Religion in Human Evolution: From the Paleolithic to the Axial Age

by Robert N. Bellah

  Further, dharma is not a “universal lawfulness which applies to Hindu society as well as to other societies,” because it does not apply at all to mlecchas, foreigners, non-aryas. One reason for this, among others, is that it can only be communicated in Sanskrit, the “correct” language for “correct” forms of behavior.146 Thus it seems clear that dharma remains archaic even in historical times.

  This understanding of dharma was not without universalistic challenges even within the Hindu tradition. There are passages in the Mahabharata where ahimsa, “non-injury” or “the sparing” of living beings, is taken to be “the core and essence of dharma.“147 Later bhakti (devotional) movements would move in the same direction. One of the most memorable examples of an ethical and universalistic concept of dharma is the edicts of the (Bud dhist) Emperor Asoka (third century BCE). But these tendencies never until modern times gained ascendency. One must remember that the most sacred section of the Mahabharata, the Bhagavadgita, teaches not ahimsa, but svadharma-the dharma appropriate to one’s caste. When Arjuna quails before the task of fighting and probably killing his relatives, his charioteer and teacher, Krsna, tells him to do his duty as a Ksatriya-that is, as a warrior, for whom killing is part of his dharma-but without attachment to the results, and so without bearing the karmic consequences of his action.148

  It is not that there was no philosophy of dharma. The Mimamsa school of philosophy devoted a great deal of effort to defending dharma as absolute commandment, as a set of injunctions, while arguing that the content of the injunctions is simply given in the Veda and cannot be derived from rational reflection. Indeed, rational reflection may lead to unfortunate consequences: the idea that avoiding evil and doing good is the essence of dharma could lead a student to have sexual relations with his teacher’s wife (one of the most severe violations of Vedic injunctions) in order to give her pleasure. Better by far to obey the injunctions as they stand without giving them rational justifications.149 One could call the Mimamsa school an axial school of philosophy (because of its sophisticated use of argument) in defense of an archaic ethical system against axial rationalization. Mimamsa is by far not the only school of thought in Indian history, but it is a central and influential one, and it would be hard to argue that any other position gained a similar degree of hegemony.”’

  This discussion of dharma leads inevitably to the vexed problem of caste. On the whole I have avoided the use of the term “caste” because of its pejorative implications and also because there is a tradition of using “class” to translate the word varna and “caste” only to refer to the word jati, a term I have not used before.151 The four varnas have been discussed in a variety of contexts. Jati has in recent times been used for the thousands of hereditary endogamous groups, usually differentiated by occupation, and also usually classified in terms of varna, although varna assignments have been often uncertain, contested, and, possibly, changing. Halbfass, however, argues that though the two terms are not quite identical, they overlap in the early texts to a degree that they were almost the same-that is, varna was used for what we normally think of as jati, and vice versa. Further, because both are endogamous hereditary groups and have grown out of the same way of classifying, the varna system is “the prototype for important aspects of the `real’ castes [jati],” and so “caste” can also be used for varna.152

  The central point for my argument is that caste in India has no basis in rational argument. One can talk about division of labor, and there was always a degree of relation between caste and occupation, but not a tight one. Impoverished Brahmins and Ksatriyas as well as prosperous Sudras (who were by no means “servants”) are found throughout Indian history. The classification of varnas is based on religious qualifications and is simply given in the Veda, beginning with RV 10.90, no more based on rational argument then the dharma that defines them. In simple terms, the Brahmins can study and teach the Veda and perform sacrifices; the Ksatriyas can study and teach the Veda; the Vaisyas can study the Veda; the Sudras can neither perform sacrifices, nor teach the Veda, nor even study it; needless to say, what applies to Sudras applies equally to those who are beyond the caste system. Even in the late Vedic period, ideas about rebirth, about samsdra and karma, were far from fully developed, but once they had become so ingrained as to be taken for granted, then the caste system could be seen as perfectly just, even though not rationally explicable, because everyone’s position in the present life is determined by actions in previous lives, even though we can have no idea what those actions were.

  It is also clear that caste is not a marginal concern in the Vedas. The revealed texts are saturated with the idea of caste, not just as a social classification, but as a way of thinking about everything in the universe. Brian K. Smith has devoted a book to a description of the many ways varna was used to make sense of the world: it was used to classify the gods, space, time, flora, fauna, and scripture, as well as society. He sums up:

  The varna system was, in sum, a totalistic ideology, by which I mean a system of ideas or categories that account for the cosmos and its parts in such a way that the interests and concerns of those who do the accounting are established, protected, and furthered … The exclusive concentration on the social application of varna can prevent us from grasping its real ideological persuasiveness as a universalistic classificatory system. We [have] surveyed the ways in which varna can be applied to classify the universe in many of its realms. The fact that the reach of varna is much more extensive than the social theory embedded within it should not, however, divert our attention from the powerful case that is being made for social differentiation and privilege. The varna system, a multifaceted and generalized classificatory scheme, had as its first and foremost goal to rationalize and represent an ideal form of hierarchical social structure by projecting that form into the domains of the supernatural, the metaphysical, the natural, and the canonical.‘53

