Pramāṇa · Search · About · API

Caste in India: The Facts and the System c7: Part II: The Past

Caste in India: The Facts and the System c7

Language
English

Contents

Part II: The Past

c7.2

Introductory. I. The Brahmanic system of Castes : the Dharmaśāstras and the Epic. II. The Brāhmaṇas and the Vedic Hymns. III. Character and Origin of the System : Classes and Castes. IV. Origins according to Legend : Rivalries between Priests and Nobles

Introductory

c7.s1.1

THE past history of caste is only intelligible in the light of present conditions, as we shall shortly see, for I am anxious to face two problems with which we are confronted, namely, what was the early condition of the Hindu castes in historic times? and what are the beginnings of the institution, if these can be traced back through those obscurities which envelop all origins?

I. The Brahmanic System of Castes: The Dharmaśāstras and the Epic

c7.s2.1

WE must then, first of all, ascertain in what light the castes are presented to us by tradition ; having observed them in practice, let us see what the literary documents have to say. I do not intend to give here more than a very summary idea of this past, for successfully to define the scope and character of the evidence is a matter of some delicacy.

c7.s2.2

The social life of the Hindus is in theory regulated by books whose authorship is attributed to more or less legendary sages, Manu, Yājñavalkya, Vasisṭha, and many others. The place which they accord to civil organization and to criminal repression has contributed, together with a too slavish translation of their Sanskrit title, to make them known as ‘Law-books’, or, to be more exact, ‘Books of the Law’—*dharmaśāstras*. We must not expect to find that they are codes, for of codes they have neither the origin, the form, nor the authority. We are in a country where the religious inspiration which presided over the early organization of societies has been less than elsewhere supplanted by the birth of a secular government. Hindu society is regulated by religious custom, and the law-books are essentially collections of religious precepts. In the absence of a true legislation and under the ever-increasing authority of the Brahmans, they have in the course of time received a sort of official and public sanction. They have, however, only acquired this late in the day, and then not without reservations; it represents a secondary evolution of their history, and in no wise offers conclusive evidence of their primitive character.

c7.s2.3

Concurrently, the stream of epic tradition flows on. Very archaic in origin, less archaic in language, it covers a whole period still somewhat imperfectly determined but in any case vast. By its nature it clearly springs from a totally different section of the population. Nevertheless, its huge canvas comprises not only stories in a heroic and legendary vein, but also gives free scope to doctrinal teaching. It was, moreover, formed at a time when the supremacy of the Brahmans and the authority of their precepts in every human concern were irrevocably established. Epic, by its style, can be directly traced to Brahmanic influence. This is perceptible in its many resemblances, often literal, to the ‘Law-books’, in its copious quotations from them, and especially from the most celebrated of them, the code of Manu. Thus, although by its subject, which is national, if not by its language, which is learned, it is addressed to the people as a whole, and although it draws its principal subject-matter from warlike legends, the Epic belongs to sacerdotal tradition. Its field is so wide and the stories so varied that some inconsistency could not fail to creep in; but on the whole the rules laid down and the chief authority are definitely the same in both cases.

c7.s2.4

Allowing for slight divergences, we can include in a single survey, without fear of any essential incongruity, the picture which unfolds itself in these two series of documents.

c7.s2.5

The theory which emerges shows us a society divided into strictly isolated castes governed by rules very similar to those which are in force to-day. The occupations assigned to each caste are distinct and limited. Marriage is carefully regulated. Only a wife of the same caste can assist her husband in family and sacrificial rites; she alone secures to the son a status equal to that of his father. Born of a woman of a lower caste, the son sinks to his mother’s caste, and his position in the partition of the father’s property is greatly weakened. At the very least the first wife must therefore be of the same caste as the man. He is, moreover, forbidden to marry either within the *gotra* of his father or among his mother’s near relations. As regards food, the distinction between foods allowed and disallowed is set forth with profuse detail; and the use of fermented liquors is condemned as one of the most inexpiable of crimes. The mere glance of a man of low caste is sufficient to pollute a meal, and it is only by exceptional authorizations that food is sometimes allowed to be received from his hands. His very gifts—frankly, I fear this rule has been strained more than once—must be rigorously refused by the Brahman. Several of the most peculiar customs are here established: the marriage of girls before puberty is enjoined, and widows are forbidden to contract a second marriage.

c7.s2.6

The extreme sentence is exclusion from the caste, but it is not as a rule without appeal: a whole code of graduated penances permit those who undergo them to re-enter their own social sphere. But the very name of the grave offences (*pātaka*—‘that which causes to fall’, and *upapātaka*) clearly asserts that their natural effect is to cause the guilty man to fall from his place in the caste assigned to him by birth.

c7.s2.7

The conformity of the data with those which can be collected by an observer at the present day is striking. There is, however, one important difference: if there is one fact which obtrudes itself in the real life of India, it is the enormous number of castes, their overlapping, and the confusion in which they are involved. In theory there are only four castes, *varṇas*: the Brahmans, priests, and learned man; the Kshatriyas, warriors and nobles; the Vaiśyas, agriculturists and merchants; and the Śūdras, the servant-class, on whom devolve all the menial offices. The Brahmans' only duties are to study and teach the Vedas, to offer sacrifices, and to offer, and above all to receive, gifts. It is the duty of the Kshatriyas to give orders, to protect the people, to offer sacrifices through the medium of the Brahmans and to study the Vedas; the Vaiśyas must raise livestock, cultivate the soil, engage in trade, and give alms, not neglecting either the sacred rites or the study of the written word; the Śūdras have only one essential task—to serve the higher castes. Outside this system there are only barbarous or despised peoples who have no access to the religious and social life of the Brahmanic world, that is to say, foreigners or Mlechchhas.

