11

The Second Mahabharata

This essay is about reading the Mahabharata, but it is also partly about the innate complexities of reading. Theories about aesthetics can be read in two ways: historically and structurally/ theoretically. In one sense, it is more appropriate, while reading texts from the past, to treat them historically—in terms of the intellectual context in which they appeared; and, if the theory is a contribution to a tradition of theoretical debates, to look at the precise emendations and additions that one theory makes to the stock of concepts and arguments it had received from earlier ones.

To follow that procedure, we should look at Abhinavagupta’s views about rasanispatti, and compare them with the preceding doctrines against which he developed his own precise concepts and arguments. This will involve looking at earlier theorists’ attempts to solve the problem of conceptual underdetermination—in the crucial term rasa-nispatti—by the concepts of irruption (utpatti), inference (pratiti), expression (vyanjana), or universalization (sadharanikarana)¹. Such analysis will closely examine which concepts he rejects, on what grounds; and which, like sadharanikarana, he continues to use; and if he uses concepts in the same sense or not, or extends them; and whether the exact arrangement in which he places them is the same or not: in other words, it would explore the conceptual structure in an Althusserian sense.² The other way is to deliberately abstract from the historical genealogies of concepts, arguments, and theories, and to compare them, without regard to historical specificity, with other theories which might have a structural, not family resemblance to them. This essay offers an exercise of the second, more structural kind.

The Mahabharata (hereafter MB) is a text that inevitably produces bafflement at various levels, by its very scale. All kinds of usual activity regarding a text—writing, reading, understanding—are exceeded by its scale, both by the scale of its simple extent, and the scale of its complexity. Faced with a text of this kind, the appropriate response is bafflement. Before we read the text, we have to read the myths which surround it, and which assist in understanding the kind of text we face. The legends then usefully split the activity of ‘writing’ into two—writing in the sense of composing the meanings, and the more external sense of actually inscribing those thoughts on pages. Nor surprisingly, these required the inhuman powers of Vyasa, and Ganesa, and miraculously, the legend goes, Vyasa avoided a burn-out and went on to compose other serious stuff—like the Harivamsa and the Brahmasutra. Hearing the MB is also meant to produce effects unobtainable by the reading of more mundane texts: even the hearing of this story is meant to cleanse us of sins. But the MB does nothing in a straightforward or conventional fashion. On some philosophic readings, the MB’s ability to cleanse us is not through some mysterious powers of chanting these verses, but by the fact that understanding them enhances our ability to grasp the meaning of a moral life, and to lead it. It produces in us a pratyavigna, an anamnetic effect on our powers of moral insight and recognition.

Are these legends, in the guise of asserting some things about the writing end of the text, really telling us something about the reading end? Are these ideas not really about Vyasa and Ganesa, but about us, those who have embarked on an equally vast, equally improbable enterprise? Is it the suggestion that the text was dictated by a divinely inspired sage to a divine scribe really a warning to its undivine readers that they are undertaking to grasp a meaning of improbable vastness, difficulty, and complexity? This presents to us the enormous single problem: how does one make some meaning of a text like the MB? It is not surprising that the extraordinary scale and complexity of the text has elicited equally extraordinary attempts to deal with it. I shall be concerned with one of these efforts in the ninth-tenth century which has left one of the most lasting and fruitful interpretative legacies regarding the aesthetics of the MB.

It is a tribute to the philosophic fecundity of the MB that the question of its meaningfulness could be viewed in two radically different ways. I shall present both arguments: for, although I wish to develop the second, I am not entirely immune to the powerful intellectual allure of the first line of reasoning. It is commonly said that the MB mirrors life itself. This is a remark of colossal ambiguity, and it can be read in two entirely different ways. One reading could be that it is pointless to search for a meaning in the vast, disorderly, fascinating complexity of the MB for something singular like a meaning: it mirrors life precisely because it is as vast, endless, formless as life itself. Therefore there are many ways in which we can learn from it—by reading single episodes, reflecting on its vast narrative expanse, its gallery of characters, its astonishing construction of situations: but that is not a search for a meaning. The MB does not have a meaning, just as life does not: there are episodes, events, experiences, both in life and in MB, in which meaning can be found.

This is a powerful and sophisticated form of cognitive pessimism, but I wish to develop an argument from the opposite side—that, despite its scale and complexity, it is possible to find a meaning in the MB. The argument I want to offer can be made in two distinct forms. It can be presented as an historical argument in the strict sense, trying to situate a new understanding of the meaning of the MB in the aesthetic and philosophic discussions of the ninth-tenth-century Sanskrit culture. I lack the historical and technical scholarship to make the argument in that form with any rigour. But I believe the same argument can be made more theoretically, and inevitably more speculatively, by simply contrasting the two most probable and compelling conceptions of what the story could mean to its readers. I also believe that for anyone who has spent time with the MB, which must mean that he has been reading different versions and different parts of the text at different points and intervals of his life, this is likely to be a personal evolution as well: we move, if we engage with it for a long time, from the first meaning of the story to the second.

