (Chauhan, Panwar, Solanki, Suryavahsi), Maratha marriages the installation of the others named after animals and plants or devak is an important part of the marriage natural objects (iguana; date-palm), and still ceremony. The nature of the ritual, however, others of a titular nature, varies in details among the various castes 136 JOHN V. FERREIRA

that practise it. In general it takes the Brahminic fashion. But the majority of the following form: The persons concerned go caste do not know to which gotra they to the woods, worship the tree involved and belong, and there is a great confusion in the cut a twig from it. A visit is then made assignment of kuls to gotras. Furthermore, with the twig to the temple of Maruti where the sameness of gotras is not necessarily a the devak is formally worshipped. In some bar to marriage, whereas the sameness of kuls instances a potter's house is visited also. The and devaks is. Therefore, Enthoven concludes twig is then brought to the marriage booth that the kuls and gotras were adopted after and tied to one of its posts or placed among the tribe rose in social status (Enthoven 1909; the household gods. Lights are waved before 1911; 1922: III: 3-40; 1924a; 1924b; 1924c: it, and a lamp is kept burning in front of 18-20, 208-221. For a variation in the form it for the entire period of the ceremony. No of devak worship, see Enthoven 1924b: 210- mourning is observed during this period. 211). After the marriage ceremony the devak is once again worshipped. Two or four days To turn lastly to an Indian investigator later it is requested to depart; and then of kinship organisations in India: In her book together with the marriage coronet is taken on that subject, Irawati Karve points out to a distant place and left there. Often the that some clan names of the Marathas are devak is also installed at the bride's house of Rajput origin, others are names of animals, with a similar ceremony; and sometimes an qualities or artefacts (the tiger, the peacock, image of it takes its place. Of the 80 devaks black, white, the pick, the axe and so first listed, 60 are trees, plants, fruits and on), and still others of other types. A totemic flowers, 11 are animals, and the rest of other connection between the clan and the animal kinds, many being instruments of industry. name it bears does not always seem to be In a later article the number is put down present, although sometimes the flesh of the at 120, of which 80 are trees and the like, animal is not eaten by the clan concerned. 24 animals and the rest other objects. The rule of exogamy, however, depends not on the clan name, but on the symbol or There is also evidence of totem taboos. devaks connected with the clan. A The predominance of trees among the devaks devaka may be a living thing or an inanimate may be due to the belief that the ancestral object like a kind of grass, a javelin or the spirits live in trees, and thus the trees leaves of five particular trees. Worshipped become guardians or devaks. The Kanarese at the time of marriage, the devaks have an castes do not have the composite totem or exogamous force, but are borne as family panchpalvi, a combination of separate devaks. names only by a few. Nor do many people In their attempts to rise in the social scale know what their devaka is, the elders in the the tribes and castes manifest a common main being repositories of this knowledge. process of evolution by discarding their The Maratha system of exogamous, totemic totemic divisions for kuls or family stocks, clans is analogous with that of the south and which in turn are replaced by Brahminic might have been derived from there. On the gotras. The devaks which possess or once other hand, like the Rajputs and the Khatris possessed all the attributes of real totems are, of northern India, the Marathas also have therefore, survivals, and are hence a a hypergamous arrangement of clans. valuable indication of racial origin. In While the family in Maharashtra is patri- addition to their kuls or surnames, the local and inheritance is in the male line, there Marathas also claim gotras named in the are many customs in the area which are THE PROBLEM OF MARATHA TOTEMISM 137

unknown in the north of India, but are their wood. Like the other Marathi-speaking found universally in the south (Karve 1953: castes, they are patrilineal and profess 156-163). Hinduism (Enthoven 1920: I: 347-350). There are several occupational castes in the The Gavadas a caste of salt-makers in Marathi-speaking area, which are related to Kanara and Ratnagiri, are, says Enthoven, or influenced by the Marathas, and which in the course of passing from "a primitive display the devak phenomenon with varia­ totemistic organisation by devaks into the tions. We shall describe the evidence among Brahminical one of gotras through the a number of them as illustrative material. Maratha system of kuls". Some families still show reverence towards the trees representing The Bhandaris are found everywhere in their devaks by not cutting them or touching the Bombay Presidency, have eight endo­ them or dining on their leaves. But devak gamous divisions and several exogamous exogamy is fast dying out, its place being- kulas. Some kula names are shared with the taken by kul exogamy. The kuls are identical Marathas. Members of the kulas show with the surnames. Most reports state that reverence towards the devaks—the banyan Kashyap is the gotra of the entire caste; but tree, the pipal tree, the kadamba tree, the another gives the gotras Vatsa and Bharad- umbar tree and the mango tree. The kulas waja in addition, and adds that marriages are referred to by the names of such trees. are prohibited between members of the same The devaks are reverenced by not being cut, kul and also of the same gotra. Members of burnt, or injured, and by being worshipped the Kashyap gotra, however, are allowed to at marriages. In some places, however, inter-marry. One account states that the persons with the same devak nevertheless entire caste has the same devak—the kalamb marry. Some Bhandaris have now begun to (Anthocephalus cadumba) tree; others, on claim Brahminic gotras. They are toddy- the other hand, assert that every kul has a tappers by occupation, drink alcohol and eat separate devak. In Vergula the devaks are flesh, are.Shaivites and are patrilineal in the the sword, seven kinds of earth, seven kinds Hindu fashion (Enthoven 1920: I: 96-103). of flowers, rice grains, cocoanuts, betel nuts, The Maratha Chambars of the Deccan, a piece of cloth, and a twig of the kalamb Konkan and Karnatak have no exogamous tree; in Shiroda it is the kalamb tree. They divisions other than families. But marriages follow the Hindu religion and are patrilineal are also prohibited among those who worship according to Hindu usage (Enthoven 1920: the same devaks which are mainly trees and I: 359-363). the panchpalvi. They profess Hinduism The Kumbhars or potters have 23 endo­ (Enthoven 1920: I: 263-268). gamous divisions. The chief of these are The fishermen caste of Gabit in Kanara the Maratha, the Konkani, the Gujarati and and Ratnagiri has recently adopted gotras, the Pardeshi. The Maratha Kumbhars but traces of kuls and devaks