I

The Bhagavad-Gita, 'Song of the Blessed One", forms part of the great Indian epic Mahabharata (1). Its 18 adhyaya chapters contain the report by Sanjaya of a dialogue between the Pandava hero Arjuna and his Yadu Charioteer Krsna, the eighth incarnation of Visnu. The actual fighting is about to begin when Arjuna feels revulsion at the leading part which he must play in the impending slaughter of cousins and kinsmen. The exhortations of Lord Krisa answer every doubt through a complete philosophical cycle, till Arjuna is ready to bend his whole mind, no longer divided against itself, to the great killing. The Gita has attracted minds of bents entirely different from each other and from that of Arjuna. Each has interpreted the supposedly divine words so differently from all the others that the original seems far more suited to raise doubts and to split a personality than to heal an inner division. Any moral philosophy which managed to receive so many variant interpretations from minds developed in widely different types of society must be highly equivocal. No question remains of its basic validity if the meaning be so flexible. Yet the book has had its uses.

If a Mahabharata war had actually been fought on the scale reported, nearly five million fighting men killed each other in an 18-day battle between Delhi and Thanesar; about 130,000 chariots (with their horses), an equal number of elephants and thrice that many riding horses were deployed. This means at least as many camp-followers and attendants as fighters. A host of this size could not be supplied without a total population of 200 millions, which India did not attain till the British period, and could not have reached without plentiful and cheap iron and steel for ploughshares and farmers' tools. Iron was certainly not available in any quantity to Indian peasants before the 6th century BC. The greatest army camp credibly reported was of 400,000 men under Chandragupta Maurya, who commanded the surplus of the newly developed Gangetic basin. The terms patti, gulma etc., given as tactical units in the M'bh did not acquire that meaning till after the Mauryans. The heroes fought with bows and arrows from their chariots, as if the numerous cavalry did not exist; but cavalry which appeared comparatively late in ancient Indian warfare, made the fighting chariots obsolete as was proved by Alexander in the Punjab.

The epic began, like the early Homeric chants, as a series of lays sung at the court of the conquerors. The lament was thinly veiled, presumably by irony; the defeated Kurus survived in legend (e.g. the Kuru-dhamma-jataka) as unsurpassable in rectitude and nobility of character. Krsna-Narayana had no role to play even in the first connected epic narrative. Should the reader doubt all this, let him read the final cantos of the extant M'bh. The Pandavas come in the end to disgraceful old age, and unattended death in the wilderness. Their opponents are admitted to heaven as of right, but the heroes are only transferred there from the tortures of hell, after a long and stubborn effort by the eldest brother Yudhisthira. It strikes even the most casual eye that this is still the older heaven of Indra and Yama; Krsna-Narayana is not its dominant figure, but a palpable and trifling insertion in a corner.

Those legendary Utopians, the pure and unconquerable Uttara-Kurus of the Digha- Nikaya (DN 32) and the Aitareya Brahmana (AB 8.14; 8.23) are not to be confused with the Kurus who survived in historical times near Delhi-Meerut. The Buddha preached several of his sermons at the settlement Kammasa-damma in Kuru- land (Majjhima Nikoya 10; 75; 106) while their capital seems to have been at Thullakotthita (MN 82), the seat of the nameless petty tribal Kuru chief, presumably descended from the Pandava conquerors whom the epic was to inflate beyond all limits. This negligible kingdom either faded away or was among the tribal groups systematically destroyed by the Magadhan emperor Mahlapadma Nanda, a few years before Alexander's raid into the Punjab. The memory, however, remains- as of a tribe, but not a full- fledged kingdom with a class structure in the eleventh book of the Arthasastra, along with similar oligarchies like the Licchavis and the Mallas known to have been destroyed about 475 BC. As for Narayana, it might be noted here that the famous benedictory initial stanza Narayanam namaskrtya, which would make the whole of the extant M'bh into a Vaisnava document, was stripped off by V. S. Sukthankar's text-criticism in 1933 as a late forgery.

We know that the Gita exercised a profound influence upon Mahatma Gandhi B. G. Tilak, the 13th century Maharashtrian reformer Jnanesvar, the earlier Vaisnava Acarya Ramanuja, and still earlier Samkara. Though both fought hard in the cause of India's liberation from British rule, Tilak and the Mahatma certainly did not draw concordant guidance for action from the Gita. Aurobindo Ghose renounced the struggle for India's freedom to concentrate upon study of the Gita. Lokamanya Tilak knew the Jnanesvari comment, but his Gita-rahasa is far from being based upon the earlier work. Jnanesvar himself did not paraphrase Samkara on the Gita, nor does his very free interpretation follow Ramanuja; tradition ascribes to him membership of the rather fantastic natha sect. Ramanuja's Vaisnavism laid a secure foundation for the acrid controversy with the earlier followers of Siva who came into prominence with the great Samkara. But then, why did Samkara also turn to the Bhagavad-gita?

What common need did these outstanding thinkers have that was at the same time not felt by ordinary people, even of their own class? They all belonged to the leisurely class of what, for lack of a better term, may be called Hindus. The consequent bias must not be ignored, for the great comparable poet-teachers from the common people did very well without the Gita. Kabir, the Banaras weaver, had both Muslim and Hindu followers for his plain yet profound teaching. Tukaram knew the Gita through the Jnanesvari, but worshipped Visnu in his own way by meditation upon God and contemporary society in the ancient caves (Buddhist and natural) near the junction of the Indrayani and Pauna rivers. Neither Jayadeva's Gita- govinda, so musical and supremely beautiful a literary effort, (charged with the love and mystery of Krsna's cult) nor the Visnuite reforms of Caitanya that swept the peasantry of Bengal off its feet were founded on the rock of the Gita. I have yet to hear that the heterogeneous collection which forms the Sikh canon owes anything substantial directly to the Gita, though it preserves verses due to Jayadeva, and the Maharashtrian Namdev. Jnanesvar ran foul of current brahmin belief at Alandi, and had to take refuge about 1290 AD on the south bank of the Godavari, in the domains of Ramacandra Yadava, to composed his famous gloss in the common people's language.