Interpretative Conditions for the Second Epic Addition of the Santa Rasa

The standard understanding of the MB narrative regards it as a virarasa text, a narrative of great heroism, although it is immediately apparent to any reader that it strives to present a picture of heroism—the life of vira in all its complexity and extremity, two qualities which constantly dominate the narrative imagination of the epic. The fundamental relations of human life, which should have an ineffaceable simplicity—the relation between a mother and a son, between wife and her husband—are bent into unrecognizably bizarre shape, like Kunti’s relation to Karna, or Draupadi’s marriage to five husbands. Human conditions and experiences are constantly pushed narratively to points of amazing extremity: Draupadi suffers a degradation worse than rape, Bhima extracts a strangely fascinating combination of just retribution and inhumanity. Eighteen aksauhini of soldiers gather on the battlefield, with few survivors. But the standard interpretation is to view it as a story of disputes attendant on royal succession and the enjoyment of vast imperial power, of utter degradation and bravery, and eventually an end that declares the restoration of a just dharmic order in which truth triumphs (satyameva jayate) and order is eventually restored. The Gita promises a destruction of evil and dharmasamsthapana, the re-establishment of a just order. After the development of the distinctive aesthetic philosophy of the Kashmir Saivas, their interpreters claim that, contrary to these semblances, the MB is a text of the santa rasa. This involves a major innovation at two levels: the suggestion at the level of narrative interpretation that the MB should be seen as a santa rasa text; but that interpretative innovation is attendant on a more fundamental theoretical innovation in aesthetics. In previous thinking of rasa theory, commentators agreed with Bharata that eight rasas existed, corresponding to the eight sthayibhavas in human emotions. Kashmiri theorists, since the time of Udbhata, speculate about santa rasa as a separate theme, and following this tradition of reflection Abhinavagupta adds a ninth rasa, the santa (tranquil) to this register—astonishingly through a commentarial operation on the Natyasastra’s original text. This was a fundamental revision of aesthetic theory; and its success can be measured from the subsequent dominance of the idea of navarasa in Indian artistic reflection.

Bharata’s text prosaically states there are eight rasas:

 

Srngara-hasya-karuna-raudro-viro-bhayanaka

Vibhatsadbhuta cetasca astah sastre rasah smrtah³

 

The Abhinavabharati maintains the pretense of a commentary, and offers an ostensible gesture of assuming the subordinate relation that the commentary bears to the principal text. Yet, in a remarkable instance of intellectual daring, where the actual innovation defies the formal stance of subservience, Abhinava makes the really ‘abhinava’ (new/unprecedented) claim that Bharata’s intention is to present nine rasas, including the santa as the ninth. But he does not state it openly, because he expresses his ideas through aesthetic suggestion—dhvani. Explicit suggestion of an idea by its own name—svasabda—is inferior to indirect suggestiveness. By a long and dazzlingly arcane demonstration of reading virtuosity, he suggests that the apparent absence of the santa is subtly permeated with its suggested presence.

Historically, the time and place of this innovation is intriguing. V. Raghavan, the distinguished Sanskrit scholar, suggests in one of his essays that Abhinava derived the ninth rasa from Buddhist philosophical reflection regarding the moral complexities of human life.⁴ Since the predominant attitude that Buddhists recommended towards the sufferings of life was an attitude that could be called santa, or imperturbable tranquillity, even if the Buddhists did not have a pronounced and properly elaborated philosophical aesthetics it is a credible hypothesis that the philosophical idea or the moral ideal of the santa was taken through an aesthetic translation by the foremost Saiva commentator on aesthetic matters. The Abhinavabharati simply enumerated the rasas as nine, in simple, utter, and undeniable defiance of the primary text. Abhinava does not offer any expostulation for this startling emendation of the text he is supposed to be merely elucidating. He simply goes on to offer a rather unconvincingly recondite argument that it is clear from some embedded signals in textual Bharata’s text that the santa is both implied and dominant.

What is the Commentarial Function?

This raises an obvious question: how was the commentarial function viewed by these philosophers? Of course, commentaries were of many different types. Those on Kalidasa’s poetic works could be said to enhance readers’ understanding of the narrative complexities, or the subtle aesthetic points of individual turns of phrase, but not alter the contents of the text itself. Commentaries were also required for philosophical texts, for a different reason. Works of speculation regarding major issues required both elucidation and a different activity which started with philosophic evaluation of the text’s claims. They could go on to either elaboration of arguments which were ambiguous, unclear, or undeveloped; or go into an argument with the textual propositions if they were questionable. Abhinava could thus go on to elaborate what was undeveloped in the sutras of the Natyasastra (hereafter NS), and add elements which were absent earlier, but without contradicting its major ideas or produce incoherence.⁵

Abhinava does not stop with just an insertion, an addition of another rasa to the palette of elementary emotions. Partly because he has to claim that it is implied, he is forced to enlarge the claim and make it much broader. Santa cannot be a rasa which sits laterally with all the other eight rasas; it must have a different kind of relation to this basic taxonomy. In Abhinava’s hands, it becomes a superordinate rasa which can encompass, embrace, and override all the rest.⁶ Once this extension of the rasa repertoire is complete, it becomes possible to turn to the revision of specific aesthetic judgements. Clearly, one of the most startling revisionist suggestions about classical literature is the one that, to a truly philosophic reading, the great epic reveals itself as a work of the santa rasa.⁷

Gary Tubb’s Analysis of the Dhvanyaloka

Numerous technical problem come in the way of establishing santa as the predominant rasa of a vast and complicated text like the MB. I am not trying to argue against Gary Tubb’s excellent, scrupulously detailed consideration of technical issues with Anandavardhana’s santa interpretation of the MB.⁸ Tubb demonstrates convincingly the difficulties of accepting the MB as a text in which santa is the dominant rasa if we follow the technical requirements of rasa theory. First, there are serious difficulties in inserting the santa inside the textual intentionalities of the NS, to assert, as the Kashmiri dhvanikaras do, that the santa is stated precisely through its absence. The idea that the sthayin of the santa—the bhava of nirveda—is suggested in the NS by its first place in the list of the vyabhicaris is too far-fetched a reading. But the obvious objection to this gloss is that the concept of dhvani is appropriate to literary texts. NS is a text about literature, but not a literary text. Ambiguity and indirectness are enhancements of a poetic language, but major deficiencies in a philosophical reasoning. We can have as much of dhvani as possible in literature, but it causes confusion if philosophical texts start working through dhvani.

