Historical Traditions in Early India: 1000 BCE to 600 CE (PDF)

Chapter 23 Historical Traditions in Early India: c. 1000 bc to c. ad 600 Romila Thapar REPRESENTING THE PAST A couple of hundred years ago it was stated that Indian civilization was unique in that it lacked historical...

Chapter 23 Historical Traditions in Early India: c. 1000 bc to c. ad 600 Romila Thapar REPRESENTING THE PAST A couple of hundred years ago it was stated that Indian civilization was unique in that it lacked historical writing and, by implication therefore, a sense of history. This was said to apply generally to the pre-modern period, but more so to the earlier period. With rare exceptions there has been little attempt to examine this generalization, as it has been taken as given. Nevertheless, while there may not be historical writing of a conventional form as we know it now, there are many texts that reflect historical consciousness which later became the basis for historical traditions. We need to look for the nature and assumptions of these traditions and the historical concerns that underlie them. In recognizing the historical traditions of early times the criteria are likely to differ from those of modern times. Modern notions of history have been governed by definitions that emerged from the Enlightenment, with an emphasis on sequential narrative and chronology, and a focus on political authority. In India this was overlaid by colonial views on representing the past. Subsequent to this, nationalisms asserted that history was the single narrative truth about the past, and generally one that contributed to the ideological foundations of the nation. However, assessing these narratives as genres of texts and evaluating their claims to being a record of the past requires them to be judged by an accepted historical method in order to demarcate the credible from fantasy, although fantasy may be a way of disguising reality. The concern today is less with whether historical writing was absent in early India, as has been frequently maintained, and more with the nature and assumptions of its historical traditions. Further- more, in examining the texts to find such traditions, the aim is not to claim historicity for each event they recount, but rather to search for the historical concerns of societies. 554 The Oxford History of Historical Writing For a variety of reasons it is now necessary to re-examine the question of the sense of history in early India.1 The definition of history has undergone change and is no longer confined to the Enlightenment view. A sense of history implies a construction of the past, but it does not have to be identical with modern historical writing. Before considering the texts that claim to represent the early Indian past, two aspects need to be considered. Since historical traditions of diverse cultures inev- itably differ in their form, comparative studies have to be more precise and differentiated than they have been so far. Equally important is the question of why it was necessary to argue that Indian civilization lacked a sense of history. This was largely, but not entirely, a colonial argument with emphases that derived partially from the Enlightenment definition of history, but more from the obvious vantage point of a colonial administration constructing an entirely new history for the colony. As a preliminary step it might be useful to explain what I mean by historical consciousness, historical tradition, and historical writing. Historical conscious- ness is an awareness of events and persons from the past, with the claim that what is being narrated happened, as is implicit in the term itiha-sa: ‘thus indeed it was’. Historical consciousness is often embedded in compositions that have other functions—frequently ritual functions. We tend to dismiss historical consciousness by applying the rules of verification, but historical consciousness is not identical with history and it also differs from historical tradition. The latter is a construction representing the past—a specific mode of making sense of the past—and can be used to orient the present. Historical writing in various forms marks a departure. It is not embedded in texts with other functions. It draws from historical traditions and creates categories of texts specific to its requirement. The narrative seeks verification by indicating its sources—explicit or implicit—provides reasonable causal explanations of the events, and is gener- ally concerned with those in authority. DIVERSE HISTORICAL TR ADITIONS AND THE RECOGNITION OF THE INDIAN TR ADITION The historiography most frequently taken as the measure of historical writing is the Judaeo-Christian. It has been argued that this had a clear teleology, forming part of a larger eschatology, and time was linear. These characteristics were not so evident in Graeco-Roman, Chinese, and Indian traditions, where eschatology 1 David Gordon White, ‘ “Digging Wells While Houses Burn?” Writing Histories of Hinduism in a Time of Identity Politics’, History and Theory, 45:4 (2006), 104–31; and Velcheru Narayana Rao, David Shulman, and Sanjay Subrahmanyam, Textures of Time: Writing History in South India, 1600–1800 (Delhi, 2005).  Historical Traditions in Early India: c. 1000 BC to c. AD 600 555 was weak and where time took various forms—linear, spiral, and cyclic. The need to explain the occurrence of past events leads to the explanation contrib- uting to the formulation of historiography. Much has been made of the lack of history in India being tied to a cyclic concept of time—an insistence which continues despite research to the contrary.2 Linear time in India is evident from the extensive incorporation of genealogies and the use of eras and precise systems of dating; and even more so in the shift in astronomy from lunar to solar reckoning. A sharp dichotomy between linear and cyclic time is not feasible, since some elements of each intersect, although pertaining to different functions. Cyclic time in the concept of the four yugas (cycles or ages) is often viewed as cosmo- logical time, whereas the more measured time in individual chronologies is linear. Where cyclic time takes a spiral form it can be seen as a wave, and ulti- mately almost linear when stretched. Such variations in the Indian texts suggest a heterogeneous time calculation. Even in the span of four yugas, the present is not a repetition of the past, since each age differs from the previous one. Eschatology is known to the Indian traditions even in cyclic time, but the pattern is unlike the more linear eschatology of the Judaeo-Christian tradition. The perfection of the first yuga gradually declines, largely because of the distancing from norms and beliefs, as well as the increase in non-meritorious behaviour. Eventually the catastrophic end is so imminent that it requires a saviour-figure. This was the bra-hman.a Kalkin, the final incarnation of Vis.n.u, in the Puranic tradition; or the Buddha Maitreya, the Buddha to come, in the Buddhist tradition. In each case it was believed that the universe, or the faith, would be restored to its pristine condition. Even this well-known information is seldom brought into the discussion on possible historical traditions. Why there was a refusal to concede the existence of any historical tradition requires enquiry. THE ABSENCE OF HISTORICAL WRITING IN INDIA The search for indigenous histories of early India began in the late eighteenth century. European scholars, familiar by this time with historical writing as a distinct category of literature, looked for the same in the Sanskrit articulation of what came to be called Hindu/Indian civilization, and were unable to find it. For instance, when the philologist William Jones suspected that there might be history in the myths and legends of the Pura-n.as, most disagreed.3 Even Jones 2 Mircea Eliade, Cosmos and History: The Myth of the Eternal Return (New York, 1959). For a critique of this position, see Romila Thapar, Time as a Metaphor of History: Early India (Delhi, 1996). 