κρῖν’ ἄνδρας κατὰ φῦλα, κατὰ φρήτρας, Ἀγάμεμνον, ὡς φρήτρη φρήτρηφιν ἀρήγῃ, φῦλα δὲ φύλοις. Il. II. 362-3.
Social divisions: the tribe
UP to this point I have been dealing with the racial divisions of the people of India, with ethnology properly so called. I turn now to their social divisions, to the ethnographic data as distinguished from the ethnological. These divisions are either tribes or castes, which in their turn are further subdivided with reference usually to matrimonial considerations. A tribe as we find it in India is a collection of families or groups of families bearing a common name which as a rule does not denote any specific occupation; generally claiming common descent from a mythical or historical ancestor, and occasionally from an animal, but in some parts of the country held together rather by the obligations of blood feud than by the tradition of kinship; usually speaking the same language and occupying, professing, or claiming to occupy a definite tract of country. A tribe is not necessarily endogamous; that is to say, it is not an invariable rule that a man of a particular tribe must marry a woman of that tribe and cannot marry a woman of a different tribe.
Types of tribes.
We may distinguish several kinds of tribes in various parts of India, and although it cannot be said that each of the seven racial types has its own distinctive form of tribe, nevertheless the correspondence between the two sets of groupings is sufficiently close to warrant the conjecture, that each type was originally organized on a characteristic tribal basis and that, where tribes have disappeared, their disappearance has been effected by caste insensibly absorbing and transforming the tribal divisions which it found in possession of particular localities. In describing the varieties of tribes I shall therefore follow the ethnic types already determined by physical characters.
The Dravidian tribe.
The Dravidian tribe exists in its most compact and vigorous form among the people of Chutia Nagpur. Descriptions of two typical instances are given in the Appendix under the heads of Munda and Santāl. Such a tribe is generally divided into a number of exogamous groups, each of which bears the name of an animal or plant common in the locality. Usually also there is a distinct village organization comprising in its most developed forms a headman with his assistant and a priest with various acolytes whose business it is to propitiate the various undefined powers from whom physical ills are to be apprehended. Another remarkable instance of the tribal organization of the Dravidians is to be found among the Kandhs or Kondhs of the Orissa Kandh Māls, once infamous for the human sacrifices which they offered to the earth goddess with the object of ensuring good crops and immunity from disease and accidents. A grim memorial of these forgotten horrors is to be seen in the Madras Museum in the form of a rude representation in wood of the head and trunk of an elephant pivoted on a stout post. To this the victim was bound head downwards and the machine was slowly turned round in the centre of a crowd of worshippers who hacked and tore away scraps of flesh to bury in their fields, chanting the while a ghastly hymn, an extract from which illustrates very clearly the theory of sympathetic magic underlying the ritual—
As the tears stream from thine eyes, So may the rain pour down in August ; As the mucus trickles from thy nostrils, So may it drizzle at intervals ; As thy blood gushes forth, So may the vegetation sprout ; As thy gore falls in drops, So may the grains of rice form.
A number of these wooden elephants, which had been used at sacrifices, were found and burnt by the British officers who put down human sacrifice in the Kandh country. The worm-eaten specimen at Madras is probably unique.* The Kandhs are divided into 50 gochis or exogamous sects, each of which bears the name of a muta or village, believes all its members to be descended from a common ancestor, and as a rule dwells as a body of blood-relations in the commune or group of villages after which it is called. The Kandh gochi appears, therefore, to represent the nearest approach that has yet been discovered to the local exogamous tribe believed by Mr. McLennan to be the primitive unit of human society.
The Mongoloid tribe.
The Mongoloid type of tribe as found in the Naga Hills is divided somewhat on the same pattern as the Kandhs into a number of khels, each of which is in theory an exogamous group of blood relations dwelling apart in its own territory and more or less at war with the rest of the world. Each khel fortifies the locality which it inhabits with a stockade, a deep ditch full of bamboo caltrops, and a craftily devised ladder, and raids are constantly made by one upon the other for the purpose of capturing wives. So far as our present researches have gone no very clear traces have been found of totemism among the Mongoloid races of India,* but the Mongoloid people of the Eastern Himalayas and the Chittagong Hills have a singular system of exogamous groups based upon their real or mythical ancestors. Instances of this grotesque variant of eponymy are the Chakma clans Ichāpochā, “the man who ate rotten shrimps,” Pirā bhāngā, “the fat man who broke the stool,” Aruyā, “the skeleton,” and so forth.†
**The Turko-Iranian tribes: the Afghan type. **
Among the Turko-Iranians there seem to be two distinct types of tribe :—
The Baloch and Brahui type.
(a) Tribes based upon kinship like the Afghan group of tribes, otherwise known as Pathāns or speakers of the Pashtu language, who trace their lineage to one Qais Abdul Rashīd who lived in the country immediately to the west of the Takht-i-Sulaimān and was thirty-seventh in descent from Malik Tālūt (King Saul). In theory, says Mr. Hughes-Buller in his admirable account of the tribal system of Baluchistan,‡ “an Afghan tribe is constituted from the number of kindred groups of agnates ; that is to say, descent is through the father, and the son inherits the blood of his father. Affiliated with a good many tribes, however, are to be found a certain number of alien groups known as Mindūn or Hamsayah. The latter term means ‘living in the same shade.’ These groups are admittedly not united to the tribe by kinship.” They do not, indeed, even claim descent from the common ancestor, and the nature of the tie that binds them to the tribe is best expressed in the picturesque phrase which describes them as Neki aur badi men sharik, “partners for better or worse”; in other words, active participators in any blood-feud that the tribe may have on their hands. Yet such is the influence of the idea of kinship upon which the tribe is based that the alien origin of the Hamsayah is admitted with reluctance, and although for matrimonial purposes they are looked upon as inferior, the tendency is continually to merge the fact of common vendetta in the fiction of common blood. These are the two leading principles which go to the making of an Afghān tribe. There are also —Mr. Hughes-Buller explains—“two other ties which unite the smaller groups: common pasture, or, more important still, common land and water, and common inheritance. The area occupied by each section can be pretty easily localized, and a group which separates itself permanently from the parent stock and makes its way to a remote locality, where it either sets up for itself or joins some other tribe, ceases to have any part or portion with the parent stock. Here the test question is: ‘Has the individual or group on separating from the parent stock, departed only temporarily or permanently?’ For, among a population largely composed of graziers, there must be constant fission, groups leaving the locality of the majority for other places as pasture or water are required for the flocks. Where the change is only temporary, groups retain as a matter of course their union with the group to which they belong. There are others, however, who wish to sever their connection with the parent group permanently, and, once this has been done, the idea of participation in the common good and ill of the parent stock disappears. Common inheritance can, in the nature of things, only be shared by the more minute groups, and this, in the absence of blood-feud, is the bond of unity in the family or Kahol. And this leads me to explain that all the four principles which I have mentioned do not affect every group equally. Thus, the smaller groups or Kahols, which in most cases correspond with the family, are united by kinship and common inheritance, but within the family group there can be no blood-feud. For blood-feud can only be carried on when help is given from outside, and no one will help the murderer within the R PI. 5 family. Leaving the lowest group, we find that common good and ill, merging in the fiction of kinship, is the influence affecting all the groups, even the largest unit, of the tribe. Common land and water are only shared by comparatively minute groups, i.e., by the Khel or Zāī, but the groups united by common locality, and possibly by common grazing, are both numerous and large.”
(b) The second type of Turko-Iranian tribe is based, primarily, not upon agnatic kinship, but upon common good and ill; in other words, it is cemented together by the obligations arising from the blood-feud. There is no eponymous ancestor, and the tribe itself does not profess to be composed of homogeneous elements. In the case of the Marri tribe of Baloch Mr. Hughes-Buller has shown that “Brāhūis, Baloch from the Punjab, Baloch from other parts of Afghanistan, Khetrāns, Afghāns, Jāts, all gained easy admission to the tribe. As soon as a man joined the tribe permanently he became a participator in good and ill. Then, having shown his worth, he was given a vested interest in the tribal welfare by acquiring a portion of the tribal lands at the decennial division, and his admission was sealed with blood by women from the tribe being given to him or his sons in marriage. Starting, therefore, with the principle of participation in common good and common ill, participation in the tribal land came to be the essence of tribesmanship among the Marris. The process is easy to follow: Admission to participation in common blood-feud; then admission to participation in tribal land; and lastly admission to kinship with the tribe. It was not until after a man or group had been given a share of tribal land at the decennial distribution that women were given to him or them in marriage.” The same principles hold good in the case of the Brāhūi, who, like the Baloch, appear both by their history and by their physique to be of Central Asian or Scythian origin, though their numbers have been recruited from among Afghāns, Kurds, Jāgdals, Baloch, and other elements, all probably belonging to the same ethnic stock.
Both Baloch and Brāhūi possess an elaborate organization for offensive and defensive purposes, based in each case on the principle that the clan or section must provide for the service of the tribe a number of armed men proportioned to the share of the tribal land which it holds. The Brāhūi system, introduced by Nasir Khān about the end of the seventeenth century, is somewhat the more complete of the two, and binds together all the Brāhūi tribes in a regular confederacy which is now, according to Mr. Hughes-Buller, beginning to regard the British Government as its effective suzerain. A full account of the Brāhūi taken from Mr. Hughes-Buller’s report on the first census of Baluchistan will be found in the ethnographic volume of the Imperial Census Report for 1901.
Marriage in Baluchistan.
None of the numerous tribes comprised in the names Afghān, Brāhūi, Baloch are strictly endogamous, and stalwart aliens, whose services are considered worth having, are admitted into the tribe by the gift of a wife, or perhaps one should rather say the loan, for, in the absence of stipulations to the contrary, a woman so given goes back to her own family on the death of her husband. Among the Baloch and Brāhūi, however, a distinct tendency towards endogamy results from the practice of marrying a woman of the same group, a near kinswoman, or, if possible, a first cousin. This seems to be due partly to the feeling that a woman’s marriage to an outsider deprives the tribe of the accession of strength that may accrue to it from her offspring; and partly also to the belief that “while among animals heredity follows the father, among human beings it follows the mother. It is argued, therefore, that there is more hope of the stock remaining pure if a man marries a woman who is nearly related to him.” In marked contrast to the Baloch and Brāhūi, the business instincts of the Afghān lead him to regard women as a marketable commodity, and under the system of walwar or payment for wives “girls are sold to the highest bidder, no matter what his social status.” It is possible, however, that in a tribe of comparatively homogeneous descent the sentiment in favour of purity of blood may operate less strongly than in a tribe of admittedly composite structure.
The word “caste.”
We shall see in a later chapter how the word fetish, which has had a great vogue in the history of religion, owes its origin to the Portuguese navigators who were brought into contact with the strange religious observances of the natives of West Africa. In the same way caste, which has obtained an equally wide currency in the literature of sociology, comes from the Portuguese adventurers who followed Vasco de Gama to the west coast of India. The word itself is derived from the Latin castus and implies purity of breed. In his article on caste in “Hobson-Jobson,” Sir Henry Yule quotes a decree of the sacred council of Goa dated 1567, which recites how in some parts of that province “the Gentoos divide themselves into distinct races or castes (castas) of greater or less dignity, holding the Christians as of lower degree, and keep these so superstitiously that no one of a higher caste can eat or drink with those of a lower.” From that time to this it has been assumed without much critical examination that the essential principle of caste is mainly concerned with matters of eating and drinking. It need not surprise us to find foreign observers laying stress upon the superficial aspects of a social system which they understood but imperfectly, and overlooking the material fact that the regulations affecting food and drink are comparatively fluid and transitory, while those relating to marriage are remarkably stable and absolute.
Definition of caste.
A caste may be defined as a collection of families or groups of families bearing a common name; claiming common descent from a mythical ancestor, human or divine; professing to follow the same hereditary calling; and regarded by those who are competent to give an opinion as forming a single homogeneous community. The name generally denotes or is associated with a specific occupation. A caste is almost invariably endogamous in the sense that a member of the large circle denoted by the common name may not marry outside that circle, but within the circle there are usually a number of smaller circles each of which is also endogamous. Thus it is not enough to say that a Brāhman at the present day cannot marry any woman who is not a Brāhman; his wife must not only be a Brāhman, she must also belong to the same endogamous division of the Brāhman caste.
M. Senart’s desecripton.
By the side of this rigid definition I may place the general description of caste which is given by M. Emile Senart in his fascinating study of the caste system of India. After reminding his readers that no statement that can be made on the subject of caste can be considered as absolutely true, that the apparent relations of the facts admit of numerous shades of distinction, and that only the most general characteristics cover the whole of the subject, M. Senart goes on to describe a caste as a close corporation, in theory at any rate rigorously hereditary; equipped with a certain traditional and independent organization, including a chief and a council; meeting on occasion in assemblies of more or less plenary authority, and joining in the celebration of certain festivals; bound together by a common occupation; observing certain common usages which relate more particularly to marriage, to food and to questions of ceremonial pollution; and ruling its members by the exercise of a jurisdiction the extent of which varies, but which succeeds, by the sanction of certain penalties and above all by the power of final or revocable exclusion from the group, in making the authority of the community effectively felt.
An English parallel.
These, in the view of one of the most distinguished of French scholars, are the leading features of Indian caste. For my own part I have always been much impressed by the difficulty of conveying to European readers who have no experience of India even an approximate idea of the extraordinary complexity of the social system which is involved in the word “caste.” At the risk of being charged with frivolity I shall, therefore, venture on an illustration, based on one which I published in Blackwood’s Magazine a good many years ago, of a caste expressed in terms of an English social group. Let us take an instance, and, in order to avoid the fumes of bewilderment that are thrown off by uncouth names, let us frame it on English lines. Let us imagine the great tribe of Smith, the “noun of multitude,” as a famous headmaster used to call it, to be transformed by art magic into a caste organized on the Indian model, in which all the subtle nuances of social merit and demerit which Punch and the society papers love to chronicle should have been set and hardened into positive regulations affecting the intermarriage of families. The caste thus formed would trace its origin back to a mythical eponymous ancestor, the first Smith who converted the rough stone hatchet into the bronze battleaxe and took his name from the “smooth”* weapons that he wrought for his tribe. Bound together by this tie of common descent, they would recognize as the cardinal doctrine of their community the rule that a Smith must always marry a Smith, and could by no possibility marry a Brown, a Jones, or a Robinson. But over and above this general canon three other modes or principles of grouping within the caste would be conspicuous. First of all, the entire caste of Smith would be split up into an indefinite number of “in-marrying” clans based upon all sorts of trivial distinctions. Brewing Smiths and baking Smiths, hunting Smiths and shooting Smiths, temperance Smiths and licensed-victualler Smiths, Smiths with double-barrelled names and hyphens, Smiths with double-barrelled names without hyphens, Conservative Smiths, Radical Smiths, tinker Smiths, tailor Smiths, Smiths of Mercia, Smiths of Wessex—all these and all other imaginable varieties of the tribe Smith would be as it were crystallized by an inexorable law forbidding the members of any of these groups to marry beyond the circle marked out by the clan-name. Thus the Unionist Mr. Smith could only marry a Unionist Miss Smith, and might not think of a Home Rule damsel; the free-trade Smiths would have nothing to say to the tariff reformers; a hyphen-Smith could only marry a hyphen-Smith, and so on. Secondly, within each class enquiry would disclose a number of “out-marrying” groups, bearing distinctive names, and governed by the rule that a man of one group could in no circumstances marry a girl of the same group. In theory each group would be regarded as a circle of blood-kindred and would trace its descent from a mythical or historical ancestor like the Wayland-Smith of the Berkshire hills, the Captain Smith who married Pocahontas, or the Mr. W. H. Smith of the railway bookstalls. The name of each would usually suggest its origin, and marriages within the limits defined by the group-name would be deemed incestuous, however remote the actual relationship between the parties concerned. A Wayland could not marry a Wayland, though the two might come from opposite ends of the kingdom and be in no way related, but must seek his bride in the Pocahontas or bookstall circle, and so on. Thus the system, the converse of that just described, would effect in a cumbrous and imperfect fashion what is done for ourselves by the table of prohibited degrees at the end of the Prayer-book—cumbrous because it would forbid marriage between people who are in no sense relations, and imperfect because the group-name would descend in the male line and would of itself present no obstacle to a man marrying his grandmother. Thirdly, running through the entire series of clans we should find yet another principle at work breaking up each in-marrying clan into three or four smaller groups which would form a sort of ascending scale of social distinction. Thus the clan of hyphen-Smiths, which we take to be the cream of the caste—the Smiths who have attained to the crowning glory of double names securely welded together by hyphens—would be again divided into, let us say, Anglican, Dissenting, and Salvationist hyphen-Smiths, taking regular rank in that order. Now the rule of this series of groups would be that a man of the highest or Anglican group might marry a girl of his own group or of the two lower groups, that a man of the second or Dissenting group might take a Dissenting or Salvationist wife, while a Salvationist man would be restricted to his own group. A woman, it will be observed, could under no circumstances marry down into a group below her, and it would be thought eminently desirable for her to marry into a higher group. Other things being equal, it is clear that two-thirds of the Anglican girls would get no husbands, and two-thirds of the Salvationist men no wives. These are some of the restrictions which would control the process of match-making among the Smiths if they were organized in a caste of the Indian type. There would also be restrictions as to food. The different in-marrying clans would be precluded from dining together, and their possibilities of reciprocal entertainment would be limited to those products of the confectioner’s shop into the composition of which water, the most fatal and effective vehicle of ceremonial impurity, had not entered. Water pollutes wholesale, but its power as a conductor of malign influence admits of being neutralized by a sufficient admixture of milk, curds, whey, or clarified butter—in fact, of anything that comes from the sacred cow. It would follow from this that the members of our imaginary caste could eat chocolates and other forms of sweetmeats together, but could not drink tea or coffee, and could only partake of ices if they were made with cream and were served on metal, not porcelain, plates. I am sensible of having trenched on the limits of literary and scientific propriety in attempting to describe an ancient and famous institution in unduly vivacious language, but the parallel is as accurate as any parallel drawn from the other end of the world can well be, and it has the advantage of being presented in terms familiar to European readers. The illustration, indeed, may be carried a step further. If we suppose the various aggregates of persons bearing the two or three thousand commonest English surnames to be formed into separate castes and organized on the lines described above, so that no one could marry outside the caste-name and could only marry within that limit subject to the restrictions imposed by differences of residence, occupation, religion, custom, social status, and the like—the mental picture thus formed will give a fairly adequate idea of the bewildering complexity of the Indian caste system.