  As with dharma, caste in India was never totally taken for granted, even aside from the principled criticism of the Buddhists and Jains. In the orthodox view, salvation was open only to those with knowledge of the Veda, and Sudras were denied that knowledge. Not only was moksa (salvation) denied them, but their rebirth chances were impaired: knowledge of Vedic injunctions was imperative if one was to be reborn in a “pleasant womb,” that is, a twiceborn womb.154 So for those at the bottom in terms of religious qualification, there was a kind of vicious circle. If they behaved well, they could be reborn in a higher group, but without the knowledge of how to behave well, that possibility was seriously diminished. There were those, particularly in the later bhakti movements, who went out of their way to spread the teachings to everyone, Sudras and outcastes included. And what they taught was not the immense corpus of Vedic injunctions but reliance on the grace of god, Siva, Visnu, Krsna, or the goddess.

  In spite of resistance and variation by time and place, the caste system has remained basic to Indian social organization until recent times. The conservative Mimamsa position has remained hegemonic. Thus, I would argue, though Upanishadic religion was axial, and many forms of rational discourse (linguistics, logic, mathematics, and so on) developed, the foundation of ethics and society remained archaic. This is a position somewhat similar to one that S. N. Eisenstadt and I developed in connection with Japan, where we have argued that the basic premises of society remained non-axial even though elements of axial culture have had a rich history of development.155 The case of India is even more striking, in that one of the great religious breakthroughs of the axial age occurred there, yet the premises of society did not follow suit, except, of course, in the extremely important case of Buddhism that will be discussed below. It is also true that every axial society has had what Eric Voegelin called “an archaic mortgage.“156 Every historical post-axial society has been a combination of axial and non-axial elements, and perhaps could not otherwise have functioned. So India must be seen as an extreme case on a continuum, not as unique. And we must remember that Buddhism, so much more
axial ethically than the Vedic, Brahmanic, Hindu tradition, is also totally a product of Indian history, even if, in the end, it did not survive in India.157

  I am aware that the position I am taking will make me liable to the accusation of Orientalism, of “essentializing” caste. If such a charge implies that I view all “Oriental” societies as inegalitarian, that is obviously not the case: Chapter 8 describes the profound egalitarianism of classical Chinese civilization. I am convinced that Islamic societies are also profoundly egalitarian. Of course, here I speak of ideology, as I do in the case of India-in practice no society since the hunter-gatherers has been very egalitarian. And even in ideology neither Chinese nor Islamic societies were egalitarian when it came to gender.

  Ronald Inden, however, in his book Imagining India has probably made the best case against essentializing caste, giving some substance to what he means by essentializing. He argues that those who see caste as the “essence” of Indian society deny to Indians “agency” and the capacity to change.”’ Instead of viewing human beings as actors determining their own fate, the essentializers, in his view, have given agency to institutions and/or internal ideas (culture?) and not to human beings. But like others in the sociological argument about agency and institutions, he doesn’t tell us what role institutions and culture have once we recognize human agency. In this entire chapter I have been dealing with ideology, including the ideology about basic social premises, but I have never viewed ideology as exercising agency. If, as I am inclined to believe, the basic caste premises of Indian society have survived for a very long time and through many major changes, it is surely because of the vigorous agency of the Brahmin intellectuals who defended caste and of the rulers who on the whole upheld their views, in spite of many protests and other forms of resistance. Neither institutions nor ideas have any agency of their own-they must constantly be “sustained,” “upheld” (dhr), by human actors, yet no society can operate without them. As is so often the case, we cannot see ideas/institutions and agency in a zero-sum way-it is both/and, not either/or.

  India has no monopoly on the history of oppression: every human society so far-except for hunter-gatherers, and even there it was better to be an adult male than a child or a woman-has been oppressive toward significant portions of its population. Democracy and slavery went together in democracy’s two greatest exemplars, ancient Athens and modern America, for a long time at least, and who can say that the United States has not been one of the most oppressive societies in history in its treatment of people both within and without it. Caste is simply the form oppression takes in India.

  But Inden makes a further move that I am inclined to take quite seriously. He argues that castes and other forms of association, both in the villages and in the cities, had a kind of “subject-citizenship,” that is, they recognized the suzerainty of the ruler but they also claimed certain “rights” (Inden’s word) to be heard and taken seriously. Local assemblies operated in away that “disenfranchised” none but the lowest castes, and such assemblies were able to be represented at court.159 All of this makes a great deal of sense to me. It gives effective political form to the many kinds of particularism characteristic of Indian society, of which caste is only one, though the most important one.