3 Asiatic Researches, 4 (1807), xvii ff. 556 The Oxford History of Historical Writing quotes only one example of historical writing, the Ra-jataran· ginı̄, written by Kalhan.a in the twelfth century ad. Indian civilization was therefore defined as ahistorical. Comparisons with the Chinese accounts of Sima Qian, or the writing in Arabic of Ibn Khaldun, or even the biblical genealogies, not to mention Grae- co-Roman narratives, strengthened, if only in contrast, the axiom of Indian society denying history. The officers of the East India Company, primarily interested in law and reli- gion to assist them in administering their Indian colonies, derived information from their bra-hman.a informants. Inevitably, the texts of Vedic Brahmanism, such as the Vedas, setting out ritual and belief, and the Dharmaśa-stras, the codes governing caste and social obligations, had priority. Other systems of knowl- edge, especially the Buddhist and the Jaina, were assessed as inferior branches of Hinduism, particularly since they were regarded as deviant by bra-hman.as. There was little attempt at placing texts in a wider discourse of alternative systems of knowledge. In Europe, German Romanticism made much of what came to be called the Oriental Renaissance.4 Religion and mysticism were said to be characteristic of Indian culture to the virtual exclusion of rational ways of organizing knowledge. The argument that in India caste, viewed as civil society, overwhelmed the state meant that without a state there could be no history. For Hegel therefore, India was a land without recorded history.5 The enthusiasm for Sanskrit in some circles fed into an influential theory of language, race, and culture—that of the Aryan race. Applied to India it became the explanation for the Aryan origins of Indian civilization, and this is turn was equated with the Brahmanism of the Vedas. Since the latter had little concern for history, the notion of an absence of history was reinforced. In the nineteenth century, a different reconstruction of Indian history drew on premises that precluded the need for an historical tradition. It underpinned the requirements of colonial policy in a changing relationship between the colo- nial power and the colony.6 A denial of a sense of history was implicit in its major theory—that of Oriental Despotism. This was articulated at length in what became the hegemonic text of the colonial construction of Indian history: James Mill’s The History of British India, published 1819–23.7 Indian society was said to be static and, since it did not register change, it had no use for recording the past, one of the functions of the past being to legitimize the present. This stasis could only be broken by British administration legislating change. Mill’s History was 4 Raymond Schwab, La Renaissance Orientale (Paris, 1950). 5 G. W. F. Hegel, The Philosophy of History (1837; New York, 1958), 140–1, 162–4. 6 Romila Thapar, ‘Ideology and the Interpretation of Early Indian History’, in ead., Inter- preting Early India (Delhi, 2000); and Ronald B. Inden, Imagining India (Oxford, 1990). 7 Javed Majeed, Ungoverned Imaginings: James Mill’s The History of British India (Oxford, 1992).  Historical Traditions in Early India: c. 1000 BC to c. AD 600 557 defining a new idiom for imperial control. Other arguments attributed the absence of a sense of history to a lack of sub-continental political unity, or to the subordination of the human will to the divine, or to obsessive religion and the control of the bra-hman.a over intellectual activities. Arguing for the absence of indigenous history had the practical advantage of allowing the formulation of a history for the colony that would underpin colo- nial policy. Colonial attitudes to knowledge pertaining to their colonies assumed that such knowledge enhanced control. Thus William Jones wrote of the itihasas and puranas being ‘in our power’, and a century later Lord Curzon saw the intel- lectual discovery of the Orient as the necessary furniture of the empire. The collection of manuscripts and artefacts for the reconstruction of history became an avid activity. Equally impressive were the decipherment of scripts and archae- ological discoveries. However, the oral compositions of the bards, collected and written about by James Tod and L. P. Tessitori, were generally bypassed by histo- rians. In the larger flow of explaining the past, colonial preconceptions tended to colour the narrative. Even the intellectually challenging discussions in Europe on the nature of history as an emerging discipline had little impact on Indolo- gists and colonial historians. Aspects of modern history focusing on demands for democracy and political freedom in the West were deliberately excluded from the prescribed texts.8 The Orient as ‘the Other’ of Europe became almost obsessive among those theorizing on histories beyond Europe. Karl Marx emphatically denied the exist- ence of a sense of history in India. Max Weber attributed the lack of transition to capitalism—as a manifestation of Otherness—to a failure of economic ration- alism. Indian historians initially subscribed to the colonial view and accepted that Indian society was ahistorical. More recently the subject has occasioned passing comments. The first collection of essays on the subject tended to repeat the premises of the older theory.9 An attempted introduction to categories of texts claiming to be historical was unaccompanied by a discussion of the claim.10 An important discussion of historical biographies as historical tradition has remained without a follow-up.11 Other arguments maintain that an historical tradition existed but was weak because of the decentralized nature of political institutions and the exclusive control of the bra-hman.as over the transmission of the tradition.12 The bifurcation of the bra-hman.as, who might have made a critical assessment, and 8 Aravind Ganachari, ‘Imperialist Appropriation and Disciplining the Indian Mind’, Economic and Political Weekly, 43:5 (February 2008), 77–87. 9 C. H. Philips (ed.), Historians of India, Pakistan and Ceylon (London, 1961). 10 Anthony Kennedy Warder, An Introduction to Indian Historiography (Bombay, 1972). 11 Vishwambhar Sharan Pathak, Ancient Historians of India (Bombay, 1966). 12 Burton Stein, ‘Early Indian Historiography: A Conspiracy Hypothesis’, Indian Economic and Social History Review, 6:1 (1969), 40–60. 558 The Oxford History of Historical Writing the scribal castes has also been suggested.13 More defensive views state that history has been formatted through modernization processes, and Indian civili- zation has been unconcerned with these.14 Such explanations are inadequate. That the idea of an historical tradition in pre-modern India is now attracting some attention is suggested by a few studies that are concerned with demon- strating the existence of such traditions and the manner of their use in society and politics. It emanates from the redefinition of history in recent times, through locating history as a social science, from studies of memory to a defence of history faced with a sentence of death pronounced by post-modernism. Most of these studies focus on the second millennium ad and have not gone further back in time.15 HISTORICAL CONSCIOUSNESS AND THE HISTORICAL TR ADITION Research of the last half-century has made it apparent that early Indian society was not static and was subject to change, and the change was not uniform in space and time. Two processes of change are evident and have continued through the centuries. One was the mutation of clan societies or lineage-based societies into castes, as part of their assimilation into state systems in the form of kingdoms. The other was the transformation of early kingdoms into more complex state systems. With historical change, new identities emerged and the past was reformulated. Historical consciousness implicit in the fi rst mutation was converted into an historical tradition in the second. Both processes are reflected in what comes under the rubric of the itiha- sa-pura- n.a tradition. Itiha- sa literally means, ‘thus indeed it was’, and has come to be used now as ‘history’, but earlier it was not history in any modern sense of the term. Pura- n.a refers to that which is old, and includes what we would now call legend or even myth. The itiha- sa-pura- n.a, or early Indian historical tradition, has two distinctly different historiographies, both of which came to be established by the mid first 13 H. Kulke, ‘Geschichtschreibung und Geschichtsbild in Hindustechen Mittelalter’, Secculum, 30 (1979), 100–13. 14 Ashis Nandy, ‘History’s Forgotten Doubles’, History and Theory, 34:2 (1995), 44–66; and see also Vinay Lal, History of History (Delhi, 2002), 14–16, 58–60. 15 Daud Ali, ‘Royal Eulogy as World History’, in Ronald B. Inden, Jonathan S. Walters, and Ali (eds.), Querying the Medieval: Texts and the History of Practices in South Asia (Delhi, 2000); Rao, Shulman, and Subrahmanyam, Textures of Time (see also the review articles of this book and the authors’ response in History and Theory, 43:3 , 366–427); Prachi Deshpande, Creative Pasts: Historical Memory and Identity in Western India, 1700–1960 (Delhi, 2007); and Sumit Guha, ‘Speaking Historically: the Changing Voices of Historical Narration in Western India, 1400– 1900’, American Historical Review 109:4 (2004), 1084–103.  