We know as little of the historic action taken or instigated by Samkara and Ramanuja as we should have known of Tilak's had only his Gita-rahasya survived. Yet, about the year 800, Samkara was active in some manner that resulted-according to tradition-in the abolition of many Buddhist monasterics. That this was achieved by his penetrating logic and sheer ability in disputation is now, the general Hindu belief. The mass of writing left in his name, and what is given therein as the Buddhist doctrine which he refutes, make only one thing clear: that he had not the remotest idea of Gotama Buddha's original teaching. Buddhism as practised in the monasteries had in any case degenerated into Lamaism with opulent vihara foundations which were a serious drain upon the economy of the country. That Samkara's activity provided a stimulus to their abolition, and Ramanuja's some handle against the wealthier barons whose worship of Siva was associated in the popular mind with their oppressive land- rent, seems a reasonable conclusion on the evidence before us. Otherwise, it would be difficult to explain why the richer, aristocratic landholders opted for Siva, the poorer, and relatively plebeian overwhelmingly for Visnu, in the bitter smarta-vaisnava feuds it is difficult to believe that they could come to blows because of differing religious philosophy. Samkara managed to discover a higher and lower knowledge in the Upanisads which allowed him "to conform to the whole apparatus of Hindu belief"-whatever that may mean- "on the lower plane, while on the higher he finds no true reality in anything; his logic, it has been well said, starts by denying the truth of the proposition, 'A is either B or not B' ...At death the soul when released is merged in the absolute and does not continue to be distinct from it". According to Ramanuja, "if in a sense there is an absolute, whence all is derived, the individual souls and matter still have a reality of their own, and the end of life is not merger in the absolute but continued blissful existence. This state is to be won by bhakti, faith in and devotion to God." It is not possible to imagine that subtle arguments on tenuous ideas gripped the masses, that people could be whipped up to a frenzy merely by the concept of restricted dualism (visnuadvaita) or thoroughgoing dualism (dvaita ).Yet frenzied conflict there was, for centuries. Neither side objected to rendering faithful service at the same time to beef-eating Muslim overlords, who knocked brahmins off without compunction or retribution, and desecrated temples without divine punishment.

The main conclusion is surely the following: Practically anything can be read into the Gita by a determined person, without denying the validity of a class system. THE GITA FURNISHED THE ONE SCRIPTURAL SOURCE WHICH COULD BE USED WITHOUT VIOLENCE TO ACCEPTED BRAHMIN METHODOLOGY, TO DRAW lNSPIRATION AND JUSTIFICATION FOR SOCIAL ACTIONS IN SOME WAY DISAGREEABLE TO A BRANCH OF THE RULING CLASS upon whose mercy the brahmins depended at the moment. That the action was not mere personal opportunism is obvious in each of the cases cited above. It remains to show how the document achieved this unique position.

That the song divine is sung for the upper classes by the brahmins, and only through them for others, is clear. We hear from the mouth of Krsna himself (G.9.32): "For those who take refuge in Me, be they even of the sinful breeds such as women, vaisyas, and sudras." That is, all women and all men of the working and producing classes are defiled by their very birth, though they may in after-!ife be freed by their faith in the god who degrades them so casually in this one. Not only that, the god himself had created such differences (G.4.13) : "The four-caste (-class) division has been created by Me" ; this is proclaimed in the list of great achievements.

The doctrines are certainly not timeless (3). Ethics come into being only as they serve some social need. Food-producing society (as distinct from conflicting aggregates of food-gathering tribal groups) originated in the fairly recent and defined historical past, so that the principles upon which it may work at some given stage could not have expressed from eternity. The Gita sets out each preceeding doctrine in a masterly and sympathetic way without naming or dissecting it, and with consummate skill passes smoothly on to another when Arjuna asks 'why then do you ask me to do something so repulsive and clearly against this?" Thus, we have a brilliant (if plagiarist) review- synthesis of many schools of thought which were in many respects mutually incompatible. The incompatibility is never brought out; all views are simply facets of the one divine mind. The best in each system is derived, naturally, as from the high God. There is none of the polemic so characreristic of disputatious Indian philosophy; only the Vedic ritual beloved of the Mimamsakas is condemned outright. The Upanisads are well-if anonymously- represented, though the svetasvalara upanisad alone contains the germ of bhakti, and none the theory of perfection through a large succession of rebirths. This function of karma is characteristically Buddhist. Without Buddhism, G.2.55-72 (recited daily as prayers at Mahatma Gandhi's asrama) would be impossible. The brahma-nirvana of G. 2.72 and 5.25 is the Buddhist ideal state of escape from the effect of karma. We may similarly trace other-unlabelled- schools of thought such as Samkhya and Mimansa down to early Vedanta (G. 15. 15 supported by the reference to the Brahma-sutra in G. 13.4). This helps date the work as somewhere between 150-350 AD, nearer the later than the earlier date. The ideas are older, not original, except perhaps the novel use of bhakti. The language is high classical Sanskrit such as could not have been written much before the Guptas, though the metre still shows the occasional irregularity (G. 8. 10d, 8. 11 b, 15. 3a, &c) in tristubhs, characteristic of the M'bh as a whole. The Sanskrit of the high Gupta period, shortly after the time of the Gita, would have been more careful in versification.