The narrative difficulties are also numerous. There are obstacles in the way of regarding Yudhisthira as the main protagonist of the epic, as Tubb shows through his ingenious comparison of the role of Yudhisthira in the MB with the figure of the Buddha in Asvaghosa’s Buddhacarita. Kashmiri theorists try to get round this by repeated observations in the last section of the Dhvanyaloka that the main protagonist of the MB is neither Arjuna nor Yudhisthira but Vasudeva, whichever way you gloss the meaning of that name.⁹ We are left eventually with the powerful suggestion, again at the end of the Dhvanyoaloka, that ‘the miserable end of the Vrsnis and the Pandavas’ indicates that the overriding narrative purpose of the epic is to point to the futility of all worldly enthusiasm—for power, wealth, love, loyalty, victory. Since all these states are marked by fulfilled desire, the lesson of the MB is to teach the value of what Abhinava terms trsnaksayasukha: the happiness that arises from a cessation of desire. The Dhvanyaloka uses a sloka from the Anukramanika to gloss it, rather distantly: ‘For the meaning intended to be hereby suggested is as follows. The adventures of the Pandavas and others which are recounted, since they come to a miserable conclusion, represent the elaboration of worldly illusion, whereas it is the blessed Vasudeva, representing ultimate truth, who is here glorified.’¹⁰

Dhvanyaloka’s major reinterpretative statement is not based on signals embedded in textual fragments like the Anukramanika phrases; it lies in a general observation about the narrative concatenation itself: ‘in the MB, which has the form of a didactic work although it contains poetic beauty, the great sage who was its author, by his furnishing a conclusion that dismays our hearts by the miserable ends of the Vrsnis and the Pandavas, shows that the primary aim of his work has been to produce a disenchantment with the world, and he has intended his primary subject to be liberation from worldly life and the rasa of peace.’¹¹

I wish to suggest a reconstruction of this claim by means of arguments that are obviously modern, using the readings of the dhvani theorists as our materials, rather than our methodological guide. I select elements from their analyses but recompose them in other ways, and, of course, we are free to add arguments of our own from the intellectual culture we inhabit. I wish to offer mainly two suggestions: the first is about the historical direction of change in Sanskrit literary hermeneutics, and the second is a reading of the signification of the narrative ending. This will, I hope, yield two sets of interesting implications: the first set about narrative intentionalities, or the play of intentions in narrative texts, which need to be brought to ‘re-presentation’ in Gadamer’s sense;¹² and second, about the process of moral knowledge in the MB narrative.

Texts as a Field of Intentionalities/Textual Intentionalities

To a lay modern observer,¹³ it appears that the revision in aesthetic theory attempts to shift the emphasis in interpretative theory to the readers’ side—to the sahrdaya—although most theories make a further distinction between the aesthetic address of the reader and the spectator, because the general category of kavya is subdivided into drsya (visual) and sravya (aural) kavyas. Against earlier theories, which are content with a vague and unclear notion of aesthetic pleasure, new theories require a rigorous conception of the exact psychological event occurring in the process of aesthetic enjoyment. Through successive philosophic elaborations on what exactly happens in the case of rasanispatti,¹⁴ the dhvani theorists aquire a clearer picture of the reading-event as a complex happening on what could be termed an interpretative field stretching from author to character to actor to reader.¹⁵ As the theory evolves and gets more elaborate, it gives more detailed attention to what is happening between the actor on the stage and the spectator in the theatre, or to the poetic-literary text and its reader. From the angle I want to develop, the fact that the discussion has to deal with the slightly awkward distinction between theatrical and poetic texts, between natya and kavya, really serves to bring out an essential feature of aesthetic interpretation. Natya brings out the act of mediation in the process of representation more clearly than kavya. Thus, there is some advantage in thinking about the literary text by way of a detour through drama. The argument appears similar to Gadamer’s analysis of textuality.¹⁶ Minimally, a text stands between an author and a reader/recipient. It is misleading, in Gadamer’s view, to think of understanding the meaning of a text exclusively as an imaginative recapture of the authorial meaning, because reading is also, if not equally, a meaning-creating activity. To recapture meanings readers have to work through language, and because of the ineradicable historicity of the lingual, the means of the capture—the language of the reader’s time—is partly a means of help, partly an obstacle. A reader creates the meaning of a text by working through resources available in the historicized formation of his culture, which includes the aesthetic language, historically specific intellectual formations, and cultural sensibilities.¹⁷

Interestingly, Gadamer too illustrates his point about the literary intention of the receiver by an example from drama. Without an act of representational mediation—by a complex function of an actor/ director—the text cannot exist as drama: only when it is enacted is it truly a play, not the printed text. The text, in other words, contains the possibility of meaning, but not, in a fundamentally literal sense, meaning itself. In another evocative example, we cannot have music until it begins to sound: a musical score is a text which contains the potential of musical meaning, but not meaning in the real sense. There is a direction in Gadamer’s theory of textual meaning: it seeks to shift the emphasis in the analytics of interpretation from an excessively author-sided conception of texts into a more reader-sided one.

Although the earlier stratum of rasa theory was not author-sided in the same sense, the dhvani elaboration appears to move in the same direction as Gadamer’s. It elaborates a theory which assigns a proper and specific function to each of these figures in the field of literary meaning. Each of these functions—the author’s composition (A), the character’s function as the ‘vessel’ of the emotion (B),¹⁸ the actor’s evocation(C),¹⁹ and the spectator’s/reader’s reception/appreciation (D).²⁰ In the AB, in the long commentary on the rasasutra involving debates with Bhattalollata, Sankuka, and Bhattanayaka, theoretical attention is almost exclusively focused on relations between C and D, and secondarily on B and C. Particularly in some phases of the debates, for instance in Sankuka’s argument that the spectator goes through a process of inference starting from a necessary false cognition, the entire philosophical enterprise is to understand the conditions of possibility of spectatorial rapture.²¹ There can be two types of theories: theories of communication, which ask how the author communicates to the actor, and the actor to the spectator; or theories of reception, which ask how the spectator understands? What exactly does he see or feel? The NS, which constitutes the primary level of this theory, does not offer a theory of communication,²² and therefore it would be wrong to suggest that here, as in the case of the tradition running from Schleiermacher to Gadamer, a heavily author-sided theory is slowly balanced by a reader-sided correction. Rather, the NS states the whole point with such self-defeating terseness that the issue, which is central to an understanding of the rasa process, is left deeply obscure. The subsequent process of the dhvanikaras’ glossing produces a conception of the aesthetic field; but the distinctive form of its elaboration is from the side of the reader.

