A History of the Mahrattas v1.c5.s3: III. The Origin of the Marāthās

A History of the Mahrattas v1.c5.s3

Language: English

III. The Origin of the Marāthās

v1.c5.s3.1

The origin of the terms ‘Mahārāshtra’ and ‘Marāthā’ or ‘Mahratta’ is still the subject of speculation, nor is it yet clearly determined whether the name of the Marāthā people is derived from Mahārāshtra, the country in which they dwelt, or whether on the other hand the latter has derived its title from the leading tribes which inhabited it in early ages. The explanation given in Molesworth’s Marāthī dictionary that Mahārāshtra signifies ‘the great country’ is based upon Brahman etymology,¹ and is not now accepted as correct. Unacceptable also is the derivation, originally given by the Mahārs themselves, from ‘Mahār-rāshtra,’ ‘the country of the Mahārs’—the degraded tribe or class which serves the villages of the Deccan as watchmen, guides and scavengers.² The accepted derivation of Gujarāt from Gujar-rāshtra, ‘the country of the Gujars,’ and of Sorath from Saurāshtra, ‘the country of the Sauras,’ does not assist in solving the problem. As will be seen from the succeeding paragraphs, Mr. R. E. Enthoven is inclined to think that Mahārāshtra preceded the word Marāthā in point of time, and that the latter is a derivative signifying merely the people who lived in Mahārāshtra and spoke Mahārāshtri, or rather its popular vernacular form, Marāthī.

v1.c5.s3.2

¹ Wilson, *Indian Caste*, ii, 48.

v1.c5.s3.3

² Wilson, ibid.

v1.c5.s3.4

Other authorities, however, point out that Rock Edict V of the Mauryan emperor, Asoka, records the dispatch of Buddhist missionaries in the third century B.C. to the Rāshtrikas, who are probably correctly identified as the inhabitants of the Deccan.

v1.c5.s3.5

Again, in a rock-hewn cell at Bedsa, Poona District, is an inscription of the first century A.D. which has been deciphered as follows:

v1.c5.s3.6

‘The religious gift of Mahābhoja’s daughter Sāmadinikā, the Mahādēvī Mahārathini, and wife of Apadevanaka.’¹

v1.c5.s3.7

¹ I.G. Bom., i. 507.

v1.c5.s3.8

It is correctly argued by Professor D. R. Bhandarkar that ‘Mahā’ in the word Mahārathini is an honorific affix, as it undoubtedly is in the word Mahā-bhoja. Corroborative evidence is supplied by another early inscription in a cave at Nānāghāt, Poona District, which records that Sātakarni, ruler of the Deccan and third prince of the indigenous Āndhra or Sātavāhana dynasty, had a queen named Nāganikā, who was ‘the daughter of the Mahārathi Tranakayiro of the Angiya family.’² The Chinese pilgrim Hiuen Tsiang (A.D. 640) styled the kingdom of the early Chālukya dynasty of the Deccan Mo-ho-lo-cha, which is the Chinese transliteration of Ma-hā-rā-tha. In A.D. 753 a prince of the indigenous dynasty of Rāshtrakutas or Rattas overthrew the reigning Chālukya, a foreigner of Hūna-Gurjara descent, and thus obtained the mastery of the Deccan which his successors held for nearly two and a quarter centuries. Later still, from the middle of the eleventh until the end of the first quarter of the thirteenth century, a dynasty of Rathas or Rattas held sway in the Southern Marāthā country. Albiruni mentions (c. A.D. 1030) ‘Mahrat-dessa’ (i.e. Mahratta dēsh, Mahratta country); Friar Jordanus refers (c. A.D. 1328) to ‘the Kingdom of Maratha’; while Holwell (Hist. Events, &c., i. 105), writing in A.D. 1765, speaks of the ‘Maharattors,’ and states that the name is compounded of ‘Rattor’ and ‘Maahah.’³

v1.c5.s3.9

² Ind. Ant., xlvii, March, 1918.

v1.c5.s3.10

³ See Yule’s Hobson-Jobson, ed. Crooke, 1903, s.v. Mahratta.

v1.c5.s3.11

There is some ground for the view that the name of the people is derived from certain (Mahā) Rāshtrikas or Mahārathis who occupied a political position of importance in ancient times, and that the name of the country may have been an altered form of their tribal or official appellation.⁴ The point is, however, not free from doubt; and future research may possibly show that the Rāṣṭrikas of the Mauryan epoch and the ‘Mahārathis’ of the inscriptions were so called merely because they lived in a country known as Mahārāṣṭra and spoke Mahārāṣṭri.

v1.c5.s3.12

⁴ For information about the (Mahā) Rāshtrikas and Mahārathis, see D. R. Bhandarkar’s ‘Dekkan of the Sātavāhana Period’ in Ind. Ant., vol. xlviii, pp. 77 ff.

