The epic of the Bharatas


back to issue

THE Mahabharata calls itself itihasam-puratanam – thus indeed it was in times past. This is not a Rankean statement and there are no definitive claims to historicity of persons and events in the narrative. However, there is a hint that some of the narrative may have been an attempt to cull from the remembered tradition that which may have happened, even if what is culled is disordered in the retelling. The Ramayana, however, is most often described as a kavya, in fact the adikavya, suggesting a small distinction between the two.

There have been many commentaries in Sanskrit on the epics from the eleventh to the seventeenth century examining their meaning and intention. An analysis of these could tell us about how the epic was perceived in periods prior to ours. One wonders for instance, whether historicity as we understand it today, was of concern to the authors and to the audience of past times. Undoubtedly among the more attractive features of the epic, although historically tantalizing, are its many enigmas. Some however suggest the occasional facet of the historical past. The attempt in this essay is to point to these and indicate the problems they raise.

Epic as a genre uses narrative to represent situations from the past. It looks back nostalgically from the point in time when it is being composed and given form which is its present. The nostalgia is for a past age of heroes and the clans to which they belonged. This slowly gave way to the present where the heroes are less important in a society governed by kings and the code of castes. The nature of authority is more focused and therefore different in kingship and the determining of status and social attitudes by reference to caste gradually becomes predictable.

Given that it spans more than one kind of society, re-examining the concepts we use in interpreting the epic becomes essential. Some are inflected by the translations we use. The context of the epic was that of clan societies and small-scale kingdoms. It is as well to remember that the term ‘raja’ did not in origin mean a king, but referred simply to ‘the one who shines’, an appropriate title for a chief. As such it is likely to have been continued when the system changed to kingship. To translate samrat as ‘imperial monarch’ is to impose a later meaning on an earlier term. A more accurate meaning might refer to the one ‘who has authority over many’. The context of the term can illumine its meaning.

Part of the transition from clan to kingdom also lay in the evolving form that was being given to varna and jati. The structure of clan societies did not require that they follow the rules of the dharma-shastras but a king ruling a kingdom had to uphold caste. There were of course exceptions such as Ashoka who does not mention caste in his edicts, but this was doubtless due to his heterodox thinking.

The transition is vividly etched in the different nuances of how the clans functioned. The Vrishnis in Dvaraka in western India with their eighteen kulas, extended families/clans, conform to a distinctive way of functioning reminiscent of oligarchies and chiefships. The Kauravas and the Pandavas battling for territory in the Doab and the western Ganga plain still observe the rules of clan ethics and codes of kin relationships but are seen to be slowly succumbing to the rules of kingship and the codes of caste, the pattern emerging in the easterly region of Magadha.


Authorship of the Mahabharata is attributed to Krishna Dvaipayana Vyasa, a brahmin of uncertain antecedents, born of the sage Parashara and a fisher woman, who is also said to have sired Dhritrashtra and Pandu. Authorship and fatherhood are coalesced and in effect Vyasa is narrating the lives of his own sons and grandsons. They in turn have no blood connection with the lineage of the Bharatas although they are constantly referred to as ‘bull of the Bharatas’. The blood connection actually ended with Bhishma, unless of course his being called pitamaha – the paternal grandfather – is taken literally, despite his vow, but in which case the story would have been different.

Genealogies can be accurate over a few generations but are also available to those who wish to fabricate connections and latch on to a respectable ancestry. This fragment of the lineage has an equally ambiguous end since the lineage continues through Parikshit, who is stillborn but revived by Krishna. Possibly the invented connection was necessary since in the Vedic corpus the epithets used for the ancestor Puru, are not the most complimentary: his speech is said to be mridhra-vach, impure, and he has an asura-rakshasa ancestry.


The epic recitation had two beginnings a generation apart. It was first recited at the snake sacrifice of the kshatriya raja, Janamejaya (the son of Parikshit), by the brahmin Vaishampayana. A generation later it is recited at a sattra, a sacrificial ritual usually intended for and performed by brahmins, although the recitation this time is by the bard Ugrashravas. A brahmin recites it to the kshatriyas and a non-brahmin to the brahmins, an inversion which is curious. The epic as we know it was in origin the bard’s memorization of what he had heard from the brahmin’s recital, and there is an insistence that both renderings are exactly as composed by Vyasa. One is immediately suspicious about changes and interpolations. The interleaving of brahmin and bard as authors or the likelihood of bardic origins is not unknown.