Historical Traditions in Early India: c. 1000 BC to c. AD 600 559 millennium ad. The more commonly referred to is the Puranic, which emerges from the Pura- n.as and related texts, and draws largely on sources composed or edited by bra- hman.as. The other, which has not been given the recognition it deserves, draws from Śramanic ideologies—primarily Buddhist and Jaina.16 The events and personalities which each highlight from the past often differ or else are represented in different forms. There is an awareness of alternative views which, although not always stated as such, are nevertheless reflected either in borrowings or contradictions. The period covered in this chapter is from about 500 bc to ad 600, but with forays into earlier times where necessary. Historical concerns during this period are evident in distinctive forms. The first is historical consciousness, which comes in an embedded form in texts with other functions. The second consists of forms that embody this consciousness and give it a distinctive identity as historical traditions. The embodied or externalized forms not subordinated to other functions come later as new genres of texts and expressions of the historical tradition. There is a shift from the representation of historical consciousness to the creation of an historical tradition. THE EMBEDDED TR ADITION: THE BIRTH OF THE HERO The embedded tradition goes back to the da- na-stutis (in praise of gift-giving), which were hymns in the R· gveda, generally dated to about 1400–1000 bc.17 Priest-poets composed hymns in Vedic Sanskrit in praise of the god Indra protecting his worshippers from their human and demonic enemies, and also assisting them to carry out successful raids to capture cattle-wealth and pastures. Indra became the model hero emulated by the aspiring human hero.18 The human hero was eulogized not only for his heroism, but also for making gifts (da- na) for the poet who had composed the hymn. Victorious rajas by these accounts gave generous, if not exaggeratedly large, gifts to the composers.19 The praise is both for the act of heroism that brought about the victory, and for the gift to the composer of the eulogy. Lauding the hero and the gift-giving as an act of patronage were to remain in tandem. The gift, symbolic of success and status, provided a precedent for future occasions, and became a bond between the giver and the recipient. Locating the narrative in a text that was memorized for ritual purposes ensured that the event and the hero would become part of the 16 Sramanic is derived from s´ramana, the term used for Buddhist and Jaina monks. 17 Rig-Veda-Samhita, ed. F. Max Müller (London, 1892); and Ralph T. H. Griffith, The Hymns of the Rgveda (1896–1897; 2nd edn, Varanasi, 1963). 18 Rgveda, 7.20.5; 3.30.4; 8.46.13; 3.51.3. 19 Romlia Thapar, ‘Da- na and Daks.ina- as Forms of Exchange’, in ead., Cultural Pasts: Essays in Early Indian History (Delhi, 2000), 521–535. 560 The Oxford History of Historical Writing Ancient India c.500 bc.  Historical Traditions in Early India: c. 1000 BC to c. AD 600 561 remembered tradition about the past. The composers claimed that they were immortalizing the hero, as indeed they were. The hymns reflect a society of clans organized into chiefdoms. The clan was a kin-based unit of agro-pastoralists, relatively egalitarian but accepting the authority of a chief—the raja—who was selected by the clan. The chief protected the territory of the clan, captured fresh pasture lands, conducted raids, distrib- uted the booty, and was often the patron of the sacrificial ritual.20 These were the required qualities of the hero, and continued to be in demand into later times. The hymns of the R·gveda in various incidental ways recorded relations of friend- ship or hostility between clans. Hostility often took the form of skirmishes. But a more serious confrontation was when Suda-s defended himself against a confed- eracy of ten other clans. Of the ten, there were some, such as the Yadus and Pu-rus, that feature in the later epics. There is some information on kinship links between individuals, although genealogical depth is small and does not exceed five generations. Rituals that bestow status on rajas were the abhis. eka, ra- jasu- ya, aśvamedha, and va- japeya, intended for consecration, claims to conquest and sovereignty, and rejuvenation, and described in the texts subsequent to the R·gveda. The rajas that performed these rituals in previous times are listed, and this became the remem- bered past—part history and part legend. The politics of heroism were highly competitive, and remained so until chieftainship became hereditary or was mutated into kingship. The retelling of the narrative by the bra- hman.as came to imply authenticity. This was the starting point of what was to become the frequency of praśastis (eulogies) in recording the activities of those who had authority. The composers were the poets (kavis) who could also be bards (su- tas) or ritual specialists (bra-hman.as) attached to the entourage of the chief. The elaborate narratives (akhya- nas) recited in the course of the aśvamedha became the prototypes of the narratives woven into the epic.21 Fragmentary narratives of heroes and clans were common currency and part of a large floating oral tradition, probably maintained by bards. Some remained separate entities, as in the Buddhist Ja- taka collection. Others were stitched together in epic forms such as the Maha- bha-rata and the Ra-ma-yan.a. The dates of the epics are controversial given the likelihood of interpolations requiring restructuring of the composition. Their present forms have been dated to anywhere between circa 400 bc to circa ad 200. What is of interest is not only their claim to be recalling past events, but also how these were modified to suit specific current perspectives. 20 Romila Thapar, From Lineage to State: Social Formations in the Mid-First Millennium B.C. in the Ganga Valley (1984; 2nd edn, Delhi, 1996), 26 ff. 21 Aitareya Bra- hman.a, 7.18.10; 3.21.1; and Śatapatha Bra- hman.a, 13.4.3.2. 562 The Oxford History of Historical Writing The embedded tradition becomes somewhat more visible in the epics.22 The Maha- bha- rata was the knitting together of many akhya-nas, gathered from Vedic and oral traditions. Its sheer size suggests a compendium rather than a unitary composition, and the boxing in of narratives allows for many events and persons to be accommodated—not unusual in the compiling of epics. It is said to have been composed in Sanskrit by the bra- hman.a Vya-sa, but also recited by his disciple, the su- ta (bard) Romahars.ana.23 The relationship of bard to bra-hman.a was ambiguous. The latter gradually edged out the former as the one who legit- imized the patron and recorded events. Vya-sa had an uncertain status, with a bra-hman.a father and a low-caste mother. In effect, the epic was a bardic compo- sition with didactic additions.24 These additions, together with explanatory legends, are so substantial that the narrative tends to become submerged. The epic was recited at a sattra, a ritual of sacrifice, and this echoes the recital of akhya-nas as part of the a-śvamedha. It is described most often as an itiha-sa, thus emphasizing the belief in the events having happened. The core story involves the rivalry of two co-lateral lineages—that of the Pa-n.d· avas and of the Kauravas—and their claims over territory in the western Ganges plain. Ruling lineages were generally, but not invariably, of the ks.atriya caste, that began as a group identified with chiefs and warriors, and evolved into a land-owning aristocracy in later times. The territory under dispute is divided into two parts, and is ruled by the Kauravas from Hastinapur and by the Pa-n.d· avas from Indraprastha. (On the assumption that the current place names are locations for the earlier sites, both have been partially and vertically exca- vated, but do not provide close co-relations with epic descriptions, which is generally in keeping with the archaeology of epic sites.) The prosperity and status of the Pa-n.d· avas riles the Kauravas and, through a series of competitive events, the Pa-n.d· avas are exiled for fourteen years. Inevitably this terminates in a battle between the two at Kuruks.etra, which becomes a major time marker in viewing the past. The battle is said to have drawn in most of the clans of northern India as allies of one of the two antagonists, and virtually marks the termination of clans. It is essentially an epic of clan polities. The patrons were the rajas, especially the Kuru-Pañca-la clans, also referred to in the earlier Vedic texts, who were asserting their authority as major chiefdoms. The epic is a kind of praśasti of some of the Pa-n.d· avas, and in a left-handed way, of a few Vr· s.n.is. The two co-lateral lineages are said to be descended from the Pu-ru line, and the Vr·s.n.is from the Yadu line. This ancestry links them to the earlier texts, and the Maha-bha-rata 22 J. L. Brockington, The Sanskrit Epics (Leiden, 1998). 