suggest that resemble the Maratha Kunbis in dress and the Gabits are of Maratha origin. Among appearance. They have exogamous units the 8 devaks listed, 7 are trees and one the based on surnames. In Poona the devaks or panchpalvi. Although similarity of devaks is marriage guardians wear a wristlet of a creep­ not necessarily a bar to marriage, they still ing plant, the maryadvel (Ipomoa biloba) retain traces of totem taboos; for instance, tree, the potter's wooden platter and the hoe; by not dining on the leaves of their devak in Satara the mango tree, the jambul tree and trees or by not cutting them or burning the vad (Ficus bengalensis) tree; in Sholapur 138 JOHN V. FERREIRA the potter's platter and the panchpalvi; and Kurmi in the United Provinces. The in other places only the panchpalvi. In Nasik Kunbis are found all over the Bombay when the leaves of the five trees or panch­ Presidency and have five territorial groups palvi are not available, mango leaves are —the Maratha, the Konkani, the Khandeshi, substituted. Enthoven, therefore, maintains the Talheri of Thana and the Kale of that this fact supports the hypothesis of one Kanara. Together with the Marathas, the object as the original devak, the composite Maratha Kunbis belonged originally, says devak , being adopted later as the caste Enthoven, to the same caste; and both their advanced socially and took up the custom exogamous kuls and exogamous devaks are of prohibiting marriages between persons identical with those of the Marathas bearing the same surname. He adds that Enthoven opines that the totemic nature of the totemic origin of the devak is further their devak system suggests that they are strengthened by the fact that the Ahir largely of a non-Aryan origin. The Konkani Kumbhars who have borrowed several Kunbis in north Ratnagiri and south Colaba customs from the Marathas have separate do not claim to be Marathas or Kshatriyas, devaks for each family even today. Similarity but are content to be known as Sudras. of surnames is a bar to marriage among the Smaller and darker than the Marathas, they Maratha Kumbhars, but similarity of devak have no kuls but surname exogamy. The is no longer .so. The Konkani Kumbhars Kale Kunbis in Belgaum and Kanara speak claim to be Marathas in origin and have Konkani and resemble the Halvakki Vakkal many devaks, most of which are leaves or in dress and ornaments. They have a number trees, but in Ratnagiri also earth or anthill of exogamous kuls or clans, each with special and peacock feathers. The Konkani Kum­ gods and goddesses. The Talher Kunbis or bhars hold their devaks in great reverence and low-landers comprise two sections: the local some affirm that a similarity of devaks is a who are but little different from the Son . bar to marriage. Their exogamous divisions Koli, and the foreign who are principally are local in origin. The Gujarati Kumbhars Marathas with surnames and Devak have several exogamous divisions named (Enthoven 1922: II: 284-314). after Rajput clans; and the Pardeshi Kumbhars have exogamous kuls with sur­ The Malls or gardeners (Sanskrit: Mala names. Each kul among the Pardeshi Kum­ garland) are found all over the Deccan and bhars has a separate deity called dhiradi north Gujarat and to a less extent in the which represents its devak. The custom, as Konkan, the Karnatak and the Kanara areas they themselves admit, is borrowed from the In all probability they were originally Kunbis Marathas. The dhiradis are worshipped at who turned their attention to gardening. marriages. Some of them are a clay figure There are three territorial groups—the of a cobra besmeared with red lead; the Maratha, the Gujarati and the Kanarese. fruit of the bel tree; two cocoanuts, two The Maratha Malis have 13 endogamous betel nuts and an idol of silver; cocoanut divisions, and exogamous kuls represented by and turmeric; and so on. The Pardeshi surnames. But sometimes a group of Kumbhars are said to have come from surnames appear together under one kul on northern India (Enthoven 1922: II: 276- the ground that their bearers are descendants 282). from a common ancestor. In addition some of the Maratha Malis have gotras adopted in The term Kunbi (Marathi for husband­ imitation of the Brahmins. In the beginning man) corresponds to Kanbi in Gujarat and every kul or group of kuls had a separate THE PROBLEM OF MARATHA TOTEMISM 139

devak, but today the devaks have ceased to area the most frequent devaks are trees— be exogamous among all but the Jire Malis the kalamb, the umbar, the vad, the mango, of the Nasik district; and the panchpalvi has the phanas (jack fruit), and the pipal. The become the common devak of the entire devaks may not be touched, cut or otherwise caste. Of the devaks listed for the Marathi used, and they are worshipped at marriages. Malis (not including the Jire Malis), three Placed on a new winnowing-fan or on a kuls have the maryadvel tree, two have the piece of cloth strewn with rice, the devak is red mace, one the sun-flower, one the feather installed near the house gods, where a lamp of the salunki or blue jay, one a kind of is kept burning till the end of the marriage grass, and six the panchpalvi. Of the six ceremony. The Konkani Nhavis are like the devaks of the Jire Malis listed, four are trees, Bhandaris in religion and social practices one is the feather of the blue jay, and (Enthoven 1922: III: 127-134).. another the feathers of the peacock (Enthoven 1922: II: 422-425). The Jingar and the Shimpi: The Jingars are an impure caste of leather-workers in ' The Nhavi or barber caste of the Bombay the Deccan, the Konkan, the South Maratha Presidency are mostly Hindus of the Lingayat country, Khandesh and Kanara. They claim sect, a small number being Muslims. Like to be of Kshatriya origin and have ten gotras. many other functional castes of the Deccan, But their gotra system has been recently they are identical with the Marathas in adopted to confirm their assumed Kshatriya names, surnames and religio-social practices. origin. For, in deciding marriage alliances Enthoven, therefore, believes that the attention is paid mainly to the surnames Marathas, the Maratha Nhavi and other which are exogamous. Every section has a functional castes of the Deccan belonged separate devak, and their chief devak is the originally to one tribe. The Maratha Nhavis panchpalvi. Thus, their origin, says have two territorial, endogamous divisions— Enthoven, seems to be the same as most of the Maratha in the Deccan and the Konkani the lower castes in the Deccan (Enthoven in the Konkan. The Maratha Nhavis have 1922: II: 99-100). The Shimpis or tailors kuls and devaks like the Marathas. The also claim to be of Kshatriya origin. They Konkani Nhavis have two divisions: the are found all over the Deccan, the Konkan Konkani Nhavis proper and the Shindes or and the Karnatak. They have many bastards. The Konkani Nhavis have surnames endogamous divisions, and are an occupa­ like those of the Maratha Nhavis, and their tional caste formed from numerous castes exogamous divisions are based on exogamous and tribes. They have many exogamous devaks. The kuls and their respective devaks surnames, and some claim to belong to listed are as follows: Bagkar: pankanis (red gotras. But the latter are not guides to mace); Bhagwat: panchpalvi; Bide: mango marriage. Their devaks in the Deccan and and axe; Ghavan: gulvel (Tinospora cordi- the Karnatak are: (i) a pair of scissors, folia) ; Gaikwad: Leaves of the sag (teak); needles and a measuring-rod; and (ii) the the umbar, the mango, the jambul, and the panchpalvi (leaves from the mango, the pipal; Jadhav: pankanis (red mace); umbar, the jambul, the palas and the sag) ; Kadam: the kalamb. tree; Korde: panch­ and in the Konkan (i) mango leaves and. palvi; Lad: the umbar; Pavar: sword-blade; umbar twigs, and (ii) pipal leaves. In the Raut: horse and the umbar; Shinde: velu Deccan the devak is taken to the temple of. (bamboo); Vagh: sun-flower; and Vagh- Maruti and then tied to a post of the chavare: panchpalvi. In the Savantwadi marriage booth; but in the Konkan it is 140 JOHN V. FERREIRA