Conversion of tribes into castes.
All over India at the present moment tribes are gradually and insensibly being transformed into castes. The stages of this operation are in themselves difficult to trace. The main agency at work is fiction, which in this instance takes the form of the pretence that whatever usage prevails to-day did not come into existence yesterday, but has been so from the beginning of time. It may be hoped that the Ethnographic Survey now in progress will throw some light upon the singular course of evolution by which large masses of people surrender a condition of comparative freedom and take in exchange a condition which becomes more burdensome in proportion as its status is higher. So far as my own observation goes, several distinct processes are involved in the movement, and these proceed independently in different places and at different times.
(1) The leading men of an aboriginal tribe, having somehow got on in the world and become independent landed proprietors, manage to enrol themselves in one of the more distinguished castes. They usually set up as Rajputs, their first step being to start a Brāhman priest, who invents for them a mythical ancestor, supplies them with a family miracle connected with the locality where the tribe is settled, and discovers that they belong to some hitherto unheard-of clan of the great Rajput community. In the earlier stages of their advancement they generally find great difficulty in getting their daughters married, as they will not take husbands from their original tribe and Rajputs of their adopted caste will, of course, not condescend to alliances with them. But after a generation or two their persistency obtains its reward and they intermarry, if not with pure Rajputs, at least with a superior order of manufactured Rajputs whose promotion into Brahmanical society dates far enough back for the steps by which it was gained to have been forgotten. Thus a real change of blood may take place, as indeed one is on occasion in a position to observe, while in any case the tribal name is completely lost and with it all possibility of correctly separating this class of people from the Hindus of purer blood and of tracing them to any particular Dravidian or Mongoloid tribe. They have been absorbed in the fullest sense of the word, and henceforth pass and are locally accepted as high-class Hindus. All stages of the process, family miracle and all, can be illustrated by actual instances taken from the leading families in various parts of India. The most picturesque instance of the class of legend to which I refer is that associated with the family of the Maharajas of Chutia Nāgpur, who call themselves Nāgbansi Rajputs, and on the strength of their mythical pedigree have probably succeeded in occasionally procuring wives of reputed Rajput blood. The story itself is a variant of the well-known Lohengrin legend. It tells how a king of the Nāgas or snakes, the strange prehistoric race which figures so largely in Indian mythology, took upon himself human form and married a beautiful Brāhman girl of Benares. His incarnation, however, was in two respects incomplete, for he could not get rid of his forked tongue and his evil-smelling breath. Consequently, as the story goes, in order to conceal these disagreeable peculiarities he always slept with his back to his wife. His precautions, however, were unsuccessful, for she discovered what he sought to conceal, and her curiosity was greatly inflamed. But the snake king, being bound by the same condition as his Teutonic prototype, could only disclose his origin at the cost of separation from his wife. Accordingly, by a device familiar to Indian husbands, he diverted her attention by proposing to take her on a pilgrimage to the shrine of Jagannāth at Puri in Orissa. The couple started by the direct route through the hills and forests of Chutia Nāgpur, and when they reached the neighbourhood of the present station of Rānchi the wife was seized by the pains of childbirth. Her curiosity revived, and she began to ask questions. By folklore etiquette questions asked on such an occasion must be answered, and her husband was compelled to explain that he was really the Takshak Raja, the king of the snakes. Having divulged this fatal secret he did not, like Lohengrin, make a dignified exit to the strains of slow music. He straightway turned into a gigantic cobra, whereupon his wife was delivered of a male child and died. The poor snake made the best of the trying position in which he found himself; he spread his hood and sheltered the infant from the rays of the midday sun. While he was thus occupied, some wood-cutters of the Mūndā tribe appeared upon the scene, and decided that a child discovered in such remarkable circumstances must be destined to a great future and should at once be adopted as their chief. That is the family legend of the Nāgbansi Rajas of Chutia Nāgpur.* It was received with derisive merriment by a number of genuine Rajputs who attended a conference which I held at Mount Abu in 1900 for the purpose of organizing the census of Rajputana. They had never heard of such a thing as a Nāgbansi Rajput, but they entirely appreciated the point of the story. Similar tales, associated sometimes with a peacock, sometimes with a cow, sometimes with other animals or trees, are told of various land-owning families which have attained brevet rank as local Rajputs. Any one who has the curiosity to inquire into the distribution of tenures on the estates of these manufactured Rajputs will usually find that a number of the best villages lying round the residence of the Chief are held on peppercorn rents by the descendants of the Brāhmans who helped him to his miraculous pedigree.
(2) A number of aborigines, as we may conveniently call them, though the term begs an insoluble question, embrace the tenets of a Hindu religious sect, losing thereby their tribal name and becoming Vaishnavas, Lingāyats, Rāmāyats, or the like. Whether there is any mixture of blood or not will depend upon local circumstances and the rules of the sect regarding intermarriage. Anyhow, the identity of the converts as aborigines is usually, though not invariably, lost, and this also may, therefore, be regarded as a case of true absorption.
(3) A whole tribe of aborigines, or a large section of a tribe, enrol themselves in the ranks of Hinduism under the style of a new caste, which, though claiming an origin of remote antiquity, is readily distinguishable by its name from any of the standard and recognized castes. Thus the great majority of the Kochh inhabitants of Jalpaiguri, Rangpur, and part of Dinājpur now invariably describe themselves as Rājbansis or Bhanga Kshatriyas—a designation which enables them to represent themselves as an outlying branch of the Kshatriyas of Hindu tradition who fled to north-eastern Bengal in order to escape from the wrath of Parasu-Rāma. They claim descent from Raja Dasaratha, the father of Rāma, they keep Brāhmans, imitate the orthodox ritual in their marriage ceremony, and have begun to adopt the Brahmanical system of gotras. In respect of this last point they are now in a curious state of transition, as they have all hit upon the same gotra (Kāsyapa) and thus habitually transgress the primary rule of the Brahmanical system, which absolutely prohibits marriage within the gotra. But for this defect in their connubial arrangements—a defect which will probably be corrected in course of time as they and their priests rise in intelligence—there would be nothing in their customs to distinguish them from Indo-Aryan Hindus; although there has been no mixture of blood and they remain thoroughly Kochh under the name of Rājbansi. It is right to add that, however baseless the tradition must be in the case of the tribe as a whole, it does not follow that it may not enshrine a grain of fact as applied to their Chief. The Rajputs in India, like the Normans in Europe, travelled far afield in their conquering excursions. In a country where history masquerades in the garb of legend there is nothing prima facie improbable in the conjecture that the story of the Bhanga-Kshatriyas may be really a mythological version of the true origin of the reigning family of Cooch Bihār. A Chief of the higher race ruling a people of the lower is a phenomenon too common to require explanation.
(4) A whole tribe of aborigines, or a section of a tribe, become gradually converted to Hinduism without, like the Rājbansis, abandoning their tribal designation. This is what has happened among the Bhūmij of Western Bengal. Here a pure Dravidian race have lost their original language and now speak only Bengali; they worship Hindu gods in addition to their own (the tendency being to relegate the tribal gods to the women) and the more advanced among them employ Brāhmans as family priests. They still retain a set of totemistic exogamous sub-divisions closely resembling those of the Mundas and the Santāls. But they are beginning to forget the totems which the names of the sub-divisions denote, and the names themselves will probably soon be abandoned in favour of more aristocratic designations. The tribe will then have become a caste in the full sense of the word, and will go on stripping itself of all customs likely to betray its true descent. The physical characteristics of its members will alone survive. With their transformation into a caste the Bhūmij will be more strictly endogamous than they were as a tribe, and even less likely to modify their physical type by intermarriage with other races.
By such processes as these, and by a variety of complex social influences whose working cannot be precisely traced, a number of types or varieties of caste have been formed which admit of being grouped as follows:—
Tribal castes.
(i) The tribal type, where a tribe like the Bhūmij referred to above has insensibly been converted into a caste, preserving its original name and many of its characteristic customs, but modifying its animistic practices more and more in the direction of orthodox Hinduism and ordering its manner of life in accordance with the same model. Numerous instances of this process are to be found all over India; it has been at work for centuries and it has even been supposed that the Sudras of Indo-Aryan tradition were originally a Dravidian tribe which was thus incorporated into the social system of the conquering race. Considerations of space preclude me from attempting an exhaustive enumeration of the castes which may plausibly be described as tribes absorbed into Hinduism, but I may mention as illustrations of the transformation that has taken place, the Ahīr, Dom, and Dosādh of the United Provinces and Bihar; the Gūjar, Jāt, Meo, and Rajput of Rajputana and the Punjab; the Koli, Mahār, and Marātha of Bombay; the Bagdi, Bāuri, Chandāl (Nāmasudra), Kaibartta, Pod, and Rājbansi-Kochh of Bengal; and in Madras the Māl, Nāyar, Vellāla, and Paraiyan or Pariah, of whom the last retain traditions of a time when they possessed an independent organization of their own and had not been relegated to a low place in the Hindu social system.
Functional castes.
(ii) The functional or occupational type of caste is so numerous and so widely diffused and its characteristics as.re so prominent that community of function is ordinarily regarded as the chief factor in the evolution of caste. Whatever the original impulse may have been, it is a matter of observation at the present day not only that almost every caste professes to have a traditional occupation, though many of its members have abandoned it, but that the adoption of new occupations or of changes in the original occupation may give rise to sub-divisions of the caste which ultimately develop into entirely distinct castes. Thus among the large castes shown in the maps at the end of this volume the Ahīrs are by tradition herdsmen; the Brāhmans priests; the Chamārs and Mochis workers in leather; the Chuhras, Bhangis, and Doms scavengers; the Dosādhs village watchmen and messengers; the Goālas milkmen; the Kaibarttas and Kewats fishermen and cultivators; the Kāyasths writers; the Koiri and Kāchhi market gardeners; the Kumhārs potters; the Pods fishermen; and the Teli and Tili oil-pressers and traders. But the proportion of a caste that actually follows the traditional occupation may vary greatly. It is shown in the Bengal Census Report* that 80 per cent. of the Ahīrs in Bihar are engaged in agriculture; that of the Bengal Brāhmans only 17 per cent. and of the Bihar Brāhmans only 8 per cent. are engaged in religious functions ; that not more than 8 per cent. of the Chamārs in Bihar live by working in leather, the remainder being cultivators or general labourers ; that two-thirds of the Kāyasths in Bengal are agriculturists, and that only thirty-five per cent. of the Telis follow their traditional profession. A remarkable instance of the formation of a caste on the basis of distinctive occupation is supplied by the Garpagāri or hail-averter in the Marātha districts of the Central Provinces, a village servant whose duty it is to control the elements and protect the crops from the destructive hail-storms which are frequent in that part of India. For this, says Mr. Russell, " he receives a contribution from the cultivators ; but in recent years an unavoidable scepticism as to his efficiency has tended to reduce his earnings. Mr. Fuller told me that on one occasion when he was hastening through the Chanda District on tour and pressed for time, the weather at one of his halting places looked threatening, and he feared that it would rain and delay the march. Among the villagers who came to see him was the local Garpagāri, and not wishing to neglect any chance he ordered him to take up his position outside the camp and keep off the rain. This the Garpagāri did, and watched through the night. In the event the rain held off, the camp moved, and that Garpagāri’s reputation was established for life." * Changes of occupation in their turn, more especially among the lower castes, tend to bring about the formation of separate castes. The Sadgops of Bengal have within recent times taken to agriculture and broken away from the pastoral caste to which they originally belonged ; the educated Kaibarttas and Pods are in course of separating themselves from their brethren who have not learnt English ; the Madhunāpit are barbers who became confectioners ; the Chāsadhobas washermen who took to agriculture. But perhaps the best illustration of the contagious influence of the fiction that differences of occupation imply a difference of blood is to be found in the list of Musalman castes enumerated by Mr. Gait in the Bengal Census Report of 1901.† This motley company includes the Abdāl of Northern and Eastern Bengal, who circumcise Muhammadan boys and castrate animals, while their women act as mid-wives ; the Bhatiyāra or inn-keepers of Bihar ; the butchers (Chik and Kasāi) ; the drummers (Nagārchi and Dafāli), of whom the latter exorcise evil spirits and avert the evil eye by beating a drum (daf) and also officiates as priests at the marriages and funerals of people who are too poor to pay the regular Qāzi; the cotton-carders (Dhunia or Nadāf) numbering 200,000 in Bengal; the barbers (Hajjām or Turk-Nāia); the Jolāha weavers, cultivators, bookbinders, tailors, and dyers numbering nearly a quarter of a million in Bengal and nearly three millions in India; the oil-pressers (Kalu); the greengrocers (Kunjra); the embroiderers (Patwa), and a number of minor groups. All of these bodies are castes of the standard Hindu type with governing committees (panchāyats or mātbars) of their own who organize strikes and see that no member of the caste engages in a degrading occupation, works for lower wages than his brethren, eats forbidden food, or marries a woman of another caste. Breaches of these and various other unwritten ordinances are visited in the last resort by the extreme penalty of excommunication. This means that no one will eat or smoke with the offender, visit at his house, or marry his daughter, while in extreme cases he is deprived of the services of the barber and the washerman.
Sectarian castes.