23 V. S. Sukhthankar et al. (eds.), Maha- bha- rata, critical edn, 19 vols. (Poona, 1927–66), Adiparvan, 1.1–10. 24 Edward Washburn Hopkins, The Great Epic of India (New York, 1901); and V. S. Sukthankar, On the Meaning of the Maha- bha- rata (Bombay, 1957).  Historical Traditions in Early India: c. 1000 BC to c. AD 600 563 provides a considerably longer genealogy than the Vedic corpus. The epic, in a sense, is both an enlargement and a restructuring of some earlier compositions. A more detailed past was now required. Recovering the original epic is compli- cated by the many interpolations, and by the rearranging of the narratives in the present form. It has been argued that a particular category of bra-hman.as, the Bhr·gus, altered the story in parts, added the didactic sections, and converted a relatively secular epic into the sacred literature of what was to take the form of Vais.n.ava Bhagavatism, the worship of Visnu.25 Such Bhr·guisation, as it has been called, gave an identity to the protagonists and their historical context. Rewriting the past or incorporating interpolations reflects various moments seen as histor- ically significant. Apart from its other purposes, the Maha-bha-rata was also an attempt to record the history of the chiefdoms through claims to kinship, territory, and raids. It emphasized a new feature linked to status: namely, genealogies. In the installa- tion of the Kauravas and the Pa-n.d· avas, there are contested claims. Neither was connected by blood to the previous generation, and this defied the rules of succession; therefore, genealogical links were forged to gloss over this lacuna. The battle at Kuruks.etra was a conflict over succession within the larger Pu-ru lineage. But, in effect, it also marked the final decline of the ks.atriyas as lineage groups representing clan polities. It becomes a statement on the decline of chief- doms as a political system. There is much that was remembered in the Maha-bha-rata as a suggestion of history. There was nostalgia for a past society that was largely disbanded after the war. The epic makes an attempt to construct fragments of that society, whether remembered or partially fictional, to record clans and their rajas, and to map significant settlements. Locating this past society in time involved intro- ducing measurements of time. This took the form of the cosmological cycles of time, the yugas. The introduction of the yuga theory into time-reckoning may well have been part of the process of the Bhr·guisation of the epic. The maha-yuga, great cyclic span of time, envisaged a series of four ages—Kr·ta, Treta, Dva-para, and Kali—each declining in length, and characterized by a corresponding decline in righteousness. The enormous and differentiated lengths of time in each age of the yuga theory, adding up to 4,320,000 human years (the figures were probably borrowed from those used in astronomy), make even approximate measurement in human terms impossible. The epic events are said to have occurred at the turning of the Dva-para yuga into the Kaliyuga. The start of the Kaliyuga has been dated by a later reckoning to 3102 bc.26 This is an unlikely date, given that even the locations do not reveal archaeological settlements of an appropriate kind. The actual date would not have been as relevant as the notion 25 V. S. Sukhtankar et al., ‘The Bhr·gus and the Bharatas: A Text-historical Study’, Annals of the Bhandarkar Oriental Research Institute, 18 (1944), 1–76. 26 The Aihole Inscription, Epigraphia Indica, vol. 6 (1900–1), 1 ff. 564 The Oxford History of Historical Writing that what followed after the period of the ks.atriya lineages was a radical change. The centrality of vam · śas, lineages, and therefore genealogies, taken back to a supposedly remote past, were also setting out other patterns of the past. Cosmo- logical time was seen as cyclic, but the measurement of human activity was more frequently in linear time, as in genealogies. There is therefore an intersection of cyclic and linear time.27 This recurs in the historical tradition. The Maha-bha-rata is called an itiha-sa, perhaps because it was thought to represent the presumed history of inter-clan conflict. The Ra-ma-yan.a, however, is generally referred to as a ka-vya, a poetic composition in Sanskrit, and often the a-dika-vya, the first of such. Its author, Va-lmı- ki, also had an ambiguous origin, although in a late section he is described as a Bhr·gu bra-hman.a.28 The narrative is of a hero in conflict. But the conflict is less between clans, as in the Mahab- harata, as between two types of polities—the kingdom versus the chiefdom. The triumph of the kingdom of Ayodhya and of Ra-ma, exemplifying the ideal king, moves historical consciousness into the transitional period of the conflict between the two systems—one declining and the other becoming foundational to future polities. Societies of forest-dwellers and chiefdoms were now demonized as ra-ks.asas, in contrast to the eulogizing of kingship, where eventually the king can even be the incarnation of deity. Ra-ma’s lineage boasts of many heroes of earlier times. As a political document, the Ra-ma-yan.a marks the arrival of the state as kingdom, of which the epic is a validation.29 The story is stereotypical. Ra-ma was the eldest son of the king of Kosala, and by rights should have succeeded his father. The intervention of his step-mother led to his being exiled from the capital at Ayodhya into the neighbouring forests and beyond. He was accompanied in exile by his wife, Sı-ta-, and his younger brother, Laks.man.a. During the exile, Ra-van.a, who ruled over the demons, kidnapped Sı-ta- and took her away to his home territory, Lanka-. Ra-ma had then to organize an army with the help of his allies, who were substantially a force of monkeys, and defeat Ra-vana in battle and rescue Sı-ta-. In later interpolations, when he assumed the kingship of Ayodhya he faced public suspicion of her faith- fulness to him and eventually banished her. Being pregnant she took refuge in the hermitage of the sage Va-lmı-ki, where her twin sons were born. Va-lmı-ki composed the narrative, and the two boys learned it as would have bards. Curi- ously, the names given to the twins, kuśilavah, refer to bards. They then recited the epic poem at the aśvamedha ritual, the patron of which was their father, Ra-ma. The first recitation of the epic was again at a ritual occasion. 27 Thapar, Time as a Metaphor of History. 28 G. H. Bhatt et al. (eds.), The Va- lmı̄ki Ra- ma- yana, critical edn, 7 vols. (Baroda, 1960–75); and The Ra- ma- yana of Va- lmı̄ki, trans. R. Goldman, 5 vols. (Princeton, 1984–). 29 Romila Thapar, ‘The Ra- ma- yana: Theme and Variation’, in ead., Cultural Pasts, 647–79.  Historical Traditions in Early India: c. 1000 BC to c. AD 600 565 The extremely brief Buddhist version of the story of Ra-ma differs and does not include the kidnapping of Sı-ta-.30 The version, as given in the Va-lmı-ki Ra-ma- yan.a, is questioned in a Jaina version, the Paumacariyam of Vimalasu-ri, written in the early centuries ad.31 There is an insistence on the historicity of the Jaina version, with the other versions being described as incorrect. The central events are broadly similar, but the depiction of Ra-van.a and the ra-ks.asas as demons is negated. Instead of being depicted as demons, the people of the forest are identi- fied with well-known lineages and clans, the Vidya-dharas and the Meghava- hanas. The latter in other sources are linked to the Cedis, who in turn were part of the Ya-dava/Yadu descent group. A political dialogue is implied in these contradictions. The questioning appears to be less about the historicity of the story but more about the demonizing and downgrading of the clan societies in the Va-lmı-ki version. It could be argued that at the root of the Ra-ma-yan.a story there appears to have been a political conflict between the lineages of the Iks.va-kus and those of the Ya-davas, and this conflict is mentioned in passing in a later source. The Va-lmı-ki epic was lending political support to the Iks.va-ku lineage, and referring to its earlier heroes, and annulling the status of the Ya-dava descent groups. Further- more, it was not just a narration of inter-clan conflict but also a depiction of the triumph of kingship as a political institution over the earlier form of chieftain- ships. It would seem that the confrontation between the two was intense, as indeed it actually was in the perhaps contemporary history of the long drawn-out campaigns between the kingdom of Magadha (in Bihar) and its neighbour to the north, the clan confederacy of the Vr·jjis. The demonizing of the ra-ks.asas—the forest-dwellers—whose institutions as described seem closer to those of chief- tainships than kingdoms, might also have arisen from the encroachments of kingdoms into forested areas, the clearing of which could ensure agrarian activity and its benefits. As part of the propagation of the worship of Vis.n.u, Ra-ma in the Va-lmı-ki version was converted into an incarnation of Vis.n.u and worshipped as such. This was the appropriation of a symbol for a new order and opposed to other religious sects who were not primarily worshippers of Vis.n.u, such as the Jainas. It strengthened the idea of kings having divine attributes. Some scholars have suggested that the epics in their current form carry an underlying theme of brahmanical reaction to the popularity of Buddhism—a popularity encouraged by the patronage of the Mauryan emperor, Aśoka.32 The didactic additions may have been in part a countering of Śramanic ideas. 