kept in a winnowing-fan near the house gods. (Eugenia jambolana). Further, on certain Both the Jingars and the Shimpis belong to occasions they worship the implements of the Hindu religion and follow the Hindu work. The banyan, the avali, the pipal and law of inheritance (Enthoven 1922: III: the tulsi are special objects of worship among 327-331). the Tambat women. Hence, Enthoven con­ cludes that common Maratha devaks might The Tarus, found chiefly in Ahmednagar, have once been of equal importance to them Nasik district and Savantvadi, say that they in regulating marriage restrictions and that were originally Marathas. They have no the gotras are a modern innovation which endogamous divisions, and no exogamous have displaced an earlier totemic organisa­ ones other than family surnames. Similarity tion. It is noteworthy that they eat flesh of devaks is a bar to marriage. Every 'sec­ and drink liquor (Enthoven 1922: III: tion' has a separate devak. The principal 361-364). devaks are the feathers of the peacock, the panchpalvi, the sun-flower and the maryadvel The primitive tribes of this region display tree (Enthoven 1922: III: 370). the phenomenon of the devak in varying . The Tambats or copper-smiths are found strength or not all. We shall examine them in the Deccan, the Konkan and Bombay city, in the following order—the Katkari, the and are supposed to have come from the Warli, the Thakur, the Mahar, the Koli and Karnatak and Telengana a thousand years the Dubla. ago. They may be described as a Brahmin The Katkari: Niggemeyer, basing himself caste in the making; for, they call themselves on Risley (1908: 98), concludes that Brahmins, have Brahmin gotras, wear the totemism among the Katkari is doubtful. sacred thread, and have even made efforts to Enthoven's information is also not very con­ train their own caste men as priests, but to clusive. He states that they are probably of little avail as the occupation of a copper­ Bhil origin and are believed to have come smith is more profitable than that of the from the north. They have five endogamous priest. On the other hand, Enthoven adds divisions and 32 exogamous ones represented that they have still traces of a totemic by surnames. "It is alleged", he adds further, organisation exemplified by the survival of "that many of the above names such as the devaks or gods of their exogamous sec­ More, Vaghmare, etc., are totemistic in tions. Like the Marathas they have also kuls origin, though reverence for the totem is no or family stocks. The exogamous kuls have longer observable". There are also some different devaks, and marriage is forbidden traces of the devak ceremony during between persons who worship the same marriages like the tying of umbar and jambul devaks. In the Deccan, however, this practice leaves to the marriage post. In 1934, how­ has become obsolete and its place has been ever, A. N. Weling published a fuller account taken by gotra exogamy. In their marriage of the Katkari from which we learn that ceremonies betel-nuts, palas (Butea frondosa) they have numerous clan names or surnames twigs and mango leaves are placed in a which are exogamous. Weling mentions 80, winncwing-fan and worshipped. There are and adds that the list is not exhaustive. special days for the worship of the banyan, Some of the surnames refer to animals, trees the pipal, the umbar, the avali, and the tulsi. or other objects (Bokya: a male cat; Bhople: The devaks of the Konkan Tambats are the a pumpkin; Dukre: a hog; Diva: a lamp; leaves of the mango, the pipal, the banyan, Gaikar: a cow; More: a peacock; Savra: the payari (Ficus rumphii) and the jambul Bombax malabaricum; and Phopli: a betel- THE PROBLEM OF MARATHA TOTEMISM 141 nut tree), others are names of various tribes totemism among them might thus have been and castes (Ahir, Bhoya, Jangam, Gosavi, acquired by them in this fashion; although Koli), and still others suggest territorial the contrary view that the Katkari were a origins (Ghogarkar, Mandavkar, Patkar) totemic tribe which has today lost much of and occupations (Katkar, Khopkar, its totemism is held by J. Abbott. The Waghmare). Jadhav, Nikam, Dalvi, Chavan, habitat of the Katkari is the hilly parts of Jagtap and Pawar are surnames of highly the Deccan or the Western Ghats, their placed Maratha families; and Dhulia, territory ranging from Thana in the north to Shetga, Nirguda and Gaikar are found Ratnagiri in the south. They are a people among the Kunbi. Mukna, Misal and of the jungle, and depend by preference and Murkute are surnames of the Mahar, and inclination on hunting, fishing and food- Kambdi that of the Kumbhar or potter's gathering and crude forms of cultivation for caste. None of the above surnames belong their existence. Their plough cultivation and exclusively to the Katkari. Some Katkari other economic activities are a recent have devaks. Among the Marathas of the acquisition. They live in local groups of 10 Bombay Presidency, adds Weling, the devak to 150 members. The entire community is is a sacred symbol which might have divided into two endogamous sections—Son originally been a totem; among the Katkari, and Dhor—which are again divided into however, there is a great deal of confusion exogamous septs. The Dhor Katkari are also between totems and devaks, suggesting the known as Marathas, and might have thus mixture of two systems of somewhat differing separated themselves recently. Descent is beliefs and practices. For instance, a Katkari patrilineal, and the language Marathi with the surname Wagh (Tiger) says that (Niggemeyer 1933: 611; Enthoven 1922: II: he does not kill a tiger, but at the same 170-181; Weling 1934: ' 1, 49-51, 52-57, time claims the kumbha tree as a devak. The 35-86, 121; Mandorff 1960: 41). Jadhav and the Mukna have the same devak, the mango tree, but they intermarry. A Valvi Another aboriginal tribe in the Bombay does not eat the flesh of an animal called Presidency is the Warli tribe. The Warli live valvi, but a Phopli (Betel-nut Tree) chews in the north-eastern parts of the Thana betel-nut without hesitation. Furthermore, district and are, therefore, in contact with the Katkari are not clear about the restric­ the Katkari. Enthoven opines that they are tions and ritual connected with the devaks. a sub-division of the Bhil. They have