(iii) The sectarian type comprises a small number of castes which commenced life as religious sects founded by philanthropic enthusiasts who, having evolved some metaphysical formula offering a speedier release from the tædium vitæ which oppresses the East, had further persuaded themselves that all men were equal, or at any rate that all believers in their teaching ought to be equal. As time went on the practical difficulties of realizing this ideal forced themselves upon the members of the sect; they found their company becoming unduly mixed; and they proceeded to reorganize themselves on the lines of an ordinary caste. A notable instance of this tendency to revert to the normal type of Hindu society is to be found in the present condition of the Lingāyat or Virshaiv caste of Bombay and Southern India, which numbers 2,900,000 adherents. Founded as a sect in the twelfth century by a reformer who proclaimed the doctrine of the equality of all who received the eightfold sacrament ordained by him and wore on their persons the mystic phallus emblematic of the god Siva, the Lingāyat community had begun by the close of the seventeenth century to develop endogamous sub-castes based upon the social distinctions which their founder had expressly abjured. At the recent Census the process of transforming the sect into a caste had advanced still further. In a petition presented to the Government of India the members of the Lingāyat community protested against the “most offensive and mischievous order” that all of them should be entered in the Census papers as belonging to the same caste, and asked that they might be recorded as Virshaiv Brāhmans, Kshatriyas, Vaisyas, or Sudras, as the case might be. It would be difficult to find a better illustration of the essentially particularist instinct of the Indian people, of the aversion with which they regard the doctrine that all men are equal, and of the growing attraction exercised by the aristocratic scheme of society which their ancient traditions enshrine. The legend of the four original castes may have no historical foundation, but there can be no question as to the spread of its influence or the strength of the sentiment which it inspires.
A somewhat similar case is that of the Sarāks of western Bengal, Chutia Nāgpur, and Orissa, who seem to be a Hinduized remnant of the early Jain people to whom local legends ascribe the ruined temples, the defaced images, and even the abandoned copper mines of that part of Bengal. Their name is a variant of Srāvaka (Sanskrit “hearer”), the designation of the Jain laity; they are strict vegetarians, never eating flesh, and on no account taking life, and if in preparing their food any mention is made of the word “cutting,” the omen is deemed so disastrous that everything must be thrown away. In Orissa they call themselves Buddhists and assemble once a year at the famous cave temples of Khandagiri near Cuttack to make offerings to the Buddhist images there and to confer on religious matters. But these survivals of their ancient faith have not saved them from the all-pervading influence of caste. They have split up into endogamous groups based partly on locality and partly on the fact that some of them have taken to the degraded occupation of weaving, and they now form a Hindu caste of the ordinary type. The same fate has befallen the Gharbāri Atīths, the Sānnyāsi, the Jugis, the Jāti-Baishtams of Bengal, the Bānhra of Nepal—Newārs, who were originally Buddhist priests but abandoned celibacy and crystallized into a caste—and the Bishnois and Sādhs of the United Provinces. The Bishnois of Rohilkhand, says Mr. Burn,* are divided into nine endogamous groups of sub-castes “called after the castes from which they were recruited. New converts take their place in the appropriate sub-castes.” In the case of the Sādhs “recruits are no longer admitted, and it is peculiar that no endogamous or exogamous divisions exist, the only restriction on marriage being that intermarriage is forbidden between two families as long as the recollection of a former marriage connexion between them remains. The instance is of special interest as the quality maintained by the tenets of the sect; which has developed into a caste, has not yet been destroyed, as is usual in such cases.” A still more remarkable, because a more modern, case is mentioned by Sir Henry Cotton, who states that “the more self-assertive portion of the Brahmo community” appears to be “in the course of forming” a new caste. All these curious developments serve to illustrate the comparatively insignificant part that religion has played in the shaping of the caste system, and the strength of the tendency to morcellement, to splitting up into fractional groups, that is characteristic of Hindu society. So long as the sectarian instinct confines itself to expressing a mere predilection for one god rather than another, or simply develops a new cult, however fantastic, which permits men to indulge in the luxury of religious eccentricity without quitting the narrow circle of their social environment, its operations are undisturbed and the sects which it forms may flourish and endure. But directly it invades the social sphere and seeks to unify and amalgamate groups of theoretically different origin it comes in contact with a force too strong for it and has to give way. Race dominates religion; sect is weaker than caste.
Even Christianity has not altogether escaped the subtle contagion of caste. Almost everywhere in India a tendency has been observed on the part of converts from Hinduism to group themselves according to the castes to which they originally belonged. This sometimes assumes the form of a division into two groups, the higher restricted to those who were members of the ‘clean’ castes from whom Brāhmans can take water, while the lower comprises all those of inferior rank. On the west coast the retention of caste distinctions was deliberately recognized by the Portuguese missionaries, and the results of this policy have survived down to the present day. The Indian Roman Catholic Christians of the Konkan, the low-lying strip of coast between the Western Ghāts and the sea, are divided into Bambans or Bammans (Brāhmans), Charodas or Chardos (Kshatriyas or Chhatris), Sudirs (Sudras), Renders (drawers of palm-juice), Gavids or Gavdas (salt-makers), Modvals (washermen), Kumbārs (potters), and Kāphris or Sidis (labourers), whose thick lips, slanting foreheads and curly beards suggest an infusion of Somali blood. Intermarriages among these groups, while not absolutely forbidden, are said to be rare, though in South Kanara such unions “are gradually becoming more frequent in cases in which members of castes other than the Bammans have succeeded in obtaining a good position in the official, legal, or commercial community.”* Infant marriage is forbidden among the Konkani Christians, but girls are married as soon as they are twelve years old, and sometimes even before that age under a special dispensation from the Bishop. Widow marriage, though not forbidden, “is as much condemned as among the pagans.” Many of them, especially the women, cannot bear the idea of eating beef, and they observe the characteristic Hindu prohibition against a wife addressing or speaking of her husband by his name. The marriage ceremony is performed in Church according to Christian rites, but it is preceded and followed by observances which are palpable survivals from the Hindu customs of betrothal and marriage. These include the formal bathing of the betrothed couple, the giving of a dinner to the poor for the benefit of the deceased ancestors of the family, the tying of a tāli or lucky necklace (which sometimes has a cross or a figure of the infant Jesus as a pendant) round the bride’s neck, the exchange of presents, and the formal transfer of the bride to her husband’s family.
Further south in the little State of Cochin on the Malabar coast, where Christianity has been established for many centuries and is believed by some authorities to date from apostolic times, a different principle has asserted itself. In the course of ages, disputes as to theological doctrine, ecclesiastical ritual, or spiritual supremacy have led to the formation among the non-Protestant Christians in Cochin of a number of sects—the Roman Catholics of the Latin rite, who use the Liturgy of the Romish Church in Latin, and are further subdivided into the Three Hundred, the Five Hundred, and the Seven Hundred, obscure schisms possibly derived merely from the number of families that were converted by the Portuguese missionaries on successive occasions; the Roman Catholics of the Syrian rite, who used the Romish Liturgy in ancient Syriac; the Chaldean Syrians, who are under the Patriarch of Babylon, and differ in several minute points of ritual from the Romo-Syrians; the Jacobite Syrians, who are under the Patriarch of Antioch; and the Reformed or St. Thomas Syrians, an offshoot of the Jacobites who recognize the supremacy neither of the Pope nor of the Patriarch of Antioch and obey a Bishop of their own. These last have come to some extent under Protestant influence, and they insist upon the title of St. Thomas Syrians as marking their close adherence to the teaching and ritual of the apostolic age. They deny that the Bible should be interpreted by the traditions of the Church; they reject confession, absolution, fasting, the invocation of Saints, and the veneration of relics; they object to masses for the dead and dispute the doctrine of baptismal regeneration. Of these seven sects the first five appear to have crystallized into regular castes between the members of which no intermarriage is possible. The two branches of the Jacobite Syrians still intermarry, subject to a further distinction between residents of the northern and southern divisions of the State, the former of whom claim to be superior to the latter on the ground of their descent from the first colonists from Syria.*
Castes formed by crossing.
(iv) Castes formed by crossing.—Modern criticism has been especially active in its attacks on that portion of the traditional theory which derives the multitude of mixed or inferior castes from an intricate series of crosses between members of the original four. No one can examine the long lists which purport to illustrate the working of this process without being struck by much that is absurd and inconsistent. But in India it does not necessarily follow that, because the individual applications of a principle are ridiculous, the principle itself can have no foundation in fact. The last thing that would occur to the literary theorists of those times, or to their successors the pandits of today, would be to go back upon actual facts, and to seek by analysis and comparison to work out the true stages of evolution. They found, as I infer from plentiful experience of my own, the a priori method simpler and more congenial. That at least did not compel them to pollute their souls by the study of plebeian usage. Having once got hold of a formula, they insisted, like Thales and his contemporaries, on making it account for the entire order of things. Thus, castes which were compact tribes, castes which had been developed out of corporations like the mediæval trade guilds, and castes which expressed the distinction between fishing and hunting, agriculture and handicrafts, were all supposed to have been evolved by interbreeding.
But the initial principle, though it could not be stretched to explain everything, nevertheless rests upon a residuum of historical fact. It happens that we can still observe its workings among a number of Dravidian tribes, which, though not yet drawn into the vortex of Brahmanism, have been in some degree affected by the example of Hindu organization. As regards inter-tribal marriages, they seem to be in a stage of development through which the Hindus themselves may have passed. A man may marry a woman of another tribe, but the offspring of such unions do not become members of either the paternal or maternal groups, but belong to a distinct endogamous aggregate, the name of which often denotes the precise cross by which it was started. Among the large tribe of Mundas we find, for instance, nine such groups—Khāngār-Mūndā, Kharia-Mūndā, Konkpat-Mūndā, Karanga-Mūndā, Mahili-Mūndā, Nāgbansi-Mūndā, Orāon-Mūndā, Sad-Mūndā, Savar-Mūndā—descended from intermarriages between Mūndā men and women of other tribes.* The Mahilis again have five sub-tribes of this kind, and themselves trace their descent to the union of a Mūndā with a Santāl woman. Illustrations of this sort might be multiplied almost indefinitely. The point to be observed is that the sub-tribes formed by inter-tribal crossing are from an early stage complete endogamous units, and that they tend continually to sever their slender connexion with the parent group, and stand forth as independent tribes. As soon as this comes to pass, and a functional or territorial name disguises their mixed descent, the process by which they have been formed is seen to resemble closely that by which the standard Indian tradition seeks to explain the appearance of other castes alongside of the classical four.
Within the limits of the regular caste system Mr. Gait mentions the Shāgirdpeshās of Bengal as the only true caste in this Province “which takes its origin from miscegenation, and which is still adding to its numbers in the same way. Amongst the members of the higher castes of Orissa who do not allow widow remarriage, and also amongst the Kāyasth immigrants from Bengal, it is a common practice to take as maid-servants and concubines women belonging to the lower clean castes, such as Chāsa and Bhandāri. The offspring of these maid-servants are known as Shāgirdpeshā. They form a regular caste of the usual type and are divided into endogamous groups with reference to the caste of the male parent. Kāyasth Shāgirdpeshās will not intermarry with Karan Shāgirdpeshās, nor Rajput Shāgirdpeshās (their number is very small) with those of Kāyasth origin, but intermarriage between the Shāgirdpeshās of Karan and of Khandāit descent sometimes takes place, just as such marriages sometimes occur between persons belonging to the castes to which they owe their origin. The caste of the mother makes no difference in the rank of the children, but those who can count several generations from their original progenitor rank higher than those in whose case the stigma of illegitimacy is more recent.
“The word Shāgirdpeshā, which is commonly pronounced Sāgarpeshā, means servant, and is applied with reference to the traditional occupation, which is domestic service. It is said that the word should properly be confined to the offspring of Bengali Kāyasths, and that the illegitimate children of Karans and other castes of Orissa should be called Krishnapakshi, or Antarpuā, or, again, Antarkaran, Antarkhandāit, etc. This distinction, however, is not observed in practice. The relationship between the legitimate children of a man of good caste and their bastard brothers and sisters is recognized, but the latter cannot eat with the former, hence they are called bhātāntar, or separated by rice. They are entitled to maintenance, but cannot inherit their father’s property so long as there are any legitimate heirs. They usually serve in their father’s house until they grow up and marry; male children are then usually given a house and a few bighas of land for their support. The Shāgirdpeshās are also sometimes known as Golām (slave)—a term which is also applied to the Sudras of Eastern Bengal, who appear in several respects to be an analogous caste. Another appellation is Kothā po (own son), as distinguished from Prajā po (tenant son), which formerly denoted a purchased slave. Their family name is usually Singh or Dās. Some of them have taken to cultivation, but they will not themselves handle the plough. They usually live in great poverty. It is said to be impossible for a Shāgirdpeshā under any circumstances to obtain admission to his father’s caste. If a man of that caste were to marry a Shāgirdpeshā woman he would be outcasted and his children would become Shāgirdpeshā. Persons of higher rank (usually outcasts) are admitted to the caste. A feast is given by the applicant for admission, and he is then formally acknowledged as a caste-fellow.
“In their social observances the Shāgirdpeshās follow the practices of the higher castes. They forbid the remarriage of widows and do not allow divorce. Polygamy is only permitted when good cause is shown, e.g., if the first wife is barren or diseased. They belong to the Vaishnava sect, worship the ordinary Hindu gods, and employ good Brāhmans. The binding portion of the marriage ceremony is the joining of the hands of bride and bridegroom by the officiating priest. Shāgirdpeshās of the first generation, being illegitimate, cannot perform their father’s srādh. They usually cremate their dead. In spite of their number (about 47,000) the caste is said to be of quite recent origin, and it is asserted that it did not exist a century and-a-half ago.” *
An older and more instructive illustration, dating possibly from before the Christian era, of the formation of a caste by crossing, is furnished by the Khas of Nepal, who are the offspring of mixed marriages between Rajputs or Brāhman immigrants and the Mongolian women of the country. “The females,” † says Hodgson, “would indeed welcome the polished Brāhmans to their embraces, but their offspring must not be stigmatized as the infamous progeny of a Brāhman and a Mlechha—must, on the contrary, be raised to eminence in the new order of things proposed to be introduced by their fathers. To this progeny also, then, the Brāhmans, in still greater defiance of their creed, communicated the rank of the second order of Hinduism; and from these two roots, mainly, sprung the now numerous, predominant, and extensively ramified tribe of the Khas, originally the name of a small clan of creedless barbarians, now the proud title of the Kshatriyas, or military order of the kingdom of Nepal. The offspring of original Khas females and of Brāhmans, with the honours and rank of the second order of Hinduism, got the patronymic titles of the first order, and hence the key to the anomalous nomenclature of so many stirpes of the military tribes of Nepal is to be sought in the nomenclature of the sacred order. It may be added, as remarkably illustrative of the lofty spirit of the Parbattias, that in spite of the yearly increasing sway of Hinduism in Nepal, and of the various attempts of the Brāhmans in high office to procure the abolition of a custom so radically opposed to the creed both parties now profess, the Khas still insist that the fruit of commerce (marriage is out of the question) between their females and males of the sacred order shall be ranked as Kshatriyas, wear the thread, and assume the patronymic title.” The Khas now call themselves Chhattris or Kshatriyas—a practice which, according to Colonel Vansittart,* dates from Sir Jang Bahadur’s visit to England in 1850. Allied to the Khas are the Ektharia and the Thākurs, both of Rajput parentage on the male side, the Thākur ranking higher because their ancestors are supposed to have been rulers of various petty States in Nepal. The Matwāla Khas, again, are the progeny of Khas men and Magar women, and the Uchai Thākurs are of the same lineage on the female side.
The Sūdra caste of Eastern Bengal, the Rājbansi Baruas of Chittagong, believed to be the offspring of Burmese fathers and Bengali mothers, the Vidurs of the Central Provinces, who claim Brāhman parentage on the male side and, though now marrying among themselves, still receive into their community the children of mixed unions between Brāhmans and women of other castes, are minor instances of the same process. The Boria caste of Assam is said by Mr. Allen† to comprise the offspring of Brāhman and Ganak widows and their descendants, and the children of Brāhmans who attained puberty before marriage, and so had to be married to men of lower caste. The name Boria is popularly derived from bari, a widow, but the members of the caste prefer to call themselves Sut or Suta, the Shastric designation of the children of a Brāhman woman by a Kshatriya, or Vaisya father. Borias are more numerous in Nowgong than in any other district of Assam, though the number of Brāhmans there is comparatively small. On pointing this out to an educated Brāhman of Nowgong, Mr. Allen received the singular explanation that “the Gosāins and Mohants of that district had put pressure upon householders to give away young Brāhman widows in marriage to men of lower castes to prevent the society from becoming demoralized.”
National castes.