v1.c5.s3.13

The problem of the origin of the people who are known as Marāthās has been obscured by the varying use of that term. Grant Duff, for example, applies the term to individuals like Tānājī Mālusrē, and to groups like the Hetkaris who were certainly not Marāthās in the strict caste-sense of that word, as it is used to-day. His use of the term, however, as well as such expressions as Marāthā Brahmans, by which some writers have denoted the Konkanasth or Chitpāvan sect, can be justified if one accepts the view that the word Marāthā, as suggested above, meant originally nothing more than a person dwelling in Mahārāṣṭra or using Marāthī as his home tongue. The gradual development of a Marāthā caste out of the old tribal elements of Mahārāṣṭra during the last two centuries or more necessitates an attempt to determine, so far as may be possible, the origin of the people who now describe themselves as Marāthās and claim communal representation as a distinct caste or community under the new franchise arrangements contemplated by the Government of India Bill of 1919.

v1.c5.s3.14

Sir Herbert Risley ¹ suggested a few years ago that the Marāthās are the descendants of Saka or Scythian invaders of India, who after occupying the grazing country of the western Panjab, and finding their progress eastward blocked by the Indo-Aryans, turned southwards and mingled with the Dravidian or aboriginal population of the Deccan. The modern Marāthās were thus of Scytho-Dravidian origin. The theory was based partly upon anthropometric data of a decidedly unconvincing character, and partly upon the proficiency of the Marāthā as a light cavalryman—a trait which Risley too readily accepted as evidence of their supposed Scythian ancestry. Dr. W. Crooke ² has successfully exploded Risley’s suggestion, pointing out that skill in horsemanship is equally a characteristic of the Rājpūts and of the despised Jāts, and that the Marāthās’ success as cavalrymen ‘was mainly the result of local conditions in the Deccan, where the indigenous pony, an admirable beast, had been improved by the introduction of Arab blood from the Persian Gulf. This trade flourished in the days of Marco Polo, and probably dates from a very early period.’

v1.c5.s3.15

¹ Risley, *People of India*, p. 59.

v1.c5.s3.16

² *Rājpūts and Mahrattas*, J. R. Anth. I., vol. xl, 1910.

v1.c5.s3.17

Mr. R. E. Enthoven, another member of the Indian Civil Service, who has for some years been in charge of the Ethnographical Survey of the Bombay Presidency, has kindly supplied the following note, reproduced *verbatim*, indicating the inferences and conclusions to be drawn from the researches of the Survey up to the present date.

v1.c5.s3.18

‘In the course of the Ethnographical Survey of Bombay some new light is being thrown on the nature and origin of the various groups known commonly as Marāthās. Pending the completion of the work of the Survey, the following remarks should be taken as provisional, though they roughly embody the results of the Survey up to date.

v1.c5.s3.19

‘It has been usual to aim at classifying the almost numberless amorphous social groups, which have been formed in India by varying influences, as either caste, tribe or sect. In caste the formative influence is taken to be occupation; in tribe, common descent; in sect, religion.

v1.c5.s3.20

‘In the process of evolution a religious group or sect may develop into a congeries of occupational groups or castes, as we find in the case of the Lingāyats, a religious community found in the Bombay and Madras Presidencies, peculiar for the worship and wearing of the *linga*.¹ A tribe may similarly evolve caste-divisions. Marāthās are neither a caste nor a sect. They are more correctly a tribe in origin, which has developed caste divisions.²

v1.c5.s3.21

¹ Hastings, *Encyclopaedia of Religion and Ethics*, s.v. Lingayat.

v1.c5.s3.22

² Census of India, 1901, Bombay, vol. ix, ch. viii.

v1.c5.s3.23

‘In common parlance the term Marāthā is used with varying meanings. Residents in the Deccan, the ancient Mahārāshtra, speaking the language of the Bombay Deccan, Marāthī, would be referred to when in other parts of India as Marāthās or Arers. The term “Mahārathini” is applied to a princess mentioned in the cave-inscriptions of the first century A.D.; but we hear little of Marāthās, as such, until the rise of Sivājī in the seventeenth century brought them into political prominence. At that time Sivājī and his followers were referred to by contemporary writers as Marāthās or Sivājīs, this designation probably covering Dhangars, Rāmoshis, and Kolis (Māvalis), who joined in the Marāthā crusades across India. At the present day, Marāthā in its most limited sense means a member of the superior Marāthā land-owning or fighting class, of which His Highness the Mahārāja of Kolhāpur, a direct descendant of Sivāji, is a leading example. Used in this sense the term denotes a class that claims Kshattriya rank, does not allow widow-remarriage, and will not intermarry with the very numerous occupational divisions known as Kunbīs, Sutārs, Lohārs, Parīts, &c. It is worthy of note that a member of the ordinary cultivating caste in the Bombay Deccan will ordinarily describe himself as a Marāthā. Pressed as to what kind of Marāthā he is, he will explain that he is a Kunbī, i.e. cultivator, thereby indicating a distinction, based on occupation, with its inferior social status, between his class and the land-owning and fighting class already referred to, which for lack of a better term may be described as Marāthās proper. The claims of the latter to Kshattriya rank based on alleged Rājpūt descent are of very doubtful validity; and it is probable that they and the Kunbīs do not differ greatly in racial origin. When the Kunbīs are more closely examined, it appears that on the one hand a Kunbī may, if successful, rise to the superior Marāthā status, while on the other hand there is by no means an insuperable barrier between Kunbīs and the tribe of Kolis, an undoubtedly aboriginal tribe of the Western Ghāts. A Kolī taking to settled cultivation in the plains may become a Marāthā Kunbī. Proof of such a development in certain villages of the Poona district was furnished to me during the census of 1901. From this it is not unnatural to assume that there is to be found in Marāthās of all grades a greater admixture of aboriginal tribal elements than it has been usual in the past to admit.