Two aspects of the theory appear remarkable: its specification of the series of relations of reception/communication in the aesthetic-textual field, and its close attention to the exact nature of the intentional acts performed by the different figures. It is not content by simply stating that these are intentional acts, or states. As intentional acts are directional (Searle),²³ the theory is trying to ascertain the exact direction, and the exact nature of the intentional act or state in each case, particularly B, C, and D.²⁴ The theory suggests that the idea of textual intention needs to be revised, and in fact there is a play of intentionalities, in the plural, on each text. There is an intentional node at each of these points. The fundamental structure of an aesthetic text is such that it allows the intervention of intentionalities at these vital points of the text coming into meaning—through the different kinds of intentionalities exercised by the author, the actor, musician, and the receiver. We are then saved from the need to appeal to the intention of the author in that narrow and absolute sense.

Another significant advantage of this revision is that this can accommodate the effects of history much more easily,²⁵ to deal with the ironic fate of all classic works—which float ‘in the waters of time’,²⁶ from their moorings in one age down to transient resting places in very different ages possessing radically different sensibilities. This leads to an ‘historical’ form of the ‘death of the author’—the original context of meaning becomes so distant that the methodological demand for a reconstruction of authorial intention becomes practically meaningless. Frequent legends of divine authorship rhetorically endorse this sense that the author and the mundane world of his intentions are irrecoverable. The reader has to find a strategy to deal with a text without an author.

If we decide not to accept their exact move to attribute the santa rasa intention to the author, but to work with the textual field of complex intentionalities—in which the actor’s intentionality plays upon the author’s, and the receiver’s upon the actor’s, it becomes possible to maintain that there can be various readings of the text, which are all simultaneously on offer. In case of the MB, the two contenders for interpretative primacy would be a conventional vira rasa reading and a santa rasa one.

Reading a Narrative Structure: ‘The Last of Life for which the First was Made’

To partially justify the santa rasa reading, we could use a modern technique: look at the narrative meaning of the epic’s ‘miserable end’. Endings are particularly significant in a narrative structure, if we take the meaning of the term narrative in a strict sense. Arthur Danto argued in his Narration and Knowledge that there is a peculiar structure of temporal relations which define a ‘narrative sentence’.²⁷ Two events which happened at different points in time but are linked by a peculiar relation are captured by a narrative sentence; but its narrative character, as opposed to other qualities, lies in the fact that the later event governs the understanding of the earlier one.²⁸ To state that the first prime minister of India was born in 1889 is to use a sentence with a narrative structure: because in 1889 the conditions for the existence of a prime minister were not there. A mere child was born to Motilal Nehru, who was of course, a successful barrister and leading figure of the Indian National Congress. Only after Independence in 1947 are the conditions for the existence of a prime minister of India in place. Statements about Nehru which seek some historical understanding of his life after 1947 must recognize this unavoidably significant fact. In a sense, the second event affects indelibly our understanding of the first one.

This can be easily generalized for historical and narrative series. A history of the Russian revolution written in the 1950s would be quite different from one written today: the events of that phase of Russian history have not changed, what has changed is the significance of those events by the emergent relation with new events like the fall of communism. Remarkably, this change is not due to an evaluative change in the observer, but to the occurrence of a subsequent set of events, which make it impossible to write the history of the earlier event without taking into account the latter ones. This is made evident in the autobiography, because in that form the writing of the text and the temporality of the last segment coincide. What happens in a narrative of fiction is identical to what happens in a real human life, and in autobiographic retelling: in a sense, narrative unfolding shows that it is for everybody, not Rabbi Ben Ezra alone, that it is the last of life for which the first was made: and we should trust the author, read all, not be afraid.²⁹

If we characterize this as the invariant structure of narrative judgement, we can see something similar occurring in the process of fictional unfolding. It is common knowledge that the meaning of a story is determined by not merely what happens inside the series of events constituting the story, but by how it ends. A story about a mutiny which ends at the moment of the mutineers’ triumph is a story that tells a very different story from one that ends with their eventual defeat. In fact, the idea that literature/poetry is niyatikrtaniyamarahita and ananyaparatantra stresses precisely this truth. Literary narrative is not conditional on historical truth, it is ananyaparatantra. A literary sequence of events is not obliged to follow on an historical sequence: it is unconditional in two senses: it is not conditioned by any sequence of events because of its facticity; it is also not constrained to follow the syntax of natural things. Within the universe of aesthetics, the rebellion is successful, however much and however finally it might be defeated in the universe of history. This is the sense in which the language of the poet (bharati kaverjayati) triumphs over the language of the chronicler tied to the rules and conditions of historical happenings. By shifting the end of the story to a point chronologically before the defeat of a rebellion the poet can help the rebels win—in a different universe.

Following this view of a narrative statement, we could offer a similar argument about the narrative structure in a story. A narrative structure can exist in both factual and fictional accounts, connecting these two by their common narrativity, which consists in a movement over time in which events that happen later add significant consequents to the earlier processes and modify their character. The overall meaning of the existing earlier series is modified by a significant later event, because the earlier series then becomes a ‘preparation’ or a series leading to the last segment in the chain. What the meaning of the chain of events is, is determined by the nature of the last segment. This will also throw some light on a commonly observed but analytically neglected process. The narratives of the Ramayana and the MB of course exist in hundreds of versions, and one of the major differentiating features between these stories is precisely the matter of their ending segment. As a child, I read some versions of the MB and the Ramayana which end with the conclusion of the Kuruksetra battle and with Rama’s coronation in Ayodhya. Often, dramatic enactments in the Ramlila use the same narrative device. It is inexact to view these recensions as a shorter form of the story; they are a different version. These do not stop the narrative series at a slightly earlier stage; by making a different segment the last one, they offer us a story with a different structure which then bears a seriously different meaning. In fact, fiction is teleological in a way that history is not.