(v) Castes of the national type.—Where there is neither nation nor national sentiment, it may seem paradoxical to talk about a national type of caste. There exist, however, certain groups, usually regarded as castes at the present day, which cherish traditions of bygone sovereignty and seem to preserve traces of an organization considerably more elaborate than that of an ordinary tribe. The Newārs, a mixed people of Mongoloid origin, who were the predominant race in Nepal proper until the country was conquered and annexed by the Gurkha Prithi Nārāyan in 1768, may be taken as an illustration of such a survival. The group comprises both Hindus and Buddhists. The latter are at present slightly more numerous, but the former are said to be gaining ground by more frequent conversions. The two communities are quite distinct, and each is divided into an elaborate series of castes. Thus, among the Hindu Newārs, we find at the top of the social scale the Devabhaja, who are Brāhmans and spiritual teachers; the Surjyabansi Mal, members of the old royal family; the Sreshta, consisting of ministers and other officials; and the Japu, who are cultivators. Then comes an intermediate group including, among others, the Awā, masons; the Kawmi, carpenters and sweetmeat-makers, an odd combination of trades; the Chhīpi, dyers of cloth; the Kāu, blacksmiths; and the Nāu, barbers. Lowest of all are the Pāsi, washermen; the Jugi, tailors and musicians; the Po, sweepers, burners of dead bodies, and executioners; and the Kulu, drummakers and curriers.
If the Marāthas can be described as a caste, their history and traditions certainly stamp them as a caste of the national type. They number five millions at the present census, 3,279,000 in Bombay, 1,538,000 in Hyderabad, 79,000 in Madras, 45,000 in Mysore, 93,000 in the Central Provinces and Berar, 28,000 in Central India. According to Mr. Enthoven,* the Bombay Marāthas “may be classified as a tribe with two divisions, Marātha and Marātha Kunbi, of which the former are hypergamous to the latter, but were not originally distinct. It remains to be explained that the Kunbis also consist of two divisions, Desh Kunbis numbering 1,900,000, and Konkani Kunbis, of which there are 350,000 recorded. Intermarriage between these divisions is not usual. The barrier, however, seems to be purely geographical. It may not withstand the altered conditions due to improvements in communications, and it is not apparently based on any religious prohibition of intermarriages. The fact that the Kunbis consist of two branches must, however, be borne in mind in attempting to arrive at a correct description of the tribal configuration.” The highest class of Marāthas is supposed to consist of ninety-six families, who profess to be of Rajput descent and to represent the Kshatriyas of the traditional system. They wear the sacred thread, marry their daughters before puberty, and forbid widows to marry again. But their claim to kinship with the Rajput is effectually refuted by the anthropometric data now published, and by the survival among them of kuldevaks or totems, such as the sun-flower, the kadamba tree (Nauclea Kadamba), the mango, the conch-shell, the peacock’s feather, and turmeric, which are worshipped at marriages and at the ceremony of dedicating a new house, while their close connexion with the Kunbis is attested by the fact that they take Kunbi girls as wives, though they do not give their own daughters to Kunbi men. A wealthy Kunbi, however, occasionally gains promotion to and marries into the higher grade and claims brevet rank as a Kshatriya. The fact seems to be that the ninety-six superior families represent Kunbis who came to the front during the decline of the Moghal Empire, won for themselves princedoms or estates, claimed the rank of landed gentry, and asserted their dignity by refusing their daughters to their less distinguished brethren.
Castes formed by migration.
(vi) Castes formed by migration.—If members of a caste leave their original habitat and settle permanently in another part of India, the tendency is for them to be separated from the parent group and to develop into a distinct caste. The stages of the process are readily traced. In the first instance it is assumed that people who go and live in foreign parts must of necessity eat forbidden food, worship alien gods, and enter into relations with strange women. Consequently, when they wish to take wives from among their own people, they find that their social status has been lowered, and that they have to pay for the privilege of marrying within the parent group. This luxury grows more and more expensive, and in course of time the emigrants marry only among themselves and thus become a sub-caste usually distinguished by a territorial name, such as Jaunpuria, Tirhutia, Bārendra, and the like. Mr. Gait has pointed out that “the prolonged residence of persons of Bihar castes in Bengal generally results in their being placed under a ban as regards marriage,” * and I had observed some years earlier that up-country barbers who settle in Bengal are called khotta and practically form a separate sub-caste, as Bengali barbers will not intermarry with them, while they are regarded as impure by the barbers of Upper India and Bihar by reason of their having taken up their residence in Bengal. If the process of differentiation is carried a step further (as indeed usually happened before the potent influence of railways had made itself felt), and the settlers assume a distinctive caste-name, all traces of their original affinities disappear and there remains only a dim tradition of their migration “from the West,” the quarter whence, in Bengal at any rate, promotion is believed to come. Owing to this loss of identity the number of instances in which we can point with certainty to the formation of castes by migration is comparatively small. Mr. Russell, writing of the Central Provinces, tells us how a native gentleman said to him, in speaking of his people, that “when a few families of Khedāwal Brāhmans from Gujarat first settled in Damoh, they had the greatest difficulty in arranging their marriages. They could not marry with their caste-fellows in Gujarat, because their sons and daughters could not ’establish themselves,’ that is, could not prove their identity as Khedāwāl Brāhmans; but since the railway has been opened, intermarriages take place freely with other Khedāwāls in Gujarat and Benares.”* So the geographical isolation of Chhattisgarh, the country of the “thirty-six forts” of the Haihaibansi dynasty of Ratanpur, has led to the social isolation of the inhabitants. “The Chhattisgarhi Brāhmans,” says Mr. Russell, “form a class apart, and up-country Brāhmans will have nothing to do with them.” The contempt in which the people of this tract are held by their neighbours, finds expression in the following depreciatory verses:
Wah hai Chhattisgarhi desh, Jahān Gond hai naresh, Niche bursi upar khāt, Lagā hai chongi kā thāt, Pahile jutā pichhe bāt, Tab āwe Chhattisgarhi hāt.
Which may be rendered thus:—
“This is Chhattisgarh, where the Gond is king of the jungle, Under his bed is a fire, for he cannot pay for a blanket; Nor for a hookah indeed,—a leaf-pipe holds his tobacco. Kick him soundly first and then he will do what you tell him.” †
The verses reflect the intolerant and domineering attitude of the Indo-Aryan towards the Dravidian, of the high-caste man towards the low, that has been characteristic of Indian society from the earliest times down to the present day.
A good illustration of the formation of a caste by migration is to be found in the traditions of the Nambudri or Namputiri Brāhmans of Malabar. These Brāhmans claim to have come to the west coast from various sacred localities in Kathiawar and the northern Deccan ; Mr. Fawcett describes them as “the truest Aryans in Southern India ;”* and their complexion and features seem to lend some support to the tradition which assigns to them a foreign origin. Whatever their original stock may have been, they are now an entirely separate caste differing from the Brāhmans of other parts of India by their systematic practice of polygamy ; by their rejection of infant marriage ; by their restriction of marriage to the eldest son, the other brothers entering into polyandrous relations with Nāyar women ; and by the curious custom of ceremonial fishing which forms part of their marriage ritual. Another instance of the same process is furnished by the Rārhi Brāhmans of Bengal. The current legend is that early in the eleventh century A.D., Adisura or Adisvara, Raja of Bengal, finding the Brāhmans then settled in his dominions too ignorant to perform for him certain Vedic ceremonies, applied to the Raja of Kanauj for priests conversant with the sacred ritual of the Aryans. In answer to his request there were sent to him five Brāhmans of Kanauj, one of them a son of the Raja, who brought with them their wives, their sacred fire, and their sacrificial implements. It is said that Adisura was at first disposed to treat them with scanty respect, but he was soon compelled to acknowledge his mistake and to make terms with people who had a monopoly o the magical powers associated with the correct performance of ancient ritual. He then made over to them five populous villages, the number of which was subsequently increased to fifty-six. The tradition seems to chronicle an early brahmottar grant, the first perhaps of the long series of similar transactions that has played so important a part in the history of land tenures, in the development of caste influence and custom, and in promoting the spread of orthodox Hinduism throughout Bengal. Adisura did what the Rajas of outlying and unorthodox tracts of country (such as Bengal was in the eleventh century) have constantly done since and are doing still. A local chief, far removed from the great centres of Brahmanical lore, somehow becomes aware of his ceremonial shortcomings. In many cases, as indeed is narrated of Adisura himself, a wandering priest brings home to him that his outlandish ritual is not up to the orthodox standard. He sends for Brāhmans, gives them grants of land near his own residence, and proceeds at their dictation to reform his ways on the model of the devout kings whom Brahmanical literature holds up as the ideal for a Raja to follow. The Brāhmans find for him a pedigree of respectable antiquity and provide him with a family legend, and in course of time, by dint of money and diplomacy, he succeeds in getting himself recognized as a member of the local Rajput community. But that does not mean that the real Rajputs will acknowledge his pretensions; nor will Brāhmans who have attached themselves to his fortunes retain their status among the community from which they have broken off. It will be said of them, as is said of the Brāhman immigrants into Bengal, that they have married local women, eaten forbidden food, adopted strange customs, and forgotten the endless details of the elaborate ritual which they set forth to teach. As priests in partibus infidelium they will be regarded with suspicion by the Brāhmans of their original stock; they will have to pay high for brides from among their own people, and eventually will be cut off altogether from the jus connubii. When that stage has been reached they will have become to all intents and purposes a separate caste retaining the generic name of Brāhman, but forming a new species and presenting a distinctive type. And this great change will have been brought about by the simple fact of their abandoning the habitat of their original community.
Occasionally it may happen that social promotion, rather than degradation, results from a change of residence. In Chanda, a remote district of the Central Provinces, a number of persons returned themselves as Barwaiks and the designation, being unknown in the Census office, was referred to the district officer for explanation. It was stated in reply that the Barwaiks were a clan of Rajputs from Orissa who had come to Nāgpur in the train of the Bhonsla Rajas and had taken military service under them. Now in Chutia Nāgpur the Baraiks or Chik-Baraiks are a sub-caste of the Pans—the helot weavers and basketmakers who perform a variety of servile functions for the organized Dravidian tribes and used to live in a kind of Ghetto in the villages of the Kandhs (Khonds) for whom they purveyed children destined for human sacrifice and, when they had failed to steal other people’s children, sold their own for this ghastly purpose. Mr. Russell observes that “though it is possible that the coincidence may be accidental, still there seems good reason to fear that it is from these humble beginnings that the Barwaik sept of Rajputs in Chanda must trace its extraction. And it is clear that before the days of railways and the half-anna post an imposture of this sort must have been practically impossible of detection.”* The conjecture seems a plausible one, and the fact that Baraik is a title actually in use among the Jadubansi Rajputs may have helped the Pans to establish their fictitious rank.
Castes formed by changes of customs.
(vii) Castes formed by changes of custom.—The formation of new castes as a consequence of the neglect of established usage or the adoption of new ceremonial practices or secular occupations has been a familiar incident of the caste system from the earliest times. We are told in Manu † how men of the three twice-born castes, who have not received the sacrament of initiation at the proper time, or who follow forbidden occupations, become Vrātyas or outcasts, intercourse with whom is punished with a double fine, and whose descendants are graded as distinct castes. Living as a Vrātya is a condition involving of itself exclusion from the original caste, and a Brāhman who performs sacrifices for such persons has to do penance. The idea of such changes of status is inherent in the system, and illustrations of its application are plentiful. Sometimes it figures in the traditions of a caste under the form of a claim to a more distinguished origin than is admitted by current opinion. The Skanda Purana, for example, recounts an episode in Parasu Rāma’s raid upon the Kshatriyas, the object of which is to show that the Kāyasths are by birth Kshatriyas of full blood, who by reason of their observing the ceremonies of the Sudras are called Vrātya or incomplete Kshatriyas. The Bābhans or Bhuinhārs of the United Provinces and Bihar are supposed, according to some legends, to be Brāhmans who lost status by taking to agriculture, and the Mongoloid Kochh of Northern Bengal describe themselves as Rājbansis, or as Vrātya or Bhanga (broken) Kshatriyas—a designation which enables them to pose as an outlying branch of that exalted community who fled to these remote districts before the wrath of Parasu Rāma, and there allowed their characteristic observances to fall into disuse. At the present day the most potent influence in bringing about elevations or depressions of social status which may result ultimately in the formation of new castes is the practice of widow marriage. With the advance of orthodox ideas that may plausibly be ascribed to the extension of railways and the diffusion of primary education it dawns upon some members of a particular caste that the custom of marrying widows is highly reprehensible, and with the assistance of their Brāhmans they set to work to discourage it. The first step is to abstain from intermarriage with people who practise the forbidden thing, and thus to form a sub-caste which adopts a high-sounding name derived from some famous locality like Ajodhya or Kanauj, or describes itself as Biyāhut or Behutā, “the married ones,” by way of emphasizing the orthodox character of their matrimonial arrangements. Thus the Awadhia or Ayodhya Kurmis of Bihar and the Kanaujia Kurmis of the United Provinces pride themselves on prohibiting the remarriage of widows and are endeavouring to establish a shadowy title to be recognized as some variety of Kshatriya, in pursuance of which, with singular ignorance of the humble origin of the great Marātha houses, they claim kinship with Sivaji, Sindhia and the Bhonsla family of Nāgpur. In Bihar they have succeeded in attaining a higher rank than ordinary Kurmis. Brāhmans take water from their hands; the funeral ceremony is performed on the twelfth day after death, according to the custom of the higher castes; and kachchi food prepared by them is eaten by Kahārs, Bhāts, and other castes who would refuse to accept food of this kind from Sudras. They have abandoned domestic service, and the wealthier members of the group exchange presents with the higher castes and are invited by them to ceremonial functions. But although the Awadhias have achieved complete practical separation from the main body of Kurmis no one accepts them as Kshatriyas or Rajputs, nor are they recognized by Hindu public opinion as forming a distinct caste. In the Punjab Sir Denzil Ibbetson* wrote in 1881 that the Gaurwa Rajputs of Gurgaon and Delhi, though retaining the title of Rajput in deference to the strength of caste-feeling and because the change in their customs was then too recent for the name to have fallen into disuse, yet had, for all purposes of equality, communion, or intermarriage, ceased to be Rajputs since they took to karewa or widow marriage. And the distinction between the Jāts and Rajputs, both sprung from a common Indo-Aryan stock, is marked by the fact that the former practise and the latter abstain from a usage which more than any other is regarded as a crucial test of relative social position.
In allusion to this fact one of the rhyming proverbs of the Punjab makes a Jāt father say—“Come, my daughter, and be married; if this husband dies there are plenty more.” The same test applies in the Kāngra Hills, the most exclusive Hindu portion of the Punjab, where Musalman domination was never fully established, and the Brāhman and Kshatriya occupy positions most nearly resembling those assigned to them by Manu. Here the line between the Thakkar and Rāthi castes, both belonging to the lower classes of Hill Rajputs, is said to consist in the fact that Rāthis do and Thakkars do not ordinarily practise widow marriage.
In Southern India movements of the same sort may be observed. Among the begging castes which form nearly one per cent. of the population of the Tamil country in Madras, the Pandārams rank highest in virtue of their abstention from meat and alcohol and more especially of their prohibition of widow marriage. The Pancharamkatti division of the Idaiyan shepherd caste allow widow marriage but connect it with the peculiar neck ornament which their women wear, and say that “Krishna used to place a similar ornament round the necks of the Idaiyan widows of whom he was enamoured, to transform them from widows into married women to whom pleasure was not forbidden.”* The story seems to be an ex post facto apology for the practice. The Jātāpu again, a branch of the Kandh (Kondh) tribe which has developed into a separate caste, are beginning to discourage widow marriage by way of emphasizing the distinction between themselves and their less civilized brethren.† In Baroda, according to Mr. Dalāl,‡ widow marriage is allowed by some degraded sub-castes of Brāhmans, Tapodhan, Vyās Sārasvat, Rajgor, Bhojak, Tragala and Koligor, which are virtually distinct castes, and also by the Kāthis, Marāthas, Rajputs, Tāghers, and Vadhels. “The higher families, among castes allowing remarriage of widows, do not, as a rule, have recourse to it, as such a marriage is considered undignified for grown-up women. It is this sense of honour and a desire to pass for superior people which has put a stop to widow remarriage among an influential section of the Lewa Kunbis and Sonis.”