v1.c5.s3.24

‘Certain aspects of the social organization of Marāthās of all occupational divisions, when examined in the light of the evidence available from the records of the Ethnographical Survey seem to lend support to this assumption. Thus, we find that in the extreme south of the Presidency, where Aryan influence on social customs is naturally less marked than in the north, a vigorous totemistic organization survives in the case of most tribes and castes. These are divided into sections known as balis that have a common object of worship in a tree, plant, animal, implement, &c. This object they will in no circumstances injure. Further, no intermarriage is allowed between members of a group with a common object of worship. Instances of such totems are: the elephant, elk, wolf, pig, tortoise, hog-deer, the coco-nut, pepper-vine, turmeric, the mango-tree, the tamarind, an axe, a white ant-heap, &c. Certain trees are of specially common occurrence in this connexion, notably the Kadamba (Nauclea cadamba), the Rui (Calotropis gigantea) and the Shami (Prosopis spicigera).

v1.c5.s3.25

‘As we pass northwards into the Deccan, the totemistic grouping is found to survive in a less vigorous form, as exemplified by the devaks of the Marāthā and Marāthā occupational groups. This devak, which seems in origin to correspond to the bali, is usually a plant or group of five plants, worshipped at the time of marriage or when entering a new house. Those who have the same devak do not usually intermarry. It is true that Marāthās of the upper class, i.e. Marāthās proper, claim to belong to one of eighty-four families or kuls, and that in their case the kul is more important as regulating intermarriage than the obsolescent or obsolete devak. Yet the survival of these devaks is of no little significance. It seems to lend support to the theory that there are among the Marāthās of all degrees traces of a kindred social organization to that of the Kolīs and other aboriginal units.

v1.c5.s3.26

‘The transitional stage at which groups possess both kuls similar to those which regulate marriage among Marāthās proper and totemistic sections such as we have discovered in the south of the Presidency, is to be observed in such castes as the Marāthā Parīts or washermen. These are divided into kuls bearing the well-known Marāthā kul names of Pawār, Gaikwād, Sūrvē, Dalvē, Kālē, Sālunkē, and Kadam. But to these kuls are attached devaks as follows:

v1.c5.s3.27

v1.c5.s3.28

and marriages are strictly forbidden between families with the same devak.

v1.c5.s3.29

‘It is interesting to note that in an early stage of the process of transition from *devak* to *kul*, the *devaks* are frequently found to be grouped into five-fold totems or *pānch-pālvi*. Since in these five-fold totems the trees, implements, &c., are clearly recognizable as those found in more primitive castes as regular totems, it would appear that the multiple *devak* is a stage in the decay of the more primitive system.’¹

v1.c5.s3.30

¹ *B.E.S.*, Monograph 32.

v1.c5.s3.31

‘It is possible that fuller developments of the results of the Ethnographical Survey may tend to show that Marāthās are merely a heterogeneous population largely aboriginal in origin, living in Mahārāshtra and using the Marāthī language.² There is probably little racial distinction between Marāthās proper, Marāthā Kunbīs, and the numerous other occupational groups in the Deccan, that describe themselves as Marāthās, such as Marāthā Sutārs, Marāthā Kumbhārs, Marāthā Lohārs, Marāthā Parīts, &c. &c.³ Among the Parīts the term Marāthā is held to be equivalent to Kunbī Parīt.⁴ These numerous occupational groups are very similar in physical type to the Kunbīs. Language, residence and social organization are common to all of them. Occupation and the resulting social status form the existing lines of cleavage.’

v1.c5.s3.32

² D. R. Bhandarkar, *Ind. Ant.*, xlviii. 80, connects Marāthās with Rāshtrikas through Mahā-Rāstrikas on the analogy of Mahā-Bhojas from Bhojas. But the term Mahārāshtra seems to be the earlier in origin.

v1.c5.s3.33

³ In Khāndesh the Marāthā Sutārs admit descent from Marāthā Kunbīs, *B.E.S.*, Monograph 175.

v1.c5.s3.34

⁴ *B.E.S.*, Monograph 32.