It is known in any case that the M'bh and the Puranas suffered a major revisions in the period given above. The M'bh in particular was in the hands of Brahmins belonging to the Bhrgu clan, who inflated it to about its present bulk (though the process of inflation continued afterwards) before the Gupta age came to flower. The Puranas also continued to be written or rewritten to assimilate some particular cult of Brahminism. The last discernible reaction of the main Purana group refers to the Guptas still as local prince between Fyzabad and Prayag(4). This context fits the Gita quite well. The earliest dated mention of anything that could possibly represent the Gita is by Hsiuen Chuang (5), early in the seventh century, who refers to a Brahmin having forged at his king's order such a text, (supposedly of antiquity) which was then 'discovered', in order to foment war. The fact does remain that the M'bh existed in two versions at the time of the Asvalayana Grhya Sutra, which refers both to the Bharata and the Mahabharata (6). The prologue of the present M'bh repeats much the same information in such a way as to make it evident that the older 24,000-sloka Bharata was still current at the time the longer version was promulgated. Every attempt was made to ascribe both to the great 'expander', Vyasay to whom almost every Purrana is also ascribed. A common factor is the number 18, which had some particular sanctity for the whole complex, and for the Brahmins connected therewith. There are 18 main gotra clan-groups of brahmins (7)though the main rsi sages are only seven in number; many of the 18 (e.g. the kevala Bhargarvas and kevala Angirasas) are difficult to fit into a rational theme. Correspondingly there are 18 main Puranas, and 18 parvan sections of the M'bh, though the previous division was into 100, as we learn from the prologue. The very action of the Bharatan war was fought over 18 days between 18 legions. The Gita has also 18 adhyoyas, which is surely not without significance. That the older Bharata epic had a shorter but similar Gita is unlikely. One could expect some sort of an exhortation to war, as is actually contained in G. 2.37: "If slain, you gain heaven; if victorious, the earth; so up, son of Kunti, and concentrate on fighting". These lines fit the occasion very well. Such pre-battle urging was customary in all lands at all times (advocated even by the supremely practical Arthasastra, 10.3) through invocations and incantations, songs of bards, proclamations by heralds, and speech of captain or king. What is highly improbable- except to the brahmin bent upon getting his niti revisions into a popular lay of war-is this most intricate three-hour discourse on moral philosophy, after the battle-conch had blared out in mutual defiance and two vast armies had begun their inexorable movement towards collision.

To put it bluntly, the utility of the Gita derives from its peculiar fundamental defect, namely dexterity in seeming to reconcile the irreconcilable. The high god repeatedly emphasizes the great virtue of non-killing (ahimsa), yet the entire discourse is an incentive to war. So G.2. 19 ff. says that it is impossible to kill or be killed. The soul merely puts off an old body as a man his old clothes, in exchange for new; it cannot be cut by weapons, nor suffer from fire, water or the storm. In G. 11, the terrified Arjuna sees all the warriors of both sides rush into a gigantic Visnu- Krshna's innumerable voracious mouths, to be swallowed up or crushed. The moral is pointed by the demoniac god himself (G. 11. 33) : that all the warriors on the field had really been destroyed by him; Arjuna's killing them would be a purely formal affair whereby he could win the opulent kingdom. Again, though the yajna sacrifice is played down or derided, it is admitted in G. 3. 14 to be the generator of rain, without which food and life would be impossible. This slippery opportunism characterizes the whole book. Naturally, it is not surprising to find so many Gita lovers imbued therewith. Once it is admitted that material reality is gross illusion, the rest follows quite simply; the world of "doublethink" is the only one that matters.

The Gita was obviously a new composition, not the expansion of some proportionately shorter religious instruction in the old version. I next propose to show that the effort did not take hold for some centuries after the composition.

The lower classes were necessary as an audience, and the heroic lays of ancient war drew them to the recitation. This made the epic a most convenient vehicle for any doctrine which the Brahmins wanted to insert; even better than rewriting the Puranas, or faking new Puranas for age-old cults. The Sanskrit language was convenient, if kept simple, because the Prakrits were breaking apart into far too many regional languages; Sanskrit was also the language which the upper classes had begun to utilize more and more, Kusana and Satavahana inscriptions are in the popular lingua franca used by monk and trader. But from 151 AD, there appears a new type of chief (oftener than not of foreign origin like Rudradaman) who brags (8) in ornate Sanskrit of his achievements, including knowledge of Sanskrit. The Buddhists had begun to ignore the Teacher's injunction to use the common people's languages; they too adopted Sanskrit. The high period of classical Sanskrit literature really begins with their religious passion-plays and poems, such as the written by Asvaghosa (9). A patrician class favouring Sanskrit as well as the Sanskrit-knowing priestly class was in existence.

No one could object to the interpolation (10) of a story (akhyana) or episode. After all, the M'bh purports to be the recitation in the Naimisa forest to the assembled sages and ascetics by a bard Ugrasravas, who repeated what Vyasa had sung to Janamejaya as having been reported by Sanjaya to Dhrtarastra! The brahmins were dissatisfied with the profit derived from the Gita, not with its authenticity. So, we have the Anu-gita (11) as a prominent sequel in the 14th Canto (Asvamedha- parvan). Arjuna confesses that he has forgotten all the fine things told before the battle, and prays for another lesson. Krsna replies that it would be impossible even for him to dredge it out of his memory once again; the great effort was not to be duplicated. However, an incredibly shoddy second Gita is offered instead which simply extols brahminism and the brahmin. Clearly, that was felt necessary at the time by the inflators though no one reads it now, and it cannot be compared to the first Gita even for a moment.

Secondly, the Gita as it stands could not possibly help any ksatriya in an imminent struggle, if indeed he could take his mind off the battle long enough to understand even a fraction thereof. The ostensible moral is : "Kill your brother, if duty calls, without passion; as long as you have faith in Me, all sins are forgiven." Now the history of India always shows not only brothers but even father and son fighting to the death over the throne, without the slightest hesitation or need for divine guidance. Indra took his own father by the foot and smashed him (RV 4.18.12), a feat which the brahmin Vamadeva applauds Ajatasatru, king of Magadha, imprisoned his father Bimbisara to usurp the throne and then had the old man killed in prison. Yet, even the Buddhists (12) and Jains as well as Brahadrayaka Upanishad (2.1) praise the son (who was the founder of India's first great empire) as a wise and able king. The Arthasastra (A.1.17-18) devotes a chapter to precautions against such ambitious heirs-apparent; and shows in the next how the heir- apparent could circumvent them if he were in a hurry to wear the crown. Krsna himself at Kuruksetra had simply to point to the Yadava contingent, his own people, who were fighting in the opposite ranks. The legend tells us that all the Yadavas ultimately perished fighting among themselves. Earlier, Krsna had killed his maternal uncle Kamsa. The tale gains a new and peculiar force if it is remembered that under mother-right, the new chief must always be the sister's son of the old.