four A Jadhav who has the mango tree as a devak endogamous divisions and 200 exogamous eats mangoes, but a Bhoplya (Pumpkin) says clans or kuls. Among the clan names some that he must avoid eating pumpkins. Nor refer to animals (tiger, deer, jackal, bear, are the Katkari particular about hanging a crocodile, peacock and cock), others to trees branch or the fruit of the devak on the and their products (six are listed), and still marriage booth or worshipping it during the others to titles (five mentioned), occupations marriage ceremony. Weling, therefore, con­ (five), localities (three), and so on. Some cludes that the Katkari are not a distinct are of Rajput origin. There are no totem tribe, but judging from their surnames, their taboos or any other indications of respect or practice of admitting members of the reverence towards the animals, trees and surrounding tribes and lower Hindu castes other objects, after which some kuls are into their fold, and some other evidence, arc named. There are also no devaks as among a considerably mixed group. The traces of the Marathas. Therefore, K. J. Save questions the existence of totemism among 142 JOHN V. FERREIRA them. They are patrilineal. In the north were recorded in the Central Provinces. their dialect is largely influenced . by Most of the common animals (tiger, cobra, Gujarati and in the south by Marathi peacock, jackal, lizard, . elephant, lark, (Enthoven 1922: III: 445-447; Save 1945: scorpion, calf and so on) have septs named 1,.5, 12-17, 189). after them. Other objects which lend their The Thakur, an 'early tribe' living in or names to the septs are mirror, sword and near forests, are also neighbours of the shield, spur milk, and cocoanut kernel. Still Katkari. Their language is Marathi. They other sept names are Rat-Killer, Incendiary, have two endogamous divisions, and had Blind Man, Dog-Killer and Vegetable Eater. originally three exogamous surnames, each Enthoven regards the Mahar as "an assembly with a separate devak—Jadhav: gold; of tribal units" comprising "the broken Nargude: sun-flower; and Pavar: the blade residue of many former aboriginal tribes". of a sword. In course of time, however, they The Koli the Bhoi, the Katkari, the Ramoshi adopted other surnames, all of which have and the Bedar have much in common with the common devak comprising the twigs of them. They are found throughout the the mango, the umbar (Ficus glomerata), Marathi-speaking areas of the Presidency, the jambul (Eugenia jambolana), and the have several endogamous divisions and teak (Tectona grandis). They have, says numerous exogamous surnames.. The Enthoven further, a nurnber of exogamous evidence, adds Enthoven, points to the fact divisions or kuls, each consisting of a group that each of, their exogamous units at one of families bearing different surnames. A time owned and worshipped a devak or more recent investigator, L. N. Chapekar, totem closely corresponding to the bali of the does not mention totemism as such, but says Kanarese 'tribes'. "The object represented that they have certain kulis or families which by the devak is worshipped, protected from cannot marry into certain others, and in their injury by the section owning the devak and marriage ceremonies they establish the devak, brought into prominence at the time of the which is "a collective term for the deities marriage ceremony". Families possessing a worshipped on an auspicious occasion", in common devak are exogamous. Of the both the bride's and the bridegroom's homes. devaks listed, 7 are animals, 16 trees and their They have certain taboos relating to animals products, and 2 inanimate objects. In many and trees. They are patrilineal and instances the devak has become obsolete and patripotestal, and have today thoroughly has been replaced by a 'composite totem' assimilated the plough culture (Enthoven or panchpalvi—the leaves of five trees 1922: III: 376-381; Chapekar 1960: 2, 15- which are similarly worshipped in the 16, 53, 56-57, 96, 104, 215). marriage ceremony. In Poona and Nasik the The Mahar are an impure or untouchable trees concerned are the mango, the pipal, the caste of menials, labourers and village watch­ rui, the shami and the umbar; in Khandesh men in the Maratha country. Russell and the arkathi, the borkathi, the jambul, the Hiralal state that their name was probably mango and the ruchkin; and in Satara the derived from Maharashtra, that they are an babul, the banyan, the jambul, the mango oboriginal, pre-Aryan tribe, and that they and the rui. The Mahar profess Hinduism were the oldest residents of the plains of and are patrilineal after the Hindu manner. Berar and Nagpur. They have a number of Robertson who wrote a book-length study of territorial sub-divisions and numerous the tribe reports that they have many exogamous, totemic 'groups', of which 57 endogamous sub-castes and numerous THE PROBLEM OF MARATHA TOTEMISM 143 exogamous kuls or clans marked by surnames of different elements, but reports no often identical with those of the Marathas, totemism. In his recent monograph on the and devaks or clan totems. "So it usually Mahadev Koli, Ghurye points out that the happens", says Robertson, "that parties of only traces of totemism among them are that the same surname cannot marry; but the the women of the Bote family whose family final basis of judgment as to who may marry deity is Kolaba are forbidden to wear black is the Kuladevata or the clan totem. That clothes and that the families which have is to say, no two persons who have the same Ghorapad (iguana) as their deity do not eat devaks may marry. Some families bearing the iguana. The sept names mentioned by different surnames have the same devaks". Mackintosh are no longer in existence, but The devaka or godling is represented by a their 'spirit and content'—the exogamous small metal taka or plate bearing the like­ restrictions—are still in force. A group of ness of the clan totem. The devakas are as exogamous families is known as gotribans a rule animals (the buffalo, the tortoise, the (gotra brothers), and according to tradition mouse-deer, the crab, the cobra, the but not in actuality, each gotriban consists peacock) and trees (the umbar, the palm, of twelve families. Ghurye points out further the nandruk, the mango, the champa). The that no marriage takes place between indivi­ taka are becoming obsolete and are being duals having the same family name. The replaced by the panchpalvi or bunch of five District Gazetteer of Ahmednagar, however, leaves. Many Mahar surnames belong to the affirms that the exogamy of the Mahadev Kunbis and Marathas, some to Brahmin Koli is based on devaks rather than on families, and others are shared by the family names, the devaks being certain Katkari. Kula descent is patrilineal. totemic objects towards which a number of Robertson concludes that the Mahar are "the Marathi-speaking castes maintain certain remnant of an ancient people widely spread observances and on which they base their and divided into clans or tribes on whose exogamous restrictions. Persons having the social structure there supervened the orga­ same devak or the same number cannot nisation of Brahminical society and culture" marry one