v1.c5.s3.35

The two main deductions to be drawn from Mr. Enthoven’s note are (*a*) that the modern Marāthās of the upper class are in origin identical with the great agricultural Kunbī tribe, and are now differentiated from them merely by the accident of social status, and (*b*) that in both the Marāthā and the Kunbī there is a distinct aboriginal strain, which facilitates the absorption into the Kunbī ranks of clearly aboriginal people like the Kolīs, and which justifies the conjecture that the Marāthās, whether of high or low status, are descended without any appreciable foreign admixture from the primeval tribes of the Deccan and Southern Marāthā country. The first deduction has also been arrived at by Dr. W. Crooke in his paper on ‘Rājpūts and Mahrattas,’⁵ in which he draws an analogy between the position of the Marāthās and Kunbīs and that of the Rājpūts, Jāts and Gujars in northern India. He points out that the claim of the upper-class Marāthās to a Rājpūt origin is unfounded; that in modern times some families may have absorbed Rājpūt blood; but that the claim is really the natural result of a rise in social status.¹ ‘The Rājpūts, Jāts and Gujars,’ remarks Dr. Crooke, ‘are ethnically akin; their position depends upon status, the Rājpūts having become the aristocratical body, while the Jāts and Gujars have remained agriculturists and cattle-raisers.’ In precisely the same way a rise in social status, resulting from the ‘kingdom-seizing’ raids of the seventeenth and eighteenth centuries or from later economic causes, has resulted in a spurious distinction being drawn between the upper-class Marāthā and the lowly Kunbī cultivator, to whom he is shown by scientific investigation to be ethnically akin. This view finds curious corroboration among the Marāthās themselves. In a Memorandum of July 31, 1919, submitted by the All-India Mahratta League to the Parliamentary Joint Committee appointed to consider the Indian Constitutional Reforms Bill of 1919, it is definitely stated (pp. 8, 9) that ‘There is no difference whatsoever between a Mahratta and a Kunbī. . . . Since days long ago in the antiquarian history of this country they are related by blood to each other. . . . The terms “Mahratta” and “Kunbī” are synonymous; the educated and well-to-do portion of the community styles itself (as) Mahrattas, while the ignorant and the rural passes under the name Kunbī, though the former is akin to the latter as members of the same caste.’²

v1.c5.s3.36

⁵ *J. R. Anth. I.*, vol. xl, 1910.

v1.c5.s3.37

¹ The salient example is Sivājī himself, who for political reasons hid the truth of his Kunbī origin in a tale of descent from the great Rājpūt house of Udaipur. See Prof. J. Sarkar, *Shivaji and His Times*, pp. 267 ff.

v1.c5.s3.38

² Elphinstone (*History of India*) arrived at the same conclusion. ‘The chiefs (Mahratta) in those days were men of families who had for generations filled the old Hindu offices of heads of villages or functionaries of districts, and had often been employed as partisans under the governments of Ahmadnagar and Bijāpur. They were all Sūdras, of the same caste with their people, though some tried to raise their consequence by claiming an infusion of Rājpūt blood’ (*O.H.I.*, 431).

v1.c5.s3.39

The second deduction that the Marāthās represent, or are mainly descended from, the aborigines of Mahārāshtra, seems to draw some indirect support from certain features of their history and customs. In 1798 Colonel Tone, who commanded a regiment of the Pēshwā’s army (see note on p. 275, vol. ii, post), wrote that the three great tribes which compose the Marāthā caste are the Kunbī or farmer, the Dhangar or shepherd, and the Gaulī or cowherd, and that this accounted for the great simplicity of manners and rusticity which in his day distinguished both the rich and poor Marāthās.¹ ‘In the Central Provinces,’ writes Russell, ‘the names of the exogamous septs among the Kunbī sub-castes confirm the hypothesis that the Kunbīs are largely recruited from the pre-Aryan or aboriginal tribes.’²

v1.c5.s3.40

¹ Waring, *History of the Mahrattas* (1810), p. 104; Russell, *T. and C.C.P.*, s.v. Maratha, p. 200.

v1.c5.s3.41

² Russell, *T. and C.C.P.*, s.v. Kunbī, p. 21.

v1.c5.s3.42

Again, one of the salient traits of the upper-class Marāthās of historic times was their profound attachment to the small estates or villages in the Deccan, in which they had originally held the hereditary office of pātel or village headman; and this attachment persisted long after they had carved out for themselves principalities and states in other parts of India. The Bhoslē of Nāgpur adopted a personal title from the village of Deor in Sātāra District; the Rājā of Sātāra, after he had become a mere puppet of the Pēshwās, clung to the watan lands which he had inherited from Sivājī. Sindia preferred to be styled Pātel in the revenue accounts of the villages he acquired in Nimār; while Holkar and the Powārs of Dhār struggled desperately after the British conquest to recover the pātelki rights in Deccan villages which had belonged to their ancestors. Lastly we read that Rāghuji Bhoslē, the first Rājā of Nāgpur, used to celebrate his Dasahra on the ninth instead of the tenth day, as is usual among the martial classes, in order to emphasize the fact that he was originally an agriculturist of Mahārāshtra, and only incidentally a warrior.³ It seems impossible to regard this trait as arising merely from land-hunger or political considerations. Was it not rather the manifestation of a natural instinct, of an innate conviction that they were the real autochthones of Mahārāshtra, and that their connexion with the very soil of the Deccan dated back to misty ages long before the rise of Saka viceroys, Chālukya overlords and Muhammadan Sultāns?

v1.c5.s3.43

³ Elliott, *Hoshangabad Settlement Report*, quoted in Russell, ibid.