Thirdly, Krsna as he appears in the M'bh is singularly ill-suited to propound any really moral doctrine. The most venerable character of the epic, Bhisma, takes up the greatest of M'bh parvans (Santi) with preaching morality on three important questions: King-craft (raja-dharma), conduct in distress (apad-dharma), and emancipation (moksa-dharma). As regent, he had administered the kingdom to which he had freely surrendered his own right. He had shown irresistible prowess and incomparable knightly honour throughout a long life of unquestioned integrity. The sole reproach anyone can make is that he uses far too many words for a man shot full of arrows, dying like a hedgehog on a support of its own quills. Still, Bhisma seems eminently fitted to teach rectitude. But Krsna? At every single crisis of the war, his advice wins the day by the crookedest of means which could never have occurred to the others. To kill Bhisma, Sikhandin was used as a living shield against whom that perfect knight would not raise a weapon, because of doubtful sex. Drona was polished off while stunned by the deliberate false report of his son's death. Karna was shot down against all rules of chivalry when dismounted and unarmed; Duryodhana was bludgeoned to death after a foul mace blow that shattered his thigh. This is by no means the complete list of iniquities. When taxed with these transgressions, Krsna replies bluntly at the end of the Salya-parvan that the man could not have been killed in any other way, that victory could never have been won otherwise. The calculated treachery of the Arthasastra saturates the actions of this divine exponent of the Bhagavad-gita. It is perhaps in the same spirit that leading modern exponents of the Gita and of ahimsa like Rajaji have declared openly that non-violence is all very well as a method of gaining power, but to be scrapped when power has been captured: "When in the driver's seat, one must use the whip."(13)

Just as the M'bh could be used as a basis only because people came to hear the war-story recite, Krsna could have been of importance only if his cult were rising in popularity, yet sufficiently unformed for such barefaced remoulding. The cult, however, is clearly synthetic. The identification with Narayana is a syncretism, taking originally distinct cults as one. In the same direction is the assimilation of many sagas to a single Krsna legend, whether or not the original hero bore the epithet of Krsna. There would, however, be no question of creating a new cult out of whole cloth; some worship or set of similar worships must already have been in existence among the common people before any brahmins could be attracted thereto. The best such recent example is that of Satyanarayana, the true Narayana, so popular all over the country, but which has no foundation whatever in scripture, and which is not even mentioned 200 years ago. Indeed, the origin seems to be in the popular legends of one Satya Pir, (14) in Bengal; the Pir himself became Satyanarayana.

The speech-goddess Vag-ambhrni, in a famous but late hymn of the Rigveda (RV. 10.12.5), declares that she draws Rudra's bow, and is her Soma and the substance of all that is best. The original god whose misdeeds are never sin is surely the upanishadic Indra who says to Pratardana Daivodasi: "Know thou Me alone; this indeed do I deem man's supreme good-that he should know Me. I slew the three- headed Tvastra, threw the Arurmagha ascetics to the wolves, and transgressing many a treaty, I pierced through and through the Prahladiyans in the heavens, the Paulomas in the upper air, and the Kalakanjas on this earth. Yet such was I then that I never turned a hair. So, he who understands Me, his world is not injured by any deed whatever of his: not by his killing his own mother, by killing his own father, by robbery, killing an embryo, or the commission of any sin whatever does his complexion fade" (Kaus .Brah. up. 3.2). The 'breaking many a treaty' is again the Arthasastra king's normal practice, though that book mentions that in olden days even a treaty concluded by simple word of mouth was sacred (A. 7.17). Indra performed all these dismal feats in vedic tradition, but that tradition nowhere makes him proclaim himself as the supreme object for bhakti; papa and bhakti are not vedic concepts. No vedic god can bestow plenary absolution as in G.18.66 : "Having cast off all (other) beliefs, rites and observances, yield to Me alone; I shall deliver you from all sins, never fear". The reason Krsna could do this and not Indra was that the older god was clearly circumscribed by immutable vedic suktas and tied to the vedic yajna fire- ritual. He was the model of the barbarous Aryan war-leader who could get drunk with his followers and lead them to victory in the fight. His luster had been sadly tarnished by intervening Buddhism, which had flatly denied yajna and brought in a whole new conception of morality and social justice. The pastoral form of bronze- age society with which Indra was indissolubly connected had gone out of productive existence.

One feature of the Krsna myth, which still puzzles Indians, would have been quite familiar to the Greeks. The incarnate god was killed-unique in all Indian tradition- by an arrow shot into his heel, as were Achilles and other Bronze-age heroes. Moreover, the archer Jaras is given in most accounts as Krsna's half-brother, obviously the tanist of the sacred king who had to kill the senior twin. Krsna himself consoles the repentant killer, and absolves him by saying that his own time had come; the sacred king's appointed term had ended. One might venture the guess that the original unpardonable sin committed by Indra and perhaps by Krsna as well was the violation of matriarchal custom, unthinkable in the older society, but which they managed to survive triumphantly, and in comparison to which all other sins paled into insignificance. Certainly, the gokula in which Krsna was brought up would be patriarchal, as a cattle-herders' commune. But the Vrindavana where he played his pranks was sacred to a mother- goddess, the goddess of a group (vrnda) symbolized by the Tulasi (Basil) plant. Krsna had to marry that goddess, and is still married to her every year, though she does not appear in the normal list of his wives; originally, this meant a hieros gamos with the priestess who represented the goddess, and the annual sacrifice of the male consort. Inasmuch as there is no myth of Krsna's annual sacrifice, but only of his having substituted for the husband, as he seems to have broken the primitive usage, as did Herakles and Theseus.

This is "in the Indian character", and we have remarked that a similar attitude is reflected in the philosophy of the Gita. No violence is done to any preceding doctrine except vedic yajna. The essential is taken from each by a remarkably keen mind capable of deep and sympathetic study; all are fitted together with consummate skill and literary ability, and cemented by bhakti without developing their contradictions. The thing to mark is that the Indian character was not always so tolerant. There are periods when people came to blows over doctrine, ritual, and worship. Emperor Harsa Siladitya (circa 600-640 AD) of Kanauj found no difficulty in worshipping Gauri, Mahesvara-Siva, and the Sun, while at the same time he gave the fullest devotion to Buddhism (23) His enemy Narendragupta- Sasanka, raided Magadha from Bengal, cut down the Bodhi tree at Gaya, and wrecked Buddhist foundations wherever he could. What was the difference? Why was a synthesis of the two religions, actually practised by others besides Harsha (as literary references can show) not successful?