The dasharajna, the battle of the ten rajas, described in the Rigveda has been viewed by some modern scholars as the seminal event for the war at Kurukshetra. Some of the clans from the Vedic corpus such as the Bharatas, Purus, Yadus and Kuru-Pancalas reappear in the Mahabharata as lineage ancestors of the epic protagonists, but not necessarily in the same situations. Panini in the fourth century BC refers indirectly to grammatical constructions associated with words such as mahabharata, yudhishthira, arjuna and vasudeva.


Epic personalities have parallels in the Buddhist Jataka stories where persons with the same names occur in events that are often dissimilar to the epics. The secreting away of Krishna as a baby, his hostility as a young man to his uncle Kamsa at Mathura and the migration of his clan to Dvaraka, is part of the narrative of his clan, the Andhaka-Venhu/Vrishni. Its end came through a drunken massacre, a story repeated in the Vaishnava texts. Neither Krishna nor his clan is particularly attractive in these stories. Draupadi is depicted as a nymphomaniac who is not satisfied with five husbands and takes a hunchback lover. But as a contrast to this, stories of the descendents of Yudhishthira ruling the Kuru realm from Indraprastha and advised by the minister Vidhura-pandita, are characterized by what is called Kuru righteousness, particularly noted for its practice of virtue. Whether these variant versions were taken from the epic or contributed to the making of the epic remains a debated subject. The ideological underpinning also differs.

The stories are not identical but may go back to a common source. The boxing-in of stories as in the epic is also a technique of adding to a narrative. Subsequently when they were given an ideological gloss they served the needs of Buddhist ethics and later of Jaina versions, or the sectarian beliefs of Vaishnava Bhagavatism, each in disagreement with the other. The diversity of these initial fragments seems to be echoed in the varied regional versions and recensions of the epic, a process that continued through the centuries. This multiplicity makes it necessary sometimes to identify the version one is referring to.

The composition was doubtless in the nature of a slow accretion as is characteristic of the epic genre virtually anywhere. Nevertheless at some point it was brought together. Tradition has it that a small text, the Jaya, became larger, the Bharata, and the larger became still larger, the Mahabharata. It is, therefore, virtually impossible to calculate a date for the events and to define an ‘epic period’ in ancient history as was once done.

This would also be one reason for the difficulty in identifying a particular archaeological site as the archaeological equivalent of a location in the epic. Excavations have been conducted at locations that are called Hastinapur and Indraprastha. These could be ancient sites although we know that place names also travel especially with migrant populations. Distances on the ground do not always tally with the text. In today’s identifications these two places were a considerable journey away from the battlefield. Equating a site with a text can raise problems of reconciling material culture with evocations of poetic licence. Will we ever find the fantasy palace of the Pandavas built by Asura Maya/asura maya even if we believe it existed and dig up the whole of Indraprastha? Homer’s epic met with similar problems after the extensive excavations at Troy and other sites.


The epic maintains that the war at Kurukshetra took place on the cusp of the Kaliyuga. This has been dated as equivalent to 3102 BC. Dates calculated on the basis of the genealogical lists of kshatriya lineages in the Puranas generally work out to between 1200-1000 BC. The archaeological evidence of a heavy silt deposit at the site of Hastinapur which has been linked to the epic reference to a flood after the war points to a date of c. 800 BC. Mention of Yavanas, Shakas, Hunas, would be interpolations in the late BCs and early centuries AD.

V.S. Sukthankar, to whom the Critical Edition of the epic owes much, argued that its composition ranged from 400 BC to AD 400. More recently scholars have suggested that the oral epic was put together as a text in about 150 BC and that this may have taken a century or so. Attempts have been made to correlate readings in astronomy with references to planetary configurations in the epic. These differ and would relate not to the text as a whole but to particular segments carrying the reference. Diversity in dates also makes it problematic to attribute authorship to a single author. The meaning of Vyasa is interestingly, one who edits and arranges.


To add to the complexity it has been argued that the epic had two intentions. The earliest narrative was constructed out of fragmentary stories of the heroic: tales of combats, marriages, games of dice, exile in the forest – and were probably formulaic on occasion. The same process has been noticed in the Homeric epics. Exile is a marvellous background for the bard as it can be stretched with add-ons heightening the reach of the imagination. Heroic exploits can be enlarged with every recitation, whether the hero is wandering across the ‘wine-dark seas’ of the Mediterranean as in the Odyssey or in the tangled forests of northern India as in the Mahabharata. Exile to the forest in India reinforced the dichotomy of grama, settlement, and aranya, forest, central to cultural perceptions.

It was probably the immense popularity of the epic, both as a linear narrative of the heroic as well as in the embroidered intricacies of the stories, that led to the second intention and reformulation – its conversion into a Bhagavata text. Sukthankar argued that this was done by the Bhrigu brahmins who also ‘Bhriguised’ the Ramayana. Both Krishna and Rama became avataras of Vishnu and this changed the character of the epic. It is worth noting that on the second occasion of its recitation the bard was required to proclaim the descent of the Bhrigu lineage, perhaps to legitimize the Bhrigus and their appropriation of the epic.

The Bhrigus, were often linked to the Angiras brahmins, associated with the Atharvaveda and among other things, were regarded as the practitioners of sorcery and magic. Neither group was the most highly respected among learned brahmins. The Bhrigus were called brahma-kshatra, when they married kshatriya women. Their learning extended into knowledge and custom beyond the conventional and in addition they were said to know niti and dharma.


The question could be asked that if the Bhrigus were not pre-eminent brahmins why were they permitted to reformulate the epic. Epic origins lay in popular compositions not in divine revelation; therefore they were not as sacred as the Vedas. The Bhrigus may have wished to convert the epic into a Bhagavata sectarian text, if they were associated with early Bhagavatism. The didactic section was the palatable way of teaching the various brahmanical dharmas. The Mahabharata can be viewed as a civilizational text not because it reflects the propagation of a particular view of these dharmas but because, among other things, it speaks to the debate on social ethics, especially between the brahmanical perspective and those that question it – a debate that has continued over many centuries.

The Bhagavata religion was not identical with the Vedic even if the Mahabharata is referred to as the fifth Veda – as indeed was the itihasa-purana itself and many other bodies of knowledge. The Bhrigu brahmins may have had associations with pre-Vedic and non-Vedic ideas as has been suggested. In one Upanishad the Bhrigu-Angirasas are linked to the itihasa-purana. This would, up to a point, make them appropriate editors of texts pertaining to the past. Nevertheless the controversy on ‘Bhriguisation’ continues.

Why the Bhagavata reformulation was necessary needs an explanation. Perhaps the realization that the Buddhist gloss on popular stories was a successful way of propagating sectarian belief may have encouraged Bhagavatas to do the same. Buddhism was subordinating the worship of clan deities to the higher ideal of the ethic of dharma as defined by the Buddha. Bhagavatism required the worship of Vishnu as the supreme deity and the varna dharma was its social ethic, both of which incidentally reinforced the requirements of kingship. This subsequent addition has left a heavy imprint on the Mahabharata and the epic genre of the text has tended to be subordinated. It is to this genre that I would like to give more space.


The first few books narrate the epic of the Kauravas and the Pandavas and take the story, together with whatever is tagged onto the narrative, to the point where war is imminent. The description of person and event veers towards the functioning of a clan-based society. It conforms in the main to the pattern of a segmentary system of lineages functioning as segments of an extensive network. The Puranas a little later called it the Chandravamsha, the Lunar Lineage. Identity is through being born into a clan, kinship controls behavioural relations and social functions, governance is through an assembly of the heads of families and status is relatively egalitarian within the clan.


Agro-pastoralism is the major source of income with an emphasis on cattle herding and the occasional cattle-raid, described more than once in the epic. The Vedas also refer to the Kuru-Pancala clans going out in ‘the dewy season’ to raid cattle. The more spectacular sacrificial rituals asserting the political authority of the patron, such as the rajasuya performed by Yudhishthira, are occasions for gathering-in tribute and exchanging gifts. These are largely in the form of the produce of hunting and herding, of weaving textiles, of mining gold and gems and maintaining domestic slaves – mainly women. Agricultural activity is of course a necessity but tends to be low key. Dependence on agriculture increases in the later parts of the text. Items of wealth placed as stakes in the game of dice are similar but of larger amounts. Ultimately Yudhishthira stakes the town, the territory and eventually himself.

The listing of wealth is in substantial if not exaggerated terms. Such occasions have been thought to be rituals but also gift-giving ceremonies, characteristic functions of heads of clans. On special occasions, the chief collects, consumes, distributes and if need be even destroys what remains of his wealth. The rajasuya was one such occasion with inevitable consequences. The intention is to assert status which in turn leads to a competition among clan chiefs each trying to outdo the previous one. Therefore gift-giving is not a one-way process. It assumes that the holding of these ceremonies turn by turn ensures the circulation of whatever is produced. It also prevents a chief acquiring excessive power through accumulating enormous wealth. Should this happen conflict may be unavoidable. The items brought to the yajna, come as gifts and tribute and not as tax.

The clans were described as kshatriyas, with status dependent not on caste but on the hierarchy among the clans. Identity came from the clan. Hence the emphases on genealogies, all of which may or may not have been taken literally, the social forms reflected in their structure also being significant. Breaks in the genealogy can be indicators of change.


That the rules of caste were not strictly observed would explain why three disparate systems of marriage are adopted over two consecutive generations in the family of Pandu. In Pandu’s marriage to Madri it would seem that a bride-price was involved, even if not in material goods, making it an asura marriage according to the dharma-shastras. The fraternal polyandry of the five Pandavas marrying Draupadi is outside any dharma-shastra scheme and is discussed at some length in the epic before it is accepted. Draupadi questioning Yudhishthira’s right to stake her in the dicing match is not the kind of statement that matches the patriarchal values advised for wifely behaviour in the social codes. The three women who command the narrative are Draupadi who poses the question of the legality of Yudhishthira’s right over her, Kunti who chooses the deities she wants as surrogate husbands and Gandhari who insists on being permanently blindfolded after her marriage to a blind husband.

The third variant form was that of Arjuna’s marriage to Subhadra (cross cousin) which is a rakshasa form according to the dharma-shastras yet the lineage continues through their stillborn grandson, Parikshit. Such flexible social practices suggest societies where alien custom could be incorporated to accommodate a new situation.

Krishna’s clan, the Andhaka-Vrishni had a lower status than the Kurus. Political functioning among them lay in the sangha, assembly, where the senior kinsmen of the eighteen kulas sat to take decisions. In kingdoms, the assembly was reduced to an advisory body whose views were not binding on the king. The concentration of power required in kingship perhaps accounts in part for the hostility of the Vrishni clans towards Magadha. The kingdom of Magadha was a challenge to the gana-sanghas, the clans of the middle Ganga plain. Eventually the clan confederacy of Vaishali was destroyed through devious means resulting in the consolidation of kingship as a polity.


Beyond the clans was the non-caste ‘Other’, treated by the heroes as virtually bereft of human value, as is apparent at various points of the narrative. These were the people of the forest such as the Nishada and the Shabara, the excluded and impure mleccha and therefore dispensable. The episode that has been commented on is that of Ekalavya, who being a Nishada, had to give his thumb as a fee to the brahmin guru thus terminating his skill as an archer. But equally traumatic is the reference to how a Nishada woman and her five sons were left in the house of lac, which was set on fire to mislead the Kauravas about the presence of the Pandavas. As a comment on clan society this requires explanation, unless it can be argued that such episodes were introduced later when the mleccha were treated as less than human.

The war at Kurukshetra forms a substantial section of the epic. As a time-marker it touches many dimensions. It marks the end of clan societies. Krishna’s comment is telling: sarvam kshatram kshayam gatam – it is the destruction of all the kshatriyas. It is said that the end of the war marked the end of the Dvapara age – the third of the four ages – and the start of the new and final Kaliyuga. The theory of the four yugas is referred to in various texts of the post-Mauryan period and was probably not in the early epic. It seems to have been a later reflection on a substantial historical change expressed literally as the coming of a new age. As with exile, war is another occasion for the bard-poet to show his skill, so the narrative extends over many parvans, books, of the epic. Despite references to formations of standing armies familiar from warfare in later periods, the battle was essentially heroic warfare and often single combat – thus allowing for some of the dubious manoeuvres attributed to Krishna’s advice.


That the war was pushed by Draupadi’s demand for revenge is an epic-heroic feature. So too was the fact that succession was contested by the absence of unchallenged primogeniture, both brothers being physically disabled. It was probably in origin a clan conflict involving those that had claims on the territory and their friends or enemies. Enlarged into a massive eighteen-day event it was said to have involved clans from all over. That this was an exaggeration is suggested by Arjuna arguing the futility of war prior to the event and Yudhishthira doing so after the event. Validating violence becomes a necessity and even the simile of the battle being a yajna, ritual of sacrifice, is used as it was in later times as well.

Although the Bhagavad-gita in the narrative is placed just before the start of the war, its teaching seems more appropriate to the society that emerged after the war. Arjuna, dismayed by the thought that he would have to kill his close kinsmen, questions the ethics of such an act. The killing of kinsmen seems to have been more heinous in a society where kinship was a primary identity than in a society where kin ties were subsumed in caste. Krishna speaking from the perspective of a caste society explains to Arjuna that as a kshatriya it is his svadharma, the social obligation of one’s caste, to fight against evil even if it means killing kinsmen. Is the moral dilemma being subordinated to caste duty?


As has been pointed out in recent studies the urgency of this discourse most likely had to do with the current debate on ahimsa provoked by Buddhist and Jaina teaching and by the alternative ideal of kingship propagated by Ashoka. In the latter the social ethic is not dependent on caste but on the quality of human behaviour as encapsulated in dhamma. This may have been the context to Yudhishthira questioning the kshatriya model. Judging by inscriptional evidence, support for Buddhism and Jainism at all levels of society was extensive in the post-Mauryan period and there would doubtless have been at least ideological confrontations with the code of the dharma-shastras. Is the Shantiparvan then a polemic in the debate with the heterodoxy? Yet there is also the play on Yudhishthira being the son of the deity Dharma which would have enhanced his sensitivity to the question of ethics.

The later part of the epic, sometimes described as the didactic sections, although these were scattered briefly elsewhere as well, begins effectively after the end of the war and consists substantially of the much-quoted Shantiparvan and Anushasanaparvan. These are important in themselves but in some ways represent a discourse outside the epic. To discuss the epic largely on the basis of these sections as is sometimes done is to do it an injustice.

Yudhishthira, appalled by the violence of the war is reluctant to assume kingship and wishes to retreat to the forest. Bhishma wounded in the war and lying on his bed of arrows persuades Yudhishthira not to renounce kingship and instead to govern as an established king in a kingdom. His long peroration on various categories of dharma and even on situations where it fails is significant in these political transitions. Kingship as a political form is viewed implicitly as superior to what came before or which continued as an alternate system. The discourse by Bhishma is almost certainly a later interpolation of the time when the gana-rajyas, chiefships and oligarchies, had declined – but not disappeared – and in many places had been replaced by kingdoms.


The discussions on rajadharma – the code for kingship and administration – on punishment, on times of distress, run parallel to themes in texts such as Kautilya’s Arthashastra and the Manu Dharma-shastra; not to mention the continuing memory of the Mauryan state. Kautilya’s list of what constitutes a state system, assumes a kingdom ruled by a king with decision-making powers, through an administration manned by non-kinsmen, located in a capital city to which the revenue comes via taxes, and where the kingdom is identified by demarcated territory, defended by a regular standing army and where other kings are allies. These are features more familiar from the later books.

The didactic sections were in a sense looking back at the epic past, but were legitimizing the coming change to kingdoms. The change was not linear and determined. It was somewhat meandering with various offshoots and in some cases earlier forms may have continued as we know that they did historically until much later. But that such a change was represented in the eventual version seems apparent. What I am suggesting is not a dichotomy between clan-society and the kingdom, nor a textbook version of each, but the difference in societies with more of one than the other.


This essay is only one passing historical perspective of the Mahabharata. Even from historians alone there are many more. And beyond them the perspectives multiply still further. The narrative of the epic is sequential, set in a frame of linear chronology, nevertheless there are substratum layers of structures, order, legitimation, and claims that seem extraneous to the narrative but are insightful. What is perhaps being indicated in all this is that historicity should not be sought for only in person and event for it may lie at a deeper level.

In the interface of the two kinds of societies that I have sketched, partially sequential and partially concurrent, there lies what might be seen as a historical tradition – the tradition of a later society remembering and reconstructing what it believes to be the earlier one, where the reconstruction becomes the perceived past. Perhaps it is because of this that the Mahabharata can call itself, itihasam puratanam.


* My thanks to Kumkum Roy, Kunal Chakrabarti, Robert and Sally Goldman for comments on an earlier draft.



V.S. Sukthankar, On the Meaning of the Mahabharata, Bombay, 1957/Delhi, 1998.

R. Goldman, Gods, Priests and Warriors: The Bhrigus of the Mahabharata, New York, 1977.

R.N. Dandekar (ed.), The Mahabharata Revisited, New Delhi, 1990.

J. Brockington, The Sanskrit Epics, Leiden, 1998.

A. Hiltebeitel, Rethinking the Mahabharata, New Delhi, 2002.