c7.s2.8

What does this fine ordinance amount to in reality? The whole authority and the whole meaning of tradition are involved in this question.

c7.s2.9

One preliminary remark: In spite of their dogmatic tone and systematic appearance, we do not need to scan the prescriptions very closely to perceive that a meticulous pursuit of very slight details covers a great many uncertainties and deficiencies. The lofty exclusiveness of the language hides weakness of authority and laxity of practice. This is to be seen in other places besides India. The authorizations are often uncertain and the degree of precision always low. More serious still are the contradictions which, explicit or implicit, are constantly met with from one passage to another.

c7.s2.10

The system only admits of four castes : there is no fifth, we are assured. And thus, from the mixture of these castes, considered according to the various imaginable hypotheses, new castes are brought forth, the ‘ mixed castes ’. The degree of respect assigned to each is less in proportion as it postulates the association of a woman of higher caste with a man of lower caste. This is not all. Children, although the offspring of a couple in the same caste, may forfeit their status if the obligatory ceremonies are neglected. They form the class of the Vrātyas, but according as they derive from Brahmans, Kshatriyas or Vaiśyas, the Vrātyas divide, with a symmetry which clearly shows an artificial ordinance, into exactly the same number of distinct castes. We are given the name of all these sections and the trades which are proper to each. But these are certainly only specimens ; these mixtures and complications imply the existence of many others, and the Viṣṇusmṛiti¹ is certainly right in describing as ‘ innumerable ’ the sections which originate in this manner.

c7.s2.11

¹ Viṣṇusmṛiti, XVI, 7.

c7.s2.12

How far we still are from the simple truth !

c7.s2.13

The four castes might at least appear to be isolated by their special functions, but reservations at once begin to pour in. Each of the higher castes is first of all authorized to embrace the mode of life peculiar to the caste which follows it in order of gradation. This declension is limited to cases where it is imposed by lack of means. But make no mistake : we are dealing not with an exceptional case of necessity but with perfectly ordinary facts thinly disguised, simply for the sake of the principle, by a decent pretext, by a reservation which it is hoped may survive the shipwreck of the theory. Let us run through the list of Brahmans who in theory are regarded as unworthy to be invited to funeral feasts : robbers, butchers, hirelings, actors, singers, keepers of low gaming-houses, along with other more harmless professions, figure in the list as very common species. It is evident, therefore, that at that time the ways of gaining a livelihood were as varied among the Brahmans as they are to-day.¹ Manu acted prudently in declaring that a Brahman must always be regarded as a high divinity ‘ whatever trade he may follow ’.²

c7.s2.14

¹ Hopkins, *Mutual Relations of Castes*, pp. 39, 52-3.

c7.s2.15

² *Mānava Dh. Ś.* IX, 319.

c7.s2.16

All Brahmans excluded from the model caste must, however, have been distributed as they are to-day, somewhat extensively among special castes. Manu seems to know nothing of them. He never breathes a word about these castes. He therefore does not pretend so to dispose the facts as to form a true picture, but confines himself to presenting the type of Brahmanic caste in its ideal integrity.

c7.s2.17

Regular marriage should only be contracted between members of the same caste. The rules, however, concerning certain marriage ceremonies and matters of inheritance, and the permission to marry, at any rate as second wives, women of lower castes—in fact, the whole theory of mixed castes—proves that the rule was not applied with such uniform severity as the general formula would lead us to suppose. The prohibition against marrying a Śūdra, which in the case of Brahmans and Kshatriyas is repeated with insistence, obviously admits of a great many accommodations. Still more so is this true of the precepts which govern food. Lastly, except for certain reservations, the use of meat itself is tolerated. Abstinence from spirits, which is strongly enjoined elsewhere, appears in certain passages as nothing but a mere counsel of perfection.

c7.s2.18

Despite the divine authority behind which it shelters, tradition is not inviolable. Its formulae appear to be absolute, but in a dozen places it warns us that the true rule resides in custom, that it is the usage proper to each region and caste which is law. It is in accordance with this usage that a king who is careful of his duty must regulate his acts and decrees.¹ To a great extent this is still true to-day. It is a trait which characterizes the whole past of India, and on which the most experienced authorities have rightly insisted.² There are texts which show that lofty descent is most easily recognized by purity of conduct—to such a degree has the intermixture of castes obscured all lineal descent.³ Others ascribe to an earlier and more perfect terrestrial age the time when the ordinance of castes was strictly maintained,⁴ thus recognizing that the rules of theory are in fact strangely elastic.

c7.s2.19

¹ Mānava Dh., VIII, 41.

c7.s2.20

² V. N. Mandlik, op. laud., p. 438

c7.s2.21

³ Mahābhār., Vanap. 12481, Śāntip, 2440; cited by Muir, Sanskrit Texts, I, 134, 484.

c7.s2.22

⁴ Mahābhār., Vanap. 11240, ap. Muir, Sanskrit Texts, I, 143.

c7.s2.23

As was to be expected, the same impression emerges from the Epic. The discrepancies here are of precisely the same nature as in the law-books. Some of them are interesting; but what strikes us most forcibly is the number of cases in the epic stories where facts are in contradiction to doctrine. We have been prepared for the strict distinction of all professions, and yet all castes take part in an armed conflict.⁵ Droṇa, although a Brahman, is one of the chief heroes of the struggle, and, although the son of a shepherd, Karṇa is one of the most celebrated military leaders. The prestige of Yajatra and Vidura does not suffer although they are descended from a Śūdra.¹ Alliances between Kshatriyas and Brahmans, even between these high personages and the lowest castes, are frequent.² We scarcely find any traces of general religious education for young nobles, although this is enjoined by law;³ still less do we see abstinence from meat or liquor observed by warriors.⁴ And yet the rule is known. More than once the theoretic condemnation is given a prominent place in the same story that bears witness to its violation.⁵ After this, it is not surprising to meet with kings of any and every caste,⁶ when Manu himself admits the possibility of a Śūdra enjoying kingly power.⁷

c7.s2.24

⁵ Cf. Hopkins in the Journ. Amer. Orient. Soc., XIII, p. 220.

c7.s2.25

¹ Ibid., p. 95.

c7.s2.26

² Ibid., p. 344.

c7.s2.27

³ Ibid., p. 109.

c7.s2.28

⁴ Ibi l., pp. 119, 120.

c7.s2.29

⁵ Ibid., p. 120, 352-3.

c7.s2.30

⁶ Ibid., p. 137.

c7.s2.31

⁷ Māṇava Dh. Ś., IV, p. 61.

c7.s2.32

The Epic is by its nature so closely connected with the nobility that it tends to attribute to kings the supremacy which is claimed by the law-books for the Brahmans.⁸ It is none the less explicit at times concerning the incomparable grandeur of the sacerdotal class.⁹ Take the story of Mataṅga. He thinks himself the son of a Brahman; as a matter of fact, he is a child of sin, his father having been a Śūdra; he is in reality only a miserable *outcast*. Informed of his shame by a miracle, he attempts to regain his lost dignity by undergoing great hardships. For centuries he labours in vain; for the space of a hundred years he piously balances himself on one foot, but without avail. Indra on his throne is moved and hastens to him; he lavishes on him the most desirable gifts and promises him exceptional favours. But the one which the penitent solicits is impossible! Thousands and millions of successive births are necessary to obtain the ascent from a lower to a higher caste. Rāma does not hesitate to cut off the head of a young Śūdra whose sole crime is to indulge in religious rites which are in theory forbidden to his caste.¹ Insolence such as this threatens to upset the whole equilibrium of public order, so essential is the maintenance of the prerogatives which belong exclusively to the various castes.

c7.s2.33

⁸ Hopkins, loc. cit., p. 73.

c7.s2.34

⁹ See the passages cited by Muir, Sanskrit Texts, I, p. 120 et seq.

c7.s2.35

¹ See passages cited by Muir, *Sanskrit Texts*, I, pp. 118-20.

c7.s2.36

Our main object being to re-establish historical sequence, we are at first disposed to accept these early testimonies regarding the social state of India as the complete and sincere expression of the actual conditions. Are we entitled to do so in the circumstances? Both in epic tradition and in sacerdotal teaching the system is identical;

c7.s2.37

but in both it is equally riddled with uncertainties and contradictions, which are tantamount to confessions of failure. Everything reveals the system as artificial and speculative. The facts have no legal basis. At every turn they give the lie to the system—by contradicting it or failing to coincide with it. The system does not even lay claim to be authoritative, since it clearly recognizes the stronger claims of custom. It is, in a word, nothing more than the focusing of an existing situation whose incoherences and complications it sets out to reconcile in an attempt to bring the whole into conformity with an ideal type.

c7.s2.38

The explanation of mixed castes can never have deceived any one; ¹ flagrant impossibilities condemn it at sight.

c7.s2.39

¹ Cf. Max Müller, *Chips*, II, p. 343.

c7.s2.40

In the presence of a number of groups whose multiplicity completely destroyed the exclusive principle of the four castes it became necessary to justify their existence, and the idea was conceived of deriving them from the very principle which they invalidated. The system itself could logically serve but once to explain the first origin of these groups. Also the number of sections which it was possible to explain in this way was certainly inadequate; the geographical names that many of them bore gave the lie at once to the genesis attributed to them. This was of little moment;

c7.s2.41

the Hindu mind, once drunk with classifications, is not likely to stick at scruples such as these.

c7.s2.42

Actuality, moreover, lent colour to this attempt. There was the case, doubtless often noticed, where the local grouping of people excluded from the paternal caste by their irregular birth and relegated to a lower rung of the social ladder, gave rise to a new subdivision. On this basis, with the deceptive mental discipline of the Hindu mind, more than dubious hypotheses were converted into positive assertions. They offered a twofold advantage: that of creating an appearance of symmetry, which is irresistibly attractive to Indian theorists, and that of eliminating from the principle laid down as the root of the social organization the very confusion which seemed likely to compromise it.

c7.s2.43

The propensity was so strong that it manifested itself in several ways. Is it not Manu himself¹ who represents as Kshatriyas reduced to the condition of Śūdras by diverse offences—omission of rites and contempt for Brahmans—the tribes of the Pauṇḍrakas, Choḍas, Drāviḍas, Kāmbojas, Yavanas, Śākas, Pāradas, Pahlavas, Cīnas, Kirātas, Daradas—that is to say, all the non-Hindu warrior populations of India or of foreign countries—Dravidians and Chinese, Persians and Greeks, Scythians and aborigines? They are, of course, bound by no tie of origin to the Brahmanic organization; but it was essential at all costs to fit them into the preconceived scheme!

c7.s2.44

¹ X, 43-4.