Meanings of the Mahabharata

What is the meaning of the MB? Can we turn this question, given the two readings, into meanings of the two MBs? Modern literary theory often claims that narratives in modern literary cultures entertain a more pronounced aspiration towards cognitive understanding of the world and the characters’ place in it. Sanskrit aesthetic texts routinely assume that literary narratives contain a strong cognitive element: but they refine the exact tone of this cognitive function. Stories offer upadesa to their readers, but in the way of a prabhu, of a suhrt or of a kanta.³⁰ The Dhvanyaloka directly admits that the MB speaks to us in two styles: as sastra (as doctrine) and as kavya (as art). Literary cultures bear a close relation with interconnected cultures of religious and moral beliefs, but the MB has its own peculiar way of connecting these two spheres.

Ostensibly, the most effective way of teaching moral rules is to enunciate them clearly, and to offer arguments in favour. But such assertoric and argumentative presentation of ethical rules is not always effective, because by stating the rule blandly this teaching does not prepare us for the constant surprises of real moral life. For, in actual life the situations of ethical choice arise in complex, unpredictable, unrecognizable ways. To take an example from the MB, it is simple and easy to command the telling of truth; but this can work in case of a relatively simple choice between telling the truth and lying. The MB carefully constructs the tale of the hermit facing a band of robbers who ask him to tell the truth about a man they have pursued and who is hiding in his ashram.³¹ Simple sastric (assertoric) command does not take into account the situational complexity in which pursuing a good act has to occur; the complementary function of narrative is to warn readers about it. The narrative does not merely add a diverting example of a rule that is already clear; rather, it adds an indispensable cognitive dimension to our knowledge of moral life by suggesting that although the rule in its assertoric form is clear, it is misleading in its contextual bareness. The MB teaches ethical rules by a combination of these two functions: of enunciating principles, as in the Santiparva, and of narrative complication.

To understand the sastric uptake of the MB, we should caution ourselves that we are dealing with a pre-modern culture which does not presuppose the typically modern enunciation of moral rules as universal injunctions. These are not rules meant for ‘individuals’ who are all alike, living in a shared ethical universe.³² Characters live in a social world deeply segmented along varna lines. The story gives a narrative exposition of social ideals, particularly the norms of warrior courage, brahminic wisdom, and an aspect the epics never neglect—the corresponding virtues of women in these statuses. In elaborating both the norms and the attendant dangers of high status, the quite different feminine encounter with the world is elaborated with equal vividness: Sita and Draupadi are as indispensable to the narratives’ putative norms as Rama or the Pandavas. The virtues of heroism and contemplation do not exist separately, they belong together in a social world in which these two groups are dominant: therefore the ‘complete’ story of their society must be their story.

The First Mahabharata and its Inadequacies

On the conventional view, if we accept there are eight rasas, it is hard to avoid a conclusion that the predominant rasa of the MB must be the vira, if a text of that size and complexity can be said to have a dominant rasa at all. The narrative illustrates the virtues required by a social world of this character, especially three kinds of virtues—the intellectual ideals required by brahmanical life; the warrior ideals demanded by the truly Ksatriya life served in fearless service of a just order; and not least the life of a woman who has to pass through the uncertainties of such a life with undiminished dignity. They have to make their way in a world in which they do not take decisions, yet are most affected by them. According to this reading, the lives of the warriors are marked by uncertainties which are nonetheless faced with courage—dangers like facing a numerically larger army,³³ the unavoidable temptations of using unjust means.³⁴ The lives of the Pandavas show how true Ksatriyas can emerge out of such trials with honour. Even in this reading, the MB has much narrative complexity, but the global meaning of the expanse of narrative incidents consists in an eventual assurance that, although the proper order of the universe is sometimes seriously challenged, it is eventually restored: and Ksatriya glory lies in not pointless victories in mundane battles, but a great defining battle which has this restorative character. A conventional vira rasa reading would confirm a confidence in the unshakable ‘truth’ of this social order. The narrative is one that ends in triumph—not of the characters, the Pandavas, but of the principle (of satyameva jayate).³⁵

The MB has, however, strangely disturbing episodes, too numerous to recount, which put these virtues and the order which they reinforce into radical questioning. Kunti’s curiosity about her boon leads to the birth of her first, secret child, who is abandoned, but he is destined for glory and returns to trouble her peace and the lives of her other children. Or consider the little episode of the Vanaparva where the heroism of the four martially more accomplished brothers are mocked: they are brought back to life by a martially feeble, intellectual elder brother. Think of the beautiful description of Draupadi’s delight at Arjuna’s success,³⁶ and her dismay at Kunti’s thoughtless order which determines the rest of her life. This is revealed right at the end when Yudhisthira explains to Bhima the reason for Draupadi’s fall on the mahaprasthana.³⁷ Instead of confirming the central rules of the moral and social order, the story seems to take pleasure in constructing situations of transgression. There are of course the last incidents of their ‘miserable end’—not just the end of the Vrsnis in a drunken catastrophe which Krsna was powerless to stop, but the strange inability of Arjuna, the victor of Kuruksetra, to lift his bow when attacked by common thieves while escorting Vrsni women to safety.³⁸ There are two ways of looking for the dark end of the narrative. One is to follow the explicit instruction from the Dhvanyaloka where the dark reminder of catastrophe is contained at the end, narratively governing the meaning of the whole preceding sequence. An alternative way is to be sensitive to the signals of imminent failure, of small disasters that accompany overt victories, of the recurrent signals of moral unmaking constantly shadowing the march of heroic events.