30 Dasaratha Jātaka, no. 461. See The Jataka, ed. E. B. Cowell, 6 vols. (Cambridge, 1895–1907). 31 Paumacariyam of Vimalasūri, ed. Hermann Jacobi (1914; Varanasi, 1962). 32 J. L. Fitzgerald, ‘Introduction to Book 12’, in The Maha- bha- rata, trans. Fitzgerald, vol. 7 (Chicago, 2004), 114 ff.; and Alf Hiltebeital, Rethinking the Maha- bha- rata (Chicago, 2001), 15–7, 205–6. 566 The Oxford History of Historical Writing THE REFORMULATION OF THE EMBEDDED TR ADITION Historical consciousness, as expressed in the different Vedic texts and the epics, was in part redesigned into what gradually became an historical tradition, as available in the Pura-n.as. A characteristic of the latter from the perspective of history is that it claims greater authenticity for itiha-sa and is a statement about the past; and this statement has a function in later historical traditions. The Pura-n.as, as a genre of texts, were composed over the first millennium ad. The Vis.n.u and the Matsya Pura-n.as are among the earlier ones. Possibly as popular texts, the earlier ones were first composed in Pra-krit, the more widely used language, and later rendered into Sanskrit by bra-hman.a authors. Each Pura-n.a focused on the rituals and myths required in the worship of a particular deity, and most incorporated cosmology. Some had a section on the succession of line- ages and of dynasties, the vam. śa-nucarita section, which is of importance to the historical tradition. This gave visibility to the embedded historical consciousness through reformulating what was inherited and constructing from it an historical tradition.33 This was a watershed between historical consciousness as embedded in ritual and epic texts and the texts that were to be overtly historical. The vam. śa- nucarita (lists of succession) was placed in what was otherwise a Vais.n.ava sectarian text.34 The association with a ritual text continued. However, this was virtually an independent section, added on, it would seem, to a religious sectarian text to ensure preservation. It was not integrated into the ritual and religious functions of the Pura-n.as. Authorship of the texts seems to alternate between the su- ta (bard) and the bra-hman.a. Presumably, when it was realized that control over the past gave access to power in the present, this control was taken over by bra-hman.a authors. Yet the earlier authorship of the su- ta, or at least an attribution to the su- ta, is indirectly stated where the su- ta participates in the dialogue. The emphasis was less on testing veracity and more on presenting the past in a manner consistent in its own terms. The vam. śa-nucarita is a continuous list of succession with few comments. Nevertheless it is possible to discern a pattern incorporating three distinct sections.35 The first is the narrative that is relatively brief and relates to the pre- diluvian period and the reigns of the fourteen Manus, each ruling for many thousands of years. A cataclysmic flood occurred during the reign of the seventh Manu, Vaivasvata, and is described in the Matsya Pura-n.a.36 There is an earlier 33 F. E. Pargiter, The Ancient Indian Historical Tradition (London, 1922); F. E. Pargiter, The Puranic Texts of the Dynasties of the Kali Age (London, 1913); and Romila Thapar, ‘Genealogical Patterns as Perceptions of the Past’, in ead., Cultural Pasts, 709–53. 34 Vis.n.u Pura- n.a (Bombay, 1963); H. H. Wilson, The Vis.n.u Pura- n.a (London, 1840); Matsya Pura- n.a, Anandasrama Series (Poona, 1907); and Vasudeva Sharana Agrawal, Matsya Pura- n.a: A Study (Varanasi, 1963). 35 Vis.n.u Pura- n.a, Book 4. 36 Matsya Pura- n.a, 1.1–34.  Historical Traditions in Early India: c. 1000 BC to c. AD 600 567 reference to this story with some small deviations in the Śatapatha Bra-hman.a and the Maha-bha-rata,37 and it parallels the Mesopotamian flood legend.38 The Flood was a time marker separating the somewhat undefined account of the Manus from the precisely set out lineages claiming to be historical. The inter- vention of deities decreases after the Flood and the heroes are the protagonists. Subsequent to the Flood, Manu had ten sons who were the ancestors to various lineages of the ks.atriyas, and were the heroes, the warrior aristocracy, and the chiefs of earlier times. According to these succession lists, they were persons of relatively equal status and linked through origins and kinship. They were divided into two major lineages: the Su-ryavamśa, the solar line, and the Candravamśa, the lunar line. The former observed the rules of primogeniture, and therefore only the descent of the eldest sons was recorded. The Ra-ma-yan.a is the story of an exemplary hero of the Iks.vaku lineage of the Su-ryavamśa. The Candravamśa was more a collection of segmentary lineages, recording all the sons—in effect the clans—who are largely those involved in the events related in the Maha- - bha-rata. The quality of the hero differed from the chief within the clan, to the king above and outside the clan. These lineages were continuous lists of generations, and therefore covered more than the smaller segments referred to in the epics.39 The lineages petered out after the two battles that had each been the foci of epic events. The more powerful dynasties that succeeded them were generally assigned a lower s´u- dra caste.40 That genealogies were important is evident, not only from their centrality to the epics, but also from the statement of Megasthenes, the Seleucid ambas- sador to the Mauryan court in the late fourth century bc, that 153 generations up to this period were recorded in India.41 The third section subsequent to epic events lists the dynasties of kings.42 They are not referred to as ks.atriyas but as bhu- pa-las and nr· pas—protectors of the earth and of men. This is a statement of historical change in the perception of authority. The kings were identified, not by lineage, but by dynasty and caste, registering a new perception of authority. The lists of dynasties began with the kings of Magadha in about the sixth century bc, and continued with the more important - dynasties such as the Śiśuna-ga, Nanda, Maurya, Śun·ga, Kan·va, and Andhra/ Sa-tva-hana, up to the Gupta dynasty in the fourth century ad. They were presented as being of unequal status since they are of varying varn.as (castes), such as s´u- dra and bra-haman.a, from the lowest to the highest. This sets them out as different from the earlier ks.atriya rajas. Assigning a low caste to a dynasty 37 Śatapatha Brāhman. a, 1.8.1.1–10; and Maha- bha- rata, 3.185. 38 W. G. Lambert and A. R. Millard, Atra-Hasis: The Babylonian Story of the Flood (Oxford, 1969). 39 F. E. Pargiter, The Ancient Indian Historical Tradition. 40 F. E. Pargiter, The Puranic Texts of the Dynasties of the Kali Age. 41 Pliny, Hist. Nat. VI. xxi. 59–60 (Loeb Classical Series). 