c7.s2.45

First of all, the theory of mixed castes makes a troublesome breach in the system. But what is to be said of the four principal castes? We cannot doubt that the relegation of all the Śūdras to the condition of a mere mass of slaves is purely arbitrary. It is contradicted by the very status assigned to them from the civil point of view by parallel texts. Is it possible to believe that the three higher castes have ever formed those exclusive, compact, and ordered unities which one has pictured to oneself? The Brahmanic caste is working out its destiny under our eyes. But how? Not, as we have already seen, as a true caste, but as an agglomeration of innumerable castes, unequal in rights and social rank, separated in this respect by immense distances. We have only to recall the long lists of degraded and fallen Brahmans enumerated by tradition. It was not otherwise when the law-books were drawn up. As to the Kshatriyas and the Vaiśyas, the few traces which have survived even of their name are as doubtful as they are rare. Where it does appear, the name may have been taken afresh from tradition during recent times—a proceeding of which we have proved examples—in order to serve the arbitrary pretensions of a given group. They can nowhere be encountered as separate, authentic castes. Up to the present we cannot see in them more than generic names, an enormous frame destined to hold and conceal the actual existence of an infinite number of divisions.

c7.s2.46

I have elsewhere had occasion to show,¹ in an article on the Indian theatre, how the Hindus proceed in order to establish theoretic teaching. A liking for classifications and a contempt for facts, a disregard of our idea of logic, and a superstitious respect for formulae—all these characterize their mental outlook, which conspires with the tyranny of the scholastic mind and the uncontested domination of a sacerdotal class to foster the growth of systems and to lend to the most artificial of these an entirely unmerited prestige. What is true of literature is no less true of religion and law, and we must not be shocked in India by opinions which elsewhere might be regarded as rash. The Hindu never hesitates to make generalizations regardless even of those reservations which to us seem most indispensable.

c7.s2.47

¹ *Revue des Deux Mondes*, May 1891.

c7.s2.48

To take one example out of a hundred : There are four conditions of life for the Brahman faithful to his caste-duties. As a novice he must study the scriptures and the rules of sacrifices ; later he marries and founds a family of sons who will carry on the tradition of family ceremonies.

c7.s2.49

Some retire into solitude in order to give themselves up to a life of austerities. Others, still more detached from the world, become mendicant ascetics. In the hands of the theorists these four conditions become regular stages which are to all appearance obligatory in the career of a Brahman. Are we to take these requirements seriously? We should be very wide of the mark if, following the texts, we were to imagine all Brahmans addicted solely to study and penance, dividing their career into four periods and dedicating the two last to the life of a hermit and to the profession of a wandering fakir! The editors of the books have simply welded into a system certain isolated and more or less exceptional facts, and lent an imperative aspect to what was no more than a rarely achieved ideal of perfection. In the same way, have we not seen a literary theorist creating a dramatic category out of a single play,¹ and converting a particular case into a universal rule?

c7.s2.50

¹ *Revue des Deux Mondes*, art. cit., pp. 93 seq.

c7.s2.51

Thus these religious and moral legislators obeyed a natural inclination which is very powerful in the Hindu mind. They also obeyed, more or less willingly, a personal and selfish tendency, the operation of which is everywhere patent, and finally deprived their work of the authority of a straightforward document. Before all else their aim was to establish the absolute supremacy of the Brahmans. With them everything bears on their own glorification and is intended to further their interests. The books which come from their schools are designed to exalt their power and strengthen their pre-eminence. Sole purveyors of literature, it is they also who gave form to the epic traditions. It is therefore only natural that, in spite of accidental discrepancies, these latter reflect with as much insistence as the sacerdotal literature itself the pretensions of the Brahmans, and point out with equal emphasis the privileges to which they lay claim.

c7.s2.52

The law-books not only reserve for the Brahmans all the influential functions and privileges, but also invariably graduate the scale of criminal punishment in their favour. We have seen how the representative council or general assembly of the caste, under the direction of its appointed chief, is empowered to maintain internal law and order, to pronounce the necessary exclusions, or to assess the terms on payment of which the delinquent may escape exclusion. In this connexion Manu and Yājñavalkya speak only of assemblies of Brahmans versed in the sacred texts.¹ The idea of extending Brahmanic power is apparent here; and indeed even to-day a Brahman, alone or co-opted to the council of the caste, often takes the preponderant part in these decisions. The whole of this literature is imbued with and animated by the spirit of Brahmanic ambition. It may be difficult to show in any detail the extent of its arbitrary action. It is certain that it colours more than one part of the picture, and that it misrepresents several of the motive-forces of the social organization as it has been sketched for us. The testimony of literary tradition must then clearly be accepted, but with many reservations.

c7.s2.53

¹ Concerning these parishads, cf. Hopkins, Mutual Relation of Castes, pp. 43 seq; V. N. Mandlik, Vyavahāra Mayūkha, p. 160, note.

c7.s2.54

This tradition seems to alternate strangely between correspondence with existing facts and divergence from them. Everything is easily understood, and contradictions are explained as local diversities, impossible symmetries as attempts at systematic explanation—provided we admit that the tradition corresponds to an absolutely analogous situation, some which still existed in recent times. This situation is only presented in a false perspective, with generalizations, corrections, and interpretations such as might be inspired either by the turn of mind peculiar to the Hindu or by a supreme preoccupation with Brahmanic interests.

c7.s2.55

I say analogous, but I should not venture to say identical. If any one fact stands out clearly from a general review of the castes, it is mobility of formation under the continuous action of fairly stable principles. This process is certainly not new;

c7.s2.56

the causes which produce it have been at work for many centuries. The early situation which corresponds to the editing of the law-books and the Epic may therefore have diverged more or less in matters of detail from the present state of affairs, but it was wholly similar in its broad outlines. It is merely necessary, in the organization of castes as elsewhere, to reserve the possibility and probability of modifications such as time never fails to introduce into human institutions,¹ even after that first and decisive evolution which has determined their individuality.

c7.s2.57

¹ It has been readily admitted (e.g. H. Mayne, *Hindu Law and Usage*, pp. 67-8, 79-80) that the restrictions which prohibit marriage from caste to caste may have been less strict formerly, and that, in a general way, the rules of the caste have probably become increasingly exclusive. The completion in a final form of the *dharmaśāstras* and their growing authority would sufficiently explain the incessant tendency towards the spread of more rigorous and uniform practice. It is clear that our sources are far from allowing us to fix with precision, in the historic and documental period, the stages of a decisive evolution in this matter.

c7.s2.58

In short, it is not theory which can give an account of facts; it is facts which help us to see theory in its true light, and confine it within its proper limits.