Let me put it that the underlying difficulties were economic. Images locked up too much, useful metal; monasteries and temples after the Gupta age withdrew far too much from circulation without replacement or compensation by adding to or stimulating production in any way. Thus, the most thoroughgoing iconoclast in Indian history was another king Harsa (1089-1101 AD) who broke up all images(24) in Kasmir, except four that were spared. This was done systematically under a special minister devotpatana-nayaka, without adducing the least theological excuse, though one could easily have been found. The Kasmirian king remained a man of culture, a patron of Sanskrit literature and the arts; he presumably read the Gita too. But he needed funds for his desperate fight against the Damara group of local barons. The particular campaign was won, at the cost of making feudalism stronger than ever.

To sum up, writing the Gita was possible only in a period when it was not absolutely necessary. Samkara could not do without the intense polemic of theological controversy. To treat all views tolerantly and to merge them into one implies that the crisis in the means of production is not too acute. FUSION AND TOLERANCE BECOME IMPOSSIBLE WHEN THE CRISIS DEEPENS, WHEN THERE IS NOT ENOUGH OF THE SURPLUS PRODUCT TO GO AROUND, AND THE SYNTHETIC METHOD DOES NOT LEAD TO INCREASED PRODUCTION. Marrying the gods to goddesses had worked earlier because the conjoint society produced much more after differences between matriarchal and patriarchal forms of property were thus reconciled. The primitive deities adopted into Siva's or Visnu's household helped enlist food-gathering aboriginals into a much greater food-producing society. The alternative would have been extermination or enslavement, each of which entailed violence with excessive strain upon contemporary production. The vedic Aryans who tried naked force had ultimately to recombine with the autochthonous people. The Gita might help reconcile certain factions of the ruling class. Its inner contradictions could stimulate some exceptional reformer to make the upper classes admit a new reality by recruiting new members. But it could not possibly bring about any fundamental change in the means of production, nor could its fundamental lack of contact with reality and disdain for logical consistency promote a rational approach to the basic problems of Indian society.

However, the Gita did contain one innovation which precisely fitted the needs of a later period: bhakti, personal devotion. To whoever composed that document, bhakti was the justification, the one way of deriving all views from a single divine source. As we have seen from the demand for the quite insipid Anu-Gita sequel, this did not suffice in its own day. But with the end of the great centralized personal empires in sight, Harsa's being the last- the new state had to be feudal from top to bottom. The essence of fully developed feudalism is the chain of personal loyalty which binds retainer to chief, tenant to lord and baron to king or emperor. Not loyalty in the abstract but with a secure foundation in the means and relations of production: land ownership, military service, tax-collection and the conversion of local produce into commodities through the magnates. This system was certainly not possible before the end of the 6th century AD. The key word (26) is samanta which till 532 at last meant 'neighbouring ruler' and by 592 AD h'ad come to mean feudal baron. The new 'barons were personally responsible to the king, and part of a tax-gathering mechanism. The Manusmrti king, for example, had no samantas; he had to administer everything himself, directly or through agents without independent status. The further development of feudalism 'from below' meant a class of people at the village level who had special rights over the land (whether of cultivation, occupation, or hereditary ownership) and performed special armed service as well as service in tax-collection. To hold this type of society and its state together, the best religion is one which emphasizes the role of bhakti, personal faith, even though the object of devotion may have clearly visible flaws.

Innumerable medieval rustic 'hero' stones (27) commemorate the death in battle usually a local cattle-raid-of an individual whose status was above that of the ordinary villager. In older days, the duty of protecting the disarmed villages would have been performed by the gulma garrisoning the locality. The right to bear arms (with the concomitant obligation to answer a call to arms) was now distributed among a select class of persons scattered through the villages. Many inscriptions vaunt the Ganga baron's sacrifice of their own heads in front of some idol, to confer benefit upon their king. More than one epigraph declares the local warrior's firm intention not to survive his chief (28). Marco Polo (29) reported of the 13th century Pandyas that the seigneurs actually cast themselves upon the king's funeral pyre, to be consumed with the royal corpse. This suits the bhakti temperament very well. Though barbarous, it is not the type of loyalty that a savage tribal chief could expect or receive from his followers, unless his tribe were in some abnormal situation.

Though bhakti was the basic need in feudal ideology, its fruits were not enjoyed equally by all. By the 12th century, feudal taxation had begun to weigh heavily upon the peasantry, who paid not only for the luxurious palace but also its counterpart- the equally rich and even more ornate temple. Brahminism had definitely come to the top, as may be seen from two monumental collections of the period, namely the Krtyakalpataru of Bhakta Laksmidhara (minister of Govindacandra Gahadavala of Kanauj, circa 1150 AD); and a century later, Hemadri's quite similar Caturvargacintamani. The latter was chancellor of the exchequer (maha-karan- dhipa) under the last Yadavas of Devagiri (Daulatabad). He is described as the outstanding computer (ganakagrahi). A few tables for quick assessment survive in Hemadri's name; the name is also (wrongly) coupled in Marathi tradition with the general use of bajri as cultivated food-grain, the cursive Modi alphabet, and the numerous close-jointed mortarless Yadava temples that had been built centuries earlier, to develop from little shrines of matchless proportion and balance into rank, clumsy, richly endowed structures by the 12th century. Yet his magnum opus, far from being another Arthasastra, or an Ain-i-Akbri, or an Indian Corpus Juris Civilis, is concerned almost entirely with brahminical rites and ritual codified from Puranas and other accepted religious books. The published seven volumes contain perhaps three fifths of the original. Anyone who performed even a tenth of the ritual rites prescribed for any given deity, lunar date, transgression, celebration, worship, festival or occasion would have no time for anything else; as a document of a superstitious leisure class, none other known today will bear comparison with it. A section on jurisprudence preserved in Laksmidhara's compendium shows that common law was practised and decisions for each caste, tribe and locality based upon their particular custom; but the work repeats smrti doctrine without mention of the innovations in practice, or discussion of a single case.