another. The Maruti or monkey- (Russell and Hira Lal 1916: IV: 129-144; god cult connects the Mahadev Koli with the Enthoven 1922: II: 401-415; Robertson Kunbis, the Marathas and other high castes 1938: 9, 17, 52-55, 57; Karve 1953: 16-17). of Maharashtra; and the Veghdeya or tiger- god cult connects them with the Thakur, the The term Koli, Enthoven maintains, is a Warli and other tribes of Maharashtra, and vague one and covers a number of tribes also with the Gond. In Maharashtra the of low status which have little in common Son Koli are fishermen, but the Mahadev beyond a position inferior to the Kunbi or Koli are agriculturists and cattle-breeders. cultivating caste. Thus the Gujarati Koli According to Vijaya B. Punekar, the Son differ very much from the Deccan or Konkan Koli are the highest of the Koli groups and Koli. In the Deccan and Konkan there are have no devaks, no clans or totems. All the many endogamous divisions. The Mahadev Koli groups are of course patrilineal Koli had 24 clans or kuls originally, but these (Enthoven 1922: II: 245-255; Hassan 1920: were later sub-divided into 218 according to I: 332-336; Ghurye 1957: 2-3, 6-7, 26, 40- Mackintosh. Similarity in devaks is also a 48, 96-97, 105, 123, 215; Russell and Hira bar to intermarriage. Hassan tells us that Lal 1916: III: 534-536; Koppers 1943; their sept names show a very curious mixtur Punekar 1959), 144 JOHN V. FERREIRA

The Dubla of the Thana district and the North Kanara it is a common practice to talukas of Surat and Broach (South Gujarat) keep and worship in the village-temples are today a group of serfs and landless carved tablets depicting the vansh or five labourers. Enthoven says that they have ancestors of the village community. Enthoven about 20 clans, but a recent monograph by identifies these five ancestors with the panch- P. G. Shah shows that they have nowadays palvi or leaves of five trees, the most frequent no 'definite idea of clans'. Nor is there any of the devaks in the Marathi-speaking area, reliable evidence of totems or devaks. which however, does not occur as such among Although largely oriented towards Gujarat, the Kanarese castes. Enthoven, therefore, the Dubla have been considered here to concludes that these facts point to a simi­ illustrate the proposition that the further larity of origin between certain tribes and northwards that one moves, the rarer castes in the Marathi-speaking area and becomes the devak phenomenon (Enthoven those further south in the Kanarese area. 1920: I: 341-342; Shah 1958: 11, 26, 30, 33, 171). Saldanha (1909, 1912) gives the following derivations of the term bali: i. from balli, The most zealous advocate of the idea tadbhava in Sanskrit, a creeper, a line; hence that the devaks among the Marathas and that which traces descent; ii. from bale, a the surrounding castes and tribes are bracelet of glass worn by females; the female in essentials totemic in nature, and were inheritance of the bali in many cases in the once much more so is R. E. Enthoven. As Kanarese area is adduced in support of this the empirical evidence presented above derivation; iii. bali, a Dravidian word, . demonstrates, devaks are found among the means a way, a road, and hence that which Marathas, among a number of occupational traces one's lineage, sept or clan. Enthoven castes in the Marathi-speaking area, and to gives the following meanings of the term: a less extent among some of the primitive i. way, road, ii. place, spot, iii. vicinity, tribes like the Katkari and the Thakur. nearness, company, iv. way order, v. race or From Surat northwards into Gujarat and lineage, vi. navel. Enthoven adds that it is Kathiawar the devak phenomenon declines further the term for the exogamous sections considerably in frequency of occurrence, and and also for the totems among the castes occasional manifestations of it in these areas of the Kanarese tracts, and that it corresponds may be ascribed to migrations or influences to the term bari in Tamil and bedagu in from the area of Maratha domination. Telugu. The balis or exogamous sections may Towards the south, in Kanara, balis take be named after some well-known animal, the place of devaks, and Enthoven sees fish, bird, tree, fruit or flower; and the affinities between the two. He notes that members of the sections not only worship the there is a remarkable resemblance between animal or object after which they are named the guardians or devaks of the Maratha but also follow strict rules framed to protect castes in the Deccan and the balis of North them from injury. Either the bali or an Kanara. The following trees, animals and image of it in stone or wood is usually in­ inanimate objects, for instance, occur as both stalled in a rude shrine near the village, and devaks and balis: the nagchampa, the cocoanuts and other offerings are made to jambul, the banyan, the screw-pine, secure its favour or protection. At certain Messua ferrea, Pandanus odoratissimus, seasons members of the section gather to­ the mouse-dear, the tortoise, the axe. gether with the caste priest and make special gold and turmeric. Furthermore, in offerings. The inheritance of the bali through THE PROBLEM OF MARATHA TOTEMISM 1.45 the mother was probably the earliest in turn displaced by the eponymous rishi of practice, and this was later displaced by a gotra in imitation of the Brahmins. descent through the father. Enthoven thus maintains that the Mara­ Further north, in the Marathi-speaking thas contain a much stronger pre-Aryan area, says Enthoven, a similar organisation racial element than was hitherto believed. is also to be found. Of the 120 devaks dis­ The fact that the Marathas have a totemic covered the most common are the mango, organisation, that they and the Maratha the kadam (Anthocephalus cadamba), the Kunbis share some of their devaks with peacock, the pipal (Ficus religiosa,), the rui primitive tribes and low castes like the Bhoi, (Galotropis), the shami (Prosopis spicegera), the Bhil, the Mahar, the Chambhar, the the umbar (Ficus glomerata) and the ved Burud and the Koli, and that neither the (Ficus indica). A large number of the devaks Vedas nor the subsequent orthodox Hindu are trees. Among many groups the possession texts contain any mention of the worship of of identical devaks is a bar to marriage; trees, animals and other objects regarded as among some the exogamous kuls arc named ancestors of groups which are exogamous are after devaks; in Sholapur persons bearing the pointers to a pre-Aryan origin of the Mara­ names of animals or trees (Wolf, Tiger, Pig, thas; and this despite their Aryan language, Buffalo, Parrot, Cormorant, and Snake- their quasi-Rajput stocks and claims of a Gourd) hold that they are descended from Kshatriya origin. Many of the occupational them; and a significant point of the devak castes like the Parit, the Nhavi, and the worship is that it must not be injured by Kumbhar were originally Marathas who lost those who acknowledge it. It is, however, status as a result of their occupation. The at marriages and other special occasions that Kunbi cultivators are also Marathas but of the devaks receive special worship. a somewhat lower social standing. The Mara­ tha claim to