42 As set out in Pargiter, The Puranic Texts of the Dynasties of the Kali Age. 568 The Oxford History of Historical Writing could indicate that they patronized what the bra-hman.as regarded as heterodox sects. Unlike the lineages, the dynasties are unrelated to each other. Kingdoms accommodated this new stratification. These kings are not the subject of pras´astis (eulogies) as were the ks´atriyas. Nevertheless, the projection is of the authority of the king and the state. The dynastic section was written in the future tense and technically, there- fore, was foretelling the future, although actually it was ‘foretelling’ the past from the point in time when the Pura-n.a was composed. Projecting past events into the future was a claim to authority. The intention was to construct and conserve a past from the brahmanical perspective, but using the data of the su- ta. The ks.atriyas were lauded because, even though some were of ambiguous origin, such as the Pu-rus, they were said to have observed the brahmanical social code, the varn.a-a-s´rama-dharma, performed the s´rauta (public rituals), and made lavish donations to the bra-hman.as. The ks.atriya genealogies were to be useful in later periods when claims were made to ks.atriya status by non-ks.atriya rulers. This is a contrast to pre-Gupta times when non-ks.atriya kings were accepted as virtually normal. Nevertheless, the Pura-n.as established a flexible ks.atriya identity. Thus it is stated that in the period just prior to the Gupta, a king of Magadha will uproot the existing kings and establish other varn.as (castes) such as Kaivartas and Pulindas (both recog- nized low castes) as kings, and that he will uproot the ks.atram and create another ks.atram.43 This is generally taken to mean ks.atriya, but could be a reference to ks.atr (power/supremacy/dominion). Parallel to the bra-hman.a authors of the Pura-n.as were court poets and scholars, generally from the elite, and writing in Sanskrit and Pra-krit, who introduced new genres of writing, which were not histories as such but incorporated some degree of historical tradition. Justifying the politics of those in authority by invoking actions from the past was an aspect of this change. This is also reflected in two historical plays by Vis´a-khadatta, the Devicandragupta and the Mudra- ra-ks.asa.44 The author was associated with the court, his grandfather having been a sa-manta—an intermediary who had received a grant of land from the king—and his father took the title of maha-ra-ja, asserting greater status. The date of the playwright is not certain, but he is generally thought to have lived in the Gupta period or soon thereafter. This is suggested by the contents of the first play, which unfortunately survives only in fragments. It concerns the defeat of Ra-ma Gupta in western India by the Śakas/Scythians who also take away his queen. This angers his younger brother who rescues the queen, eventu- ally defeats the Śakas and has his brother assassinated, and comes to the throne 43 Ibid., 53. 44 Mudrara- ks.asa, ed. A. Hillebrandt (Breslau, 1912); and Three Sanskrit Plays, trans. M. Coulson (Harmondsworth, 1981).  Historical Traditions in Early India: c. 1000 BC to c. AD 600 569 as Candragupta II. The play appears to have been written to exonerate the actions of Candragupta II. The story is related in other texts as well. The Mudrara-ks.asa focuses on the important historical transition from the Nanda dynasty to the Maurya in circa 321 bc. The action involves complicated intrigue, but since the protagonists are ministers and advisers to the kings, the dialogue is often a comment on how the politics of the situation were viewed in terms of a later age looking back on an earlier one, but perhaps with an eye on its own times as well. Viśa-khadatta was well read and familiar with sources that narrate events relating to the earlier period, such as the Arthas´a-stra of Kaut·ilya—a major work on political economy, whose author Kaut·ilya is a protagonist in the play—various Buddhist and Jaina narratives in which he also features, and the vams´a-nucarita of the Pura-n.as. Initially yajñas (sacrificial rituals), in addition to propitiating the gods, were statements of status when performed on a lavish scale. In the Gupta period, and later with yajñas becoming less important, genealogical connections became a new form of legitimation. The su- tas, as earlier genealogists, gradually gave way to the bra-hman.as, appropriating this source of controlling the past. Genealogies and descent lists are indicators of linear time. Additions, subtrac- tions, and spurious names which are normal to maintaining such lists do not affect the linearity. Linear time is further confirmed when there is mention of regnal years for members of a dynasty, or else the length of the dynasty is specif- ically stated. Historical chronology is also expressed in terms of eras, which occur in inscriptions. The two most frequently used are the Vikrama samvat of 57 bc and the Śaka samvat of ad 78. In the Pura-n.as, time is also measured in cycles, as in the theory of the yugas which is described in detail.45 Cycles are often dated on the position of constellations, as at the start of the Kaliyuga. The vams´a- nucarita crosses the yugas from the Dva-para to the Kaliyuga, with the Kuruks.etra war as the approximate time marker. Cosmological cyclic time is the larger frame. Within this, the descent groups and dynasties which measure time with reference to persons and events linked to itiha-sa conform to a linear pattern of time. In a sense, linear time becomes a segment within the arc of the cycle. The reconstruction of the past in the vams´a-nucarita section of the Vis.n.u Pura-n.a is a reformulation of the data on the past, now made accessible to the newly emerging royal families. Incorporating it into the Pura-n.as was a handy device to provide both authority and preservation. A new ruler seeking ks.atriya status would prefer to link himself to a Puranic lineage. The need for such a reordering of the past was doubtless also a reaction to the challenge of the contrasting reconstruction in the Buddhist tradition, and more particularly the Theravada tradition. The Puranic historical tradition avoids mentioning the version of events as given in heterodox sources. These were tied to a different 45 Vis.n.u Pura- n.a, I. 3. 570 The Oxford History of Historical Writing ideological tradition. For example, in describing the emergence of political authority and government, brahmanical theory sought legitimacy through divine sanction in the appointment of a raja, whereas Buddhist sources negated divine intervention and ascribed authority to a social contract, in an effort to terminate conflict over the institution of family and of private property. This latter was seen as a solution to the fear of anarchy. It suggests an evolving diachronic view of history, and provides some clues to the differing nature of causality in the two traditions. THE JAINA AND BUDDHIST HISTORICAL TR ADITION Perceptions arising from the itiha-sa-pura-n.a traditions were predictably different in the parallel Jaina and Buddhist writing, although the difference narrows in later times. The teachers of these religions, Maha-vı-ra and Gautama Buddha, were both historical persons of the Jña-trika and Śakya gan.a-san· ghas (clan-based societies). This underlay the assumption of historicity in the Śramanic tradition. The traditional date for the death of Maha-vı-ra was 527 bc, and for the Buddha (referred to as the maha-parinirva-n.a) was 486/483 bc. There has been some debate about these dates among modern scholars, although the suggested alternatives have a margin of up to a century later. However, the chronology within the two traditions was generally consistent in using their respective chronologies. The early texts of the Jaina tradition, such as the a-ca-ran· ga and the Kalpasu- tra, narrate episodes in the life of Jaina teachers,46 encouraging an interest in biog- raphy as a genre, as well as the attempt to record the succession of Jaina pontiffs at major monasteries. A concern for chronology is also expressed in the use of the Vikram era of 58 bc and the Śaka era of ad 78. In critiquing brahmanical versions of the past, attempts were made by Jaina authors to retell current narratives of the past, demonstrating thereby the biases of other traditions, and incidentally incorporating its own in the retelling. An early phase in viewing the history of the sects in the context of general history comes from Magadha and the middle Ganges plain, where the sect orig- inated and its teachings were initially propagated. But more effective historical writing belongs to the later period after the eighth century ad. Much of this is located in western India where the Jainas had both a religious and a political presence. Characteristic of monastic sects were the pat·t· a-valis, recording the succession of important Elders of the San·gha, Order, as well as kings and merchants—the organizers of the sects and the patrons. The early texts, frequently written by monks, were in Pra-krit, and gradually there was a change to using Sanskrit. - 46 Aca- ran· ga, ed. Walthur Schubring (Leipzig, 1910); and Jaina Sutras, ed. and trans. Hermann Jacobi (Oxford, 1884).  