An account is given in the chapter on marriage and caste of what may be called the internal structure of tribes and caste in India—the various endogamous, exogamous, and hypergamous divisions which restrict and regulate matrimony and form the minor wheels of the vast and intricate machinery by which Hindu society is controlled. From the point of view of general ethnology considerable interest attaches to one particular kind of division, to those exogamous groups which are based upon totems.
Totemism.
The existence of totemism in India on a large scale has been brought to notice only in recent years: the enquiries instituted in connexion with the census have added materially to our knowledge of the subject; and special attention is being given to it in the ethnographic survey now being conducted in all British provinces and the more important Native States. No apology therefore is needed for mentioning it at length here, since it throws an important sidelight on the development of castes from tribes. At the bottom of the social system, as understood by the average Hindu, we find in the Dravidian region of India a large body of tribes and castes each of which is broken up into a number of totemistic septs. Each sept bears the name of an animal, a tree, a plant, or of some material object, natural or artificial, which the members of that sept are prohibited from killing, eating, cutting, burning, carrying, using, etc. Well-defined groups of this type are found among the Dravidian Santāls and Orāons, both of whom still retain their original language, worship non-Aryan gods, and have a fairly compact tribal organization. The following are specimens selected from among the seventy-three Orāon and the ninety-one Santāl septs:—
| Orāon. Name of sept. | Orāon. Totem. | Santāl. Name of sept. | Santāl. Totem. |
|---|---|---|---|
| Tirki. | Young mice. | Ergo. | Rat. |
| Ekka. | Tortoise. | Murmu. | Nilgāi. |
| Kispotta. | Pig’s entrails. | Hānsdā. | Wild goose. |
| Lakra. | Hyena. | Mārudi. | A kind of grass. |
| Bāgh. | Tiger. | Besrā. | Hawk. |
| Kujrar. | Oil from Kujrar tree. | Hemron. | Betel palm. |
| Gede. | Duck. | Saren | The constellation Pleiades. |
| Khoepa. | Wild dog. | Sankh. | Conch-shell. |
| Minji. | Eel. | Guā. | Areca nut. |
| Chirra. | Squirrel. | Kārā. | Buffalo. |
The Hos of Singhbhum and the Mūndās of the Chutia Nāgpur plateau have also exogamous septs of the same type as the Orāons and Santāls, with similar rules as to the totem being taboo to the members of the group. The lists given in The Tribes and Castes of Bengal contain the names of 323 Mūndā septs and 46 Ho septs. Six of the latter are found also among the Santāls. The other Ho septs appear to be mostly of the local or communal type, such as are in use among the Kandhs,* but this is not quite certain, and the point needs looking into by some one well acquainted with the Ho dialect, who would probably find little difficulty in identifying the names, as the tribe is known to be in the habit of giving to places descriptive names having reference to their natural characteristics.
In Chutia Nāgpur.
Nearly all the Mūndā sept names are of the totem type, and the characteristic taboos appear to be recognized. The Tarwār or Talwār sept, for example, may not touch a sword, the Udbāru may not use the oil of a particular tree, the Sindur may not use vermilion, the Baghela may not kill or eat a quail, and, strangest of all, rice is taboo to the Dhān sept, the members of which, though rice is grown all round them, must supply its place with gondli or millet. It is difficult not to be sceptical as to the rigid observance of this last prohibition.
A step higher in the social scale, according to Hindu estimation, the Bhūmij of Mānbhum mark an early stage in the course of development by which a non-Aryan tribe transforms itself into a full-blown caste, claiming a definite rank in the Brahmanical system. With the exception of a few residents of outlying villages bordering on the Mūndā country of the Chutia Nāgpur plateau, the Bhūmij have lost their original language (Mūndāri), and now speak only Bengali. They worship Hindu gods in addition to the fetishistic deities more or less common to them and other Dravidians, but the tendency is to keep the latter rather in the background and to relegate the less formidable among them to the women and children to be worshipped in a hole-and-corner kind of way, with the assistance of a tribal hedge-priest (Lāyā), who is supposed to be specially acquainted with their ways. Some of the leading men of the tribe, who call themselves Bhuinhārs, and hold large landed tenures on terms of police service, have set up as Rajputs, and keep a low class of Brāhmans as their family priests. They have, as a rule, borrowed the Rajput class titles, but cannot conform with the Rajput rules of intermarriage, and marry within a narrow circle of pseudo-Rajputs like themselves. The rest of the tribe, numbering at the census of 1901, 370,239, are divided into a number of exogamous groups, of which the following are examples. It is curious to observe in a tribe still in a state of transition, that one of the Brahmanical gotras, Sāndilya, has been borrowed from the higher castes, and in the process of borrowing has been transformed from a Vedic saint into a bird:—
| BHŪMIJ. | |
|---|---|
| Name of sept. | Totem. |
| Sālrisi. | Sāl fish. |
| Hānsda. | Wild goose. |
| Leng. | Mushroom. |
| Sāndilya. | A bird. |
| Hemron. | Betel palm. |
| Tumarung. | Pumpkin. |
| Nāg. | Snake. |
At a further stage in the same process of evolution, and on a slightly superior social level, we find the Mahilis, Korās, and Kurmis, all of whom claim to be members of the Hindu community. They have totemistic exogamous sections, of which the following are fairly representative:—
| MAHILI. | KORĀ. | ||
|---|---|---|---|
| Name of section. | Totem. | Name of section. | Totem. |
| Dungri. | Dumur fig. | Kasyab. | Tortoise. |
| Turu. | Turu grass. | Saulā. | Sāl fish. |
| Kānti. | Ear of any animal. | Kāsibak. | Heron. |
| Hānsda. | Wild goose. | Hānsda. | Wild goose. |
| Murmu. | Nīlgāi. | Butku. | Pig. |
| Sāmpu. | Bull. |
| KURMI. | |
|---|---|
| Name of section. | Totem. |
| Kesariā. | Kesar grass. |
| Tarār. | Buffalo. |
| Dumuriā. | Dumur fig. |
| Chonchmukruār. | Spider. |
| Hastowār. | Tortoise. |
| Jālbanuār. | Net. |
| Sankhowār. | Shell ornaments. |
| Bāghbanuār. | Tiger. |
| Katiār. | Silk cloth. |
Of these three castes the Mahilis appear to have broken off most recently from the tribe. They still worship some of the Santāl gods in addition to the standard Hindu deities; they will eat food cooked by a Santāl; their caste organization is supervised, like that of the Santāls, by an official bearing the title of Parganāit; they permit the marriage of adults and tolerate sexual intercourse before marriage within the limits of the caste; and they have not yet attained to the dignity of employing Brāhmans for ceremonial purposes. If I may hazard a conjecture on so obscure a question, I should be inclined to class them as Santāls who took to the degraded occupation of basket-making, and thus lost the jus connubii within the tribe. In the case of the Korās there is no clue to warrant their affiliation to any particular tribe, but their traditions say that they came from the Chutia Nāgpur plateau, while their name suggests a Dravidian origin, and it seems possible that they may be an offshoot of the Mũndās, who somehow sank from the status of independent cultivators to their present position of earth-cutting and tank-digging labourers. They allow adult marriage, their standard of feminine chastity is low, and they have not yet fitted themselves out with Brāhmans. In the customary rules of inheritance which their panchāyat or caste council administers, it is curious to find the usage known in the Punjab as chundavand, by which the sons, however few, of one wife take a share equal to that of the sons, however many, of another. The Kurmis may perhaps be a Hinduized branch of the Santāls. The latter, who are more particular about food, or rather about whom they eat with, than is commonly supposed, will eat cooked rice with the Kurmis, and according to one tradition regard them as elder brothers of their own. However this may be, the totemism of the Kurmis of Western Bengal stamps them as of Dravidian descent, and clearly distinguishes them from the Kurmis of Bihar and the United Provinces. They show signs of a leaning towards orthodox Hinduism, and employ Brāhmans for the worship of Hindu gods, but not in the propitiation of rural and family deities or in their marriage ceremonies.
In Orissa.
One more instance of totemism in Bengal deserves special notice here, as it shows the usage maintaining its ground among people of far higher social standing than any of the castes already mentioned. The Kumhārs of Orissa take rank immediately below the Karan or writer caste, and thus have only two or three large castes above them. They are divided into two endogamous sub-castes—Jagannāthi or Oriya Kumhārs, who work standing and make large earthen pots, and Khattya Kumhārs, who turn the wheel sitting and make small earthen pots, cups, toys, etc. The latter are immigrants from Upper India, whose number is comparatively insignificant. For matrimonial purposes the Jagannāthi Kumhārs are subdivided into the following exogamous sections:—
| Name of section. | Totem. |
|---|---|
| Kaundinya. | Tiger. |
| Sarpa. | Snake. |
| Neul. | Weasel. |
| Goru. | Cow. |
| Mudir. | Frog. |
| Bhadbhadria, | Sparrow. |
| Kurmā. | Tortoise. |
The members of each section express their respect for the animal whose name the section bears by refraining from killing or injuring it, and by bowing when they meet it. The entire caste also abstain from eating, and even go so far as to worship the sāl fish, because the rings on its scales resemble the wheel which is the symbol of the potter’s art. The Khattya Kumhārs have only one section (Kasyapa), and thus, like the Rājbansis of Rangpur, are really endogamous in spite of themselves. The reason, no doubt, is that there are too few of them in Orissa to fit up a proper exogamous system, and they content themselves with the pretence of one. Both sub-castes appear to be conscious that the names of their sections are open to misconception, and explain that they are really the names of certain saints who, being present at Daksha’s horse sacrifice, transformed themselves into animals to escape the wrath of Siva, whom Daksha, like Peleus in the Greek myth, had neglected to invite.* It may well be that we owe the preservation of these interesting totemistic groups to the ingenuity of the person who devised this respectable means of accounting for a series of names so likely to compromise the reputation of the caste. In the case of the Khattya Kumhārs, the fact that their single section bears the name of Kasyapa, while they venerate the tortoise (kachhap), and tell an odd story by way of apology for the practice, may perhaps lend weight to the conjecture, in itself a fairly plausible one, that many of the lower castes in Bengal who are beginning to set up as pure Hindus have taken advantage of the resemblance in sound between Kachhap and Kasyap (chh and s both become sh in colloquial Bengali) to convert a totemistic title into an eponymous one, and have gone on to borrow such other Brahmanical gotras as seemed to them desirable. If, for example, we analyze the matrimonial arrangements of the Bhars of Mānbhum, many of whom are the hereditary personal servants of the pseudo-Rajput Raja of Pachete, we find the foregoing conjecture borne out by the fact that two out of the seven sections which they recognize are called after the peacock and the bel fruit, while the rest are eponymous. But this is an exceptionally clear case of survival, and I fear it is hardly possible to simplify the diagnosis of non-Aryan castes by laying down a general rule, that all castes with a section bearing the name Kasyapa who have not demonstrably borrowed that appellation from the Brāhmans, are probably offshoots from some non-Aryan tribe.
In Bombay.
In the Bombay Presidency the Kātkaris of the Konkan will not kill a red-faced monkey,* the Vaidus, or herbalists of Poona will not kill a rabbit, and the Vadars whose name is derived from the Vad (Ficus Indica), will not fell the Indian fig tree. The totemistic character of the septs which regulate marriage is, however, most pronounced in the Kanara district which borders on the Dravidian tract of the South. The rice-growing caste of Hālvakki Vakkaḷ in Kanara have a number of exogamous septs or bali (lit. a creeper) which include the tortoise, the sambhar, the monkey, the hog-deer, two sorts of fish, saffron, the acacia and several other trees, and the axe used for felling them. As we find them now, these groups are plainly totemistic. Thus the members of the screw-pine bali will neither cut the tree nor pluck its flowers, and those of the Bargal bali will not kill or eat the barga or mouse-deer. The followers of the Shirin bah, named after the shirbal tree (Acacia speciosa), will not sit in the shade of the tree, and refrain from injuring it in any way. But in Kanara, as in Orissa, there is a tendency to disguise or get rid of these compromising designations as the people who own them rise in the social scale. The Halepaikỵ once free-booters and now peaceful tappers of toddy trees, are divided into two endogamous groups, one dwelling on the coast and taking its name (Tengina) from the cocoanut tree, and the other living in the hills and calling itself Bainu after the sago-palm. Each of these again contains a number of exogamous balis. The Tengina have the wolf, the pig, the porcupine, the root of the pepper plant, turmeric, and the river; to which the Bainu add the the snake, the sāmbhar deer, and gold. The members of the Nāgchampa group will not wear the flower of that name in their hair, nor will the Kadave bali kill a sāmbhar. Two of the balis are called after the low castes Mahār and Hole, and it is curious to find that the other groups, though they will take girls from these balis, will not give them their own daughters to wife. Among the Halepāiks, unlike most of the Kanara castes, the bali descends through the female line, that is to say, the children belong to the bali of the mother, not of the father. Similar groups are found among the Suppalig (musicians), the Āger (salt workers and makers of palm-leaf umbrellas), the Ahīr (cowherds), and the Mukur (labourers and makers of shell-lime). Several of these have the elephant for a totem and may not wear ornaments of ivory.
In Central India.
Among the Bhīls of the Sātpura hills, who may be taken to represent the furthest extension westward of the Dravidian type, Major Luard* has discovered forty-one septs, all of which are exogamous. Where two distinct septs have the same totem intermarriage is prohibited. All the septs revere and refrain from injuring or using their totems, and make a formal obeisance when meeting or passing them, while the women veil their faces. Among the totems are moths (ava), snakes, tigers, bamboos, pīpal and other trees, and a kind of creeper called gaola on which the members may not tread, and if they do so accidentally must apologize by making a salaam. The Maoli sept have as their totem a sort of basket (kiliya) for carrying grain which they are forbidden to use. The basket resembles in shape the shrine of the goddess of a certain hill where women may not worship. The Mori or peacock sept may not knowingly tread on the tracks of a peacock, and if a woman sees a peacock she must veil her face or look away. The cult of the totem consists in seeking for the footprint of a peacock in the jungle and making a salaam to it. The ground is then made smooth round the footprint, a svāstika is inscribed in the dust, and offerings of grain are deposited on a piece of red cloth. The Sanyar sept worship the cat, but consider it unlucky for their totem to enter their houses and usually keep a dog tied up at the door to frighten it away. The Khangār caste of Bundelkhand, which is cited by Major Luard as an illustration of the conversion of a tribe into a caste, have among their totems horses, iguanas, snakes, cows, elephants, alligators, rice, turmeric, various trees and shrubs, and bricks. The members of the Īnt or brick sept may not use bricks in their houses and their domestic architecture is restricted to wattle and mud. The report on the census of Central India also contains a curious instance of the apparent degradation of a caste into a tribe accompanied by the adoption of totems. The Sondhiās or Sundhiās of Malwa are said to be descended from the survivors of a Rajput army who were defeated by Shah Jahan and were ashamed to return to their homes. They therefore stayed in Malwa, married Sondhiā women, borrowed some of the Sondhiā totems and the Sondhiā gods, and in course of time allowed widows to marry again. Ten of the twenty-four septs into which the tribe is divided still cherish traditions of their Rajput origin and, while taking wives from the other septs, refuse to give their daughters in return.
For the Central Provinces Mr. Russell* gives a long list of totems found among sixteen castes and tribes, including not only the primitive Gonds, Korkus, and Orāons, and the leather-working Chamārs, but also the pastoral Ahīrs, the respectable carpenter caste (Barhai) and the Dhīmars, from all of whom Brāhmans can take water, while the last named are commonly employed by them as personal servants. The list comprises elephants, lions, tigers, bears, wolves, jackals, buffaloes, goats, monkeys, peacocks, parrots, crocodiles, lizards, tortoises, porcupines, scorpions, snakes, also salt, rice, Indian corn, pumpkins, mangoes, cucumbers, lotus leaves, vermilion and a variety of trees. All of these are regarded with reverence, and members of the sept abstain from killing, using or naming them.
In Madras.