The real military strength of the Marathas, as later of the Sikhs, derived obviously from the simpler, less caste-ridden, and less unequal life. The later Maratha generals like the Sinde and Gaekwad rose from relatively obscure families, unlike the earlier and more distinguished Candrarao More, Bhonsle, and Jadhav, the last of whom might claim kinship with the Yadava emperors of Devagiri and through them perhaps with Krsna himself. Malharrao Holkar was of the Dhangar shepherd caste, and would normally not have been allowed to rise to the status of a general, duke, and eventually king. It seems to me that some of this goes back, like the bhagva jhenda flag of Maratha armies, to Varkari custom. In spite of the brahmin Badave priests, and the rampant brahminism of the Peshwa days, the Varkari pilgrims minimized caste observances and distinctions on the journey. However, the reform and its struggle was never consciously directed against feudalism, so that its very success meant feudal patronage- and ultimately feudal decay by diversion of a democratic movement into the dismal channels of conquest and rapine.

The conglomerate Gita philosophy might provide a loophole for innovation, but never the analytical tools necessary to make a way out of the social impasse. Jnanesvar's life and tragic career illustrate this in full measure. He does not give a literal translation of the divine message, but its meaning and essence in his own terms, and in words that any Maratha peasant could understand. Jnanesvar's longest comment on the original comes in the 13th adhyaya of the Gita, the chapter on 'the field and field-knower', particularly on G. 13. 7 (where he himself apologizes in J. 13. 314-338 for having been carried away far from the original) and on G. 13. 11. In the former, (J. 13.218-224), he flays the rainrnaking yajanika fire-sacrificers; yet in I. 3. 134-5, these very sacrifices were taken as normal and necessary by him as by his divine exemplar; and once again (G. 18.5; J.18. 149-152) both warn us that the yajna must not be abandoned any more than charity ( dana ) or ascetic practices (tapas). The suffocating contradictions of mixed superstition are neatly brought out in I. 13. 812-822: "The peasant farmer sets up cult after cult, according to convenience. He follows the preacher who seems most impressive at the moment, learns his mystic formula. Harsh to the living, he relies heavily on stones and images; but even then never lives true to anyone of then. He will have My (=Krsna 's) image made established in a corner of the house, and then go on pilgrimage to some god or other. He will pray to Me daily, but also worship the family's tutelary deity at need, and other gods as well, each at the particular auspicious moment. He founds My cult, that makes vows to others; on anniversary days, he is devoted to the ancestral Manes. The worship he gives Us on the eleventh (lunar date) is no more than that he renders to the sacred cobras on the fifth. He is devotee solely of Ganesa on the (annual) fourth; on the fourteenth, says he, 'Mother Durga, I am yours alone' ... At the Nine Nights (of the Mother-goddesses) he will recite the set praise of Candi, serve meals outdoors on the Sunday, and rush off on Monday with a bel fruit offering to Siva's phallic symbol. Thus he prays unemittingly, never still for a moment; like a prostitute at the town gate". In Jnanesvar's society, however, such eclectic worship was the universal practice at all levels, to the very highest people for whom Laksmidhara and Hemadri indited their monstrous compendia. To that extent, though indirectly, the commentator voices a protest against the growth of an oppressive upper class. The Gita doctrine is given a remarkably attractive turn by Jnanesvar's quite original interpretation (J. 9, 460-470): "Ksatriya, vaisya, woman, sudra and untouchable retain their separate existence only so long as they have not attained Me... just as rivers have their individual names, whether coming from east or west, only till they merge into the ocean. Whatever be the reason for which one's mind enters into Me, he then becomes Me, even as the iron that strikes to break the philosopher's stone turns into gold at the contact. So, by carnal love like the milk- maids, Kamsa in fear, Sisupala by undying hatred, Vasudeva and the Yadavas by kinship, or Narada, Dhruva, Aknira, Suka and Sanatkumara through devotion-they all attained Me. I am the final resting place, whether they come to Me by the right or the wrong path, bhakti, lust or the purest love, or in enmity". Neither the callous G. 9.32 on which this charming comment is made, nor the fundamentally brutal Krsna saga manifest such a calm elevation above jealous, exclusive bhakti. Yet, on the very next stanza, the scholiast extols brahmins as veritable gods on earth! His rejection by contemporary brahmins, which must surely have been a main reason for the decision to render the Gita into Marathi, never prevented him from striving after the brahmin vedic lore officially denied to all but initiates. That is, he embodied the inner contradictions which he discerned in contemporary society but failed to discover in the Gita. Therefore, he could launch nomovement towards their solution. Though an adept in yoga as a path towards physical immortality and mystical perfection (cf. J. on G. 6. 13-15), there was nothing left for him except suicide. That the gods remained silent at the unexpected Muslim blow which devastated their many richly endowed temples and no incarnation of Krsna turned up to save the Yadava kingdom, might have been another cause for despair.

The main social problem was violently placed upon a new footing by Alauddin Khilji and the Muslim conquest which imposed payment of heavy tribute. This intensified the need for more effective tax collection; that in turn encouraged a new powerful system of efficient feudalism. Some optimists have maintained that the poorer classes benefited because Alauddin squeezed only the rich, who were rendered powerless. This disingenuous view carefully neglects to mention that even in the Doabs (which were directly administered) none of the former burdens of the peasantry were lifted. Their dues were collected by a different agency, though it remains true that the Hindu upper classes were prevented for a while from imposing fresh exactions. The provinces had not even this consolation, for the throne of Delhi exacted harsh tribute from conquered areas, without troubling itself about how provincial magnates gathered it- and how much more besides. Local military power was reduced only to a stage where it constituted little danger to the imperial forces, but the mechanism of violence more than sufficed for its main purpose, revenue collection. Whether the tribute was actually paid or not, and even over regions not subject to tribute, the imposts and exactions grew steadily. The class that collected the surplus retained an increasing portion, so that the needs of the state could be satisfied only in the earlier period, when feudalism stimulated trade and fresh agrarian production. Then the crisis was aggravated, to be resolved by another foreign conquest that introduced a totally different form of production, the bourgeois-capitalist. The modern independence movement did not challenge the productive form; it only asked that the newly developed Indian bourgeoisie be in power.

Modern life is founded upon science and freedom. That is, modern production rests in the final analysis upon accurate cognition of material reality (science), and recognition of necessity (freedom). A myth may grip us by its imagery, and may indeed have portrayed some natural phenomenon or process at a time when mankind had not learned to probe nature's secrets or to discover the endless properties of matter. Religion clothes some myth in dogma. "Science needs religion" is a poor way of saying that the scientists and those who utilize his discoveries must not dispense with social ethics. There is no need to dig into the Gita or the Bible for an ethical system sandwiched with pure superstition. Such books can still be enjoyed for their aesthetic value. Those who claim more usually try to shackle the minds of other people, and to impede man's progress, under the most specious claims.

Individual human perfection on the spiritual plane becomes much easier when every individual's material needs are firmly satisfied on a scale agreed upon as reasonable (30) by the society of his day. That is, the main root of evil is social. The fundamental causes of social evil are no longer concealed from human sight. Their cure lies not lie in theology but in socialism; the application of modern science, based upon logical deduction from planned experiment, to the structure of society itself. Science is at the basis of modern production; and no other tools of production are in sight for the satisfaction of man's needs. Moreover, the material needs could, certainly be satisfied for all, if the relations of production did not hinder it.

2. R. G. Bhandarkar's Vaisnavism, Saivism and Minor Religious Systems (originally published in 19l3 in the Grundriss d. Indo-Arischen Philologie u. Altertumskunde ; re-issued, Poona 1929, in vol. IV of his collected works) gives a good summary of the influence of the doctrine in the classical and medieval period, but without reference to the historical context, which was indeed not known at the time. Its influence upon Bhandarkar himself led to a petty reformist movement, the Prarthana Samaj (an offshoot of the Brahmo Samaj) in whim RGB was the dominant figure; and support of widow remarriage, then unheard of for brahmins, though practised by some 85% at least of the population. That he spoke for a very narrow class in the attempt to speak for the whole of India never struck him, nor for that matter other contemporary 'reformers'. Still, the silent change of emphasis from caste to class was a necessary advance.

3. In particular, the translation of dhamma as religion, or even a universal Law for all society was "new concept with Buddhism, not accepted even after the time of the G. For example Manusmriti 8.41 reads "The (king) must inquire into the laws (dharma) of each caste (jati), district (janapad), guild (sreni), and household (kula), and only then give his own legal decision (svodharma)". A great deal of the confusion over the Gita derives from ignorance of reality, of the actual practices of large social groups; and from taking brahmin documents as representative of all Indian society.

4. The standard reference work is F. E. Pargiter's The purana text of the dynasties of the Kali age (Oxford, 1913). Some of the theories have been tested e.g. A. B. Keith's review in the JRAS, but the work has survived and gained and deserved reputation for its synoptic edition of the historical kernel in the major puranas.

5. Translated in S. Beal: Buddhist Records of the Western World (London 1884, vol. I, pp. 184.186). The equivalent of G. 237 does occur on p.185, and the association with a great battle at Dhrmasetra, where bones still whitened the earth, is explicit, in an otherwise garbled account.

6. V. S. Sukthakar: The Nala episode and the Ramayana in Festchrift F.W. Thomas, pp. 294- 303, especially p. 302, where he concludes that the two versions bracket the extant Ramayana. The paper is reprinted in his Memorial edition (Poona 1944), pp. 400-415. For the mechanism of inflation see his Epic Studies, VI; and my notes on the Parvasamgraha, in the JAOS. 69.110-117; for the Bhismaparvan and the 745 stanzas of the Gita, ibid 71.21-25.

7. J. Brough : The early Brahmanical system of gotra and pravara (Cambridge, 1953), p. 27, notes that the kevala Angirasas are completely omitted by Hiranyakesi- Satyasadhna, but takes this to be a casual lacuna. So great an omission is highly improbable. My review in JAOS 73.202-208 was mistaken for a polemic, when the point being made was that theoretical works on gotra need to be checked by independent observation. For example, the segrava (=saigrava) gotra found in Brahmi; inscriptions at Mathura is not known to the books. Even more striking are the innumerable local brahmin groups whose confounding to theory has never been tested. City people in Maharashtra take brahmins to be primarily of the Saraswat, Citpavan, Desastha and Karhada groups. The 1941 Census caste tables for Bombay province, as published show that such categories are together outnumbered by the "Other Brahmans" and that local brahmin groups are the rule, though the books and theory are in the hands of the major groups named. The Bhrgus are specially connected with the M'bh inflation, as was shown by V. S. Sukthankar in his magnificent Epic Studies VI (ABORI 18.1-76; Mem. Ed. 1.278-337). It is important to note that the Bhargava inflation was independent of though not hostile to the Narayana inflation, which continued after the first had tapered off. So much so, that the famous benedictory stanza Narayanam namaskrtya of the popular editions drops out of the critical text, but most of the property Bhargava inflations (e.g. needless emphasis upon Parasurama) all remain. In G. 10.25, the Lord reveals himself as Bhrgu among the great sages (maharsinam Bhrgur aham), though that sage occupies no position in vedic tradition, and a trifling one even later.

8. Epigraphia Indica 8.36 ff.

9. Asvaghosa's Buddhacarita and Saundarananda still exist, not to speak of subhasita verses scattered through anthologies in his name. The fragments of a play Sariputra-prakarana were arranged in order by H. Luders, from Central Asian (Turfan) finds. This or another play of the same name was acted by hired actors in Fa Hsien's time in the Gupta heartland, as were also similar plays on the conversion of Moggallana and Kassapa; note that all three disciples and Asvaghosa himself were brahmins.

10. The M'bh diaskeuasts proclaim their desire to include everything. In M'bh 1.1-2, the work is successively an itihasa, a purana, an upanisad, a veda, and outweighs all four vedas together. It is the storehouse for poets. M'bh, 1.56.33 boasts: yad ihasti tad anyatra, yan nehasti na tat kva-cit: whatever is here might be elsewhere, but what was not here could hardly ever be found!

11. Translated by K. T. Telang, see Note 1. There is an Uttaragita, quite modern apocryphal work.

12. This is the second sutta of the Digha-nikaya, and has served as the model, in many ways, for the later Milindpanho, questions of king Menander.