belong to the ancient 96 Enthoven analyses the course of social Kshatriya families has no foundation in fact development as follows: At first each exoga­ and may have been adopted after the Mara­ mous unit had its own devak. With the thas became, with Shivaji a power to be progress of the primitive tribes towards reckoned with. greater social unity, the panchpalvi came into existence, and became a frequent com­ The origin of the word Maratha (Enthoven posite totem, leading to the decline of totem tells us) has been traced variously as follows: exogamy. This is evidenced by the fact that i. To Maharashtra which, according to Sans­ ol the five trees from which leaves are taken krit writers of the north, comprised the great to form the panchpalvi, one is regarded with Deccan plain; ii. To a combination of maha special reverence by the family-stock concern­ (great) and rashtrika (a Sanskrit term for ed, who do not injure it or make use of petty chiefs in the Deccan); iii. To a combi­ it in any way. After the panchpalvi come nation of maha (great) and ratha (a chariot the kuls or family stocks named after human rider or warrior). The earliest reference to progenitors, the Marathas claiming 96 of the Marathas occurs in an inscription of these kuls. In other words, social evolution in about 100 B.C. The term Maratha embraces the Bombay Presidency shows progress from three classes: i. The Marathas proper or ancestor worship in the guise of trees and chiefs, land owners and warriors of the animals to a system of family stocks in which Deccan and Konkan, who claim to be Ksha­ a human progenitor displaces the totems of an triyas and avoid widow marriage. They are earlier stage. Later, the human progenitor is socially superior to the cultivating caste of 146 JOHN V. FERREIRA

Kunbis with whom, however, they maintain in Maharashtra—the Brahmins, the low in places a hypergamous relation, ii. The castes like the Nhavi and the Parit and the Maratha Kunbis or cultivators. There is little unclean classes like the Mahar and the Mang. or no difference between the Marathas and More specifically, however, it refers to the the Kunbis. iii. The occupational castes. fighting and land-owning groups on the one Judging from the devaks all these three hand, and the agricultural Kunbis on the classes appear to have had a common origin other. Some of the higher groups appear (Enthoven 1909, 1911, 1920-1922, 1921, to be of Rajput descent (W. Crookes' Intro­ 1924a, 1924b, 1924c). duction in Risley 1915: xx-xxi; Russell and Hira Lal 1916: IV: 198-204; Hassan 1920: The views of some other writers on the I: 473-491). origin of the Marathas may be added here for purposes of contract and comparison, The next interpreter of the devak phenor Colonel James Tod regarded them as the menon to whom we now turn our attention off-spring of Rajputs and the local popula­ is H. Niggemeyer. Niggemeyer bases himself tion. Risley believed that they originated in largely on Enthoven's factual and interpreta­ groups of Scythians who were driven from tive evidence, calls the phenomenon Maratha the pasture-lands in Western Punjab towards totemism, and agrees with Enthoven that we the south where they intermixed with the are dealing here with a form of totemism Dravidians. W. Crookes, however, affirms that in decline. Maratha totemism, according to there is no historical or traditional evidence of Niggemeyer, is distinguished from the other a Scythian migration into the Deccan and forms of totemism in India by a predomi­ that the Marathas are closely connected with nance of plant totems and the worship of a mixed race of cultivators who are found the totem or its image at marriages. Using in a vast area from the Deccan to the valley these two characteristics as a criterion, of the Ganges and are known as Kunbis or Niggemeyer, then, concludes that Maratha Kurmis. In course of time they asserted their totemism has made its influence widely felt, superiority over the humbler Kunbis, but radiating all over the Bombay Presidency, took brides from them while at the same far into Central India and perhaps even into time refusing to give their own daughters in the south. Some of the castes and tribes in marriage to them. In some places they have distant areas, which display this influence are secured the right of marriage with certain the Burud, the Chador, the Basor, the Dahait, Rajput clans. Their totemism points to a the Khangar, the Chauhan, the Bhaina, the pre-Aryan origin. Russell and Hira Lai say Toreya, the Bestha and even the Gond and that they are a military caste of south India, the Bhil (Niggemeyer 1933: 431-433, 447- mainly derived from the peasant population 449). of Kunbis. They claim Rajput origin and Enthoven's view that the devaks are tote- several of their clans have Rajput names. mic survivals has been called into question They are a caste of purely military origin by J. Abbott. Abbott's criticism runs as recruited from the various castes of Maha­ follows: i. On the basis of the evidence rashtra. Some of their families might have derived from the Katkari it is impossible had Rajput ancestors. Syed Siraj ul Hassan to ascribe a totemic origin to the devak. states that they are the chief fighting, land­ There are a few examples of the survival owning and cultivating caste of the Deccan, of genuine totemism in the Bombay Presi­ Berar and the Central Provinces. The term dency, as among the Katkari for instance. Maratha comprises all the classes of society But the totemic organisation of the Katkari THE PROBLEM OF MARATHA TOTEMISM 147 is for practical purposes a thing of the past. their devaks is that they were originally The totem plays no part in the invocation objects by which the families took an oath. of the dead, in the marriage ceremony or in iii. There is some identity of names between other rituals practised by the Katkari. The Maratha families and their devaks; on the Katkari also have devaks which are more or other hand, it is easy to find families which less identical with those of the Marathas, but have devaks other than the animals or trees tradition indicates that groups with different which the family names signify. In fact, the devaks constitute one and the same original number of families with names similar to their family. Further, unlike the totem, the devak devaks is smaller than that of families with plays an important part in the marriage no such similarity. Again, there is reliable ceremony. Thus, the Katkari have both evidence to suggest that punning or play on totems and devaks. ii. The existing difinitions words and names may lead to the choice of of the term devak are many and do not have a devak, the extension of its meaning, the the same connotation. Frazer calls the devaks erection of taboos and forms of charms on sacred symbols which appear to have been the ground that there is sakti or power in originally totems. Enthoven, too, accepts the a name. iv. The source from which the connection between the devak and totemism, devak is taken is not invariably revered, and but has differing views on its inner nature. when it is, the reverence is not necessarily In 1909 he called it a marriage guardian; totemic. Enthoven's assertion that the