Historical Traditions in Early India: c. 1000 BC to c. AD 600 571 Buddhist texts, where they refer to the pre-Buddhist period, draw on some sources that were common to the itiha-sa-pura-n.a traditions. Subsequent to the establishment of Buddhism, history is determined by persons and events in rela- tion to the Buddha and the Buddhist San·gha. Connections between political authority and the Buddhist San·gha, such as co-relating the succession of monastic Elders with the reign of kings, are a departure from the Puranic tradi- tion. Whatever borrowing there might have been from any common source or from the Puranic tradition dwindles with the narrative of dynasties starting in about the sixth century bc. From this point on, the Puranic and the Sramanic traditions tend to diverge. The Buddhist tradition has a sharper understanding of the centrality of an historical perspective. The reasons can be many: the founder and the Elders so central to the narrative were historical figures; literacy meant not only writing and copying the Buddhist Canon but also commentaries and monastic chronicles, not to mention the biographies (some close to hagiog- raphies) of the historical figures; the growth of sectarian fissions and differences of doctrine needing to be recorded; and information about the properties of the monasteries having to be maintained. Events were generally related to the central date, that of the maha-parinirva-n.a, the date of the death of the Buddha, calcu- lated as 486/483 bc, although in one tradition from Sri Lanka there was a discrep- ancy of sixty years, with the date being 544 bc. Early Buddhist texts were concerned with establishing the historicity of the Buddha, the history of the San·gha, Order, and the succession of Elders in various sects as they emerged from the initial teaching. The Buddhist narratives written by monks and by commentators also start with the history of the teaching in the middle Ganges plain and the emergence of the monastic order. Written in Pa-li, a Sanskrit vernacular used extensively by early Buddhists, the teaching and initial history of the San·gha has come to be called the Pa-li Canon. Parallel to this were the vam · s´as, the registers of lineages, succession lists, and chronicles. The earliest of these was the Buddhavams´a, which narrates the lives of the Buddhas who are believed to have preceded Gautama. Claiming continuity from earlier teaching, it uses the puranas as sources, this being the older oral tradition. Chronology becomes precise from the time of Gautama and this also separates him from the earlier Buddhas. The Buddhist monastic chronicles of the Therava-da sect of Sri Lanka, the Dı̄pavamsa (fourth century ad) and the Maha-vam · sa (fifth century ad), are the important histories. Starting in the eastern Ganges plain, they move to Sri Lanka, linking the history of the island with that of parts of the subcontinent through the Buddhist connection.47 As chronicles, they have a definitive perspec- tive: to defend the Maha-viha-ra monastery of the Therava-da amidst the conflicts 47 Dı̄pavamsa: An Ancient Buddhist Historical Record, ed. and trans. Hermann Oldenberg (Berlin, 1879); and The Maha- vamsa, trans. Wilhelm Geiger (London, 1912). 572 The Oxford History of Historical Writing of various Buddhist sects. They claim the earlier Sinhalat·t· hakatha--Maha-vam · sa, now extant, as a source. The tradition of monastic chronicles was continued into medieval times, the Cu- lavam - · sa being a successor to the Mahavam · sa. It is thought that the Dı̄pavamsa may have been initially written by nuns, since no author is mentioned and there is much that is said about nuns and convents. Buddhaghos.a, writing in the fifth century ad, drew on the Dı̄pavamsa for his commentary, the Samantapasa-dika-. Maha-na-ma, the monk who wrote the Maha-vam · sa, updated the narrative as he saw fit and addressed specific themes from the past. Sri Lanka had to be prepared for the arrival of the Buddhist mission. This was initiated by a fantasy of the Buddha flying to the island and back. Maha-na-ma’s subsequent concerns were historical: narrating the peopling of the island with migrants from eastern India; the first Buddhist mission to Sri Lanka brought by the son of Aśoka Maurya which linked its history in some detail to that of the early Mauryas; the successive councils through which Ther- ava-da Buddhism was contoured and dissident groups came into existence, and the role of the Elders in this process; and the evolving of the Maha-viha-ra monas- tery to which he belonged as the centre for the Buddhist San·gha, its competition for patronage with the later Abhayagiri monastery, and the acrimonious rela- tionship between the two in the context of the politics of royal patronage. The last is a major agenda of the chronicle. The Maha-vam · sa presents the chro- nology of the Elders and their links with royalty. Much is made of the epic role of a Sri Lankan king, Dut·t·aga-mini, in forcing back south Indian rule to the main- land, and this gives a strong political intention to the text. A connection is also made between Sri Lankan royalty and the ancient solar lineage of Iks.va-ku, called Okka-ka in Pa-li, where the Śa-kya clan of the Buddha also had links with Iks.va-ku, as stated in the Pura-n.as. The attempt was not only to prove the antiquity of the local royalty, but also perhaps to take it away from south Indian connections. The Vamsatthappaka-sinı̄ was a later commentary on the history narrated in these chronicles and this encouraged a continuing return to the earlier texts. Both the Dı̄pavam - · sa and the Mahavam · sa recount the establishing of Ther- avada monastic sects in Sri Lanka. The Buddhist San·gha, like most religious - sectarian organizations, required patronage from its community of followers, and more importantly, from royalty. The history of kings became part of the history of the San·gha, and the interface between the two had political importance. There was, in effect, more than one Buddhist tradition setting out a history. The Northern Buddhist tradition was different from the Therava-da, and had its own narratives about the past. Although some texts are in Pra-krit, most are in Sanskrit and increasingly so. Some related to kings who had been patrons of the San·gha, as was the Mauryan emperor Aśoka who was the focus of the As´oka- vada-na, the glorious deeds of Aśoka, and of the Divya-vada-na.48 To demonstrate 48 John S. Strong, The Legend of King As´oka: A Study and Translation of the Aśoka- vada- na (Princeton, 1983); and Divya- vada- na, ed. P. L. Vaidya (Darbhanga, 1959).  Historical Traditions in Early India: c. 1000 BC to c. AD 600 573 the authority of the San·gha, kings are depicted as complying with its demands or those of its Elder, except when they are wickedly hostile. The As´oka-vada-na depicts Aśoka initially as such, but on becoming a Buddhist, he mutates into an exemplar of a Buddhist monarch. The treatment of events is dominated by the king’s relationship with the San·gha. The validation from history is both of the kingdom and the San·gha. The persons and events from the past in the Buddhist texts were generally different from those referred to in the Pura-n.as, where Aśoka, for instance, is just a name in a king list. The Divya-vada-na narrates the spread of Buddhism to the north-west of the subcontinent, which became a major centre. Apart from Aśoka it also introduces the Kus.a-n.a king Kanis.ka as a patron. Historicity is suggested as, for example, in discussing the causes of the rise and decline of dynasties and the moral judgements that follow. The historicity of the Buddha encouraged the writing of biography, and the earliest of the major biographies was the Buddhacarita of Aśvaghos.a, written in the early centuries of the Christian era.49 Emanating from the Northern Buddhist tradition, this was an attempt to narrate the early life of Buddha and thereby also to crystallize the legendary material associated with it. The span is from the Buddha’s birth to his enlightenment, and the style is more of hagiography than of biography. It tends to be formulaic; nevertheless it may have spurred an interest in writing about the crucial events in the life of a king—the form that was to be taken by the historical caritas (biographies) in the period after circa ad 600. The format of the Buddhacarita probably contributed to creating the genre of biography—the carita in the subsequent period. CONCLUSION The itiha-sa-pura-n.a tradition evolves from being an embedded tradition from which items are selected and used in formulating later historical traditions of the Puranic or Śramanic variety. The narratives selected are either endorsed or else contested by other traditions also claiming to represent the past. Borrowings from earlier texts are not arbitrary and come to have a functional role in the more visibly historical texts of the period subsequent to ad 600, such as in the caritas, the inscriptional annals, and the vam - · s´avalı̄s (chronicles). Structurally, the past was conceptualized in three phases. The first is a narra- tive of origins, which is largely formulaic and therefore probably not meant to be taken literally, but which provides indicators of status through the presence of deities, and includes pointers to what are regarded as important actions. This moves into the second phase which narrates the vam · s´as (lineages), referring in the main to chiefdoms, and claiming to provide the occasional ancestor to later kings. The representations of persons and events in the Maha-bha-rata and the 49 The Buddhacarita, ed. and trans. Edward Hamilton Johnston, 2 vols. (Delhi, 1972). 