II. The Brāhmaṇas and the Vedic Hymns

c7.s3.1

THE law-books and the Epic synchronize with an epoch when the Hindu world was definitely constituted. They rest on a twofold foundation of productions still more ancient : a stratum of sacerdotal literature, and below that the treasure of the Vedic hymns—that collection which, in the narrower terminology of the West, is called specifically the Veda. It is necessary for us to go back to this earlier stratum.

c7.s3.2

From ancient times the teachings proper to the sacerdotal schools were condensed into a species of aphorisms called Sūtras, of which we possess numerous collections. It is from this source that the law-books were drawn. Each is ultimately connected, in a manner more or less direct, with one or other of the schools. This technical literature itself crowns a long preparation represented by the Brāhmaṇas, the most remote testimony of religious reflection applied to the acts of sacrifice : singular books, where in the course of describing successive ceremonies the most rash etymological tricks and the most profound mysticism, childish deductions, and daring speculations jostle one another. Nowhere do they treat of set purpose the subject which we are considering. We find only accidental indications, which are all the more valuable as they are few and far between.

c7.s3.3

According to an eminent authority, Herr Weber, ‘ the organization of castes is fully developed at this period of the Brāhmaṇas ; from that date we find ourselves confronted by the same situation which appears to us idealized and codified in the laws of Manu.¹ In default of definite statements, the allusions and fragmentary information permit of no doubt on the subject.

c7.s3.4

¹ A. Weber, *Indische Studien*, X, 2.

c7.s3.5

The four castes appear there as already established in their respective individualities and privileges. The rights and duties of the Brahmans in particular are in complete accord with more modern descriptions ; ² the necessary purity of the race is duly inculcated.³ Members of the three high castes are bound to take a first wife of equal rank, without prejudice, of course, to other secondary unions.⁴ The caste is destroyed by persistent neglect of initiation ; ⁵ it would be destroyed by many other offences if graduated expiations did not open a way of escape from this supreme penalty.⁶ All intercourse with the excluded members, *patitas* (the fallen), is forbidden ; ⁷ no union with them is allowed ; no food may be taken from their hands. The preoccupation with impure contacts is ever present; it is forbidden to eat with people of low origin,¹ or to use the vessels of Śūdras,² and a Brahman may not be a doctor³ on account of the pollutions rendered inevitable by the profession. The use of liquors is disapproved;⁴ the use of meat is, in certain cases at least, prohibited; while the flesh of various animals is forbidden.⁵ The mixed castes themselves find a place here, in fact, if not in theory. A considerable number are enumerated by name.⁶

c7.s3.6

² Ibid., pp. 41 et seq.

c7.s3.7

³ Ibid., p. 69.

c7.s3.8

⁴ Ibid., p. 21.

c7.s3.9

⁵ Ibid., p. 21.

c7.s3.10

⁶ Ibid., pp. 101 et seq.

c7.s3.11

⁷ Ibid., p. 21.

c7.s3.12

¹ Ibid., p. 74.

c7.s3.13

² Ibid., p. 16, note.

c7.s3.14

³ Ibid., p. 64.

c7.s3.15

⁴ Ibid., p. 62.

c7.s3.16

⁵ Ibid., pp. 62, 63-4.

c7.s3.17

⁶ In the celebrated invocation of the Śatarudrīya; cf. *Indische Studien*, II, pp. 32 seq.

c7.s3.18

If the rules are tempered in numerous ways, that by no means implies that they are in course of formation. In our times also, if we wished to resolve custom into general formulae, we should be thrown back on a multitude of similar reservations. We must learn to fathom the meaning of these uncertainties and contradictions, which are open, according to circumstances, to various explanations.

c7.s3.19

Taking certain passages literally, it would seem that the dignity of a Brahman was at that time the reward of knowledge and virtue rather than the privilege of birth.⁷ But experience, verified by all subsequent literature, teaches us the true significance of the language; it is nothing but a device to glorify the supposed virtue and knowledge of the priests, and in no way involves forgetfulness of the rights conferred by birth alone. In one sense it might find literal justification; neglect of religious obligations of which ignorance or vice may become the source is sufficient to bring about expulsion from the caste.

c7.s3.20

⁷ Weber, loc. laud., pp. 70-1.

c7.s3.21

If the expiations for Brahmans are here rendered singularly mild in many cases, are we therefore to conclude that but little value was attached to the prerogatives whose integrity these were intended to restore? I am all the less inclined to believe it, since even to-day the purifications and fines are, as we have seen, exceedingly light. Think, too, of that passionate glorification of Brahman greatness which is constantly exhibited,¹ of the absurdly exaggerated fees, amounting to hundreds of thousands of cows, which are demanded for their intervention in the sacrifices,² and of the coolness with which it is declared that in face of a lowlier adversary it is the judge's duty always to decide in favour of a Brahman, whatever wrongs he may have committed! How could the author or editors of these books—all Brahmans—have been niggardly in granting to the Brahmans facilities either for evading or softening painful obligations or expiating their offences? This very indulgence proves with what prestige they were already invested.