v1.c5.s3.44

Finally, let us glance at the two chief deities of the Marāthā people, Khaṇdoba and Bhavānī. Khaṇdoba, whose name is supposed to be derived from Khaṇde-aba, i.e. ‘sword-father,’ in allusion to the sword which he is usually depicted as carrying unsheathed in his right hand, is the tutelary deity of the Deccan, just as Bhāirav, whom he resembles in several particulars, is the tutelary guardian of the village.¹ His chief temples are at Jejuri in Poona District, and at Pāl in Sātāra District; the officiating priests are Guravs, a widely distributed class of temple ministrants. It is noteworthy that in most temples of Khaṇdoba the pujāris or officiating priests are not invariably Brahmans; in the Thāna District they belong to the lower castes; and of three temples in Bombay, only one has a Brahman officiant, the duties in the other two being performed respectively by a Kamāthi and a Murlī, i.e. a woman dedicated to the god.² This fact of itself would indicate that the cult of Khaṇdoba is ancient and primarily non-Brahmanic. Khaṇdoba’s peculiar connexion with the Marāthās is shown by the fact that a large portion of the temple buildings at Jejuri are gifts from the important Marāthā families of Powār, Holkar and Vinchurkar; that Shāhjī, the father of Sivājī, visited the temple in 1662 at an important moment of his life; and that it was to Jejuri that Tārā Bāi Mohite made the Pēshwā repair and swear a solemn promise in 1755. Khaṇdoba combines the position of a household god with that of guardian deity of Mahārāshtra, for he is the chief household god of all the various Hindu castes and tribes in the Poona District; and so widely-acknowledged is his position that pilgrimages are made to his temple not only by Marāthās from other districts like Khāndesh and Berār, but also by the Kolīs or aboriginal fishing-tribe of the Konkan, while new settlers in the Deccan, whether from northern India, Gujarāt or the Carnatic, are declared to join early in paying him reverence.³

v1.c5.s3.45

¹ B.G., xviii, pt. i, 290 f.

v1.c5.s3.46

² Enthoven, *Folk Lore of the Konkan*, i.; B.G., xviii, pt. iii, 132 ff.; K. Raghunathji, *Hindu Temples of Bombay* (1900), Nos. 335, 362 and 375.

v1.c5.s3.47

³ B.G., xvi. 72. The Kolīs prior to 1885 subscribed the cost of plating with brass a representation of a tortoise in front of the Jejuri temple courtyard.

v1.c5.s3.48

Khaṇdoba or Khaṇde Rāo, as he is frequently called, is a strangely composite figure. He is usually represented on horseback, attended by a dog, with a naked sword in his right hand, and his left arm, which carries a shield, supporting his wife Mālsirās or Mālsubāī. He is shown with four arms, the two additional arms holding the trident (*trisūl*) and the drum (*damaru*). His images are usually made of metal, rarely of wood or stone.¹

v1.c5.s3.49

¹ B.G., xviii, pt. i. 290. For the drum (*damaru*), an aboriginal fetish, see Crooke, *P.R. and F.L.*, i. 28.

v1.c5.s3.50

He is now regarded as an incarnation of Siva, under the names Malhārī or Mārtand. According to the *Kshetra Khand* of the *Brahma Purān*, Siva assumed the form of Mārtand Bhāirav and accompanied by two female attendants, dogs, and a vast array of *bhūts* (ghosts), defeated and slew two demons, Mani and Mal, who had harassed the seven Rishis. The name Malhārī enshrines the story of his defeat of Mal, on which occasion Khandoba rode a yellow horse, fought under a yellow flag, and slew yellow demons. After his victory Siva remained on the hill of battle at the Rishis' request in the form of a *svayāmbhu lingam*. Hence Khandoba is worshipped at Jejuri under the form of the *lingam* (phallus), the peculiar emblem of Siva. The story may perhaps be taken to indicate the method by which the aboriginal Khandoba was absorbed into the neo-Brahmanic pantheon, in precisely the same way as tales were manufactured in order to prove the identity of aboriginal goddesses like Kālī, Chāndika and Chamūnda with the Brahmanic Mahādēvī, consort of Siva. The name Mārtand, which is a synonym of the sun, coupled with the yellow horse and yellow flag, may enshrine a reminiscence of sun-worship, superimposed upon or blended, like the worship of Siva, with the cult of an aboriginal guardian or protector.

v1.c5.s3.51

Secondly, Khandoba is a vegetable deity. Under the name of *bhandar* he is identical with, and resident in, turmeric powder. The Vāghes, or males dedicated to him, carry turmeric in tiger-skin bags, which they apply to the foreheads of Khandoba's devotees.² Marāthās swear their most binding oath upon the *Bel-bhandar*, i.e. the leaves of the Bel tree (*Aegle marmelos*) and turmeric; and it was thus that Sumar Singh, the murderer of the Pēshwā Nārāyan Rāo, was made to take oath by Ānandībāi, the jealous aunt of the prince and wife of Raghunāth Rāo.¹

v1.c5.s3.52

² Mackintosh, *Account of the Ramoossies* (1833), p. 24.