The Second Mahabharata

The addition of the ninth rasa makes it possible to register, reflect, and imaginatively expand on these signs embedded in the interstices, almost the underside, of the narrative structure. Read through the theory of the nine rasas, we get a second MB, fundamentally different from the first in its moral flavour, and perhaps its historical import. Commentaries are of two kinds: in some, meanings which are condensed or inexplicit are elaborated by expounding on the meanings of words, their syntactical connections, and their interpretative connotation; but in this case the Dhvanyaloka makes a profound but exceptionally concise gesture towards the epic’s meaning, without further elaboration. The task of the elaboration is left to the reader. The interpretative act leaves the reader free to follow its direction creatively. The ‘miserable ending’ is not a formula to think through, which absolves readers of the responsibility of reflection, but a trigger to think, an incitement to our own interpretative imagination.

In this reading, the MB is a text of the santa rasa, a tragic sense of the world,³⁹ which reveals the insubstantiality of every single component of the Brahman-Ksatriya social-ethical ideal. It shows the political order built around kingship as desperately fallible, starting from the impossibility of determination of the rights of succession. The rule of primogeniture seeks to provide a clear, unquestionable line of royal succession. But the story leads us to a situation where both sides of a clan can lay claim to rightful kingship, and can have a complaint that they lost it by an unjust turn of fate.⁴⁰ It shows the futility of valour which rarely achieves justice, particularly because sometimes the most important injustices happen not on the battlefield but inside the secure and ordered space of the court. The greatest iniquity of the MB occurs not in the blind and desperate moments of battles but in the public space of the court.

This space marked for the announcement and enforcement of justice is turned into the space for the gravest iniquity. Renowned warriors allow themselves to become parts the most inhuman degradation of a woman. In its totality, the story shows that justice is rarely achieved in the mundane world. Beauty is only rewarded by lust, vulnerability, and indignity. From the moment she enters the world, Draupadi is an object of desire and contention.⁴¹ Detachment is often indistinguishable from moral failure. There can be endless dispute on whether Bhishma and Drona were at that fateful moment detached or simply morally feeble. Revenge is peculiarly unsatisfying: because though at the end of the battle the Pandavas have extracted revenge, it is a catastrophe for the winners as much as for the losers. In the night of the tarpana, when dead warriors are reunited with their loved ones for a few hours of revival, they join in an indistinguishably common mourning.⁴²

Other signs in the text are equally profound and intriguing. If we accept that the sastric element is very significant in the MB, the manner in which this is delivered is astonishingly unusual. The great philosophical peroration in the text is delivered not by a Brahman whose socially designated function is contemplation, but by a warrior. Krishna is, however, a warrior who has given up war, and decided through a subterfuge of common kinship not to fight at Kuruksetra. Is there an ‘event’ in this apparent anomaly? Should we read this narrative fact? At the end one might wonder if this story points to any sukha at all, even trsna-ksay-asukha, which, according to this theory, only the vimalapartibhanasali readers will be able to grasp. If the purpose of the text is to produce moral knowledge—not in the sense of teaching individual moral principles, but showing what leading a moral life involves—then the darker version of the story throws a clearer light upon the world. It does not show us how to lead a moral life; it suggests, rather, that to lead a moral life is something that cannot be shown. The goodness of acts has to be invented at every significant step: and individuals should live in humility at the contingency, fallibility, and imperfection of their effort at being good.

Aesthetic and Social Ideals

Other questions, of radically non-aesthetic kinds, could be raised about the two readings of the epic. It is fascinating to ask how this aesthetic reflection might be related to the moral universe of this historical epoch.⁴³ Where did this suggestion of adding the santa to the existing rasa register come from? As noticed earlier, V. Raghavan suggests, intriguingly, that the idea of the santa did not come an internal elaboration of conventional brahminical religious thought, but is an original invention of the Buddhist tradition.⁴⁴ Buddhist philosophical reflection is much concerned with the meaning of tranquillity in conduct, and the santa is the predominant rasa of the Buddhist play Naganada.⁴⁵ Kashmiri Saiva philosophers were engaged in a fierce doctrinal battle against the Buddhists and played a significant role in the eventual decline in Buddhist religious influence. Despite this overt doctrinal conflict, on Raghavan’s view Abhinavagupta acted in a manner not uncommon in great philosophic debates: one side absorbs what it considers the most valuable, and therefore the less controvertible, elements from the adversarial theory. This strengthens the theory which can effect this subsumption; and it also ensures, for the surrounding intellectual culture, that, despite the decline of a school, the valuable elements of its thinking are never lost.⁴⁶ Saiva thinking, while refuting Buddhist arguments, enriched its aesthetic thought by incorporating the most valuable contribution of the Buddhists, probably by converting a philosophical thesis into an aesthetic theory. Historians have noted that Kashmir in Abhinavagupta’s time seems to be going through serious social and ideological transformation.⁴⁷ Buddhism, after all, presented the most fundamental challenge to the moral ideal of classical brahminical society. It is always problematic to connect the changes in aesthetic thinking with contextual social change too directly. But it is an interesting, if speculative, question to ask: does the revision of aesthetic theory and the resultant re-reading of the rasa of the epic text bear any relation to a transformation of intellectual culture in response to a fundamental social crisis?

Alexis Sanderson has suggested that Saivas were elaborating a new model of religiosity that ‘tried to transcend the disjunctions and oppositions of the brahmanical social order’ by finding through their doctrine of pratyabhijna a wider base for Hindu religious life.⁴⁸In the absence of more accurate sociological knowledge, it is possible only to speculate about what the social and historical roots of such momentous revisions were in patterns of intellectual life. But the aesthetic revision is undeniable. Kashmiri theorists deserve our eternal gratitude for their amazing gift to all future readers—without changing a single word of the text, they managed to give us a second Mahabharata.