c7.s3.22

¹ Weber, loc. laud., pp. 35 et seq.

c7.s3.23

² Ibid., pp. 52 et seq.

c7.s3.24

Without doubt this literature already rests on the foundations revealed by the law-books and the Epic. It is a pointless task to search in it for contemporary evidence of the foundation of the system, but in its exaggerations and in the completely exclusive inspiration by which it is permeated, it shows still more clearly at work the inclinations, interests, and defects whence has sprung, not the organization itself, but the system which gave it its dogmatic form. It exposes very clearly its artificial and speculative character. With its roots at least it comes into contact with the most primitive literary strata; among the Vedic hymns are several which are beyond all doubt contemporary with the period during which it was growing up under the direction of the priests.

c7.s3.25

One of these hymns is held to be the earliest document explicitly proving the existence of castes in India. It is that which describes how the universe issued in its entirety from the substance of the primitive male, Purusha. The text declares that ‘the Brahman was his mouth, the Rājanya (Kshatriya) his arms, the Vaiśya his thighs’, and that ‘the Śūdra was born from his feet’. By common consent this passage is among the most recent of the venerable collection in which it ranks, but it has profited to some extent by the prestige which attaches to the whole. On the other hand, in order to weigh up and comment on this piece of evidence, scholars have adopted the preconceived idea that the existence of castes must manifest itself in the form of the four castes of the mature system; to my mind, however, nothing could be less warranted.

c7.s3.26

One example will make my meaning clear. Haug,¹ and after him, with more precision, Kern,² have sought to show, contrary to the most generally received opinion,³ that not only were castes perfectly well known at the Vedic epoch, but that they go back farther—to the days when the ancestors of the Iranians and those of the Hindus lived side by side. They base their argument either on the texts of the Avesta or on the more recent evidence which shows the early population of Iran divided into four *pishtras*, analogous to the four *varṇas* of India. The existence of castes is nowhere attested in the history of Persia. The caste-idea is, however, in the opinion of Kern and Haug, so indissolubly connected with the quadruple division of Brahmans, Kshatriyas, Vaiśyas, and Śūdras, that in order to discover its equivalent in a kindred region they unhesitatingly conclude that the caste-system must have had a parallel existence in the two countries! I consider—and I shall return to this point—that Kern is perfectly justified in comparing the two series. I readily admit that there is more than an outward resemblance between them—there is an intimate affinity; but despite the fusion which has operated in India between this quadruple division and the caste-system nothing proves that the connexion is original or necessary, or that the one implies the other.

c7.s3.27

¹ *Brahma und die Brahmanen*, in the Memoirs of the Academy of Munich.

c7.s3.28

² *Indische Theor. over de Standenverdeeling*, in the Memoirs of the Academy of Amsterdam, 1871, pp. 24 et seq.

c7.s3.29

³ A *résumé* of the controversy on the Vedic existence of castes will be found in Zimmer, *Altind. Leben*, pp. 185 et seq.

c7.s3.30

I must make my meaning clear. The official doctrine acknowledges four castes only, but reality bursts these narrow confines and reveals them in infinite numbers. And here there is a conflict of vital importance between facts and theory—the only one indeed which is not readily settled. Is it possible to argue a difference of time? But, as we have just seen, theory observes and admits by more than one indication and also by the very contradictions in which it is involved that from an early date castes have been very much more numerous than theory appeared at first sight to assume. I have already mentioned how unlikely it is that a Kshatriya and a Vaiśya caste ever really existed. These huge categories are, moreover, incompatible with the very rules—that jealous exclusivism, that corporate and autonomous organization—which characterize caste as it is to-day.

c7.s3.31

The millions of persons in India who lay claim to the title of Brahman, and who are, in a sense, united by that name, are in reality divided into a multitude of perfectly distinct sections, each one of which possesses the attributes and features that characterize this caste. Ordinarily we speak of the Brahmanic caste; we should say the *Brahmanic castes*. We include in a single generic term a multiplicity of castes, each one of which has its own individuality. Manu’s admissions on the subject of the degraded Brahmans prove that he did exactly the same. Then, as now, the name Brahman included them all only as a general honorific title. The Mahābhārata mentions in one place¹ that the son of a Brahman is a Brahman whatever his mother’s origin may have been. This contradiction of Manu’s rules is not necessarily so absolute as it first appears; no matter what theory may say, it is quite possible for an individual to remain a Brahman although he suffers a change of caste.

c7.s3.32

¹ *Anuśāsanap*. 2515, quoted by Kern, loc. laud., p. 42.

c7.s3.33

In our own time the Rajputs, the military class of Western India, claim to correspond—as they do in fact in rank and profession—to the Kshatriyas of the system. Does this mean that they form only one caste, or that they are merely the progressive subdivision of a single caste? We have observed, on the contrary, that under our very eyes castes assume without any right a name which offers them social advantages. Why, therefore, should this be regarded as new?

c7.s3.34

Here we put our finger on the true situation: the names of Brahman, Kshatriya, Vaiśya and Śūdra represent not four primitive ‘castes’, but four ‘classes’. These classes may be exceedingly ancient; it is only in later times that they have been superimposed on the castes. Different by nature and origin, the true castes, or the organisms from which they sprang, were from the beginning more diverse and more numerous. This explanation alone accounts for the glaring incongruity apparent between fact and theory.