v1.c5.s3.53

¹ B.G., xviii, pt. i. 290; Gupte, Hindu Holidays, &c., 21.

v1.c5.s3.54

Thirdly, Khandoba is an animal god, for he is identified with the dog who runs beside his horse; and Marāthās, when calling a dog, usually cry ‘Khandi, Khandi.’² Mountstuart Elphinstone, describing a visit to Khandoba’s temple at Jejuri, wrote, ‘Candee Rao is a great god among the Mahrattas and particularly among the soldiers. He is the god of Holkar’s family, and this temple was built by the first Mulhar Rao. We were followed by many beggars. . . . Among others was a boy who instead of asking for charity barked like a dog. He said he was Candee Rao’s dog, and had been longer in the kennel than any of the present pack. A woman made a child under her arm bark in the same way, and asked charity for him, because, as she said, she was bringing up a dog.’³ These dog-servants of Khandoba are still to be found at the temple of Malhāri (Khandoba) at Gudguddapur in Dhārwār District.⁴ In this matter of the dog, Khandoba approximates very closely to Bhāiron (Bhāirav), who in Bombay is a grāmadēvata or village-guardian, and is always accompanied by a black dog. Indeed, once we descend into the Konkan, away from the sphere of influence of Jejuri, we find Khandoba degenerating into a mere village-godling, and scarcely distinguishable from the equally aboriginal Bhāirav.⁵

v1.c5.s3.55

² B.G., xviii, pt. iii. 132 ff.

v1.c5.s3.56

³ Colebrooke, Life of M. Elphinstone, i. 286.

v1.c5.s3.57

⁴ B.G., xviii, pt. iii. 132 ff.

v1.c5.s3.58

⁵ Enthoven, Konkan Folk Lore, 21. See also P.R. and F.L. (i. 107 ff.), by W. Crooke, who tells me that in his opinion Bhāiron was originally an earth-god, guardian of fertility, &c.

v1.c5.s3.59

Fourthly, Khandoba, though the recognized god of the Marāthās, is pre-eminently the god of the Rāmosīs and Dhangars. The former declare that Khandoba, before he attained divinity, was a Lingāyat Vānī. Mackintosh describes how the Rāmosīs, before setting out to commit robbery or dacoity, always made a vow to Khandoba that if the expedition was successful they would make him a suitable offering. On their way home with the loot the Rāmosīs invariably dispatched one of their band to offer one, two or five rupees to the god.⁶ Oomiah Nāik (Uma Nāyak), the famous Rāmosī freebooter, who was hunted down by Bombay troops and hanged in 1832, gave to the Jejuri temple a large portion of the sum obtained by looting the Government treasury at Bhāmburda in February 1824.¹ One of Uma’s associates was a Vāghe (male devotee of Khandoba) named Khandu; and when in 1828 Uma was summoned to a formal interview with the British officer in command of the cavalry operating against him, he borrowed for the occasion the āftābgīrs, horns and other insignia from the temple at Jejuri.² It is well known that the Rāmosīs recognize the binding power of an oath taken on the turmeric (bhandar), the vegetable abode of Khandoba³; and the war-cry of their marauding gangs in old days was ‘El-kōt, El-kōt,’ which is a synonym of Khandoba, signifying ‘70 million’ (from Kanarese yelu kōt)—an allusion to the number of the demon hosts of Mani slain by the god in the form of Mārtand. By the Dhangars also Khandoba is specially revered, and is represented in his usual equestrian form, not only with Mālsubāī his wife in front, but also with his mistress Bānāī, a Dhangar woman, behind him. Khandoba’s love for Bānāī (Bān (arrow) and āi (mother) = quiver ?) forms the theme of a Dhangar tale, which appears to enshrine some reminiscence of marriage by capture.⁴ The great temple at Jejuri contains minor shrines of Hegadi, a Dhangar, and Pradhān, a Lingāyat Vānī, who are supposed to have been Khandoba’s ministers, and Hegadi, according to Dhangar belief, was the brother of Khandoba’s mistress, Bānāī. The shrines of Mālsubāī and Bānāī are situated respectively at the top and bottom of the hill occupied by the Jejuri temple; the god, it is said, having been forced to put his mistress’ shrine at the foot of the hill in order to appease the jealousy of Mālsubāī.⁵ It is not easy to determine what facts, if any, underlie these details of the Khandoba cult. Prima facie the tale looks like an attempt to co-ordinate the ancient connexion between the Dhangars and their tribal god with the reverence paid to him by later sectaries of a higher social status. It is sufficient for the present theme to point to the close relationship between this early tutelary deity and the ancient tribe of Dhangars or graziers, who offer stone sheep to Bānāi, the god’s mistress, in the hope that their flocks and herds will multiply.¹ One might hazard the conjecture that Bānāi herself is the primeval earth-goddess, whose sexual commerce with a male god of fertility was regarded as imperative for the welfare and safety of her devotees.

v1.c5.s3.60

⁶ Mackintosh, Account of the Ramoossies, 26, 28.

v1.c5.s3.61

¹ Ibid., 132.