In Madras the Boya shikāri tribe of the Deccan is divided into 101 totemistic septs, among them chimalu, ants; eddulu, bulls; jenneru, sweet-scented oleander; jerrabutula, centipedes; yenumalu, buffaloes; and kusa, grass. The Jātapu, the civilized division of the Kandhs or Khonds, have among their totems koaloka, arrows; kondacorri, hill sheep; kutraki, wild goats; and vinka, white ants. The large agricultural caste of Kapu, numbering nearly three millions, have among their exogamous sections the cock (kodi), the sheep (mekala), and a shrub known as tangedu (Cassia auriculata). Of the 102 sections of the trading Komatis six are totemistic, the totems including the tamarind, the tulsi (Ocymum Sanctum), and the betel vine. The weaving Kurnis count among their totems saffron, gold, cummin, gram, pepper, buffaloes, and certain trees.*
In Assam.
In Assam the Gāros have monkeys, horses, bears, mice, lizards, frogs, crows, pumpkins, and a number of trees among their totems; the Kachāris recognize as totems the tree snail, the muga insect, the sesamum plant, the kumra or giant gourd, and the tiger. Members of the tiger sept have to throw away their earthenware utensils by way of atonement when a tiger is killed. The louse and the buffalo are the only animal totems on record among the Khāsi; the Kuki have the dog; the Lalung eggs, fish, and pumpkins; the Mikir totems appear to be mainly vegetable. Our information, however, on totemism in Assam is extremely scanty, and the subject requires further investigation.†
In Burma.
For Burma the facts, so far as they go, are thus stated by Mr. Lowis:—
“The question of endogamy naturally leads to that of totemism. Sir George Scott says in the Upper Burma Gazetteer: ‘All the Indo-Chinese races have a predilection for totemistic birth stories. Some claim to have sprung from eggs, some from dogs, some from reptiles.’ The Wās, like a tribe in North-West America cited by Mr. Andrew Lang in his Custom and Myth, state that their primæval ancestors were tadpoles. The Palaungs trace their beginnings back to a Nāga princess who laid three eggs, out of the first of which their early ancestor was hatched. An egg-laying Nāga princess figures in the early legendary history of the Mons or Talaings and points to an affinity between the Palaungs and the Talaings which the most recent linguistic research has done much to strengthen. Up to the present time all attempts to ascertain the original of the Kachin family names have failed. The totem of the Kachins should, if anything, be a pumpkin, for legend has it that the whole race is descended from a being who was made out of a pumpkin. So far as I can discover, however, their belief in this singular genesis does not deter Kachins from eating the vegetable to which they owe their origin. They do not even appear to be precluded from gathering it under certain circumstances or at a particular period of the year, as is the case with some of the Western Australian tribes.” The Southern Chins, on the other hand, are forbidden to kill or eat the King-Crow which hatched “the original Chin egg.” The bird is regarded in the light of a parent, but, as it is not used as a crest by the Chins, Mr. Houghton is of opinion that it cannot be looked upon as, properly speaking, a totem. The rising sun of the Red Karens is something of the nature of a totemistic badge. Mr. Smeaton refers to it as follows in his Loyal Karens of Burma:—
“Every Red Karen has a rising sun—the crest of his nobility—tattooed on his back. In challenging to combat he does not slap his left folded arm with his right palm, as the rest of the Karens and the Burmans do, but, coiling his right arm round his left side, strikes the tattoo on his back. This action is supposed by him to rouse the magic power of the symbol.”
Sir George Scott, however, seems to detect no totemistic inwardness in this tattoo mark, for he sums up the matter under consideration in the following words:—
“Totemism also shows itself in the prescribed form of names for Shān and Kachin children and in the changing or concealing of personal names, but, so far as is yet known, there is no tribe which habitually takes its family name or has crests and badges taken from some natural object, plant, or animal, though the limiting of marriages between the inhabitants of certain villages only practised both by tribes of Karens and Kachins is no doubt the outgrowth of this totem idea.”
Enough has been said to show that totemistic exogamy prevails in India on a fairly large scale, that it is still in active operation, and that it presents features which deserve further investigation in their bearing on the problems of general ethnology. On these grounds I venture to add a few comments on the striking explanation of the origin of totemism which was put forward by Sir J. G. Frazer in the Fortnightly Review in 1899.* The subject is one of special interest in India because the Indian evidence seems not only to point to conclusions different from those arrived at by Sir J. G. Frazer on the basis of the Australian data published by Messrs. Spencer and Gillen,* but to suggest a new canon for determining the historical value of ethnographic evidence in general.
Sir J.G. Frazer’s theory of totemism.
" A totem," says Sir J. G. Frazer, " is a class of natural phenomena or material object—most commonly a species of animals or plants—between which and himself the savage believes that a certain intimate relation exists. The exact nature of the relation is not easy to ascertain ; various explanations of it have been suggested, but none has as yet won general acceptance. Whatever it may be, it generally leads the savage to abstain from killing or eating his totem, if his totem happens to be a species of animals or plants. Further, the group of persons who are knit to any particular totem by this mysterious tie commonly bear the name of the totem, believe themselves to be of one blood, and strictly refuse to sanction the marriage or cohabitation of members of the group with each other. This prohibition to marry within the group is now generally called by the name of exogamy. Thus totemism has commonly been treated as a primitive system both of religion and of society. As a system of religion it embraces the mystic union of the savage with his totem ; as a system of society it comprises the relations in which men and women of the same totem stand to each other and to the members of other totemic groups. And corresponding to these two sides of the system are two rough and ready tests or canons of totemism : first, the rule that a man may not kill or eat his totem animal or plant ; and second, the rule that he may not marry or cohabit with a woman of the same totem. Whether the two sides—the religious and social—have always co-existed or are essentially independent, is a question which has been variously answered. Some writers—for example, Sir John Lubbock and Mr. Herbert Spencer—have held that totemism began as a system of society only, and that the superstitious regard for the totem developed later, through a simple process of misunderstanding. Others, including J. F. McLennan and Robertson Smith, were of opinion that the religious reverence for the totem is original, and must, at least, have preceded the introduction of exogamy."
The system of totems prevailing in Central Australia is so far parallel to that known in India that it includes, not only animals and plants, but also a number of objects, animate and inanimate. Thus while the Australians have " totems of the wind, the sun, the evening star, fire, water, cloud, and so on," we find among our Dravidians in India the month of June, Wednesday in every week, the moon, the rainbow, and the constellation Pleiades figuring as totems among a number of names which include pretty well the entire flora and fauna of the country where the tribe is settled. But while among the Australians the religious aspect of the totem is relatively more prominent than the social, in India the position is reversed ; the social side of the system is very much alive while the religious side has fallen into disuse. It is the religious side on which Sir J. G. Frazer lays stress, and he explains totemism as “ primarily an organized and co-operative system of magic designed to secure for the members of the community, on the one hand, a plentiful supply of all the commodities of which they stand in need, and, on the other hand, immunity from all the perils and dangers to which man is exposed in his struggle with nature.” In other words, totemism is a primitive Commissariat and General Providence Department which at a later stage took over the business of regulating marriage. The evidence for this proposition is derived from the magical ceremonies called intichiuma in which the members of each totem solemnly mimic the animals and plants after which they are called, and eat a small portion of them with the object of ensuring a plentiful supply of the species. Thus the men of the totem called after the Witchetty grub, a succulent caterpillar of some kind which is esteemed a great luxury, paint their bodies in imitation of the grub, crawl through a structure of boughs supposed to represent its chrysalis, chant a song inviting the insect to go and lay eggs, and butt each other in the stomach with the remark “ You have eaten much food.” The Emu men dress themselves up to resemble Emus and imitate the movements and aimless gazing about of the birds ; the Kangaroo men and the men of the Nakea flower totem go through similar mummeries. An admirable collection of the totemistic symbols of the Arunta, together with photographs of the ritual observed in the invocation of the totems themselves, may be seen in the Ethnological department of the Museum at Melbourne.
Now in the first place the doubt occurs to one whether small and moribund tribes, such as the Australians, can fairly be taken to be typical of primitive man. If they could, then man would be primitive still, and we should none of us have got to the point of vexing our souls about the origin of anything. The one distinctive feature of the Australian natives is their incapacity for any sort of progressive evolution.
Surely an atrophied or, it may be, degenerative man of that type is not the sort of ancestor we want to discover; for it is difficult to see what we can learn from him. In Europe, on the other hand, primitive man, so far as we can judge from the traces he has left behind, seems to have been an animal of an entirely different type. He had, indeed, his weaknesses—does not his vates sacer, Mr. Andrew Lang, impute to him a diet of oysters and foes—but he fought a good fight with his environment and, as events show, he came out a winner. It seems then that the quest of primitive man ready made and only waiting to be observed and analyzed may be nothing better than a tempting short cut leading to delusion, and that what we must look to is not so much primitive man as primitive usage regarded in its bearing on evolution.
Totemism and Exogamy.
It is from this point of view that I wish to put in a plea for the consideration of the Indian data. Primitive usages may, I would suggest, be divided, as Mr. Bagehot divided political institutions, into the effective and the ineffective, in other words, into those which affect evolution and those which do not. In the case of totemism we can distinguish these two pretty clearly. The magical ritual of the Arunta tribe obviously belongs to the ineffective class. No one outside the Arunta—and even among them one would think there must be augurs—supposes that by performing the most elaborate parody of the demeanour of certain animals a man can really cause them to increase and multiply. In India, on the other hand, our totemistic people have got rid of all such antics, if, indeed, they ever practised them, and retain only the unquestionably effective factor in the system, the rule that a man may not marry a woman of his own totem. They have, it is true, also the rule that people may not eat, injure or make use of their totems, but this prohibition is relatively weak, and in some cases the totems are articles such as rice and salt, which the members of the totem-kin could hardly do without.
Given then a state of things such as this, that tribes which are in no way moribund or degenerate, but on the contrary extremely full of life, retain the effective part of an archaic usage along with the traces of its ineffective parts, may we not reasonably conclude that this effective part, which has stood the wear and tear of ages and contributed to the evolution of the tribe, furnishes the clue to the real origin of the usage itself? Assume this to be so and totemism at once wheels into line and takes the place, which it appears clearly to occupy in India, of a form of exogamy. The particular form presents no great difficulty. Primitive men are like children : they are constantly saying to themselves " Let’s pretend, " and a favourite and wide-spread form of the game is to pretend to be animals. Only they play it in earnest, and very grim earnest it sometimes is, as any one will discover who has to administer a district where people believe that men can transform themselves into animals at will, or can be so transformed by the agency of witchcraft.
It will be asked, what then is the origin of exogamy ? Here again I think the Indian evidence suggests an answer. Just as the special phenomenon of totemism may be explained by reference to the general law of exogamy, so exogamy itself may be traced to the still more general law of natural selection. Nor need we strain the law. We know that there is a tendency in individuals or groups of individuals to vary their habits ; and that useful variations tend to be preserved and ultimately transmitted. Now suppose that in a primitive community, such as the Nāga khel or the Kandh gochi, the men happened to vary in the direction of taking their wives from some other community and that this infusion of fresh blood proved advantageous to the group. The original instinct would then be stimulated by heredity, and the element of sexual selection would, in course of time, come into play. For an exogamous group would have a larger choice of women than an endogamous one, and would thus get finer women, who again, in the course of the primitive struggle for wives, would be appropriated by the strongest and most warlike men. The exogamous groups so strengthened would tend, as time went on, to " eat up, " in the expressive Zulu phrase, their endogamous neighbours, or at any rate to deprive them of the pick of their marriageable girls ; and the custom of exogamy would spread, partly by imitation, and partly by the extinction of the groups which did not practise it.
The fact that we cannot say how people came to vary in this particular fashion is not necessarily fatal to the hypothesis put forward. In the case of animals other than man we do not call in question the doctrine of natural selection because we cannot trace the precise cause which gave rise to some beneficial variation. It is enough that variations do occur, and that the beneficial ones tend to be transmitted. If, however, an attempt must be made to pierce the veil which shuts off from our view the ages of pre-historic evolution, it does not seem unreasonable to suppose that here and there some half-accidental circumstance, such as the transmission of a physical defect or an hereditary disease, may have given primitive man a sort of warning, and thus have induced the particular kind of variation which his circumstances required. Conquest again may have produced the same effect by bringing about a beneficial mixture of stocks, though it is a little difficult to see, as Mr. Lang pointed out long ago, why the possession of foreign women should have disinclined people to marry the women of their own group. At the same time it is conceivable that the impulse may have been set going by some tribe from which all its marriageable women had been raided and which was thus driven by necessity to start raiding on its own account. I have elsewhere given instances, drawn from the Kandhs and Nāgas, which lend themselves to this view; but I am not sure that we need travel beyond the tendency to accidental variation which appears in all living organisms and may be assumed to have shaped the development of primitive man.
Classification of castes.
In a country where the accident of birth determines irrevocably the whole course of a man’s social and domestic relations, and he must throughout life eat, drink, dress, marry, and give in marriage in accordance with the usages of the community into which he was born, one is tempted at first sight to assume that the one thing that he may be expected to know with certainty, and to disclose without much reluctance, is the name of the caste, tribe, or race to which he belongs. As a matter of fact no column in the Census schedule displays a more bewildering variety of entries, or gives so much trouble to the enumerating and testing staff and to the central offices which compile the results. If the person enumerated gives the name of a well-known tribe, such as Bhīl or Santāl, or of a standard caste like Brāhman or Kāyasth, all is well. But he may belong to an obscure caste from the other end of India; he may give the name of a religious sect, of a sub-caste, of an exogamous sept or section, of a hypergamous group; he may mention some titular designation which sounds finer than the name of his caste; he may describe himself by his occupation or by the province or tract of country from which he comes. These various alternatives, which are far from exhausting the possibilities of the situation, undergo a series of transformations at the hands of the more or less illiterate enumerator who writes them down in his own vernacular, and of the abstractor in the central office who transliterates them into English. Then begins a laborious and most difficult process of sorting, referencing, cross-referencing, and corresponding with local authorities, which ultimately results in the compilation of a table showing the distribution of the inhabitants of India by Caste, Tribe, Race, or Nationality. The arrangement of this table is alphabetical and it consists of two parts. The first is a general list of all the groups returned, with their distribution by religion; while the second shows the distribution by provinces and states of all groups with an aggregate strength of 10,000. An analysis of the table shows that it includes 2,378 main castes and tribes and 43 races or nationalities. With the latter we are not concerned here; as to the former, the question at once arises—on what principle should they be arranged? An alphabetical system is useful for reference, and essential for the purely statistical purposes of a census table. But it does not help us in the least towards presenting an intelligible picture of the social grouping of that large proportion of the people of India which is organized, admittedly or tacitly, on the basis of caste. In this matter a new departure was taken at the Census of 1901. The classification followed in 1891 was then described as “based on considerations partly ethnological, partly historical, and partly, again, functional. The second predominate, for instance, in the first caste group, and the last throughout the middle of the return; but wherever practicable, as it is in the latter portion of the scheme, ethnological distinctions have been maintained. Then, again, it must be mentioned that the functional grouping is based less on the occupation that prevails in each case in the present day than on that which is traditional with it, or which gave rise to its differentiation from the rest of the community.” The main heads of the scheme embodying the application of these principles are given at page 188 of the Report on the Census of India for 1891, and its detailed application is shown in Imperial Table XVII.
Judged by its results this scheme is open to criticism in several respects. It accords neither with native tradition and practice, nor with any theory of caste that has ever been propounded by students of the subject. In different parts it proceeds on different principles, with the result that on the one hand it separates groups which are really allied, and on the other includes in the same category groups of widely different origin and status. It is in fact a patch-work classification in which occupation predominates, varied here and there by considerations of caste, history, tradition, ethnical affinity, and geographical position. Illustrations of these defects might be multiplied almost indefinitely, but it is sufficient to mention that the Dravidian Khandāits of Orissa are classed with Rajputs and Bābhans, Jāts, Marāthas, and Nāyars; that Brāhman priests, Mirāsi musicians, and Bahurupia buffoons fall within the same general category; that the Mongoloid Koch, Kachāri, Thāru, and Mech are widely separated; and that more than half of the Musalmans, including the converted aborigines of Eastern Bengal and Assam, are shown as “Musalman Foreign Races,” the rest being merged among a number of occupational groups purporting to be endogamous.
Method adopted in Census of 1901.