c7.s3.35

It is here that comparison with the Iranian texts assumes its full value. Between the four Iranian *pishtras* and the four Hindu *varṇas*, the symmetry is significant:¹ the Athravas or priests corresponding to the Brahmans, the Rathaeśthas or warriors to the Kshatriyas, the Vāstrya-Fshuyants or head of families to the Vaiśyas, and the Hūitis or manual workers to the Śūdras. The general resemblance is striking, and throws into the shade a few doubtful differences. The Vaiśyas are, in Brahmanic tradition, chiefly regarded as cultivators and merchants, but Buddhist literature in calling them generally gṛhapatis ¹ or ‘householders’ brings them strictly into line with the interpretation of the Iranian category. The Hūiti class is not described with a precision which allows us to institute a conclusive comparison with the Śūdras; but the very manner in which, as is often the case with Śūdras, it is left on one side, and consequently isolated from the three first classes, creates between the two, both represented as classes religiously and socially inferior, yet another and by no means a weak bond. In both cases the definite entry of the individual into the community of the higher classes is marked by an identical ceremony—the investiture of the sacred thread.² The correspondence is therefore perfect.

c7.s3.36

¹ Cf. also Ludwig, *Rig Veda*, III, pp. 243-4.

c7.s3.37

¹ To quote one example only, cf. Majjh. Nikāya, ed. Trenckner, I, 85, 30 et seq.

c7.s3.38

² Spiegel, III, p. 700. It is very probable that in Iran, as in India, it was reserved for the three first classes; the fourth appears to have enjoyed only partial rights and to have been relegated to an inferior position (Geiger, p. 479 seq.). What might seem more doubtful, although Spiegel unhesitatingly accepts it (III, pp. 548-9), is whether the Hūitis were really included among the followers of Zoroastrianism, on whom the investitures ought to be conferred.

c7.s3.39

It has been questioned, and very rightly so, whether the pishtras of Iran ever constituted castes.³ It is another matter to decide whether there existed in Iran germs whence the caste might have sprung, germs from which it has sprung in India. In any case, the four *pishtras* of the Avesta represent only *classes*, and in the beginning it was not otherwise with the quadruple Hindu division. If the partition agrees in both cases, it is because it goes back to a remote date; if castes have been evolved in India alone, it is because this partition is not inherently and indissolubly bound up with the caste-system.

c7.s3.40

³ Spiegel, Eran. Alterthumsk., II, pp. 551 et seq.; Geiger, Ostiran. Cultur, p. 485.

c7.s3.41

I know that there is a tendency to accommodate matters and to claim that the present disintegration is the result of the slow disorganization of a strict primitive unity. This is glaringly impossible. It is, however, a point to which I must return when enumerating the various theories which have been advanced in the endeavour to account for origins. For the present I shall confine myself to the scraps of information furnished by literary tradition; we must probe deeply those indications which its oldest monuments offer for our researches.

III. Character and Origin of the System. Classes and Castes

c7.s4.1

THE four castes, as I have indicated above, are not simply co-ordinated in Hindu sources. They are resolved into two groups, the one composed of the three higher, the other comprising the fourth alone.

c7.s4.2

The first embraces the *Āryas*, otherwise called the *dvijas*, persons to whose natural birth is added the religious rebirth conferred by initiation. The *Śūdras*, excluded from this kind of sacrament, have no part at all in learning or the sacred texts to which it serves as an indispensable introduction, no direct part whatever in sacrifices, nor in any of the ceremonies destined to sanctify the life of the higher castes in its various phases. At best they are allowed to celebrate certain inferior rites, and through that alone they are still, although only in a very humble degree, included within the common organization. Initiation is the door by which one enters into the great Aryan family ; as Manu expressly states, the Aryan himself is no better than the Śūdra until he has passed through this second birth. The division, then, is essential ; it is religious, not simply social. A dead man of the three upper castes, if carried by a Śūdra, could not enter into heaven.¹ The strongest form of condemnation for certain offences among Brahmans is to proclaim that thereby they have become Śūdras,² that is to say, *outcasts*. Manu states that for the Śūdra there is no deadly sin, *pātaka*.³ For him there are, in fact, no offences involving degradation, for he has no access to those heights from which it is possible to fall.

c7.s4.3

¹ *Mānava Dh. Ś.*, V, 104.

c7.s4.4

² *Baudhāyana Dh. S.*, II, 3, 6, 32.

c7.s4.5

³ Ibid., X, 126.

c7.s4.6

So sharp a distinction can hardly fail, at the period to which our researches take us, to correspond to a national secession. It is impossible to discern whether the population comprised under the name of Śūdras was solely composed of those aboriginal elements which the Āryans met when immigrating from the north-west into India, or whether it included a mixture of elements. The point is of secondary importance. Between Āryas and Śūdras there was certainly an original opposition of race, more or less absolute. The interbreeding inevitable between conquerors and conquered, between invaders and aborigines, has succeeded in modifying the essential differences, but not in effacing the consciousness of the past.

c7.s4.7

It is significant of the excessive hostility and contempt with which the Śūdra was regarded that a text ¹ puts on the same level the murder of a Śūdra and the destruction of a chameleon, a peacock, or a frog. A Brahman novice has the right to take summarily and without leave from a Śūdra what he requires to pay his master’s fees.² The most terrible punishments overtake a Śūdra who, even in external relations, should fail to keep at a proper distance from a person belonging to the three Āryan castes.

c7.s4.8

¹ Apastamba Dh. S., I, 25, 13.

c7.s4.9

² Ibid., I, 7, 20-1.

c7.s4.10

The literature of the Brāhmaṇas ³ establishes the antithesis between Śūdra and Ārya, Ārya embracing the three higher castes. A multitude of signs betrays between the two terms not merely inequalities of rank but the much deeper struggle of two religious traditions. The Vedic hymns reveal the struggle at its height.