v1.c5.s3.62

² Ibid., 155, 157.

v1.c5.s3.63

³ B.G., xviii, pt. i. 290.

v1.c5.s3.64

⁴ Note that the Gaolis, or milkmen and cattle-graziers, also claim a special connexion with Khandoba through the maiden Pālāī, to whom Khandoba appeared at Pāl in Sātāra and was married.

v1.c5.s3.65

⁵ B.G., xviii. 132 ff. ; Gupte, Hindu Holidays, &c., 21.

v1.c5.s3.66

¹ B.G., xviii, pt. i. 132.

v1.c5.s3.67

Fifthly, Khandoba is a god of fertility. Both his temples at Jejuri and Pāl are directly associated with the ancient custom of religious prostitution or sanctified harlotry. Children of both sexes are dedicated to him by lower-class Marāthās, in pursuance usually of vows made with the object of obtaining offspring. If a Kunbī has no children, or if his children are short-lived, he will make a vow to Khandoba that if the latter blesses him with a child, he will set it apart to worship and attend upon the god.² Male children thus dedicated are called Vāghes or Vāghyas (i.e. ‘tigers’), and the females Murlīs (i.e. ‘flutes’). Both the Vāghes and Murlīs comprise Marāthā, Dhangar and Mahār sub-divisions, of which the two former eat together.³ Married women also become Murlīs, and leave their husbands and children on the plea that they have made a vow to the god, or that they have been warned in a dream to become wives of Khandoba. The Murlīs, who earn their living by prostitution, are regarded with considerable respect by middle and lower-class Hindus; those who remain at the temple give their earnings to the god and perform various minor duties at the shrine; while others wander about the Deccan with a troupe of Vāghes, singing songs (lāvanis), often of a very erotic character, in honour of Khandoba, and supporting themselves by begging and prostitution. Murlīs and their attendant musicians frequently visit Bombay city, among the lower classes of which they enjoy considerable popularity; but on several occasions the indecent character of their performances has necessitated the intervention of the police. The words used at the dedication of a boy (Vāghe) and at the marriage of a girl (Murli) to the god are ‘El-kōt ghe’ (i.e. ‘Take, El-kōt’); ‘Elkōt’ being, as I have remarked above, the war-cry of the Rāmosī robber-bands in the eighteenth and early nineteenth centuries.¹ The Vāghes, as they grow up, become temple ministrants and beggars, are regarded as Khandoba’s disciples, and wander about the country with their tiger-skin bags of turmeric, singing songs in honour of the god. In old days, at a festival held at Jejuri on the bright 6th of Mārga-Shirsha (November-December), a Vāghe was obliged to run a sword through his thigh, and thus wounded perambulate the town, while the bloody sword was laid as an offering before the god. Further, at all fairs in Khandoba’s honour, the ancient practice of hook-swinging was performed, chiefly by women, who were frequently members of the Rāmosī tribe.²

v1.c5.s3.68

² In 1903, when I was sub-divisional magistrate at Poona, the police prosecuted before me a man and an old woman for dedicating an eight-year-old girl to Khandoba. The accused were Mahārs. The evidence showed that the child had been attacked by plague, then epidemic in Poona, and that after trying various remedies her uncle and grandmother vowed to Khandoba that if he spared her life, she should be dedicated to him. The child recovered, and was duly married to the god, thus becoming the victim of an offence under the Indian Penal Code.

v1.c5.s3.69

³ B.E.S., Monograph 75.

v1.c5.s3.70

¹ The Vāghes’ sacred cry is *Jāi Khande rāyacha Elkot*, i.e. ‘Victory to Khandē Rāo.’ *Enthoven, Konkan Folk Lore*, pp. 22, 23.

v1.c5.s3.71

² Mackintosh, *Account of the Ramoossies*, 53; *Bom. Gaz.*, xviii, pt. iii. 132 f. For account of hook-swinging see Thurston, *E.N.S.I.*, pp. 487 ff.

v1.c5.s3.72

In Chapter IV of his *Adonis, Attis, Osiris*, Sir J. Frazer has dealt at length with the geographical distribution and origin of the system of sacred prostitution. All that need be said here is that this feature of the cult of the Marāthā god Khandoba, as also the custom of dedicating males to his service, has no parallel in northern India, and is obviously non-Aryan in character. To find any corresponding practices one has to examine the worship of ancient southern Indian earth-goddesses, like Yellamma, Māyaka, Firangāi and Ambābāi, to all of whom girls are dedicated by the well-known process of marriage to a dagger. The system is indeed a legacy of Dravidian culture, and may be assumed to have been established as one of the most serious features of their ritual among the aboriginal tribes which occupied the Deccan and Carnatic in early ages. The custom has shown extraordinary vitality. Aryan Brahmanism not only did not reject it, but lent it countenance by adopting Khandoba into the Brahmanic pantheon as an *avatār* of Siva, and by accepting the god as one of its regular household deities; Muhammadan rule in the Deccan exercised no effect upon it; and a century of undisturbed British suzerainty, with its sustained missionary and philanthropic endeavour, has failed to reduce appreciably the numbers of women who are set apart or elect to serve as the brides of Khandoba and the mistresses of men.