References

Abhinavagupta. 1956. Abhinavabharati, in M. Ramakrishna Kavi, ed., Natyasastra of Bharatamuni with the Commentary Abhinavabharati of Abhinavaguptacarya, Gaekwad Oriental Series, No. 36 [NS]. Baroda: Sadhana Press.

Althusser, Louis. 1970. For Marx. London: Allen Lane, Penguin.

Anandavardhana. 1990. Dhvanyaloka. Cambridge, Mass.: Harvard University Press.

Bakhtin, Mikhail. 1990. Epic and Novel. In Mikhail Bakhtin, The Dialogic Imagination. Austin: The University of Texas Press.

Danto, Arthur. 1985. Narration and Knowledge. New York: Columbia University Press.

Gadamer, H.G. 1981. Truth and Method. London: Sheed and Ward.

Kavyaprakasa of Mammata. 1965. Ed. Vamanacharya Jhalkikar, 7th edition. Poona: Bhandarkar Oriental Research Institute.

Mahabharata. 1927–66. Ed. V.S. Sukhthankar, et al. Poona: Bhandarkar Oriental Institute.

Pollock, Sheldon. 2006. The Language of the Gods in the World of Men. Berkeley: The University of California Press.

Raghavan, V. 1967. The Number of Rasas. Madras: Adyar Research Centre and Library.

Ricoeur, Paul. 1976. Interpretation Theory. Fort Worth: Texas Christian University Press.

Sanderson, Alexis. 2006 Saivism and Brahamanism in the Early Medieval Period. Gonda Lectures. Amsterdam: The Royal Netherlands Academy of Arts and Sciences.

Searle, John. 1970. Speech Acts. Cambridge: Cambridge University Press.

Sri Harsa. 1992. Naganada. Delhi: Motilal Banarasidass.

Tagore, Rabindranath. 1972. Karna-kunti-samvad. In Katha o Kahini. Sancayita. Kolkata: Visvabharati.

Tubb, Gary. 1985. The Santarasa in the Mahabharata. In Arvind Sharma, ed., Essays on the Mahabharata. Journal of South Asian Literature. Winter-Spring.

Visvanatha. 2004. Sahityadarpana. Delhi: Motilal Banarasidass.

¹ These arguments are discussed and serially rejected in the first section of the Abhinavagupta (hereafter AB) commentary on the rasasutra. See Abhinavagupta 1992: ch. 6. The idea of utpatti—origin or irruption—is associated with Lollata; pratiti—inference—with Sankuka; vyanjana—coming into expression—with Anandavardhana; and the crucial concept of sadharanikarana (generalization/universalization) which Abhinava accepts, but uses somewhat differently, from the provocative Bhattanayaka who said with brilliant provocation ‘rasa na utpadyate, na pratiyate, nabhibyanjate’ (rasa neither erupts, nor becomes an object of inference, nor comes into expression). Ibid.: AB commentary, 278.

² Althusser 1970: see the chapter ‘On the Materialist Dialectic’. For instance, there is Althusser’s insistence that a concept like alienation may be used by the theories of both the young and the mature Marx, but their theoretical significance might be different, depending on exactly where the concept is placed in the general structure of the conceptual architecture.

³ This is probably the most celebrated line in Indian aesthetics: Abhinavagupta 1992: ch. 6, the rasadhyaya. But by the fourteenth century, in Visvanatha Kaviraja’s systematizing treatise Sahityadarpana, the ninth is equally authoritatively established by a deft displacement of words:

srngarahasyakarunaraudravirabhayanakah

vibhatasodbhuta etyastau rasa santastatha matah

Visvanatha 2004: ch. 3, 182, 106.

⁴ Raghavan 1967.

⁵ This is a serious question for Abhinavagupta, shown by the fact that he returns to this at several places: for instance, another famous passage in the commentary on the rasasutra, where he introduces the metaphor of the vivekasopanaparamapara (Abhinavagupta 1992: 280), the ladder of reasoning. I have discussed this passage briefly in ‘The Sudden Death of Sanskrit Knowledge’ (reproduced within the present volume).

⁶ For details of the argument, see Raghavan 1967.

⁷ This suggestion is advanced, with typical economy and audacity, in the final sections of Anandavardhana’s Dhvanyaloka, and taken up by Abhinava’s interlinked commentary, the Locana. See Anandvardhana 1990: 696.

⁸ Tubb 1985: 141–67.

⁹ Anandavardhana 1990: 691.

¹⁰ Ibid.

¹¹ Ibid.: 690–1. Emphasis added.

¹² I am referring to the well-known passages in Truth and Method, regarding the meaning of an act of representation. See Gadamer 1981: 97ff.

¹³ I wish to stress that I am not a scholar of Sanskrit or of ancient Indian aesthetics, and therefore the following discussion depends exclusively on modern, primarily Western, cultural theory.

¹⁴ The first section of the commentary on the rasasutra concisely present these successive attempts at elucidation of that event as upaciti, pratiti, vyanjana and sadharanikarana. Abhinavagupta 1992: ch. 6.

¹⁵ It is interesting to compare the pictures of the act of interpretation in modern theories like Ricouer’s and the Kashmiri one. Ricoeur’s Interpretation Theory places the text, as in Gadamer, between the author and the reader:

author...........[_text_].............reader

The Kashmiri theorists view the field of interpretation/signification as author..........character..............actor/reciter..............reader/hearer/spectator

See Ricoeur 1976.

¹⁶ Gadamer 1981: pt II, 91ff.

¹⁷ We are dealing with the narrowly aesthetic considerations of historicity, but Gadamer also advances much broader ontological arguments regarding obstacles to a perfect re-imagination of authorial meanings.

¹⁸ The term for actor patra—a vessel—captures this inflection perfectly.