13. This was clearly stated by Mr. C. Rajagopalachari in a press interview.

14. The only published source I have been able to locate for the original cult is the Satya Pirer Katha in Bengali by Ramesvara Bhattacaryya (ed. by Sri-Nagendranath Gupta, Calcutta University 1930) .

15. This paragraph and the next are treated in greater detail in a paper of mine on the avatara syncretism and possible sources of the Bhagavad- gita, ]BBRAS. vol. 24-25 -1948-9), pp. 121-134.

16. RV. 8.96.13-14, but sometimes interpreted mystically as part of the Soma legend. The traditional explanation is that this Krsna was an Asura "i.e. non-Aryan, and the fighting against Indra on the banks of the Amsumati river was real, not Symbolic of something else.

17. Ed. K. de Vreese, Leiden 1936. This particular naga cult had been virtually killed by the Buddhist monks (Rajatarangini 1.177-8), while the brahmins had also been reduced to helplessness at the time of the Buddhist teacher Nagarjuna. They made a come-back by writing the Nilamatapurana (Raj. 1. 182-6) , Kalhana informs us in passing.

18. ABORI 1.59-66; ]RAS 1909.1055-6, 1087-92; 1910.813-5, 815-7.

19. See JRAS 1960; 17-31, 135-144, or chapter IV of this book; for the cave painting (originally discovered by Carlleyle) Mrs. B. Allchin in Man, 58.1958, article 207 + plate M (pp. 153-5).

20. The assurance is unmistakable: " Whenever true belief (dharmra) pales and un- righteousness flourishes, then do I throw out another offshoot of myself". The next stanza proclaims that the god comes into being from age to age, to protect the good people, destroy the wicked and to establish dharma.. It need not be further emphasized that the superfluous incarnation in M'bh times wasted a perfectly good avatara, badly needed elsewhere.

21. The cult is coeval with the foundation of Parvati village, hence older than the Peshwa temple to the goddess who killed that demon. Cf. Bombay Gazetteer vol. 1.8, pt. 3 (Poona District), p. 388.

22. Art In The Ice Age by J. Maringer and H. G. Bandi after Hugo Obermaier (London 1953); especially figures 30, 31, 70 (with mask, and arms imitating mammoth tusks), 142, 143 and perhaps 166.

23. This shows in Harsa's inscriptions (e.g. Epigraphia Indica 7.155-60); benedictory verses at the beginning of his Buddhist drama Nagananda, addressed to Gauri; Bana's description in the Harsacarita and Hsiuen Chuang's account (Beal 1.223; the stupa, vihara, fine Mahesvara temple and the Sun-temple were all close together near Kanauj, and all constantly thronged with worshippers) .

24. For the iconoclasm of Harsha of Kasmir, Rajataratigi 7.1080-1098. He had predecessors of similar bent, though less systematic: Jayapida in the 8th century (Raj. 631-3 ; 638.9) and Samkaravarman (5.168-70) in 883-902 AD.

25. The Gupta gold coinage is impressive, but hardly useful for normal transactions. Their silver coinage is notoriously inferior to, say, pre-Mauryan punch-marked coins, and rather rare in hoards; of Harsa, only one coinage is known, and even that rather doubtful in silver. The Chinese travellers Fa Hsien and Hsiuen Chuang are emphatic in the assertion that most of the transactions were barter, and that cowry shells were also used, but very little currency. The accumulation by temples, monasteries and barons did nothing for the circu1ation of wealth or of commodities.

26. This is discussed in a paper of mine to appear in the Journal for the Economic and Social History of the Orient (Leiden), on feudal trade charters. Yasodharman of Malwa uses samanta as neighbouring, ruler, whereas Visnusena (a Maitraka king) issued a charter in 592 AD where samanta can only have the feudal meaning.

27. The hero-stones carved in bas- relief are to be found in almost any village not recently settled, throughout Maharashtra and the south. A good collection is in the National Defence Academy's Museum at Kharakwasla, near Poona. The death in fending off cattle raiders seem to be symbolized in many cases by a pair of ox-heads in the lowest panels. The story progresses upwards, to the funeral, perhaps with a sati, and going to heaven. The top of the relief slab is generally carved in the semblance of a funeral drum familiar since Buddhist days. For inscriptions, even a single volume (Epigraphia Carnatica X, for example): Kolar 79, feudal grant for family of baron killed in battle (about 800 AD); Kolar 226 (circa 950 AD), grant of a field, on account of the death of a warrior fighting against cattle raiders; Kolar 232 (150 AD), Ko!ar 233 (815 AD), Mulbagal 92, 700 AD : Mulbagal 93, 970 AD etc., with the hero-relief in every case.

28. Less well known than Ganga inscriptions are the minor ones showing how widely the custom was spread: e.g. from the Ep. Carnatica, Goribindnur 73 (circa 900 AD), the village watchman sacrifices his own head; Cintamani 31 1050 AD), when the Odeya of the village went to heaven his servant had his own head cut off -and a field was dedicated to his memory; oaths of not surviving the lord are taken in Kolar 129 (circa 1220 AD), Mulbagal 77 (1250 AD), Mulbagal 78 etc. Occasionally, a memorial was erected to a particularly able hound, as in Mulbagal 85 (975 AD), and Mulbagal 162, though the dog's prowess rather than bhakti is praised.

29. Penguin Classics L 57, Travels of Marco Polo (trans. R. E. Latham), pp. 236-8, for the cremation, and ritual suicide in front of some idol, by royal consent.

30. By 'society' is meant not only the rulers but the ruled. If the sudra should agree that he ought to starve for imaginary sins committed in some supposed previous birth, either his group will die out, or at best be unable as well as unwilling to fight against invaders. Indian feudal history, however, is full of raids and counter-raids, not only by Muslims. It follows that the expropriated class will not show by its actions that they regard the expropriation as reasonable on religious grounds, particularly when they see the very same religion unable to defend its proponents against armed heretics. My point is simply that the fulfillment of certain material needs is as essential to health of the mind as it is to that of the body. It seems to me that the Gita philosophy, like so much else in India's spiritual heritage, is based in the final analysis upon the inability to satisfy more than the barest material needs of a large number.