  Inden wants to argue that reason and will were involved in the construction of Indian polities, and uses the Rashtrakuta empire of the eighth to the tenth centuries CE as an example.160 Nonetheless, what he describes is, as far as I can see, a collection of particularistic loyalties, in the end fragile and fissile, and not a strong state by Chinese standards, for example. His description does not seriously undermine Romila Thapar’s conclusion:

  Even when the lineage system [as exemplified by jati and varna loyalties] was absorbed into the state, its identity was not entirely eliminated. Administration, except at the higher levels, remained a local concern and the absence of impersonal recruitment to office meant that kinship ties were still effective. Legal codes drew substantially on customary law and incorporated local practices. Legitimacy was frequently expressed through rituals pertaining to the lineage system such as the Vedic sacrifices … Thus it was not so much that the state was a segmentary system with a concentration of power at the centre shading off into ritual hegemony at the periphery as that the state system in itself was not a unitary, monolithic system restructuring the entire territory under its control but rather that it had a margin for flexibility in relation to peripheral areas.”

  For me the fact that “unitary, monolithic” states (such as those constructed on the basis of Chinese legalist doctrine) were rare in India is not a “bad thing.” Several features of the Indian pattern combined to limit the despotic tendencies so evident in historic societies (and not only in the “Orient”). The fact that at the top of society in the alliance between Brahmins and Ksatriyas (brahmaksatra), authority and power were divided, and that dbarma was “the ruling power standing above the ruling power,” meant that there were major restraints on arbitrary political power-despotism. The result may have been a relatively weak state-we will have to consider later the meaning of the ideal of universal rulership in the Indic tradition-but it meant that the people “on the ground,” so to speak, had a variety of defenses against rationalizing tendencies that would have seriously impinged on their form of life or even their existence. Thapar mentioned the importance of customary law in India, and we should remember that dharma included customary law in principle as long as it did not violate Vedic injunctions, and in practice often when it did. If Inden perhaps stretches his terminology too far in speaking of “citizenship” and “rights,” he is not wrong in seeing in these particularistic defenses of customary life a sort of functional equivalent to these ideas in a quite different cultural idiom. In short, as we know to our sorrow in the twentieth and twenty-first centuries, strong states are by no means an unvarnished good, even if “failed states” can be worse. In the case of traditional India we are talking about limited states, not failed states.

  Before moving to a discussion of Buddhism, a brief comparative look at India and China, particularly with respect to the ethos of their dominant classes, might be instructive. In some ways Brahmins could well be compared to Confucians: both were the keepers of the normative order, dharma in the case of India, li in the case of China, and it is worth remembering that both terms originally meant sacrificial ritual, though they were broadened to include the normative order as a whole. Yet there is a striking difference in relation to the state: the Confucian saw the state as the potential embodiment of the ideal social order and so saw public office as his primary calling; the Brahmin saw a social order to a considerable extent independent of the state, though defended by it, as the ideal order, and his primary calling was as religious teacher and priest. It is this difference that has made Westerners think of China as “secular” and India as “religious,” though the Chinese ideal state was supposed to embody religious values as much as the ideal Indian social order was.

  But there is another matter that seems to justify what I consider a skewed Western perception: the way in which the two elites related to the highest religious order. The traditional Hindu formulation of the three “ends of life”-dharma (duty), artha (success), and kama (pleasure)-held for all upper-caste householders. As Charles Malamoud has pointed out, these three ends of life do not map easily onto the varnas. The Brahmins were responsible for “dharma pronounced,” as he puts it, the Ksatriyas for “dharma protected,” the Ksatriyas also for “political artha,” and the Vaisas for “economic artha.” Kama “in the sense of (desire for) sensual pleasure” is common to all varnas, but has an especially strong affinity to Ksatriyas, though it is also a dangerous temptation for them.162 Although China did not have such a typology, the Confucians were responsible for the transmission and inter pretation of the li. But the Indian typology was complicated by the addition of a fourth category, moksa (salvation, liberation), and although that too had a Brahmin primacy in the beginning, there wa
s more than a little tension between the demands of dharma and the demands of moksa, insofar as the serious pursuit of moksa required the life of a renouncer, incompatible with the life of a householder and his primary obligation (as in China) to continue the patrilineage.‘63 In the late Vedic period there was controversy over which had priority for Brahmins, householder or renouncer, as will be discussed further below, but the compromise solution that was included in the idea of varnasramadharma was that the householder and renouncer would be successive “stages of life” (as`rama), with the renouncer stage beginning in old age after all the obligations of the householder had been fulfilled.

  If there is a term parallel to moksa in classical China, it would be one sense of dao, namely the Daoist sense of that with which one merges to attain “salvation.” However, the Confucian meaning of dao was very close to the meaning of dharma, that is, “the Way of the ancestors.” But although Daoists can be considered renouncers, in that they showed little interest in the pursuit of ordinary life, they were renouncers of a rather different type than the Indian ones, more insouciant, more aesthetic, less serious, even in a sense, less “religious.” One further parallel, however, is that Confucians and Daoists were often the same persons, indicating a division of calling within the intellectual elite similar to that of India. It is just that in China one was, as the proverbial saying has it, “Confucian in office, Daoist out of office.” What is interesting here is that renouncers turn up in the axial age in India and China, though with very different cultural emphasis. They are less obvious in ancient Israel and Greece, though not missing: the Nazirites in ancient Israel, the Cynics (in the technical sense of that term) in ancient Greece, though neither seems to have been as central at the moment of the axial transition as renouncers in China and India.