In organizing the Census of 1901 I suggested to my colleagues that an attempt should be made to arrange the various groups that had to be dealt with on some system which would command general acceptance, at any rate, within the limits of the province to which it was applied. I did not expect that the same system would suit all provinces or even all divisions of the same province; and I was quite prepared to find the preparation of a combined table for the whole of India a task of insuperable difficulty. But I was confident that the provincial results would throw light upon a variety of social movements which at present escape notice; that they would add greatly to the interest of the reports; and that they would provide a sound statistical ground-work for the ethnographic survey of India which is now in progress.
The principle suggested as a basis was that of classification by social precedence as recognized by native public opinion at the present day, and manifesting itself in the facts that particular castes are supposed to be the modern representatives of one or other of the castes of the theoretical Hindu system; that Brāhmans will take water from certain castes; that Brāhmans of high standing will serve particular castes; that certain castes though not served by the best Brāhmans, have nevertheless got Brāhmans of their own, whose rank varies according to circumstances; that certain castes are not served by Brāhmans at all, but have priests of their own; that the status of certain castes has been raised by their taking to infant-marriage or abandoning the remarriage of widows; that the status of some castes has been lowered by their living in a particular locality; that the status of others has been modified by their pursuing some occupation in a special or peculiar way; that some can claim the services of the village barber, the village palanquin-bearer, the village midwife, etc., while others cannot; that some castes may not enter the courtyards of certain temples; that some castes are subject to special taboos, such as that they must not use the village well, or may draw water only with their own vessels, that they must live outside the village or in a separate quarter, that they must leave the road on the approach of a high-caste man, or must call out to give warning of their approach. In the case of the Animistic tribes it was mentioned that the prevalence of totemism and the degree of adoption of Hindu usage would serve as ready tests. All Superintendents, except three who were either defeated by the complexity of the facts or were afraid of hurting people’s feelings, readily grasped the main idea of the scheme, and their patient industry, supplemented by the intelligent assistance readily given by the highest native authorities, has added very greatly to our knowledge of an obscure and intricate subject.
Its practical working.
The best evidence of the general success of the experiment, and incidentally of the remarkable vitality of caste at the present day, is to be found in the great number of petitions and memorials to which it gave rise, the bulk of which were submitted in English and emanated from the educated classes who are sometimes alleged to be anxious to free themselves from the trammels of the caste system. If the principle on which the classification was based had not appealed to the usages and traditions of the great mass of Hindus, it is inconceivable that so many people should have taken much trouble and incurred substantial expenditure with the object of securing its application in a particular way. Of these memorials the most elaborate was that received from the Khatris of the Punjab and the United Provinces who felt themselves aggrieved by the Superintendent of Census in the latter Province having provisionally classified them as Vaisyas, whereas in the specimen table circulated by me they had been placed in the same group as the Rajputs. A meeting of protest was held at Bareilly, and a great array of authorities was marshalled to prove that the Khatris are lineally descended from the Kshatriyas of Hindu mythology, much as if the modern Greeks were to claim direct descent from Achilles and were to cite the Catalogue of the Ships in the second book of the Iliad in support of their pretensions. In passing orders on their memorial I pointed out that they were mistaken in supposing that this was the first census in which any attempt had been made to classify castes on a definite principle, or that the selection of social precedence as a basis was an entirely new departure. As a matter of fact the scheme of classification adopted in 1891 purported to arrange the groups more or less in accordance with the position generally assigned to each in the social scale, as has been suggested by Sir Denzil Ibbetson in his Report on the Punjab Census of 1881. The result, in the case of the Khatris, was to include them as number 13 in “Group XV—Traders” immediately after the Aroras of the Punjab, ten places lower than the Agarwāls, and several places below the Kāndus and Kasarwānis of the United Provinces and the Subarnabaniks of Bengal. The Rajputs, on the other hand, ranked first in the entire scheme as number 1 of “Group I—Military and Dominant.” In the Bengal Census Report of 1891 the Rajputs were placed among “the patrician class,” while the Khatris were grouped with the Baniyas between the Baidyas and Kāyasthas in a group described as “the Vaisyas Proper or Plebeian Middle Class.” It was obviously improbable that the Khatris desired this classification to be maintained, and the evidence laid before me not only brought out the conspicuous part played by them in the authentic history of the Punjab in modern times, but seemed to make it clear that in British India at any rate they are generally believed to be the modern representatives of the Kshatriyas of Hindu tradition. For census purposes the fact that most people do hold this belief was sufficient in itself, and it would have been irrelevant to enquire into the grounds upon which the opinion was based. Superintendents of census were accordingly instructed to include the Khatris under the heading Kshatriya in their classification of castes. The decision gave general satisfaction and served to illustrate the practical working of the principle that the sole test of social precedence prescribed was Indian public opinion, and that this test was to be applied with due consideration for the susceptibilities of the persons concerned. The other memorials were disposed of by the Provincial Superintendents on similar lines.
Its general results.
As no stereotyped scheme of classification was drawn up, but every Province was left to adopt its own system in consultation with its own experts and representative men, it was clearly impossible to draw up any general scheme for the whole of India. One might as well have tried to construct a table of social precedence for Europe, which should bring together on the same list Spanish grandees, Swiss hotel-keepers, Turkish Pashas, and Stock Exchange millionaires, and should indicate the precise degree of relative distinction attaching to each. The problem, in fact, is essentially a local one. Every man has honour in his own country, and India is no more one country than Europe—indeed very much less. The Provincial schemes of classification are summarized in the Census Report of India, 1901, vol. i, p. 560 et seq. Although they cannot be reduced to common terms, they exhibit points of resemblance and difference which deserve some further examination. The first thing to observe is the predominance throughout India of the influence of the traditional system of four original castes. In every scheme of grouping the Brāhman heads the list. Then come the castes whom popular opinion accepts as the modern representatives of the Kshatriyas, and these are followed by the mercantile groups, supposed to be akin to the Vaisyas. When we leave the higher circles of the twice-born, the difficulty of finding a uniform basis of classification becomes apparent. The ancient designation Sūdra finds no great favour in modern times, and we can point to no group that is generally recognized as representing it. The term is used in Bombay, Madras, and Bengal, to denote a considerable number of castes of moderate respectability, the higher of whom are considered “clean” Sūdras, while the precise status of the lower is a question which lends itself to endless controversy. At this stage of the grouping a sharp distinction may be noticed between Upper India and Bombay and Madras. In Rajputana, the Punjab, the United Provinces, the Central Provinces, Bengal, and Assam the grade next below twice-born rank is occupied by a number of castes from whose hands Brāhmans and members of the higher castes will take water and certain kinds of sweetmeats. Below these again is a rather indeterminate group from whom water is taken by some of the higher castes but not by others. Further down, where the test of water no longer applies, the status of a caste depends on the nature of its occupation and its habits in respect of diet. There are castes whose touch defiles the twice-born, but who refrain from the crowning enormity of eating beef ; while below these again, in the social system of Upper India, are people like Chamārs and Doms who eat beef and various sorts of miscellaneous vermin. In Western and Southern India the idea that the social status of a caste depends on whether Brāhmans will take water and sweetmeats from its members is unknown, for the higher castes will, as a rule, take water only from persons of their own caste and sub-caste. In Madras especially the idea of ceremonial pollution by the proximity of a member of an unclean caste has been developed with much elaboration. Thus the table of social precedence attached to the Cochin report * shows that while a Nāyar can pollute a man of a higher caste only by touching him, people of the Kammālan group, including masons, blacksmiths, carpenters, and workers in leather, pollute at a distance of twenty-four feet, toddy-drawers (Iluvan or Tiyan) at thirty-six feet, Pulāyan or Cheruman cultivators at forty-eight feet, while in the case of the Paraiyan (Pariahs) who eat beef, the range of pollution is stated to be no less than sixty-four feet. Where these fantastic notions prevail and the authority of the Brāhman is unquestioned, it follows as a necessary consequence that the unhappy people who diffuse an atmosphere of impurity wherever they go are forbidden to enter the high caste quarter of the village, and are compelled either to leave the road when they see a Brāhman coming or to announce their own approach by a special cry like the lepers of Europe in the Middle Ages. Such is the logic of intolerance in parts of Southern India.
Social precedence of Hindus in Bengal.
The subject of classification is examined fully in some of the Provincial Census Reports, to which the reader is referred for further particulars. No attempt was made to grade every caste. Large classes were formed, and the various groups included in these were arranged in alphabetical order, so as to escape the necessity of settling the more delicate questions of precedence. As an illustration of the method of procedure I may refer to the table of precedence for Bengal Proper, which was compiled by me some years ago and has been adopted by Mr. Gait for the purpose of the Bengal Census Report † after careful examination by local committees of Indian gentlemen appointed for the purpose.
The entire Hindu population of this tract, numbering twenty millions, has been divided into seven classes. The first class is reserved for the Brāhmans, of whom there are more than a million, forming six per cent. of the Hindus of Bengal. As every one knows, there are Brāhmans and Brāhmans, of status varying from the Rārhi, who claim to have been imported by Adisura from Kanauj, to the Barna Brāhmans who serve the lower castes and from whose hands pure Brāhmans will not take water. No attempt has been made to deal with these multifarious distinctions in the table. It would be a thankless task to try to determine the precise degree of social merit or demerit that attaches to the Pirāli Brāhmans, who are supposed to have been forced, some four centuries ago, to smell or, as some say, to eat the beefsteaks that had been cooked for the renegade Brāhman Pīr Ali, the dewān of the Muhammadan ruler of Jessore; to the Vyasokta Brāhmans who serve the Chāsi Kaibartta caste and rank so low that even their own clients will not touch food in their houses; to the Agradāni who preside at funeral ceremonies and take the offerings of the dead; to the Achārji fortune-teller, palmist, and maker of horoscopes; and to the Bhāt Brāhman, a tawdry parody of the bard and genealogist of heroic times, whose rapacity and shamelessness are proverbial.
Next in order, at the top of the second class, come the Rajputs, who disown any connexion with Bengal, and base their claims to precedence on their supposed descent from the pure Rajputs of the distant Indo-Aryan tract. Their number (113,405) must include a large number of families belonging to local castes who acquired land and assumed the title of Rajput on the strength of their territorial position. Then follow the Baidyas, by tradition physicians, and the writer caste of Kāyasth. The former pose as the modern representatives of the Ambastha of Manu and assert their superiority to the Kāyasthas on the ground that the latter have been pronounced by the High Court of Calcutta to be Sūdras, a Kāyasth judge concurring, and that their funeral usages confirm this finding; that the Sanskrit College, when first opened, admitted only Brāhmans and Baidyas as students; that the Kāyasths were originally the domestic servants of the two higher castes, and when poor take service still; and that native social usage concedes higher rank to the Baidyas at certain ceremonies to which members of the respectable castes are invited. The Kāyasths, on the other hand, claim to be Kshatriyas, who took to clerical work; deny the identity of the Baidyas with the Ambasthas; and describe them as a local caste, unknown in the great centres of Hinduism, who were Sūdras till about a century ago, when they took to wearing the sacred thread, and bribed the Brāhmans to acquiesce in their pretensions. The alphabetical arrangement observed in the table leaves the question an open one.
The third class, numbering three millions, comprises the functional castes originally known as Navasākha, the nine “branches” or “arrows,” and other clean Sūdras, from whose hands the higher castes take water, and who are served by high-class Brāhmans. Confectioners, perfume vendors, betel growers, pressers and sellers of oil, gardeners, potters, and barbers figure in this group, the constitution of which appears to have been largely determined by consideration of practical convenience. The preparation of a Hindu meal is a very elaborate performance, involving lengthy ablutions and a variety of ritualistic observances which cannot be performed on a journey, and it is essential to the comfort of the orthodox traveller that he should be able to procure sweetmeats of various kinds without being troubled by misgivings as to the ceremonial cleanliness of the people from whom he buys them. In matters of food and drink caste rules are wisely elastic. It has, I believe, been held that neither ice nor soda-water count as water for the purpose of conveying pollution; there are special exemptions in favour of biscuits and patent medicines, for the last of which the Bengali has an insatiable appetite; and in an outlying district where the only palanquin-bearers available were Dravidian Bhuiyas, I have known them to be given brevet rank as a water-giving (jalācharaniya) caste in order that the twice-born traveller might be able to get a drink without quitting his palanquin.
The fourth class includes only two castes—the Chāsi Kaibartta and the Goāla—from whom water is taken by the high castes, but whose Brāhmans are held to be degraded. About the former group I wrote in 1891: “It seems likely, as time goes on, that this sub-caste will rise in social estimation, and will altogether sink the Kaibartta, so that eventually it is possible that they may succeed in securing a place with the Navasākha.” The forecast has to this extent been fulfilled that at the recent Census the Chāsi Kaibartta called themselves Māhishya, the name of the offspring of a legendary cross between Kshatriyas and Vaisya, and posed as a separate caste. In Nadia, according to Mr. Gait, “the new idea gained such ground that many Chāsi Kaibarttas in domestic service under other castes threw up their work, saying it was beneath their dignity. Finding, however, that no other means of livelihood were available they were soon fain to return and beg their employers’ forgiveness.” * The higher castes, moreover, expressed their disapproval of a movement which upset their domestic arrangements by a concerted refusal to take water from the hands of a Chāsi. Notwithstanding these discouragements I have little doubt that by the next Census the Māhishya will have succeeded in establishing their claim.† Their case is of interest for the light that it throws on the evolution of a caste.
The fifth class contains a rather miscellaneous assortment of castes, including the Baishtam, the Sunri, and the Sunbarnabanik, from whom the higher castes do not usually take water. Their precedence is also defined by the fact that although the village barber will shave them he will not cut their toe-nails nor will he take part in their marriage ceremonies. Here again quaint problems of status arise. The Baishtams are a group formed by the conversion to Vaishnavism of members of many different castes, who have embraced the tenets of different Vaishnava sects. In theory inter-marriage between these sects is prohibited, but if a man of one sect wishes to marry a woman of another, he has only to convert her by a simple ritual to his own sect and the obstacles to their union are removed. The social standing of the caste is necessarily low, as it is recruited from among all classes of society, and large numbers of prostitutes and people who have got into trouble in consequence of sexual irregularities are found among its ranks.
Within the caste, however, many of them retain their old social distinctions, and a Baishtam of Kayasth origin would not ordinarily take water from the hands of one whose ancestors were Chandâls. Outsiders also recognize these differences and take water from Baishtams who are known to have belonged to one of the clean castes. Where the origin of a Baishtam is unknown, water which he has touched can only be used forwashing.
The Subarnabaniks are a mercantile caste peculiar to Bengal Proper, who claim to be the modern representatives of the ancient Vaisya. In spite of their wealth and influence, their high-bred appearance, and the notorious beauty of the women of the caste, their claim to this distinguished ancestry has failed to obtain general recognition. They are excluded from the ranks of the Navasâkha, or nine clean Sûdra castes, and none but Vaidik Brâhmans will take water from their hands. To account for the comparatively low status assigned to them, the Subarnabaniks cite a variety of traditions, some of which, however unsupported by historical evidence, deserve to be briefly mentioned here as illustrations of the kind of stories which tend to grow up wherever the business talents and practical ability of a particular community have advanced it in the eyes of the world conspicuously beyond its rank in the theoretical order of castes. These people, for example, say that their ancestors came to Bengal from Oudh during the reign of Adisura, who was struck by their financial ability and conferred on them the title of Subarnabanik, or trader in gold, as a mark of his favour. They then wore the Brahmanical thread, studied the Vedas and were generally recognized as Vaisyas of high rank. The stories of their degradation all centre round the name of Ballal Sen, who was Raja of Eastern Bengal in 1070 A.D. His intrigue with a beautiful Patni girl is said to have been ridiculed on the stage by some young Subarnabaniks, while the entire body refused to be present at the penance whereby the king affected to purify himself from the sin of intercourse with a maiden of low caste. Another cause of offence is said to have been the refusal of a leading Subarnabanik to lend Ballal large sums of money to carry on war with Manipur. Authorities differ concerning the method by which the Raja obtained his revenge. Some say that in the course of the penance already referred to, a number of small golden calves had been distributed to the attendant Brâhmans. One of these Brâhmans was suborned by Ballal Sen to fill the hollow inside of a calf with lac-dye, and to take the figure to a Subarnabanik for sale. In testing the gold the Subarnabanik let out the lac-dye, which was at once pronounced to be blood. Having thus fastened upon the caste the inexpiable guilt of killing a cow, Ballal Sen publicly declared them and their Brahmans to be degraded, deprived them of the right to wear the sacred thread, and threatened with similar degradation any one who should eat or associate with them.