devaks in 1922 he spoke of it as a 'god of the are totemic, as the objects they represent exogamous section' and identified the devaks are not touched, cut or otherwise used, of the Pardeshi Kumbhars with their family contains several inaccurate surmises. Some deities, and of the Baris with their 'house of the objects reverenced acquire their sanc­ goddess'; and in 1926 he phrased his view tity as the result of a general Hindu belief of the matter as follows: the devak is 'the and practice; and there are many instances totemistic spirit contained in some tree, of trees from which devaks are cut which animal or material object which in addition are not reverenced except when the devak to being the subject of special worship, is taken from them. v. Other considerations regulates the marriage laws of many primitive which discredit the hypothesis that the devak sections of the population. In origin it ap­ is or was a totem are the facts that the pears to have been an ancestral spirit'. To family sometimes extends the meanings of Campbell the devak was a marriage guardian. its devak when it divides (the Powars have But the kula devata or family gods of the a sword or knife as devak, the Dhar Powars Hindus are not ancestral spirits; nor do the stick a lemon at its point, the Har Powars Marathas identify their devaks with ancestral put a garland of onions or umbar around spirits. The offerings that are made to the it and so on) ; that many families have more puri or ancestors are not made to the devaks. than one devak which they use alternatively, There is also a ceremony in which an ances­ one of them being an easily obtainable tral spirit is invoked at a marriage. It is called object; that families which claim on inter­ Mulapurusa and is practised among castes relationship at all have the same devak and which have devaks; but it has no connection vice versa; that the substitution of one devak with the installation of the devak and may for another is a frequent occurrence when be even performed together with the devak there is a similarity of names; that there are ceremony at marriages. Finally, the only villages in which a number of racially diffe­ explanation which the Marathas offer of rent groups and different sects have a 148 JOHN V. FERREIRA common devak; that when a family migrates sakti. The devak institution is found mainly from one village to another it often adopts in the Deccan and Konkan, and probably the devak which is common in the area; and had its origin in the latter area. that the similarity of devaks is not a uni­ versal bar to intermarriage, vi. The panch- Hence Abbott sums up the matter as palvi and craft tools can be called devaks follows: "To go from the manda, the manek­ only in a loose sense, and the reasons quoted stambha, and the vedikhamb through the to justify the view that they are totems are halgamba to the Maratha devak is merely to irrelevant. The number of trees used in the proceed in orderly stages from the use of a panchpalvi is limited in relation to the num­ single symbol by a whole caste to a differen­ ber of devaks known. Nor does the panchpalvi tiation of symbols among families which act as a bar to consanguineous marriages. reaches its most perfect form in the devak. Again, there is no evidence to suggest that To assume a priori that the devak is a totem one leaf in the panchpalvi receives special is to neglect altogether these other customs reverence. Further, the panchpalvi occurs which are so alike in ritual detail and frequently in Hindu ritual, but its use as a meaning and to which no totemistic origin devak is precisely restricted. So far as the can be attributed. The whole theory of tote­ craft tools are concerned, many craftsmen mism in fine as applied to the devak insti­ tie their tools to the marriage booth, but do tution is a loose construction based on not call them devaks; and when they do, it deduction from a few premises, moulded by- very likely signifies that the tools are imbued assumed analogy and framed without consi­ with sakti or divine power, vii. The devak deration for one of the principal and funda­ is, therefore, not a totem or a family god, mental axioms of Indian thought" (Abbott but a symbol into which the sakti of a deity 1932: 448-460, 461-464, 526-527). is temporarily invoked, according to the Hindu practice of avahana. Among the We have thus a corpus of factual and Marathas and related castes the sakti of interpretative evidence before us, which the often identical village or family god enables the drawing of certain conclusions. is invoked into the devak to ensure the presence of the deity in the marriage booth So far as our sources are concerned, we during the marriage ceremony. The ritual are largely indebted to R. E. Enthoven for followed in the installation of the devak is in our factual data on the Marathas and the accordance with the ordinary canons which occupational castes related to or influenced guide the practice of avahana. There are by them. Enthoven's survey of the tribes and also parallel customs in the Karnatak (the castes of the Bombay Presidency belongs to halgamba), among the Jains in Gujarat (the the series of compilations, which was first manekstambha or ruby pillar), among the attempted in a more or less systematic manner Tamil Christians (the arasani kallu), among by E. Dalton for the north-eastern parts of the Rajputs (the vedikhamb), and among India, which was later given an official the Ahir Gauli immigrants from, Mysore imprimatur by H. Risley who himself under­ (the manda), which have no connection at took such a compilation for Bengal, and all with totemism, but with the invocation which was carried further for other provinces of sakti. Thus both internal and external of the country by other officials. These evidence indicates that the devaks are not compilations (as also the various Govern­ totems but temporary repositories of divine ment Gazetteers and other similar documents) have been both praised and criticized by THE PROBLEM OF MARATHA TOTEMISM 149 a number of scholai's with Socio-cultural Abbott indicates, is untenable. There is a interests. They have been praised because residue of evidence which demonstrates the they are valuable sources of information on identity of surnames and devaks, of taboos groups of people, which have otherwise not akin to totemic ones, and of exogamous been systematically studied; and they have prescriptions based on them. Therefore, the been criticized because they suffer from conflicting interpretations of Enthoven and several inaccuracies and from tentalizing in­ Abbott can be properly taken to imply that complete accounts of various phenomena. the devak phenomenon, as it appears in our Enthoven's lists of devaks and practices sources, is a blend of two originally differing connected with them are, for instance, rarely series of beliefs and practices. As Weling, full and exhaustive, and not always reliable. referring to the Katkari, suggests, we are Indeed, Elwin has pointed out somewhere facing a mixture of two somewhat different that Enthoven's methods of garnering infor­ systems. The confusion of elements which mation, dependent as they were on the the evidence makes manifest