574 The Oxford History of Historical Writing Ra-ma-yana often differ in Buddhist and Jaina versions, suggesting that the historicity of a particular version was less important than the projection of particular ideologies. The last phase attempts historicity in the succession of dynasties and the kings within each dynasty, or the history of the San·gha as an equivalent focus of authority. This pattern suggests the influence of the vam - · s´anucarita format, since the latter was a deliberate attempt to give a structure to the past and create an historical tradition. Genealogies become important and therefore can be fabri- cated, as they were earlier. The concept of vam · s´a (lineage) is crucial, referring as it does to descent and succession, as well as providing continuity from societies prior to kingship moving towards becoming kingdoms. The creation of the new ks.atriyas, as referred to in the Pura-n.as, required genealogies. In theory, the observance of the normative social code is insisted upon. The highlighting of particular dynasties or individual kings differs considerably for this period in the Puranic and Sramanic sources, and is dependent on the politics of the relation- ship between the sect and the king. In the Sramanic tradition, the history of the San·gha has priority, and in some ways subsumes the history of the Elders, who for this tradition were the source of authority. Political power, as reflected in these texts, has many foci, and therefore many forms of validation. There is a hint or a reference to a variety of confrontations— clan contestation, political factions at the courts, and sectarian competition. Chiefdoms as pre-state societies, with a premium on birth and kinship as arbi- ters of status, express claims to power and social legitimation in their own way. The coming of the state as a political form requires other inputs into the construc- tion of the past, as is reflected in Buddhist historical writing, with its closeness to state forms and the appurtenances of the state. The royal court becomes the focus of an identity that binds the region. Its counterpart is the San·gha. There is a consciousness of historical change through both the new genre of texts and in their contents, and this is related to historical needs. Historical change is implicit in the transition from chief to king, which was not an absolute dichotomy, but nevertheless registered a difference. A justification was required for this new identity, and a further validation from the past for the shift in form. This might perhaps be a partial explanation for why history came to be written in ways more familiar to us now, subsequent to the vam - - · s´anucarita -of the Puran.as. The traditions shift to different categories of authors. The su tas (bards) asso- ciated with the hero-lauds of the da-na-stutis of the R·gveda and the original composition of the epics, had chiefdoms as their primary context. This gives way to bra-hman.as, such as the Bhr·gus, as authors and editors whose concern was to reorder the epics to convert them into texts promulgating the Vais.n.ava Bha-ga- vata religion. These authors in turn give way to the bra-hman.a authors of the Pura-n.as, some of whom may have been Bhr·gus and thus provided continuity. None of these categories were court poets, and even the bards were itinerant and not attached to courts. The court poet as the author of historical writing was to  Historical Traditions in Early India: c. 1000 BC to c. AD 600 575 come later. Monks and scholars who were members of the San·gha had yet another perspective, in which the state had an interface with the San·gha. Nevertheless there was a degree of networking among these groups, and particularly where a wider coverage of sources was required. Bards functioned as such in small polities and where the identities of clans remained an undercur- rent, as indeed they still do in some parts of the subcontinent. The bifurcation becomes deeper if there is a distinction between the remembered oral tradition and the textual tradition. The su- ta keeps the oral tradition, and his record need not be identical with the text, nor his mandate and his audience. The official version of the documents from the court could replace the popular version. The continuing importance of each was dependent on the patronage they received from various social groups. The ideology and agenda of the author has to be recognized. Where religious functionaries were the authors, ideology became written into the texts more firmly. Where they were attached as functionaries to the royal court, even when they were bra-hman.as, their writing incorporated more of politics and the perspectives of courtly society. Patrons change from chiefs to kings, and since this was an ongoing historical process it forms a continuous thread in the perception of the past. A marked difference is noticeable in Buddhist and Jaina historical writing, since in these texts patronage is dual: the king and the San·gha. Religion intervenes in the narrative of the historical, but the narrative is soon released from ritual texts, although in some cases its end purpose can be religious edification. The history, or attempted history, focuses on those in authority, whether political or reli- gious, and an attempt is made to weave the two together. This is not the case with the earlier texts that have been discussed here, since the claim to historicity is not as urgent as in the Buddhist and Jaina traditions. An historical tradition in early societies is seen as necessary by those who maintain that a tradition that comes from the past must have centrality in the present, by those who are in authority and whose aura increases if they can vali- date and legitimize themselves by a connection with the past, and by those who are dictating the parameters of identity. The historical tradition in early India was addressing these concerns. TIMELINE/KEY DATES c.1400–1000 bc R·gveda c.400 bc – ad 200 Maha-bha-rata and Ra-ma-yana - c.200 bc – ad 200 A ca-ran· gasu- tra c. ad 200 As´oka-vada-na; Divya-vada-na c.100 Buddhacarita of Aś vaghos.a c.300 Dı̄pavam. sa c.400 Vis.n.u Pura-n.a 576 The Oxford History of Historical Writing c.400 Maha-vamsa of Maha-na-ma c.500 Devicandragupta and Mudrara-ks.asa of Viśa-khadatta KEY HISTORICAL SOURCES Bhatt, G. H. et al. (eds.), The Va-lmı̄ki Ra-ma-yana, critical edn., 7 vols. (Baroda, 1960–1975). The Buddhacarita, ed. and trans. Edward Hamilton Johnston, 2 vols. (1935; Delhi, 1972). Dı̄pavamsa: An Ancient Buddhist Historical Record, ed. and trans. Hermann Oldenberg (Berlin, 1879). Griffith, Ralph T. H., The Hymns of the Rgveda (1896–1897; 2nd edn, Varanasi, 1963). Jaina Sutras, ed. and trans. Hermann Jacobi (Oxford, 1884). The Maha-bha-rata, trans. J. A. B. van Buitenen, vols. 1–3 (Chicago, 1973–). The Maha-bha-rata, vol. 7, trans. J. Fitzgerald (Chicago, 2005). The Maha-vam · sa, trans. W. Geiger (London, 1912). Mudrara-ks.asa, ed. A. Hillebrandt (Breslau, 1912). Pargiter, F. E., The Puranic Texts of the Dynasties of the Kali Age (London, 1913). The Ra-ma-yan· a of Va-lmı̄ki, trans. R. Goldman et al., 5 vols. (Princeton, 1984–). Rig-Veda Samhita, ed. F. Max Müller (London, 1892). Strong, John S., The Legend of King As´oka: A Study and Translation of the As´oka- vada-na (Princeton, 1983). Sukthankar, V. S. et al. (ed.), Maha-bha-rata, critical edn, 19 vols. (Poona, 1927–66). Three Sanskrit Plays, trans. M. Coulson (Harmondsworth, 1981). Vis.n.u Pura-n.a (Bombay, 1963). The Vis.n.u Pura-n.a, trans. H. H. Wilson (London, 1840). BIBLIOGR APHY Brockington, John, The Sanskrit Epics (Leiden, 1998). Inden, Ronald B., Imagining India (Oxford, 1990). Pathak, Vishwambhar Sharan, Ancient Historians of India (Bombay, 1966). Philips, C. H. (ed.), Historians of India, Pakistan and Ceylon (London, 1961). Rao, Velcheru Narayana, Shulman, David, and Subrahmanyam, Sanjay, Textures of Time: Writing History in South India, 1600–1800 (Delhi, 2005). Sukthankar, V. S., On the Meaning of the Maha-bha-rata (Bombay, 1957). Thapar, Romila, From Lineage to State: Social Formations in the Mid-First Millennium B.C. in the Ganga Valley (1984; 2nd edn., Delhi, 1996). —— Time as a Metaphor of History: Early India (Delhi, 1996). —— Interpreting Early India (Delhi, 2000).