¹⁹ The dhvani tradition produces the most elaborate, intricate, and subtle analysis of the relation between the actor and the spectator, through its argument, developed particularly in Abhinavagupta, that true understanding must avoid the errors of svagatatva and paragatatva. Abhinavagupta 1992: 281–3. It also offers astonishing subtle reflections on even apparently mundane things like our understanding of the idea that Krsna was a beautiful person through the suggestive presence of a handsome actor. We must not believe that Rama was handsome literally like actor X, who is also handsome; we, in any case, have no means of knowing how handsome Krsna really was; we must submit ourselves to a free-floating conception of ‘handsomeness’ which is generalised. Abhinavagupta 1992: 281–3.

²⁰ Appreciation is probably a better translation, because although there is a trend towards sadharanikarana or universalization, the offer of the text is not indiscriminate; there is an opposite tendency at work when it specifies that appreciation can be achieved only by those who are vimalapratibhanasali receivers of art.

²¹ Rapture, although awkward in other respects, is appropriate because the theory emphasizes the spectators’ absorption (raptness) which the Sahityadarpana puts beautifully as ‘vedyantara-sparsa-sunya’, untouched by any other perception. Visvanatha 2004: chs. 3, 2, 48.

²² It is remarkable how little the rasasutra commits itself to: ‘vibhavaanubhava-vyabhicari-samyogat rasa-nispattih’—the sentence is interpretatively primitive.

²³ Searle 1970.

²⁴ Paradaoxically, the Dhvanyaloka discussion on the MB, after spending a great deal of effort to disentangle the intentional stances of A, B, and D, which opens up the possibility of developing a theory which protects us from using authorial intention as the final court of appeal, relapses into an argument which critically depends on authorial intent. By ‘the miserable end of the Vrsnis and the Pandavas’, they maintain, the author wishes to show us that the cognitive purpose of the epic is a cessation of desire.

²⁵ Again remarkably similar to Gadamer’s ideas about effective historicity. Gadamer 1981: 305–41.

²⁶ The persistent Sanskrit metaphor of kalasrotah—the river of time.

²⁷ Danto 1985.

²⁸ Ibid.: ch. VIII.

²⁹ Browning’s poem, ‘Rabbi Ben Ezra’, Stanza 1.

³⁰ See Mammata’s Kavyaprakasa, Ist Ullasa, 2.

³¹ Arjuna’s demeanour of untroubled confidence in his unequal battle with the Kuru army during the uttaragograha (the capture of Virata’s cattle from the northern side) at the end of the ‘goharanaparvadhyaya’, in Virataparva, Mahabharata.

³² See Bakhtin’s argument that epic characters come from a social universe which are segmented between groups, but homogeneous within them. See Bakhtin 1990.

³³ Arjuna’s demeanour of untroubled confidence in his unequal battle with the Kuru army during the uttaragograha (the capture of Virata’s cattle from the northern side) at the end of the ‘goharanaparvadhyaya’, in Virataparva, MB.

³⁴ A celebrated example is Yudhisthira’s eventual agreement to use a partial lie in the death of Drona, his teacher. Dronaparva, MB.

³⁵ In various parts of the narrative this is indicated directly. For instance, when Yudhisthira asks for Drona’s blessings before the commencement of the battle, Drona states ‘yato Ksrna stato dharmah yato dharma stato jayah’. Bhismavadhaparvadhyaya, Bhismaparva, MB.

³⁶ Viddhantu laksyam prasamiksya krsna

parthantu sakrapratimam niriksya

svabhyastarupapi naveva nityam

vinapi hasam hasativa kanya

madadretehpi skhalativa bhavair

baca-byaharativa drastya. Svayamvaraparvadhyaya, Adiparva MB 1.179.22

³⁷ Mahaprasthanikaparva, MB.

³⁸ Mausalaparva, MB.

³⁹ The use of the term tragic is both helpful and misleading: that is the closest term of technical aesthetics by which we can partly elucidate what the santa means; but it is, in the strict sense, quite different from the tragic sense. It is a form of tranquillity that is achieved after going through the experience of suffering, when its intensity is stilled and the immediate suffering is distanced. In Tagore’s poem, Karna describes the end of the war in similar terms: heritechhi santimay sunya parinam (I can see the end—empty, and tranquil), and he asks Kunti: ‘je pakser parajay, se paksa tyajite more koro na ahvan’ (Do not ask me to leave the side of those destined for defeat). Tagore 1972: 403.

⁴⁰ Dhrtarastra was blind, but should this disability cause the royal line of pass to Pandu’s lineage, or is it a disability which is personal and should not affect his line?

⁴¹ It is doubtful if any modern narrative can rival the MB in its demonstration of the terrible effects of regarding women as objects of desire, and as tokens of the indirect exchanges of male malevolence.

⁴² Putradarsanaparvadhyaya, Asramavasikaparva, MB. It is interesting to reflect on the nature of this narrative element. It is for the men to destroy, and for women to try to heal, for men to kill, and for the women to mourn.

⁴³ For a magisterial survey of the history of Sanskrit literature, see Pollock 2006, and the central question of the work is the connection between culture and power.

⁴⁴ Raghavan 1967. ‘Udbhata recognizes the santa as can be seen from his Kavyalamkarasarasamgraha. He is thus the first commentator on the NS and the first alankarika now known to have definitely begun to speak of Rasas as nine in number. He may therefore have made the necessary alteration in the text of the NS as shown above and pointed out by Abhinavagupta.’ (p. 13). The revised text is:

srngarahasyakarunah raudravirabhayanakah

vibhastadbhutasantasca nava natye rasah smrtah (p.13).

⁴⁵ Sri Harsa 1992.

⁴⁶ It is possible to find examples of this kind from contemporary social and political theory. Rawls’s theory of liberalism could be viewed as one which subsumes elements from socialist thought—like its concern for justice and egalitarianism—to construct a version of liberalism that is much harder to argue against.

⁴⁷ Pollock 2006.

⁴⁸ Sanderson 2006: 5.