may be taken efforts of untrained clerks and other petty as a proof that the synthesis between these officials, have produced results which leave two systems has not been carried to the much to be desired in point of accuracy and point of completion. Perhaps one important authenticity. The moral of this fact is the factor which inhibited the process of synthesis urgent necessity of more systematic investi­ was the desire to rise in the social scale by gations by trained personnel of the castes imitating the Socio-cultural forms of the which occupy the lower rungs of Hindu topmost castes. society, before valuable data disappear Syntheses between totemism as such and altogether in the course of time. other beliefs and practices like dynamism On the primitive tribes of the Marathi- (mana of the Melanesians, for instance), speaking area we are better informed through ancestor cults and even idol worship have ethnographic works (Weling 1934, Save often occurred in the course of cultural 1945, Ghurye 1957, Punckar 1959 and development and are amply documented in Chapekar 1960) accomplished under the the totemic literature. Therefore, it is not auspices of the Department of Sociology extraordinary to witness in the Marathi- (University of Bombay) and guided by the speaking area the occurrence of something doyen of Indian sociologists, G. S. Ghurye. similar—the convergence of the belief in the devak as a symbol of sakti and of totemic When we turn to the devak phenomenon stimuli stemming from elsewhere. This itself, we are confronted by the striking fact absorption of totemic traits implies a that two diametrically opposed interpretations certain measure of mental flexibility which hold the field—the interpretation of Enthoven is always a creditable feature, and, therefore, and that of Abbott. Abbott's thesis may be redounds to the credit of the Marathas. divided for the sake of arriving at the extent of its truth into two parts: i. that the devak In a larger work on totemism in India is a symbol of sakti or divine power; and I have sought to demonstrate that the centre ii. that it has, therefore, nothing to do with of Indian totemism lay in the habitat of the totemism. A careful consideration of the Munda-speaking tribes, although these tribes evidence marshalled by Abbott shows that themselves were not in their entirety its the first part of his thesis is well made. The original bearers. From this centre totemism second part, however, as the evidence presen­ radiated outwards in several directions. A ted by Enthoven and not fully refuted by southward drift with an ever-weakening 150 JOHN V. FP:RREIRA impulsion is testified by the totemism of the element than was hitherto supposed cannot Telugu-speaking castes and a westward . be sustained, if Abbott's interpretation of drift by the Tulu-speaking and Kanarese- the devak holds water. This is not to deny speaking castes. Both the Telugu-speaking that a small primitive component has gone and the Tulu-speaking and Kanarese- into the make-up of the Maratha population. speaking castes here referred to occupy the Essentially, however, the Marathas are a lower rungs of the Hindu social ladder. 'Mediterranoid people' with a large compo­ Now, as Enthoven points out, there are nent of "some broad-headed strain which parallels between the bali system of the Kana- cannot be identified at present" (Karve rese-speaking castes in the south and the 1948: 71). devak system of the Marathas and the Niggemeyer's claim that Maratha influence occupational castes associated with them. has made itself widely felt among the castes Therefore, it is likely that the totemic stimuli and tribes of Central India and elsewhere, and elements found in the Marathi-speaking if one judges from the distribution of area emanated from this source, that is to predominantly plant totems and from the say, from the Kanarese castes. A corrobora­ worship of the totem or its image at marria­ tion of this view comes from I. Karve, who ges, can be supported from another angle. in another context, that of kinship organi­ The Marathas, a virile and dynamic popu­ sation, points out that the Marathi-speaking lation, who, in their expansive force, resemble area occupies an intermediate position, a the Germans in Europe, have often tended position between the northern and southern to radiate outwards from their homeland, zones, and that many features of kinship and as a conquering and progressive commu­ organisation found frequently in the south nity must have appeared to the tribes and occur also in the Maradii-speaking area. If lower castes as a fitting model for Socio- the primitive tribes examined above likewise cultural imitation from fairly early times. contributed their share of totemic stimuli and If, then, the devak phenomenon is, indeed, elements, it must have been a relatively an example of a partial synthesis of differing small one; for the tribes of this region arc ideologies, its study should prove of some not at all strongly totemic. significance in the India of today where Enthoven's affirmation that the totemic there is much talk of and hankering after evidence, as found among the Marathi- a synthesis of the old and the new, of the speaking castes is a proof that the Marathas socio-religious traditions and the mounting contain a much stronger, pre-Aryan racial influence of science and technology.

REFERENCES.

Abbott J. 1932. The Keys of Power (London). Chapekar L. N. 1960. Thakurs of the Sahyadri (Bombay). Enthoven R. E. 1909. "The Study of Ethnography in the Bombay Presidency." four. Bom. Anthro. Soc. VIII: 433-445. Enthoven R. E. 1911. Totem Theories (Silver Jubilee Memorial Number). Jour. horn. Anthro. Soc. 63-86. Enthoven R. E. 1920-1922. The Tribes and Castes of Bombay (Bombay). Enthoven R. E. 1924a. "The Devaks of the Deccan and Konkan," Jour. Bom. Anthro. Soc. XIII. 1-14. Enthoven R. E. 1924b. "Some Further Notes on the Devaks of the Bombay Presidency," Jour. Bom. Anthro. Soc. XIII: 649-657. THE PROBLEM OF MARATHA TOTEMISM 151

Enthoven R. E. 1924c. The Folklore of Bombay (Oxford). Frazer J. G. 1910. Totemism and Exogamy (London). Ghurye G. S. 1957. The Mahadev Kolis (Bombay). Hassan S. S. 1920. The Castes and Tribes of H.E.H. the Nizam's Dominions (Bombay) Karve I. 1948. Anthropometric Measurements of the Marathas (Poona). Karve I. 1953. Kinship Organisation in India (Poona). Koppers W. The Kolis in North-West Central India (Ethos I-II: 1-18). Mandorff H. 1960. Notes on Some Primitive Hunting Tribes of Southern and Central India (Bui. International Committee on Urgent Anthropological and Ethnological Research III: 40-44). Niggemeyer H. 1933. "Totemismus in Vorderindien." Anthropos XXVIII: 407-461, 579-619. Punekar V. B. 1959. The Son Kolis of Bombay (Bombay). Risley H. H. 1908, 1915. The Peoples of India (Calcutta, Bombay). Robertson A. 1938. The Mahar Folk (Calcutta). Russell R. V. and Hira Lai 1916. The Tribes and Castes of the Central. Provinces of India (London). Saldanha J. A. 1909, 1912. "Balis and Totems in Kanara". Jour. Bom. Anthro. Soc. VIII: 382-387 IX: 253-280. Save K. J. 1945. The Warlis (Bombay). Shah P. G. 1958. The Dublas of Gujarat (Delhi). Weling A. N. 1934. The Katkaris (Bombay).