v1.c5.s3.73

No less aboriginal in character is the figure of Bhavānī, Sivājī’s family goddess. Like Khandoba, she is the patron deity of the Rāmosī tribe; for the Rāmosīs of Purandar always made vows to her shrine at Kōndanpur, near Singarh, whenever they were about to set forth plundering in the Konkan; and, like Khandoba, the goddess received a share of the Government treasure looted at Bhāmburda in 1824. She approximates in this matter to Kālī, who was the patron goddess of the Thugs. In the Deccan Bhavānī is one of the old *grāmadēvatas* or village-guardians; she can cause sickness and misfortune, but if pacified by the slaughter of a goat or cock will fend off troubles and give advice regarding the future.¹ Nor is her worship confined to the Marāthā Kunbīs and allied tribes; the upper-class Marāthās reverence her also. In the old palace of the Rājās of Sātāra the audience-hall, built by Rājā Shāhjī in 1844, is dedicated to Bhavānī, and in former days the Dasahra festivities always commenced with the slaughter of a buffalo at Bhavānī’s shrine.² Sivājī himself, as Grant Duff and others have recorded, claimed to be inspired on important occasions by the goddess; he built her a new shrine at Pratāpgarh; he named his trusty sword after her. Her famous shrine at Tuljāpur in the modern Hyderābād State is believed to be the scene of her victory over the buffalo-demon³; and the brass images of Bhavānī, still procurable in the Deccan, represent her in the form *Mahīshāsūra-mārdini*, i.e. ‘the slayer of the buffalo demon.’⁴ Bhavānī is without doubt one of the ancient *Mātas*, *Mātris* or mother-goddesses of the Dravidian or pre-Aryan population of India, who could only be placated by bloody sacrifices, in which in ancient times human beings, as well as buffaloes or goats, figured as the victims. She belongs to the same order of ideas and the same level of civilization as Kālī, Bhāiravi, and Chāndika, who were originally local earth and jungle goddesses, and were transformed by the contact of later Aryan ideas into manifestations of Mahādēvi, the Sakti or personified energy of the great god Siva.¹ Like other leading deities, Bhavāni of Tuljāpur has her own special vāhan or steed in the shape of the tiger. Crooke has suggested that these animal steeds may be relics of totemism, which was undoubtedly prevalent in India in past ages, or may be deities of the aboriginal tribes imported in later times into Brahmanic Hinduism.² Whatever be the origin of these ‘vehicles,’ Bhavāni herself is clearly one of the pre-Aryan mother-goddesses, whose energies are usually recruited in two ways, viz. by the rites of the sacred marriage and by the blood-sacrifice, often specially of male victims.³

v1.c5.s3.74

¹ B.G., xviii, pt. i. 289.

v1.c5.s3.75

² B.G., xix. 565, 567.

v1.c5.s3.76

³ See Crooke, *P.R.* and *F.L.*, ii. 236, 237, for buffalo-worship.

v1.c5.s3.77

⁴ I have three of these images in my possession. A similar image is represented on plates 34 and 35 of Moor’s *Hindu Pantheon*, London, 1810.

v1.c5.s3.78

¹ Barth, Religions of India; Macnicol, Indian Theism; Crooke, P.R. and F.L.

v1.c5.s3.79

² Crooke, P.R. and F.L.

v1.c5.s3.80

³ Crooke, The Cults of the Mother Goddesses in India, read before the British Association, 1919.

v1.c5.s3.81

To sum up therefore, it is permissible to hold that the inferences deducible from the data hitherto collected by the Ethnographical Survey regarding the origin of the Marāthās, both of the upper and lower classes, find some corroboration in the deep and instinctive attachment of the Marāthās to their ancestral village hakks (rights), and in the character of their two most widely honoured deities. The more exalted position now occupied by Bhavāni as the result of the catholic labours of the heralds’ college of Brahmanism cannot disguise the fact that she was originally one of the savage mother-goddesses worshipped by the older tribes of the hills and forests of India. Khandoba likewise, the patron of jungle-robbers and of old nomadic tribes of sheep and cattle graziers, combining in himself ancient plant-worship and non-Brahmanic animal-worship, and associated at his chief shrines with a custom of male-dedication and sanctified harlotry which finds its only parallel in the cults of obviously pre-Aryan figures like Yellamma, the universal mother, and of the ancient goddesses of Assyria, Babylonia and Egypt, seems to deride the hypothesis that his devoted worshippers can justifiably lay claim to Scythian, Rājpūt or indeed any strain of foreign blood. Although his cult has been to some extent overlaid by later accretions arising out of the Brahmanic practice of absorbing the local gods of the older tribes into post-Vedic Hinduism, Khandoba still remains primarily a non-Aryan spirit-scaring god of fertility, and maintains his pristine importance simply because his devotees, the upper and lower-class Marāthās and allied tribes and castes, are the direct descendants of the people who established his cult in prehistoric Mahārāshtra.

Linked texts (1)

Brahmapurāṇa — Vyāsa, Sanskrit