This essay has been extracted from The Historian and her Craft Collected Essays and Lectures of Romila Thapar recently published by the Oxford University Press and republished here with permission.
Image courtesy Governance Now
The historian’s interest in epic literature arises from the argument that epic literature generally documents a period which is prior to that of early recorded history. The epic does not exclude other data where it may be available but tends all the same to intrude on the consciousness of the historian claiming to be the primary source. However, the epic is essentially a literary crystallisation of the heroic ideal. By its very nature, therefore, it is not to be taken as factual evidence but as the representation of an ideal—a caution which some historians either ignore or forget when using the epic as source material.
Epic literature is intrinsically a part of the oral tradition, composed, compiled and collated over many centuries from bardic fragments and enriched with interpolations of later times. 1 In the Indian tradition, shorter compositions such as the gāthās, dānastutis, nārāśaṁsīs and ākhyānas were the precursors of the epic. That the Mahābhārata was put together at various times is clear from the internal evidence, for in one case it commences with Vaiśampāyana, the disciple of Vyāsa, reciting the epic at the Janamejaya sacrifice and in another with the recitation by the sūta Ugraśravas at the sacrifice of the ṛṣi Śaunaka which took place a generation later than the first occasion. An earlier and shorter text than the present one also appears to have been known. The Aranyaka-parvan provides an excellent illustration of the adding on of bardic episodes to the main narrative. Ultimately the entire collection was redacted, probably by Bhṛgu authors, and this imbues the text with a relative unity of theme and literary form.
The making of the epic inevitably creates problems for the historian, who, in the past, has tended to treat the epic as a single text of a particular period and has attempted to test it for historical authenticity. Epic events can rarely be precisely dated nor can one speak of an epic period nor indeed can the narrative of the epic be treated necessarily and without cross-evidence as authentic history. What is of significance to the historian are the assumptions made about the past in the epic and about the changes which earlier societies have undergone, assumptions which eventually will lead us to understand the historical function of the epic rather than to limit our exploration to its historicity. This is especially relevant with the Indian epics since one of the characteristics of Indian social history is the extensive range of social formations which co-existed in all periods, even to the present. The appeal of the epic varies in accordance with the particular society to which it relates at particular points in the text, and it is in the sifting of these segments that the historian can contribute. In such a situation even the critical edition may eventually be found to require some further pruning.
Historians have made many attempts to date the culminating event of the epic, the war between the Pāṇḍavas and the Kauravas. The suggested dates range from the fourth millennium bc to the seventh century bc and the evidence for these draws on a variety of sources, astronomy, archaeology and genealogical history; but none of these dates are universally acceptable. To search for a conclusive date is in a sense to search for the impossible since the purpose of epic literature is not a concern with a chronological sequence of events. Even the central event may not be historical since it is pivotal to the action of the epic but not necessarily to the reconstruction of history. In the vision of the bard, the event, which may once have been historical, can transcend its historicity and become symbolic. It would seem to me, therefore, that it is more relevant for the historian to analyse the assumptions of the epic which would undoubtedly provide an insight into the society of a people, which insight may not be chronologically exact but which would nevertheless enrich our understanding of the past.
Many decades ago scholars argued that there were at least two traditions incorporated into the epic, the original epic and the pseudo-epic. One was the earlier, narrative layer reciting a series of stories based on bardic material. The other consisted of a number of didactic sections relating to the rāja-dharma , mokṣa-dharma and the practices of different sections of society, which drew ostensibly on the dharma-śāstra literature and which were interpolated into the epic at later periods. The rationale for these interpretations would be that with the conversion of the epic into sacred literature, it was necessary that it also incorporate discussions on ethical norms and the definitions of authority both temporal and sacred. I would like to support this distinction and argue that there is a difference in the depiction of society in these two layers and that difference is important to the historical understanding of the epic.
The narrative sections seem to depict societies of tribal chief ships moving towards the change to a state system with monarchy as the norm. There is a strong emphasis on lineage rights and functions and a fairly flexible inclusion of a variety of kinship forms. The economy tends to be pastoral-cum-agrarian in which cattle raids and gift-exchange are important components. Heroism is wrapped up in the defence of territory and the honour of the kinsfolk. The didactic sections in contrast assume a highly stratified society with frequent reference to caste functions rather than lineage functions. The political system assumes well-established monarchies and an increasing concentration of authority in the hands of the king. The economy is essentially agrarian with a familiarity with urban centres as commercial units. Gift-giving involves the granting of land in addition to other forms of wealth. In contrasting the diff erences between the two sections it can be argued that since the evidence from the narrative section is drawn from the actual happenings as described and the information which they provide on social formations is gathered through inference by the historian, there is, therefore, less likelihood of deliberate interpolation in the text. The didactic sections concentrate on the legalities of social functioning and leave little to historical inference. I would like to examine what appear to me to be two diverse traditions in the light of archaeological, anthropological and historical evidence.