c7.s4.11

³ Indische Studien, X, pp. 4 et seq.

c7.s4.12

The word varṇa, literally colour, is generally accepted as being in Sanskrit the name of the caste. In regard to this I shall have to make reservations. It is at any rate certain that it serves regularly to designate the four theoretic castes. This meaning is not known in the Veda. On the other hand, the word is there employed in two opposing phrases: Ārya varṇa and Dāsa varṇa, the ‘ Āryan race ’ and the ‘ enemy race ’.⁴ They have still more transparent synonyms, such as *black skin*, the *black people*. In more modern literature, the *black race* (*kṛṣṇa varṇa*) continues sometimes to be opposed to the Brahmans.¹ This antithesis is therefore the completely equivalent prototype of that which later came to be expressed by *Ārya* and *Śūdra*, *Ārya varṇa* and *Śaudra varṇa*.²

c7.s4.13

⁴ Zimmer, Altind. Leben, pp. 113-14.

c7.s4.14

¹ *Āpastamba Dh. S.*, I, 28, 11.

c7.s4.15

² *Śatap. Brāhm.*, XIV, 4, 2, 26, cited *Indische Studien*, X, p. 10. *Ārya varṇa* and *Dāsa varṇa* are found in the *Śāṅkhāyana Brāhmaṇa*, cited *Indische Studien*, X, p. 4, note.

c7.s4.16

Through whatever evolution its use may later have passed, the word *varṇa* was therefore first employed to distinguish between two different and hostile peoples; it characterized the one by the relative whiteness, the other by the blackness of its skin. If the ‘Āryan Varṇas’ signify in later literature the three castes reputed to be of Āryan stock, the expression was originally used in the singular: the ‘Āryan Varṇa’ was the collective designation of the whole light-coloured race of invaders.

c7.s4.17

It is therefore certain that the terminology of the system rests on a past which was different. It envelops in its network and fashions to its own ends divisions which originally were connected with quite different ideas, and terms which it has diverted from their first significance. Let us bear this warning in mind.

c7.s4.18

Whatever idea we may form of the state of things during the Vedic period, it is incontestable that the hymns distinguish three main categories among the Āryan population—priests, chiefs, and people. We find priests, under various titles, incessantly occupied with sacrificial acts and the composition of the songs that accompany them ; we follow the chiefs in combat and assemblies ; while the people are always spoken of in a plural which generally shows the ‘ clans ’ of which it is composed gathered round the chiefs in warfare.

c7.s4.19

That sacerdotal functions were by this time strongly organized and protected against over-easy intrusions is warranted by their complicated nature alone. It is universally admitted that the royal power, or more generally the dignity of the nobles, had here, as everywhere else, a marked tendency to become established by a more or less strict heredity. But certainly only a mind strongly biased and determined in advance to find in the past the teachings of the Brahmanic system can discover in this triple classification ‘ castes ’ in the strict meaning of the word and as implied by the more modern doctrine and usage. None of the characteristics which constitute the caste is expressly mentioned.

c7.s4.20

The three terms are sometimes found together ; they palpably comprise the entire Āryan people. One verse asserts that the Viśas (clans) bow spontaneously to the chief (rājan), who is preceded by a Brahman (*Brahman*)¹; that is to say, in technical language, before the king who has a *purohita*, or family-priest. While proving that the claims of sacerdotal power are already established, it presents the situation in its true light: the king and the priest, in the function proper to them, on the one side, and on the other the people. We are then in the presence of classes, more or less exclusive and jealous, but not of castes.

c7.s4.21

¹ *Rig-Veda*, VIII, 35, 16-18.

c7.s4.22

We cannot, however, fail to recognize that this triple division corresponds exactly to the three first castes of the Brahmanic theory. How are we to explain this fact?

c7.s4.23

Of the names borne by the three castes, *Brāhmaṇa* alone is to be found in the hymns (excepting, as is proper, the hymn to Purusha, of which I have spoken above, which assumes as complete this system whose origins we are attempting to fathom). Even this is rare. The primitive form *Brahman* is frequent; in the neuter it is the accepted word used collectively for all the sacerdotal functions. Of the two names which at a later date designate the warriors, *Kshatriya*, which as an epithet expressing power is often appended to the names of certain gods, is only applied to chiefs once or twice in passages suspected of belonging to a recent period; *Rājanya* is unknown. On the other hand, the simple form *Rājan* is the common title of the nobles. The term *Kshatra* summarizes the idea of royal power. The name of the *Vaiśyas* is foreign to the hymns, while the primitive *Viś* is, in the plural, the invariable name of those ‘ clans ’ which constitute the mass of the nation. The trinity *Brahma*, *Kshatra*, and *Viś*, embracing the whole *Āryan* people, is not encountered only in the Veda. In the later literature of the *Brāhmaṇas* it is still an accepted thing.¹ It is repeated with an insistence which testifies to an ancient and respected origin in the very books which know and employ on occasion the definite terms : *Brāhmaṇa*, *Kshatriya*, *Vaiśya*.

c7.s4.24

¹ Cf. *Indische Studien*, X, pp. 18-19, 27.

c7.s4.25

It is clear, I think, that these terms are derived from the ancient formula, a technical and academic derivation. The social order with which they are associated is not merely the spontaneous and organic extension of the situation reflected by the Veda ; like the words in which it is embodied, it represents a carefully thought-out system, adapted to conditions either entirely new or at least very different from those whence the primitive triple division originated.

First 126 of 191 passages. Read on

Linked texts (17)