In default of independent testimony to the accuracy of this tradition we can hardly accept it as a narrative of historical events. It is no doubt conceivable that a despotic monarch might order the social degradation of a particular class of his subjects provided that it were not too numerous or too influential ; and it is generally believed that Ballal Sen did effect some changes of this kind in the relative status of certain families of Brahmans. Notwithstanding this, the story of the depression of an entire caste from a very high to a comparatively low rank in the social system makes a large demand on our belief, and inclines one to suspect that it may have been evolved in recent times to account for the position actually occupied by the caste being lower than that to which their riches and ability would entitle them to lay claim. From this point of view, the conjecture that the Subarnabaniks are Hindustani Baniyas, who lost status by residing in Bengal and marrying Bengali women, seems to deserve some consideration.
The sixth class includes a long list of castes, numbering nearly eight millions, who abstain from eating beef, pork, and fowls, but from whom the higher castes will not take water. They are served by degraded Brahmans ; the regular barbers refuse to shave them ; and some of them have special barbers of their own. Most of them, however, can get their clothes washed by the village washerman. The typical members of the group, according to the census of 1911, are the Bagdi, (1,041,892), Dravidian cultivators and labourers, the Jaliya or fishing Kaibartta (375,936), the Namasudra or Chandal (2,087,162), the Pod (536,591), fishermen and cultivators, and the Rajbansi-Koch (2,049,454), nearly all of whom are small cultivators.
The seventh class represents the lowest grade of the Bengal system, castes who eat all manner of unclean food, whose touch pollutes, whom no Brahman, however degraded, will serve, and for whom neither barber nor washerman will work. It comprises the scavenging Doms and Haris, the leather-working Chamārs and Mochis, and the Bauris who eat rats and revere the dog as their totem because, as they told Colonel Dalton, it is the right thing to have a sacred animal of some kind, and dogs are useful while alive and not very nice to eat when dead.*
Social precendence among Muhammadans.
Islam, whether regarded as a religious system or as a theory of things, is in every respect the antithesis of Hinduism. Its ideal is strenuous action rather than hypnotic contemplation ; it allots to man a single life and bids him live it and make the best of it ; its practical spirit knows nothing of a series of lives, of transmigration, of karma, of the weariness of existence which weighs upon the Indian mind. For the dream of absorption into an impersonal Weltgeist it substitutes a very personal Paradise made up of joys such as all Orientals understand. On its social side the religion of Muhammad is equally opposed to the Hindu scheme of a hierarchy of castes, an elaborate stratification of society based upon subtle distinctions of food, drink, dress, marriage, and ceremonial usage. In the sight of God and of His Prophet all followers of Islam are equal. In India, however, caste is in the air ; its contagion has spread even to the Muhammadans ; and we find its evolution proceeding on characteristically Hindu lines. In both communities foreign descent forms the highest claim to social distinction ; in both, promotion cometh from the West. As the twice-born Aryan is to the mass of Hindus, so is the Muhammadan of alleged Arab, Persian, Afghān or Moghal origin to the rank and file of his co-religionists. And just as in the traditional Hindu system men of the higher groups could marry women of the lower while the converse process was vigorously condemned, so within the higher ranks of the Muhammadans a Saiyad will marry a Shekh’s daughter but will not give his daughter in return, and inter-marriage between the upper circle of soi-disant foreigners and the main body of Indian Muhammadans is generally reprobated, except in parts of the country where the aristocratic element is small and must arrange its marriages as best it can. Even there, however, it is only under the stress of great poverty that a member of the Ashrāf or “noble” class will give his daughter to one of the Ajlāf or “low people,” as converts of indigenous origin are called in Bengal. Of course, the limits of the various groups are not defined as sharply as they are with the Hindus. The well-known proverb, which occurs in various forms in different parts of Northern India—" Last year I was a Jolāha; now I am a Shekh; next year if prices rise, I shall become a Saiyad “—marks the difference, though analogous changes of status are not unknown among Hindus, and, as Mr. Gait observes, " promotion is not so rapid in reality as it is in the proverb.” But speaking generally, it may be said that the social cadre of the higher ranks of Muhammadans is based on hypergamy with a tendency in the direction of endogamy, while the lower functional groups are strictly endogamous, and are organized on the model of regular castes with councils and officers who enforce the observance of caste rules by the time-honoured sanction of boycotting.
According to Mr. Gait the Bengal Muhammadans " recognize two main social divisions: (1) Ashrāf or Sharīf and (2) Ajlāf, which in Bengali has been corrupted to Atrāp. The first, which means ’ noble ’ or ’ persons of high extraction,’ includes all undoubted descendants of foreigners and converts from the higher castes of Hindus.* All other Muhammadans, including the functional groups to be presently mentioned, and all converts of lower rank are collectively known by the contemptuous term Ajlāf, ’ wretches ’ or ’ mean people ‘; they are also called Kamīna or Itar, ’ base ’ or ’ Razīl,’ a corruption of Rizāl, ’ worthless.’ This category includes the various classes of converts who are known as Nāo Muslim in Bihar and Nasya in North Bengal, but who in East Bengal, where their numbers are greatest, have usually succeeded in establishing their claim to be called Shekh. It also includes various functional groups such as that of the Jolāha or weaver, Dhunia or cotton-carder, Kulu or oil-presser, Kunjra or vegetable-seller, Hajjām or barber, Darzi or tailor, and the like. Of these divisions, the Ashrāf takes no count. To him all alike are Ajlāf. This distinction, which is primarily one between the Muhammadans of foreign birth and those of local origin, corresponds very closely to the Hindu division of the community into Dwijas or castes of twice-born rank, comprising the various classes of the Aryan invaders, and the Sūdras or aborigines whom they subdued. Like the higher Hindu castes, the Ashrāf consider it degrading to accept menial service or to handle the plough. The traditional occupation of the Saiyads is the priesthood, while the Moghals and Pathāns correspond to the Kshatriyas of the Hindu regime.
“In some places a third class, called Arzal or ’ lowest of all,’ is added. It consists of the very lowest castes, such as the Halalkhor, Lalbegi, Abdal, and Bediya, with whom no other Muhammadan would associate, and who are forbidden to enter the mosque or to use the public burial ground.”* I have described the Bengal scheme of social precedence at some length, because of the curious beliefs and traditions which it embodies and by reason of the testimony which it bears to the remarkable stability of the caste instinct in spite of the many modern influences which seem at first sight to be sapping its foundations. The scheme deals, moreover, with conditions with which I am to some extent familiar, and it represents an advanced stage of a process which appears to me to be going on with varying degrees of rapidity in all parts of India where Hindu sentiment and tradition are the dominant factors of social development. The extension of railways which indirectly diffuses Brahmanical influence and at the same time weakens trivial caste restrictions; the tendency to revive the authority of the Hindu scriptures and to find in them the solution of modern problems; the advance of vernacular education which increases the demands for popular versions of, and extracts from, these writings, and the spread of English education which encourages sceptical tendencies;—these are among the causes which, in my opinion, are tending on the one hand to bring about among the population regarded as a whole a more rigid observance of the essential incidents of caste, especially of those connected with marriage, and on the other to introduce greater laxity in respect of the minor injunctions which are concerned with food and drink.
Case of Baluchistan.
On the outskirts of the Empire there are two regions where Hindu standards of social precedence and Hindu notions of caste are neither recognized nor known. In Baluchistan, until less than a generation ago, Hindus were tolerated only as a useful class of menials who carried on the petty trade which the fighting races deemed below their dignity. They adopted the device, not unknown in mediæval Europe, of putting themselves under the protection of their more powerful neighbours, and Mr. Hughes-Buller tells us that even now a Hindu when asked to what caste he belongs " will often describe himself by the name of the tribal group to whom he holds himself attached.
Their position generally was extremely degraded, and may best be gauged by the fact that among Baloch, Brāhūi, and Afghāns there was an unwritten rule that in the course of raids and counter raids women, children and Hindus were to be spared." * Among the non-Hindu people of Baluchistan the question of social precedence is intricate and obscure and its details must be studied in Mr. Hughes-Buller’s excellent report. Of the three chief races the Afghāns rank highest in virtue of their former sovereignty; then comes the Baloch who also once bore rule, and last the Brāhūi who were in power at the time of the British occupation. The relative position of the two latter tribes is indicated by various proverbs, by the attempts of the Brāhūi to trace their descent to the Baloch, and by the fact that “no self-respecting Baloch will give his daughter to a Brāhūi.” The test of marriage, however, appears not to apply to the Afghān, who regards the question as a matter of business and will sell his daughter to any man who can pay her price. Below these races come the Jāts, a term which seems to be loosely used to denote all sorts of menial classes, including professional musicians (Langahs), blacksmiths (Loris), and leather-workers (Mochis). But even here there is no hard and fast prohibition of inter-marriage, and both Baloch and Brāhūi will take wives from among the Jāts. Within the circle of each tribe a condition of theoretical equality appears to prevail, tempered by personal considerations arising from capacity to lead, religious sanctity, age and kinship with a ruling family.
In Burma caste is so little known that the Burmese language possesses no word for it, while one of the difficulties of conducting the Census of the numerous Indian immigrants is the impossibility of making the average Burman enumerator understand the meaning of the Indian term zāt or jāt. Differences of religion he can grasp in a vague sort of way, he has a notion of what is meant by race, but caste remains to him an insoluble mystery—a thing with which his democratic spirit, regardless of social distinctions, has no sympathy whatever. Mr. Lowis assures us that there are not and never have been any true castes in Burma, though a class of landed proprietors in Minbu known as the Thugaungs appear to be endogamous, and thirty-six professional groups with hereditary occupations are said to have existed among the Chins.†
Distribution of Social groups.
No attempt can be made here to analyse and explain the distribution of the 2,300 castes and tribes which have been enumerated in the Census. The mere bulk of the undertaking would in any case ensure its failure; the mass of detail would be tedious and bewildering; while the causes which have determined the settlement and diffusion of particular groups belong more properly to local history and are, in any case, largely a matter of conjecture. In order, however, to give some idea of the facts and to provide a statistical basis for further researches, I have selected thirty-six of the principal tribes and castes and have shown their distribution by Provinces and States in the series of small maps annexed to this volume. The maps are constructed on the principle of graphic representation recommended by M. Bertillon. The strength of the caste to which a map relates is depicted in each province by a rectangle, of which the base indicates the total population of the province, while the height denotes the proportion which the numbers of the caste bear to the total population; thus the area of the rectangle gives the actual strength of the caste. Most of the names have also been entered in the large map showing the physical types.
Diffused groups.
A glance at the maps will show that some castes are diffused over the whole of India, while others are localized in particular provinces or tracts of country. The typical instance of a widely diffused caste is furnished by the Brāhmans, who number nearly fifteen millions, and represent a proportion of the total population ranging from ten per cent. in the United Provinces, Central India, and Rajputana to three per cent. in Madras, the Central Provinces and Bengal, and two per cent. in Assam and Chutia Nagpur. The distribution accords fairly well with the history and traditions of the caste. They are strongest in their original centre, numbering nearly five millions in the United Provinces, and weakest in the outlying tracts, peopled mainly by non-Aryan races, which their influence has even now only imperfectly reached. There can, however, be little doubt that many of the Brāhmans of the more remote tracts have been manufactured on the spot by the simple process of conferring the title of Brāhman on the tribal priests of the local deities. The so-called Barna Brāhmans who serve the lower castes of Bengal probably obtained sacerdotal rank in this fashion. That the priestly caste is not of altogether unmixed descent is attested by the numerous legends of Rajas, who, having sworn a rash oath to feed a stated number of Brāhmans, usually a lakh and a quarter, found the supply run out and were obliged to make them up for the occasion out of any materials that were at hand. A similar conclusion may perhaps be drawn from the well-known distich—
Kariā Brāhman, gorā Chamār, Inke sāth na utariye pār. If the Brāhman be black, Or the Chamār be fair ; At the ford of the river Let the wise man beware !
As with the Brāhmans so in the chief functional groups the tendency is towards wide diffusion, and their racial composition probably varies materially in different provinces. Owing to differences of language the maps fail to bring out the complete facts in relation to the whole of India. Thus the leather workers (Chamār and Muchi) of Upper India, numbering over eleven millions and forming twelve per cent. of the population of the United Provinces, correspond with the Chakkiliyan (486,884) and Madiga (755,316) of Madras, but the map does not include these. The large pastoral group (Ahīr and Goāla) numbering nearly ten millions in Upper India and forming eight per cent. of the population of the United Provinces appears in the South of India under the names Gola (855,221) and Idaiyan (694,829), neither of which are taken account of in the map. The same may be said of the potters (Kumhār) and the oilmen (Teli, and Tili) both widely diffused functional groups whose distribution is imperfectly exhibited in the maps. In each province such groups form, of course, distinct castes which have probably been evolved independently.
Localized groups.
Of the localized groups a large number are admittedly tribes. The Bhīl, Gond, Koli, and Santāl come within this category and are still outside the Hindu social system. The Doms, Dosādhs, Gūjars, Jāts, Kaibarttas, Nāmasūdras (Chandāls) Pods, Nayars, Pallis, Paraiyans (Pariahs), and Rājbansi-Koch represent tribes which have been transformed into castes at a comparatively recent date and retain some traces of the tribal stage of development.
The three Muhammadan maps—Jolāha, Pathān and Saiyad—are of interest for the light that they throw upon the spread of Islam. The Jolāha weavers number nearly three millions in all, and the solid blocks which they form in the Punjab (695,216), in the United Provinces (923,042), and in Bengal (1,242,049), seem to mark the area in which the lower classes of the community were converted en masse to a faith which seemed to hold out to them the prospect of a social status unattainable under the rigid system of caste. The Pathān map denotes a different order of phenomena, and may be taken to indicate roughly the degree of diffusion of the main body of the foreign Musalmān element and their descendants. It shows us the sturdy, pugnacious, enterprising Pathān pushing forward from the frontier and establishing himself among the feebler folk of India wherever there was fighting to be done or money to be made. The Saiyad map on the other hand seems to give some clue to the distribution of the upper classes of the immigrant Musalmāns.
Footnotes
- [Census Report, Bengal, 1901, vol i. p. 486.]
- Census Report of the Central Provinces, 1901, vol. i. p. 178. † [Vol. i. p. 443 et seq.]
- Census Report of the United Provinces, 1901, vol. i. p. 214.
- [See Sarat Chandra Roy, The Mundas and their Country, 1912, p. 400 et seq.]
- Notes on Nepal, 1896, p. 89. † Census Report of Assam, 1901, vol. i., p. 124, et seq.
- Census Report of Bombay, 1901, vol. i., p. 183, et seq.
[* Ethnographic Survey, Bombay, No. 134, 1909, pp. 1, 12 ; Census Report, Bombay, 1911, vol. i, p. 269.]
[† Bombay Census Report, 1911, vol. i, p. 263.]
[‡ Ethnographic Survey, Bombay, No. 12, 1904, p. 2 et seq.]
- [Census Report, Bengal, 1901, vol. i., p. 355 note.]
- [Census Report, Central Provinces, 1901, vol. i., p. 157.] † [Laws, ii. 39, x. 20, xi. 63.]
- [Census Report, 1881, para. 446.]
- Muir, Original Sanskrit Texts, IV. p. 372.
| Name of section. | Totem. |
|---|---|
| Kaundinya. | Tiger. |
| Sarpa. | Snake. |
| Neul. | Weasel. |
| Goru. | Cow. |
| Mudir. | Frog. |
| Bhadbhadria. | Sparrow. |
| Kurmā. | Tortoise. |
- Fortnightly Review, N. S., LXV, pp. 647-665, 835-852; [Totemism and Exogamy, 1910, vol. i, p. 91 et seq.].
- [Descriptive Ethnology of Bengal, 1872, p. 327.]
- In some places many of the Moghals and Pathāns are regarded as Ajlāf.
- [Census Report, Bengal, 1901, vol. i., p. 439.]