Although the epic story finally involves the clans and kingdoms of virtually the entire northern half of the subcontinent, the core of the narrative focuses on the Gaṅga-Yamuna doāb and its vicinity and more particularly the upper doāb. The territory of the Kurus extended from the Indo-Gangetic divide into the upper doāb with capitals at Hastināpura and Indraprastha. That of the Pañcālas, who play an important part in the narrative, lay in the adjoining areas to the south-east. The proximity of the Kuru and Pañcāla would suggest both hostility and close alliances and the later Vedic texts speak of an eventual alliance. This would not be incompatible with the evidence of the epic in that the Pāṇḍavas were the successors to the Kuru realm and had a marriage alliance with the Pañcālas.
The geographical location provides another possible dimension of evidence which is gradually being brought into the discussion in the form of archaeological data. The earliest settlements in the area date to the start of the second millennium bc. There is evidence now to argue that the Harappan tradition did bequeath some survivals to the later cultures as seen at sites with late Harappan and Painted Grey ware settlements. The earlier Ochre Colour Pottery Culture although indigenous to the region is by all accounts a relatively primitive culture of small settlements and inferior living standards. Its connections with the Copper Hoards which would indicate a more advanced society are by no means established. Undoubtedly the dominant culture of the region is the Painted Grey Ware or PGW variously dated from the late second or early first millennium bc. Attempts have been made to identify this culture with later Vedic literature on the basis of common features of material remains as well as with the Mahābhārata by referring to common geographical locations. 6 Excavations at Hastināpura and Indraprastha show a flourishing PGW level which in the case of the latter is also the earliest occupational level. It has further been argued that there is evidence of a flood at Hastināpura which terminates the PGW level (which at this site dates to about 800 bc ) and that this may be the flood referred to in the epic as having occurred during the period of Nicakṣu when the city was deserted and the capital shifted to Kauśāmbī.
The distribution of the PGW generally conforms to the core areas occupied by the clans and kingdoms who participated in the events of the epic. There is a noticeably larger number of settlements and the demographic increase may have had to do with an improvement in material culture. The settlements are clearly pre-urban, consisting of huts of wattle and daub with rarely any structural remains of a substantial quality. Of the material remains the most striking is the occurrence of iron artefacts, mainly items of warfare and a few used in agriculture. The presence of weights would point to some minimal degree of exchange other than primitive barter. There is a profusion of animal bones particularly of horse and cattle which judging from the literary data formed the major units of wealth. The widespread occurrence of horse bones is new to the Indian scene since horse bones from late levels of Indus civilization sites are scant. The quantity of cattle bones is also larger for this period. Many which come from domestic contexts often carry marks suggestive of the eating of cattle flesh for food. There is a striking absence of graves or burials which is in sharp contrast to the earlier Indus civilization as well as the contemporary megalithic cultures. It would seem that cremation was the normal form of disposing the dead in the PGW culture.
The culture which succeeds the PGW is the Northern Black Polished Ware culture or the NBP, dated to the sixth century bc and continuing into the Mauryan period. Here there is evidence of a qualitative and quantitative change from the preceding PGW culture. The demographic picture shows larger concentrations of populations at single sites. The mud plaster huts are replaced with well-defined structures of mud-brick and burnt brick. Brick lined drains, soakage pits and terracotta ringwells indicate a civic life of no mean proportions. There is a substantial increase in the use of iron artefacts many of which are now employed commonly in domestic use. Above all there is also the presence of punch-marked as well as uninscribed cast coins indicating a major change in economic exchange relations involving the use of coined money and an incipient commercial economy.
For the historian, the archaeological co-relation poses a dilemma. If the material culture of the epic is co-related with the earlier PGW culture with which the narrative sections of the epic seem to agree to some extent, then the date of these sections can be placed between the mid-second and the mid-first millennia bc but the culture will have to be described as pre-urban, transitional between pastoralism and an agrarian economy and probably supporting tribal chief ships on the edge of change to state forms and monarchical systems. The excavated evidence of these levels at Hastināpura and Indraprastha in neither case suggest the splendours of great kingdoms with wealthy capitals; rather, they were people with a technologically unsophisticated culture. The elaborate descriptions of material culture with references to a developed agrarian economy and prosperous towns as given in the didactic sections of the epic would perforce have to be dated to periods later than the mid-fi rst millennium bc in any co-relation with archaeological evidence.
Archaeological continuities can, in some instances, be connected back to the Harappan period in the area under discussion. This makes it possible to suggest that some at least of the traditions recorded in the epic could also, in origin, go back to this period. But this does not mean that the epic or the main events date to the Harappan period.
It would seem apparent, therefore, that the historian would have to argue for different parts of the epic having been composed at different periods and integrated into the text at a relatively late point in time. The basic question is again not that of the date of the epic but the elucidation of the fragments which have been stitched, as it were, into the epic.
Read more extracts from the series here: