5.-Vishwa1

Frame Narratives aNd Forked BegiNNiNgs:or, How to read tHe ÅdiparvaN*

Vishwa P. Adluri

1. Intoduction

The Mahābhārata contains two beginnings at 1.1 and 1.4, each beginning with the identical line lomaharṣaṇaputra ugraśravāḥ sūtaḥ paurāṇiko naimiṣāraṇye śaunakasya kulapater dvādaśavārṣike satre (1.1.1 and 1.4.1).1 In

this paper, I demonstrate the failure of text-historical methods to provide an adequate explanation for this feature. Arguing that it is no mere accident, but a meaningful duplication, I show how the double beginning is integral to the form, content, and function of the epic.

The Indian epic, the Mahābhārata, is a literary work of stunning complexity. At its core,2 this narrative recounts the story of the vicissitudes of the Kuru dynasty, but this story is located within a far more comprehensive literary and philosophical program.3 The central story of a fratricidal conflict over the throne of Hāstinapura is nothing if not a cipher for fundamental philosophical and cosmological reflections concerning issues such as time (kāla),4 fate (daivam),5 right action (dharma), the fulfillment of the goals of human life (the puruṣārthas),6 the soul and the nature of consciousness,7 and the being that transcends all becoming (Kṛṣṇa Vāsudeva).8 The complexity of this philosophical vision places tremendous demands on the epic’s structure,9 forcing it to break with the bounds of a simple linear narrative.10 In this paper, I analyze some aspects of the epic’s structure, focusing especially on the first major book of the Mahābhārata, the Ādiparvan.

I consider the frame narratives of the Ādiparvan in the next section (2. The Frame Narratives of the Ādiparvan), before showing how the outermost frame is split to include not one but two beginnings (3. The Double Beginning

143

Journal of Vaishnava Studies144

of the Ādiparvan). In the fourth section (Many Beginnings, Many Models), I then analyze some earlier responses to the problem, focusing especially on van Buitenen’s thesis of historical expansion and Oberlies’ more recent thesis that the epic is comprised of multiple “versions” laid one on top of the other. Section five (The Double Beginning in the Critical Edition) then examines the evidence for the double beginning in the Critical Edition (CE), where I show that the manuscript evidence does not support either of these historicist models. I also discuss arguments that the CE succeeds in recovering a text that is more archetypal than its editors could have imagined. Thus, on philological grounds alone, we must set aside “text-historical” speculations such as those of Oberlies and Tsuchida (on whose work Oberlies largely bases his argument), and attempt to understand the double beginning as a meaningful and necessary component of the epic’s narrative architecture.

In section six (Two Beginnings, Two Narrations), I begin the positive part of the interpretation. Through a careful reading of the opening books of the Ādiparvan, I show how Mehta and Tsuchida overlook crucial evidence that we have not one but two Ugraśravas narrations in the epic: one beginning in the Anukramaṇiparvan and corresponding to the first beginning, the other beginning in the Paulomaparvan and corresponding to the second beginning. Once we overcome “text-historical” prejudices that the Mahābhārata is a badly composed text,11 essentially an accident of history, we are then in a position to understand the specific function of the two beginnings. In section seven (Cosmological and Genealogical Beginnings: The Anukramaṇi- and Paulomaparvans), I then demonstrate how the two beginnings provide alternative ways of entering the narrative of “becoming.” In section eight (A Fork in the Beginning: Hermeneutics in the Pauṣyaparvan), I then focus on the space between the two beginnings. I show how the double beginning creates a non-narrative space between the Anukramaṇi- and Paulomaparvans, which allows for the text to convey its own hermeneutic apparatus in the form of the Pauṣyaparvan. Finally, I return to Oberlies’ “onion-skin” model of the epic in section nine (Return to the Problem of “Text-Historical” “Criticism”) and then end, in section ten (Conclusion: Establishing Criteria), with some concluding remarks on the epic as a whole.

Finally, a note on terms. The epic makes use of various terms to describe its major sections and their subdivisions, and it is important to keep these apart in the discussion that follows. The epic is primarily articulated into eighteen parvans or “major books” (as van Buitenen translates). These parvans are further divided into varying numbers of upaparvan or “minor books,” with the upaparvans again containing a number of ākhyānas (“narratives”) and being divided into a number of adhyāyas (“chapters”). Only the first two divisions, i.e.,

How to Read the Ådiparvan 145

the parvans and the upaparvans, are of significance to this paper. I will either use the Sanskrit terms or, following van Buitenen’s usage, refer to them as “major books” or “minor books.” I have tried to be as consistent as possible in this usage, except for a few instances where it would have made for awkward reading. In those cases, however, context makes it amply clear whether the reference is to a “major” or a “minor” book. This paper is mainly restricted to a discussion of the first major book, the Ādiparvan; of this book, I focus principally on the first six minor books. The following chart lays out the eighteen parvans, the number of upaparvans in each, and then expands on the six upaparvans of the Ādiparvan that are central to this study.

* All figures taken from van Buitenen 1973

1

No. Name of parvan (Major book)

Number of upaparvans (Minor books)

Upaparvans 1-6 of the Ādiparvan

1 Ādiparvan (The Book of the Beginning)

1-225

2 Sabhāparvan (The Book of the Assembly Hall)

1-72

3 Āraṇyakaparvan (The Book of the Forest)

1-298

4 Virāṭaparvan (The Book of Virāṭa)

1-67

5 Udyogaparvan (The Book of the Effort)

1-197

6 Bhīsṃaparvan (The Book of Bhīṣma)

1-117

7 Droṇaparvan (The Book of Droṇa)

1-173

8 Karṇaparvan (The Book of Karṇa)

1-69

9 Śalyaparvan (The Book of Śalya)

1-64

Ugraśravas-Śaunaka dialogical level: 1. Anukramaṇiaparvan (The Lists of Contents) 2. Parvasaṁgrahaparvan (The Summaries of the Books) 3. Pauṣyaparvan (The Book of Pauṣya) 4. Paulomaparvan (The Book of Puloman) 5. Āstīkaparvan (The Book of Āstīka)

10 Sauptikaparvan (The Book of the Sleeping Warriors)

1-18

11 Strīparvan (The Book of the Women)

1-27

12 Śāntiparvan (The Book of the Peace)

1-353

13 Anuśāsanaparvan (The Book of the Instructions)

1-154

14 Āśvamedhikaparvan (The Book of the Horse Sacrifice)

1-96

15 Āśramavāsikaparvan (The Book of the Hermitage)

1-47

16 Mausalaparvan (The Book of the Clubs)

1-8

17 Mahāprasthānikaparvan (The Book of the Great Journey)

1-3

18 Svargārohaṇaparvan (The Book of the Ascent to Heaven)

1-15

Vaiśaṁpāyana-Janamejaya dialogical level: 6. Ādivaṁśāvataraṇaparvan


2. The Frame Narratives of the ĀdiparvanOne of the most puzzling aspects of the Mahābhārata is the way some narratives are contained within other narratives. Scholars use the words “enframing” or “emboxment” to describe the way more remote narratives are used to convey other narratives. At its simplest, this technique allows the epic to answer basic questions such as “by whom was this story narrated? And to whom? Where did this narration occur?,”12 but the use of this technique is not restricted to providing appropriate context. Were its function merely literary, we would not find such a complexity of narratives, each interlocking with others in unexpected and subtle ways. Nor can a simplistic use of this principle preclude the problem of establishing a context for the outermost narration. In a linear narrative structure, the outermost narration must always remain open, as this level is attributed to the omniscient narrator.13 Let us see how the Ādiparvan addresses this problem.

According to Mahābhārata 1.53, the brahmin Vaiśaṁpāyana first recounted the story of the bheda or conflict between the two branches of the Kuru line in the presence of Vyāsa, the epic’s traditional author, at king Janamejaya’s sarpasatra or snake sacrifice. Vaiśaṁpāyana’s narration in turn is placed within a second narrative, that of how the snake sacrifice came about. The original bheda narrative is thus itself placed within a second bheda narrative, the story of a conflict between snakes14 and men—a conflict in which two descendants of the Kuru line once again play a major role.15 But who recounts this narrative? The epic tells us that the bard Ugraśravas, arriving once at the Naimiṣa forest, came across a group of brahmin seers or ṛṣis attending Śaunaka’s twelve-year sacrificial rite and recounted the Mahābhārata at their behest. Unlike Vaiśaṁpāyana’s narration, which begins with the specific details of the conflict between the young Kuru princes, Ugraśravas’ narrative is more expansive. It begins, among other things, with a story of how king Janamejaya resolved upon the snake sacrifice in order to avenge his father Parikṣit’s death, and leads up to the snake sacrifice at Takṣaśīlā (cf. 18.5.29) at which Vaiśaṁpāyana narrated the Mahābhārata to the king. Vaiśaṁpāyana’s narration is thus enfolded within a second level of narration. Ugraśravas, in recounting the Mahābhārata, actually recounts the story he overheard at the snake sacrifice. In other words, his narration is both a story about an earlier narration and a narration of the story heard at this earlier narration.

The epic thus from the very beginning establishes two levels of narration:1. Vaiśaṁpāyana’s narration of the Kuru narrative to Janamejaya in the

intervals of the snake sacrifice, and


2. The bard Ugraśravas’ narration of the events leading up to the snake sacrifice, the performance of the snake sacrifice, the arrival of sage Vyāsa and his disciples at this sacrifice, and Vaiśaṁpāyana’s narration of the Mahābhārata.

One could expect this logic to continue indefinitely—if Ugraśravas’ narration provides a setting and context for Vaiśaṁpāyana’s narration, who relates Ugraśravas’ narration?—but Ugraśravas’ narration constitutes a final level of narration beyond which there is no further regress.16 As readers, we first receive the Mahābhārata from Ugraśravas17 and only subsequently become privy to Vaiśaṁpāyana’s narration. Ugraśravas’ arrival at the Naimiṣa forest is placed in the mouth of an omniscient and nameless narrator, who introduces Ugraśravas to the reader as a bard accomplished in ancient lore (sūtaḥ pauraṇiko; 1.1.1) and describes the initial exchange between the seers and the bard. Ugraśravas then begins narrating at 1.1.20 and, except for a periodic sautir uvāca (“the bard said”) or variants thereof to remind us who is speaking, the omniscient narrator recedes into the background. Ugraśravas’ narration continues until 1.53, at which point Vaiśaṁpāyana is introduced as the narrator.18 With the exception of a few significant returns to the Ugraśravas dialogical level, the main narration remains at the Vaiśaṁpāyana-Janamejaya dialogical level throughout. Toward the end of the epic, the process is reversed: Vaiśaṁpāyana concludes his narration in the final major book (the Svargārohaṇaparvan), and the narrative returns to Ugraśravas’ narration. Ugraśravas describes the conclusion of Janamejaya’s snake sacrifice, and, mentioning a few other details, brings the great epic to a close.19

The following outline clarifies the relationship of the frame narratives to the epic:

A. Frame level: Outer (first) frame20

Setting: Naimiṣa forestDialogical level: Ugraśravas -> Naimiṣeya ṛṣis

Narration of ancillary narratives (esp. the snake sacrifice) and of Vaiśaṁpāyana’s narration of “The Mahābhārata”

B. Frame level: Inner (second) frame

Setting: Janamejaya’s sarpasatraDialogical level: Vaiśaṁpāyana -> Janamejaya

Narration of “The Mahābhārata”


3. The Double Beginning of the ĀdiparvanIt would thus seem as though the main purpose of the frame narratives is to provide a plausible context for the narration of such a long epic. As C. Z. Minkowski has argued,

the extended ritual activity of a sattra, with its cyclical daily activity and its long breaks, during which the king as dikṣita must remain in his state of consecration by following only elevating pursuits and speaking only true things, provides a believable setting for the narration of an epic as long as the Mahābhārata. From the point of view, therefore, of narrative credibility, the sattra is a good choice for a frame story. (Minkowski 1989: 403)

However, there is more to the epic’s frame narratives than just providing a believable setting. The Mahābhārata does not just duplicate the frame narratives—placing Vaiśaṁpāyana’s narration at a satra within the context of another satra—but duplicates the beginning itself. Mahābhārata 1.1 recounts Ugraśravas’ arrival at the Naimiṣa forest21 and subsequent narration of the epic but, three minor books later, the epic once again recounts the story of Ugraśravas’ arrival at the Naimiṣa forest, as though nothing had gone before! Once again, the bard greets the sages assembled at Śaunaka’s sacrificial rite and inquires of them what they would like to hear.22 The sages ask the bard to wait for their chieftain (kulapatir; 1.4.523) Śaunaka’s arrival and to narrate the stories Śaunaka asks of him. Ugraśravas complies and this time his narration continues all the way to Janamejaya’s snake sacrifice and Vaiśaṁpāyana’s arrival and subsequent narration of the Mahābhārata at this event.

However, we would be wrong to reduce one of the beginnings to the other. Even though the two beginnings share the same opening prose sentence—“The Bard Ugraśravas, the son of Lomaharṣaṇa, singer of the ancient Lore, once came to the Naimiṣa Forest, where the seers of strict vows were sitting together at the Twelve-year Session of family chieftain Śaunaka”24—the two beginnings do not simply replicate each other’s contents. In the first minor book, Ugraśravas begins his narration with a benediction to the Primeval Person (ādyaṁ puruṣam) at 1.1.20 and continues with a cosmology that describes the birth of the grandfather Prajāpati from the “large Egg” (bṛhad aṇḍam), the undecaying seed (bījam akṣayam; 1.1.27) of all beings. The fourth minor book, in contrast, begins with a genealogical narrative: Śaunaka, having arrived and seated himself, asks the bard whether he too has learned the “entire stock of ancient Lore” (purāṇam akhilaṁ; 1.5.1) which his father Lomaharṣaṇa used to narrate. Then he continues, “Now from among all the tales, I would first like to hear the one of the Descent of the


Bhṛgus [vaṁśam ... bhārgavam; 1.5.3].25 Tell the tale—we are eager to hear you.” We also cannot dismiss the first beginning as irrelevant to this narration.

Besides enclosing the all-important Lists of Contents (the Anukramaṇiparvan, the first minor book) and the Summaries of the Books (the Parvasaṁgrahaparvan, the second minor book), the first beginning introduces Janamejaya in the third minor book and sets up a triad of sacrifices culminating in the great sarpasatra of the fifth minor book.26 Although modern scholars have frequently undertaken their critical “surgeries” on the text with the intent of smoothing out the epic’s narrative structure,27 one cannot appreciate the meaning of the snake sacrifice without the hermeneutic and pedagogical background provided in the first three minor books and, especially, in the third minor book.28 In the next section, I discuss some of the limitations of these approaches, followed by a discussion of the double beginning itself.

The following outline clarifies the “split” in the outermost frame narrative:

A1. Frame level: Outer (first) frame

Setting: Naimiṣa forestDialogical level: Ugraśravas -> Naimiṣeya ṛṣis Beginning: First (Anukramaṇiparvan)

Description of Ugraśravas’ arrival at the Naimiṣa forest; Ugraśravas requested to narrate the epic; cosmological beginning (including summaries) continues until the end of the Pauṣyaparvan.

A2. Frame level: Outer (first) frame

Setting: Naimiṣa forestDialogical level: Ugraśravas -> Śaunaka Beginning: Second (Paulomaparvan)

Description of Ugraśravas’ arrival at the Naimiṣa; Ugraśravas requested to narrate the epic; genealogical beginning continues until the introduction of Vaiśaṁpāyana in the Āstīkaparvan.

4. Many Beginnings, Many ModelsJ. A. B. van Buitenen has argued that the Ādiparvan “illustrates to perfection all the issues that the text as a whole raises.” “Parts of it,” he continues, “are man–ifestly components of the main story; others are equally obviously accretions

that have no organic relationship to the story whatever; still others are difficult to determine one way or the other.” Citing line 1.1.50—“There are brahmins who learn The Bhārata from Manu onward, others again from the tale of The Book of Āstīka onward, others again from the tale of Uparicara onward”29—in support, van Buitenen argues that “it is reasonably clear that it [i.e., the “main story”] could hardly have begun before 1.90, and all that went before, roughly half the entire book, was added at a later time” (van Buitenen 1973: 1). He thus discounts not only the snake cycle, which attains a preliminary end in the Āstīkaparvan,30 but also the genealogy of the Kuru line recounted through the stories of Śakuntalā and Yayāti.31 Other scholars such as Oberlies have advanced an “onion skin” model of the epic, according to which the epic is composed of a series of “versions” that can be peeled off like the skin of an onion. Oberlies writes,

According to Mbh 1,1.50 there are “some Brahmans who learn the Bhārata beginning with Manu, others who learn it beginning with Āstīka, and [again] others who learn it from Uparicara on in the right way” (manvādi bhāratam kecid āstīkādi thatāpare/ thatoparicarādy anye viprāḥ samyag adhīyate). The beginning from “Manu” on may refer to Mbh 1,1.27ff., where the creation of the world is reported. Āstīka and Vasu in contrast clearly target Mbh 1,3/13 and 1/57. If then this hint of the Mahābhārata is to be understood ‘historically’ [historisch], the reference here would be to three versions of the text. And if one compares these against the text present to us in the critical edition, the ‘Manu-version’ would be characterized through the fact that it ‘begins with the beginning’, the ‘Āstīka-version’ through the fact that it lacks the outermost frame (and therewith the first dialogical level) and the list of contents, and the ‘Vasu-version’ through the fact that it would be without the narrations of Āstīka and (therewith without) the inner frame, that of the ‘snake sacrifice’, (as well as without the outer frame and the first four lists of contents). In other words, the distinguishing characteristic of these three versions could be the absence of a frame—the ‘Manu-version’ would be that with two, the ‘Āstīka-version’ that with one frame, while such [a frame] would be completely lacking for the ‘Vasu-version’. (Oberlies 2008: 87-88)32

Unfortunately, Oberlies never tells us his reasons for assuming that this “hint of the Mahābhārata” is to be understood “‘historically’ [historisch].” I am unclear whether his use of scare quotes is meant to imply neutral distancing or an ironic use of the term, but in any case it suggests that Oberlies is himself not quite clear about what it could possibly mean for an epic to hint “historically.”33 Does Oberlies mean that the epic is hinting at its own historical growth? Or that we should examine its literary form for clues (“hints”) to its history? If the former, was this hint a deliberate historical addition? Or could the epic poets have intentionally



manipulated the epic’s literary form to make it appear as a historical account?34 If the latter, what guarantee is there that these “hints” actually exist given that we possess no evidence of this history other than these texts?35

In addition to these theoretical issues, Oberlies’ insistence on a historical interpretation is also textually problematic as it implies either textual or historical priority of one passage over the other. The passage, however, merely states there are brahmins who learn the Mahābhārata from this point onward (cf. -ādi), without making any claims as to an absolute (textual or historical) priority of one beginning over the other. The three beginnings are therefore best understood as three distinct points to enter the narrative, rather than as referring to the historical priority of one narrative over the other. In other words, we have a single, continuous narrative that offers us a multiplicity of beginnings. We may contrast Oberlies’ historicist approach with the 13th century Vaiṣṇava philosopher Madhvācārya, who reads the same passage allegorically. Madhvācārya writes:

The meaning of the Bhārata, in so far as it is a relation of the events with which Śrī Kṛṣṇa and the Pāṇḍavas are connected, is āstikādi, or historical. That interpretation by which we find lessons on virtue, divine love, and the other ten qualities, on sacred duty and righteous practices, on character and training, on Brahmā and the other gods, is called manvādi, or religious and moral. Thirdly, the interpretation by which every sentence, word or syllable is shown to be the significant name, or to be the declaration of the glories, of the Almighty Ruler of the universe, is called auparicara or transcendental. (Cited and trans. in Klostermaier 2007: 62)

One could, of course, dispute Madhvācārya’s suggestion on the grounds that it leaves the ground of pure history to speak about transcendent realities such as morality or an Almighty.36 The historical reading has the advantage that it is accessible to everyman, and allows for a simple model of the epic’s growth. But here, too, matters are not as simple as Oberlies’ three-stage model suggests. The outer bheda narrative—the story of the conflict between snakes and men—overlaps with the central bheda narrative in ways that are both subtle and problematic.37 For example, the narrative of the bheda or “breach” between the Kurus and the Pāṇḍavas continuously overlaps with the narrative of the “breach” between snakes and men.38 Oberlies’ forced articulation of the epic into an original “Vasu-version” and the later “enframed” versions cannot account for the continuous interaction of the two bheda narratives throughout the epic. Oberlies’ model is also unable to explain why the epic has two other “beginnings” (viz., Mahābhārata 1.1 and 1.4) that correspond to neither of the


beginnings of these three “versions.” In fact, the entire problem of the twin beginnings is elided and it remains ambiguous in Oberlies’ analysis whether Ugraśravas’ narration to Śaunaka in the Paulomaparvan belongs to the “Āstīka-version” or to the “Manu-version.”

In order to be able to account for this second beginning, Oberlies is now forced to interpret the Paulomaparvan as a “switch narrative”39 (Schalterzählung) that links the “older Āstīkaparvan” (i.e., our current Pauṣyaparvan) to the Āstīkaparvan.40 In addition to the local problem of how to fuse the Pauṣyaparvan to the Āstīkaparvan, Oberlies also has to account for the link between his so-called “Āstīka-version” and the “Vasu-version.”41 His solution is to interpret the Āstīkaparvan itself as a “neck-saving narrative” (Hals(lösungs)erzählung), i.e., a story told in order to save someone’s life such as Scheherazade’s narration in The Thousand and One Nights or, closer to home, Rumpelstiltskin’s riddle posed to the miller’s daughter, which she must answer in order to rescue her first born.42 “The snake king,” writes Oberlies, “sways in the highest danger [in the sky?], and in order to save his ‘neck’ [‘Hals’], Āstīka narrates a story [Geschichte]—the Mahābhārata—until the life-threatening danger for Takṣaka is past.”43 The actual problem of the double beginning gives way to a fanciful comparison to The Thousand and One Nights and the Mahābhārata44 is reduced to the epigraphy of an historical “king Vasu” ruler of a “golden age.”45 Besides the problem that there is no evidence that the reference is to a historical “king Vasu” or that the epic was necessarily built up around this historical core,46 there are also serious logical flaws to this argument, as I demonstrate below.

The hypothesis of three distinct “versions” of the Mahābhārata, although appealing in its simplicity, cannot account for the complexity of the text. Crucially, Oberlies’ reconstruction either begs many of the questions it is supposed to answer or does not answer these at all. Why does the epic refer to a multiplicity of origins? Why does it introduce multiple beginnings—not just from Manu or Āstīka or Uparicara onward, but also from the story of Bhīṣma’s birth or the bheda narrative forward? Why does it duplicate the outermost beginning using the same opening phrase both times? Why does Śaunaka ask Ugraśravas to narrate the ancient lore in its entirety (purāṇam akhilaṁ; 1.5.1) and does this include the material previously narrated to the sages? Or does the first beginning enclose the second, with Ugraśravas self-consciously narrating the story of an earlier visit to the Naimiṣa? Not only is Oberlies’ reconstruction unable to answer these questions, but it leads to even more puzzling questions: if Āstīka replaces Vaiśaṁpāyana as narrator in the “Āstīka-version,” where did Āstīka hear this story from? And if he arrives at


the sarpasatra while it is in progress and begins narrating the Mahābhārata to Janamejaya, where are we to locate Vaiśaṁpāyana’s narration to Janamejaya? What happens to the Vaiśaṁpāyana-Janamejaya dialogical level? Does Āstīka’s narration refer to the “original” “Vasu-version” and, if so, is this the version containing Vaiśaṁpāyana’s narration to Janamejaya? Where and when did this narration occur, if not at the sarpasatra? And if Āstīka knows of this narration, why would he repeat the same narrative to Janamejaya (albeit with additions to make it longer)? Why does the king wish to hear the story twice?47

Oberlies’ attempted reconstruction of the Ādiparvan perfectly illustrates the dangers of attempting to second-guess a text, especially one as carefully composed as the Mahābhārata.48 As an alternative to this “text-historical” approach, I propose a hermeneutic approach49 that uses the text itself to interpret the text rather than adducing pseudo-historical considerations or employing circular arguments.50 Before articulating this hermeneutic interpretation, however, I first consider the evidence of the CE in the next section.

5. The Double Beginning in the Critical EditionScholarly criticisms of the double beginning are based on a perception of the redactor as a sort of bricoleur who works with scraps of texts,51 either taping one piece to the other to extend the text backward or splicing materials here and there into the text.52 But if we set aside this prejudice for a moment and consider the manuscript evidence as presented in the CE, a different picture emerges. The evidence of the manuscript tradition is unambiguous: the double beginning occurs in all the manuscripts collated for the CE (cf. Sukthankar 1933: lxxxvii). Although Mehta has argued that this “only means that” the two beginnings “are older than the recensional ramification” (Mehta 1973: 548), this is clearly an argument ex silentio, as, in the absence of manuscript evidence, the arguments on both sides are evenly balanced. Not only is there no manuscript evidence that the double beginning is a later addition to the text, but the double beginning is also clearly a central feature as it occurs in all manuscripts of both recensions. If it were really so evident that the “two beginnings” “smack of two different redactorial agencies” (ibid., 549), one could expect that one of the two beginnings would have been removed long back at some stage in the epic’s history. Yet, remarkably, within the span of roughly 330 years covered by the manuscripts collated for the CE, not one scribe or redactor found the double beginning a serious enough problem to consider “correcting” it. Even Nīlakaṇṭha, whose version of the Mahābhārata is explicitly intended as a compendium of the “best” readings, does not excise


either of the two beginnings, perhaps indicating that he did not consider the two beginnings as alternative readings.53 Sukthankar rightly criticizes Vidushekhara Bhattacharya, who is perhaps the first to have drawn attention to the “problem” of the double beginning, for his suggestion that the first prose sentence of the Mahābhārata be “deleted from the CE, because it is intrinsically inappropriate in the context.”54

The manuscript evidence is especially strong for the double beginning, as about 70 manuscripts were either fully or partly examined and collated for the CE of the Ādiparvan, and, of those, about 60 were actually used in preparing the text.55 A number of manuscripts also preface the opening lines of 1.4 lomaharṣaṇaputra ugraśravāḥ sūtaḥ… with sautir uvāca or variant forms, thus explicitly forging a link between the preceding narration and the second beginning.56 While this does not demonstrate that the two beginnings are original,57 the onus is emphatically on the modern critic to demonstrate that the two beginnings were combined,58 as Mehta describes it, “without any attempt at organic combination—a strange patchwork!” (Mehta 1973: 547). Whatever the modern critic may have to say about the seeming incongruity of the two beginnings, the fact remains that for at least four centuries the double beginning was not felt to be something that required “correction.”

Especially if one takes the evidence of the CE seriously, it is clear that one requires a more sophisticated theory of the double beginning than those proposed hitherto. I am not here concerned with the question of whether the text as constituted in the CE is the “Ur-Mahābhārata” or the “Ur-Bhārata,” that elusive Holy Grail of philology, although I am in agreement with Hiltebeitel that the CE editors succeed in retrieving a text that is more thoroughly archetypal than they could have imagined.59 Instead, I want to emphasize that, minimally, the CE tells us something of how the Mahābhārata was read, understood, and transmitted among a pan-Indian “reading community”60 in the last four hundred years and that, in this entire period, the double beginning was felt to be a necessary and meaningful component of the text. It thus cannot be attributed either to an accident of manuscript transmission61 or to an idiosyncratic desire to make the Mahābhārata “a veritable repository of epic tradition” (Mehta 1973: 550). In contrast to these pseudo-historical explanations, however, which rely upon second-guessing the redactor, I argue that the fact that the poet(s) chose to compose two beginnings must have its reasons.62 The enigma of the double beginning thus confronts us with the problem of understanding the text as a unified whole, from both structural and philosophical perspectives.63


6. Two Beginnings, Two NarrationsIf one sets aside these text-historical prejudices for a moment and considers the text itself, it becomes clear that the text is not deficient with respect to structure but, rather, carefully and purposefully constructed. The entire Mahābhārata is arranged in 18 chapters, the Bhagavadgītā and the Nārāyaṇīya also feature 18 chapters, 18 armies encounter each other in the Mahābhārata battle, and the battle itself lasts 18 days.64 Further, as Oberlies already saw, the Anugītā is composed of 36 (18 x 2) chapters, the Pāṇḍavas are exiled for 12 years, and Arjuna must spend 12 years65 alone for intruding on Yudhiṣṭhira and Draupadī in their private quarters.66 These numeric equivalences are, of course, only the most visible sign of careful composition or redaction, but they hint at an interest in symmetry that can also be found, for example, in the stories of Ruru and Jaratkāru.67 Symmetry, doubling, and repetition are crucial elements in a narrative based on a cyclical understanding of time.68 Thus, rather than excising one of the two beginnings as a “repetition,” I argue that we must examine the text itself for clues on how to read the double beginning.69

The Ādiparvan, which van Buitenen considers one of the latest major books (cf. van Buitenen 1973: xxiii), shows signs of remarkable redactorial and philosophical activity. It is here that we must look for clues to the thematic and compositional unity of the entire epic. The first major book is of global significance to the epic, as we find here both a nexus of themes that are carried through the entire epic and the entrance into the labyrinthine narrative of the remaining 17 major books. The Ādiparvan provides the epic with the textual authority, integrity, and coherence necessary to present a narrative of this scope. Stating its fiction that the Vaiśaṁpāyana narrative had grown over time,70 the Ādiparvan provides a rationale for canonizing the text in its present form by legitimizing the original recitation at the second retelling by the bard Ugraśravas at Śaunaka’s twelve-year sacrificial session. The opening minor books of the Ādiparvan thus authorize the text in its redacted form, articulate a unified set of concerns, and seal off the canon against further “expansion.” While these “expansions” of the epic have been noted, explanations of the logic of expansion have been unsatisfactory. Ranging from the ease of insertions into loose leaf manuscripts (cf. van Buitenen 1973: xxix) to sophisticated theories of “embedding” texts according to the hierarchy of sacrifice (cf. Minkowski 1989), all theories thus far provide overtly mechanical and formal models. My analysis of the Ādiparvan, however, shows that the logic of expansion is also philosophical. The double beginning of this major book especially cannot have been an accident, as it demonstrates great textual self-consciousness and enables the text to overcome the problem of positing an absolute origin in time.71 Let us see how.


After Ugraśravas takes his place, the ṛṣis assembled at Śaunaka’s twelve-year session question him about his wanderings. The bard relates that he was present at Janamejaya’s snake sacrifice, where he heard the Mahābhārata of Kṛṣṇa Dvaipāyana Vyāsa. He then undertook a pilgrimage to many sacred fords and sanctuaries, before visiting Samantapañcaka, the place where the great battle between the Kurus and the Pāṇḍavas was fought. Having visited this “holy place sought by the twice-born” (puṇyaṁ dvijaniṣevitam; 1.1.11), the bard sought out Śaunaka’s hermitage in the Naimiṣa forest. The hermits, having completed their rituals, ask the bard to relate the Mahābhārata as he heard it from Vyāsa. The bard then clarifies the relationship of Vyāsa to the text. He recounts to the sages that Vaiśaṁpāyana narrated the Mahābhārata at the bidding of his teacher Vyāsa. The authority of the author and the authority of king Janamejaya testify to the truthfulness of the recitation: both are witness to Vaiśaṁpāyana’s recitation. The emendation of the narrator from Vyāsa to Vaiśaṁpāyana is significant, as it separates the author and the bard while simultaneously authorizing the retelling. Further, it separates the epic from that which is heard: śruti. The Mahābhārata is that which is remembered: smṛti.

As Hiltebeitel has already noted, the two beginnings present us with quite different descriptions of the same event.72 In the first beginning, Ugraśravas is welcomed by the sages assembled at Śaunaka’s twelve-year sacrifice, who offer him a seat. After Ugraśravas seats himself and is at ease, the sages, addressing him as “the lotus-eyed one” (kamalapatrākṣa; 1.1.7), ask him where he has come from and where he has whiled away his days. In response, Ugraśravas tells the sages that he was present at “the Snake-Sacrifice of the great-spirited royal seer Janamejaya, son of Parikṣit, where Vaiśaṁpāyana recounted all manner of auspicious tales of events, just as they had happened, in the presence of the king. They were tales that had first been recounted by Kṛṣṇa Dvaipāyana.”73 The sages’ interest in Ugraśravas’ doings and wanderings in the first beginning sharply contrasts with the second beginning, where the sages peremptorily inform the newcomer: “Later we shall ask you, son of Lomaharṣaṇa, and you shall relate your repertory of stories, which we shall be eager to hear. But for the time being the reverend Śaunka is sitting in his fire hall. He knows the clestial tales, the tales that are told of the Gods and the Asuras, and he knows fully the stories of men, Snakes, and Gandharvas” (1.4.3-4). As Hiltebeitel notes, Ugraśravas “can hardly feel much esteemed” at being told that “the reverend Śaunaka” “already knows ‘completely’ all such stories as Ugraśravas might tell him” (Hiltebeitel 2001: 103).

Equally significant is Ugraśravas’ response: Ugraśravas’ first words in the


entire epic relate not to the great battle, but to the retelling of that battle at Janamejaya’s sacrifice. Ugraśravas now informs the sages that, after listening to “these stories of manifold import that form part of The Mahābhārata,”74 he undertook a pilgrimage to the Kurukṣetra, the site of the battle between the Kurus and Pāṇḍavas. Ugraśravas then concludes his first speech in the epic, asking the sages what they would like to hear narrated.

In an epic that is seemingly unafraid of redundancy, the sages now recount back everything that Ugraśravas has narrated to them concerning the epic’s first narration:

Tell us [they say] that ancient Lore that was related by the eminent sage Dvaipāyana, which the Gods and brahmin seers honored when they heard it! That divine language of the sublime Histories, in all the varieties of words and books, the sacred Account of the Bhāratas, that language of complex word and meaning, ruled by refinement and reinforced by all sciences, which Vaiśaṁpāyana, at Dvaipāyana’s bidding, repeated truthfully to the satisfaction of King Janamejaya at the king’s sacrifice. We wish to hear that Grand Collection, now joined to the Collections of the Four Vedas, which Vyāsa the miracle-monger compiled, replete with the Law and dispelling all danger of evil!75

Van Buitenen’s translation, although generally excellent, does not quite succeed in capturing the passage’s significance here. Hiltebeitel is much more alert to the philosophical resonances of the Naimiṣa sages’ panegyric; he translates as follows:

We wish to hear that wonderworker Vyāsa’s collection (samhitā) of the Bhārata, the history (itihāsa), that most excellent communication (ākhyānavariṣṭha), diversified in quarter-lines and sections (vicitrapadaparvan), with subtle meanings combined with logic (sūkṣmārthanyāyukta) and adorned with Vedic meanings (vedārthairbhūṣita), which Ṛṣi Vaiśaṁpāyana properly recited with delight at the sattra of Janamejaya by Dvaipāyana’s command—holy, connected with meanings of books (granthārthasaṁyuta), furnished with refinement (saṁskāropagata), sacred, supported by various Śāstras (nānāśāstropabṛṁhita), equaled by the four Vedas, productive of virtue, and dispelling of fear and sin. (1.1.16-19). (2001: 100)

Let us see how this opening passage encapsulates the audience’s intertextual (including especially Vedic textual) interests.

Traditional Indian commentators note that the beginning of any text must contain a maṅgalam or benediction, followed by a statement of anubandha cātuṣṭya or the four criteria qualifying the text. These criteria are:

1. Adhikāri or the student who has fulfilled preparatory learning requirements and thus is qualified to study the text;


2. Viṣaya or the subject matter; 3. Prayoga or the purpose of the text; and, finally, 4. Saṁbandha or the project, i.e., the relationship between the subject matter

and the goal of the text. Remarkably, we find all four anubandhas referred to in this passage. 1. A major limitation of the śruti tradition is that one of the qualifications

refers to the varṇa system. The epic, which is called a fifth Veda and popularly referred to as a strī-śūdra-veda,76 overcomes this limitation, as the text itself guides the student through aesthetic experience into the necessary competence. Thus the “stories” of the trials of Āyoda’s students also, on a deeper level, serves as “initiations.” This could easily be one of the tasks of the epic’s project: suitably to cover up subtle and esoteric meanings of the Veda by hiding them under everyone’s nose disguised as “history.” Student qualifications are not suspended, however: the Vaiśaṁpāyana narrative occurs under the aegis of the qualified teacher Vyāsa, and the reader-student’s qualifications are underscored by the Naimiṣa frame, into which the Pauṣya preparatory apparatus is inserted. The notion of the adhikāri is thus clearly in the background of the lines: “We wish to hear that wonderworker Vyāsa’s collection (samhitā) of the Bhārata… which Ṛṣi Vaiśaṁpāyana properly recited with delight at the sattra of Janamejaya by Dvaipāyana’s command.”

2. Viṣaya is mentioned next: “Vyāsa’s collection (samhitā) of the Bhārata, the history (itihāsa), that most excellent communication (ākhyānavariṣṭha), diversified in quarter-lines and sections (vicitrapadaparvan), with subtle meanings combined with logic (sūkṣmārthanyāyukta) and adorned with Vedic meanings (vedārthairbhūṣita).”

3. Followed by a reference to prayoga: “productive of virtue, and dispelling of fear and sin.” The text also claims to lead to all four goals, i.e., dharma, artha, kāma, and mokṣa, but, of these, the first and the last are especially emphasized.

4. Finally, the sages address the saṁbandha: “… holy, connected with meanings of books (granthārthasaṁyuta), furnished with refinement (saṁskāropagata), sacred, supported by various Śāstras (nānāśāstropabṛṁhita), equaled by the four Vedas.”

Every line in the first beginning is thus rich with clues to the text’s project. But in addition to providing clues to its project, the text also contains subtle hints of the personal connections behind the narration. Consider, for example, the sages’ statement that “Vaiśaṁpāyana, at Dvaipāyana’s bidding, repeated [the epic] truthfully to the satisfaction of King Janamejaya at the king’s sacrifice.” Obviously the sages know more of this narration than they let on;


how else could they know that Vaiśaṁpāyana repeated the epic “truthfully” and to the king’s “satisfaction”? The important point here is to see how repetition in the epic is never gratuitous, but always made to reinforce a message or with subtle shifts of emphasis that nonetheless can have major effects.

Obsessed with the idea of a historical “redactor” who combined two distinct versions of the beginning, Tsuchida and Mehta fail to consider unambiguous evidence, as pointed out by Brodbeck,77 that there are, in fact, two Naimiṣa narrations: the first by Ugraśravas to the ṛṣis (1.1.1-1.3.195) and the second also by Ugraśravas but to Śaunaka rather than the sages (1.4.1-1.54.24). In the second minor book of the Ādiparvan, the Parvasaṁgraha or the “summaries of the books,” Ugraśravas himself refers to this earlier narration; he says “I shall narrate to you the full story [vistaram] of The Bhārata from The Book of Puloman onward, as it was told at Śaunaka’s Session...,”78 but he could hardly do this if this narration were ahead of him in time.79 Van Buitenen translates vistaram as “the full story,” perhaps because he thought there was only one Naimiṣa recitation, but this does not do the term justice. Vistaram rather has the sense of “I shall elaborate” or “I shall expand” and thus underscores the fact that the narration beginning at 1.1.4 is a more elaborate recitation.80 Following this comment, Ugraśravas then presents the summaries of the books arranged in 100 upaparvans or “minor books,” which he concludes with the words: “This full Century of Books, which was recited by the great-spirited Vyāsa, was later exactly so recounted by Ugraśravas, son of the Bard Lomaharṣaṇa, in the Naimiṣa Forest, but in Eighteen Books.”81

But when was this? The Parvasaṁgraha must be referring to an earlier recitation by Ugraśravas, if this “full Century of Books” (parvaśataṁ pūrṇaṁ) has already been “recounted” (punaḥ ... kathitaṁ) at the time of Ugraśravas’ statement. If one takes the Parvasaṁgraha seriously, we not only have two beginnings, but also two narrations corresponding to these two beginnings: Ugraśravas’ outer narration, which enfolds the inner one, and an inner narration that in turn enfolds Vaiśaṁpāyana’s narration of the Mahābhārata. Brodbeck has also suggested that the word “gṛhapati describing Śaunaka (at 1.4:11) might distinguish the earlier satra (where Śaunaka talks with Ugraśravas) from the (iterable) later satra (beginning 1.1) where Śaunaka is just kulapati and not necessarily personally present.”82 I find his suggestion regarding “iterability”83 especially valuable, as it fits in with my thesis that the reader is a central component of the epic84 and that discussions of the text that leave the reader’s perspective out of consideration (e.g., Tsuchida, Oberlies) are reading the epic from the wrong end, as it were.


7. Cosmological and Genealogical Beginnings: The Anukrama∫i- and Paul–omaparvans

As indicated above, the second beginning does not simply replicate the contents of the first. The first beginning is explicitly cosmological and eschatological: it begins with a triple maṅgalācaraṇa (a prayer invoking auspiciousness) to the god Hari, the seer Vyāsa, composer of the epic, and the “many meters” (chandovṛttaiś; 1.1.26) of the epic,85 followed by a cosmology that has explicit parallels to both the Hiraṇyagarbha Sūkta (ṚV 1.10.82, 1.10.121) and the Nāsadīya Sūkta (ṚV 1.10.129) of the Ṛg Veda:

When all this was without light and unillumined and on all its sides covered by darkness, there arose one large egg, the inexhaustible seed of all creatures. They say that this was the great divine cause, in the beginning of the eon and that on which it rests is revealed as the True light, the ever-lasting Brahman.86

Following the opening maṅgalācaraṇa and cosmology, the first minor book then presents two summaries (1.1.67-94 and 1.1.95-195); the second presented as the lamentation of a grieving king who has lost all his one hundred sons in the great war. In 54 quatrains all ending in the same refrain—tadā nāśaṁse vijayāya saṁjaya! “then, Saṁjaya, I lost hope of victory!”—the blind Dhṛtarāṣṭra retrospectively recounts the events leading up to the great war. Dhṛtarāṣṭra’s bhavagītā87 or lament of becoming sets up the basic problem of the epic as that of the destruction of becoming; here is how he concludes his soliloquy:

Woe! Ten, I hear, have survived the war, three of ours, and seven of the Pāṇḍavas. Eighteen armies perished in that battle, that war of the barons. Now a dullness that is all overspread by darkness seems to permeate me. No sign of sense do I see, Bard, my mind seems to go crazy.88

Dhṛtarāṣṭra’s after-the-fact realization provokes a sobering rejoinder from Saṁjaya, who reminds him of the inevitability of destruction for everything that exists in time:

All this is rooted in Time, to be or not to be, to be happy or not to be happy. Time ripens the creatures. Time rots them. And Time again puts out the Time that burns down the creatures. Time shrinks them and expands them again. Time walks in all creatures, unaverted, impartial. Whatever beings there were in the past will be in the future, whatever are busy now, they are all the creatures of Time—know it and do not lose your sense.89

The bard’s description of the work of time is immediately followed by a eulogy of Kṛṣna Vāsudeva, who is eulogized here as brahman or absolute being:


In this book, Kṛṣṇa Dvaipāyana has uttered a holy Upaniṣad… And Kṛṣṇa Vāsudeva is glorified here, the self-eternal Blessed Lord—for He is the truth and the right and the pure and the holy. He is the eternal Brahman—the supreme Surety, the Everlasting light of whose divine exploits the wise tell tales. From Him begins existence that is not yet, and the non-existent that becomes. His is the continuity and the activity. His is birth, death and rebirth.90

These passages set up the contrast between time (kāla) and eternity as the basic problem of the epic;91 following the first summary of the Kuru narrative at 1.1.67-94, Dhṛtarāṣtra’s lament and Saṁjaya’s response provide a first interpretive guideline to the meaning of the Kuru narrative. Lines 1.1.193-195, which are directly juxtaposed to this discussion of the awesome destruction of time, then introduce Kṛṣṇa Vāsudeva in terms recalling the tradition of soteriological ontology in the Upaniṣads and Kṛṣṇa’s divine manifestation in chapter 11 of the Bhagavadgītā. The first minor book concludes with a description of the merits that accrue from listening to the epic and an etymological derivation of the name Mahābhārata that underscores its function as a mokṣaśāstra:

Once the divine seers foregathered, and on one scale they hung the four Vedas in the balance, and on the other scale The Bhārata; and both in size and in weight it was the heavier. Therefore, because of its size and weight, it is called The Mahābhārata—he who knows this etymology is freed from all sins.92

The second minor book then begins with the sages asking the bard to describe Samantapañcaka.93 The bard complies, clarifying the etymology of the name Samantapañcaka. At the juncture of the tretā and dvāpara yugas (tretādvāparayoḥ saṁdhau; 1.2.3), Rāma Jāmadagnya slew the entire race of Kṣatriyas (sarvaṁ kṣatram), filling five lakes with their blood (pañca cakāra rudhirahradān; 1.2.4). Here was also fought at the juncture of the dvāpara and kali yugas (saṁprāpte kalidvāparayor) the battle between the Kurus and the Pāṇḍavas (yuddhaṁ kurupāṇḍavasenayoḥ; 1.2.9).94

The seers’ question signals a shifts away from the present moment in the Naimiṣa forest (nimiṣ = moment) to the epochal cycle as it turns at Samantapañcaka. This shift is simultaneously a shift from the reader’s individual concerns, especially foregrounded in the concluding phalaśruti (the recitation of the merits accruing on hearing the narrative) of the Anukramaṇiparvan, to the narrative of epochal time that dominates the Parvasaṁgrahaparvan. The reference to Kurukṣetra’s original name and its meaning marks Kurukṣetra as a sacrificial site and sacrifice itself is linked to the cyclical turning of


the ages. At the end of an age, all living beings go under in a dissolution (pralaya). Thus, concealed within the seers’ seemingly innocuous interest in etymology are a wealth of clues to the Mahābhārata narrative. By locating the upcoming narrative within a cosmic, geographic, and temporal context, the arbitrariness of violence is negated and violence itself elevated to a cosmic principle. Something escapes cosmic dissolution and forms the seeds of the new beginning; destruction is always incomplete. Becoming proceeds through cycles of destruction but this destruction is never total so there is no end to the cycle. Thus, just as a few kṣatriyas survive Paraśurāma’s massacre, Parikṣit will survive the Kurukṣetra battle. The turn to the history of Kurukṣetra makes it clear that the Kurukṣetra battle is no ordinary war, as it unfolds at the preeminent sacrifical and cosmological site at the juncture of dvāpara and kali yugas.

Ugraśravas’ description of Samantapañcaka in lines 1.2.2-14 is then followed by the third and fourth summaries of the epic at 1.2.34-69 and at 1.2.71-234, and the bard then concludes with a second invocation of the benefits accruing from listening to the Mahābhārata.

Let us now turn to the second beginning in the Paulomaparvan. In the second beginning, Śaunaka enters and, having taken his seat, asks the bard whether he too has learned the “entire stock of ancient Lore” which his father used to recount. “[W]e have heard them before, and long ago it was,” says Śaunaka, “from your own father.” “Now from among all the tales, I would first like to hear the one of the Descent of the Bhṛgus. Tell the tale—we are eager to hear you.”95 Hiltebeitel has contrasted the seeming “hauteur” (2001: 104) of Śaunaka and the sages in the second beginning, but, more than that, it is the shift in emphasis away from the bard as someone in attendance at Janamejaya’s sacrifice and who received the Mahābhārata directly from Vaiśaṁpāyana in Vyāsa’s presence that is notable here. In the first beginning, this fact was emphasized twice: once in the bard’s own statement (1.1.8-11) and once by the sages who recounted back to the bard the circumstances of his first hearing (1.1.15-19). In the second beginning, there is no reference either to Ugraśravas’ presence at the sarpasatra or to his having heard it from Vaiśaṁpāyana: in response to Śaunaka’s question, Ugraśravas merely says, “Whatever was so perfectly committed to memory long ago, O best of the twiceborn, by such great-spirited brahmins as Vaiśaṁpāyana and his successors, and was recited by them of yore to my father and again committed to memory by him—all that I myself learned no less perfectly.”96 Even though Vaiśaṁpāyana is once again placed at the head of the line of transmission, there is no suggestion that Ugraśravas may have heard the epic directly from him. Instead, the transmission is now mediated by his father, and this genealogical mediation


perfectly reflects the content of Śaunaka’s request.Ugraśravas now launches into the Bhṛgu genealogy, beginning with the

birth of Cyavana from ṛṣi Bhṛgu under strange circumstances. Bhṛgu begets a son on Pulomā, born of his “virility” (bhṛgor vīryasamudbhavaḥ; 1.5.11). The demon Puloman sees Pulomā and is overcome by desire for her. Assuming the form of a boar (varāharūpeṇa; 1.6.1), he carries her off, and her unborn son falls enraged from his mother’s womb (roṣān mātuś cyutaḥ kukṣeś; 1.6.2). Seeing the son aborted from his mother’s womb and radiant as the sun (ādityavarcasam; 1.6.3), the demon is burnt to ashes. Because he fell (cyutaḥ), the son is known as Cyavana (cyavanas tena so ’bhavat; 1.6.2).97 The story of Cyavana’s birth is the first in a set of genealogical narratives in the Paulomaparvan that also includes the stories of Pramati’s birth from Cyavana and Ruru’s birth from Pramati. The cycle which begins with Cyavana’s fall into becoming only comes to an end when Ruru meets his namesake, a “lizard” named Ruru, who tells Ruru about how the brahmin Āstīka rescued the snakes at Janamejaya’s snake sacrifice.98 Āstīka is the offspring of sage Jaratkāru and the snake-woman Jaratkāru. Because his father, before departing for the forest, declared “There is [asty eṣa] a child in you, fortunate woman…,”99 his name came to be Āstīka or “he who is possessed of the quality ‘there is’.”100 The Āstīkaparvan, which unites elements of cosmology (the story of the churning of the ocean) and genealogy (the Jaratkāru narrative),101 fuses the cosmological beginning of the Anukramaṇiparvan and the genealogical beginning of the Paulomaparvan: two narratives—one cosmological, the other genealogical—culminate in the great sarpasatra, the setting for the first narration of the epic.102 Ugraśravas’ narration to Śaunaka comes to an end in the next book, and gives way to Vaiśaṁpāyana’s narration to Janamejaya: the central bheda narrative of the doings of the Kurus, culminating in the Kurukṣetra battle.

Let us return to the problem of the double beginning. In the preceding section, I showed why the hypothesis of a “conflation” of two beginnings fails on the grounds of textual evidence alone. The two narrations corresponding to the two beginnings constitute distinct textual moments and therefore must be held apart while reading the epic. In this section, I have shown that these two textual levels address different themes—cosmology and genealogy—and thus are “alternative” beginnings only in the strictly defined sense that they constitute alternative points of entry into the narrative of “becoming.”103 They are not “alternatives” in the sense that we could excise or do without one of the two beginnings. The double beginning, in fact, splits the text from the very outset: like the forked tongue of a snake, we have two beginnings that


run parallel to each other until they finally come together in the Āstīkaparvan to give us the main body of the epic. Cosmology and genealogy, which are two “knowledges of becoming” or two “genera of becoming,”104 unite to create sacrifice (Janamejaya’s sarpasatra), and it is within that sacrificial setting that the raṇa or the battle of Kurukṣetra must ultimately be placed—and understood.

An analysis of the double beginning is thus not just limited to the “text-historical” question of whether both beginnings are equally original, or belong together, or represent a conflation of different manuscripts. Instead, it opens on to the global question of how the epic is to be read. Is the Mahābhārata, as 19th century German Indologists thought, no more than the story of a small conflict between tribal chieftains of the Indo-Gangetic plain only latterly inflated through the addition of extraneous material?105 Or, is it, as Indian tradition has always held, an all-encompassing text with fundamental insights into being, becoming, dharma, the puruṣārthas, etc.? One can see how incommensurable the two interpretive frameworks are. Tracing the four genera thus constitutes an essential step in demonstrating that the Kurukṣetra war is not war as ordinarily understood, but thematized under the aspect of representing a genus of becoming. Agōn is the paradigmatic human activity: even Heraclitus says “strife [polemos] is the father of all” (Πόλεμος πάντων μὲν πατήρ ἐστι; fr. 53) It would be as absurd to try to extract some “historical” Ionian battle from this statement as to identify a historical Kuru conflict on the basis of the Mahābhārata, and yet the latter has been attempted time and again in epic scholarship.

The chart below clarifies the relation of the epic’s two beginnings to its first narration and to its main narrative plot:

Chart 1: The Double Beginning of the Ādiparvan.106

Genealogy and cosmology are like the two sacrificial sticks that are rubbed together to produce the sacrificial flame;107 the Kuru conflict is the sacrifice (raṇayajña108); and


Parikṣit is the yajñaśiṣṭa or the sacrificial remainder;109 and Janamejaya is the one who, “feeding”110 on the sacrificial remainder, becomes immortal.111

Obsessed with reconstructing Indian ethnography and history on the basis of the epic, “text-historical” scholarship has rarely asked about why exactly Janamejaya should be the patron of the great snake sacrifice that constitutes the setting for the first narration of the Mahābhārata. Oldenberg, for example, based on a single Ṛgvedic reference to king Janamejaya,112 is able to imagine “the splendor of Janamejaya’s kingdom” (1922: 12) where brahmins seated on “gold-embroidered mat(s)” (1922: 15) recite the narrative to a king interested in hearing the “the heroic poem of the deeds of his ancestors” (1922: 12). We should set aside Oldenberg’s absurd attempts to secure historical antiquity for Janamejaya, to ask: why Janamejaya? Why this king whose name means “victorious over birth”? Janamejaya, I argue, is the one who hears the entire story of the Mahābhārata, that awesome narrative of the destruction of the Kuru line, and understands its meaning. Once being, in the form of Āstīka,113 arrives through this textual yajña, there is a “joyous tumult of cheers” (tato halahalāśabdaḥ prītijaḥ samavartata; 1.53.9) and king Janamejaya, much pleased (prītimāṁś; 1.53.10), gives away many gifts to the sadasyas and to Āstīka. With the monumental sarpasattra, Janamejaya has finally overcome the “unseen danger” (adṛṣṭam; 1.3.8) of death: he is one who is “victorious over birth.”

8. A Fork in the Beginning: Hermeneutics in the Paußyaparvan

Although this analysis solves the so-called “problem” of the double beginning, a few questions remain: if both beginnings are equally original, where does one begin? Does one enter the narrative via cosmology (i.e., via the first beginning in the Anukramaṇiparvan) or does one enter it via genealogy (i.e., via the second beginning in the Paulomaparvan)? In this section, I argue that the most original beginning114 corresponds to neither the cosmological nor the genealogical beginning, but to a third “beginning”: a hermeneutic beginning in the Pauṣyaparvan.

As van Buitenen already noted, the Pauṣyaparvan is exceptional among the Mahābhārata’s opening books (cf. van Buitenen 1973: 2). Not only is it one of the few portions of the Mahābhārata composed almost entirely in prose,115 but its placement is also highly significant as it separates the first beginning from the second. As we saw above, Ugraśravas himself states that he will narrate “the full story [vistaram] of The Bhārata from The Book of Puloman onward, as it was told at Śaunaka’s Session...,”116 which suggests that the first beginning somehow


comes to an end with the Pauṣyaparvan. The third minor book is also the first narrative segment of the epic following the lists of contents and summaries. It is the first time we encounter king Janamejaya and includes the first mention of Kurukṣetra, one of only three references until chapter 89 and the only one not embedded in someone’s verses. Significantly, Kurukṣetra is introduced not as the site of the great war, but as a sacrificial site.117 In fact, the entire minor book revolves around four themes: sacrifice, initiation, pedagogy, and hermeneutics. As I have already analyzed the sacrificial, initiatory, and pedagogical aspects of the Pauṣyaparvan elsewhere,118 I would like to focus here on its hermeneutic significance.

Let us see how the Pauṣya narrative, beginning with Saramā and ending with Uttaṅka, defines a hermeneutic program. Three “sacrifices” structure the Pauṣyaparvan: the sacrificial session at Kurukṣetra where Saramā appears and warns Janamejaya of an unseen danger (adṛṣṭam; 1.3.8), Janamejaya’s conquest of Takṣaśilā, and the snake sacrifice. These three form a series: at the conclusion of every sacrifice, a person appears and interprets the sacrifice, while triggering the next one. Saramā appears during Janamejaya’s first sacrifice and warns him that he has not overcome his mortality. Her warning sends him in search of fame and conquest through conquering Takṣaśilā, the next “sacrifice” in the series. While historical fame grants a limited form of immortality, it cannot lead to true salvation. For this reason, following Janamejaya’s conquest of Takṣaśilā, a further interpreter appears. Uttaṅka criticizes the king for his conquest of Takṣaśilā, and urges him to perform the third sacrifice. Janamejaya finally gains salvation through the third sacrifice with the appearance of Āstīka, the savior.119 Saramā’s warning of an unseen danger (adrṣṭaṁ) sets in motion a series of events that results in the appearance of being itself.

The name Āstīka means “he who is possessed of the quality ‘there is’.” The story of Āstīka’s birth recounted in the fifth minor book of the Mahābhārata bears important clues to his significance in the narrative. The sage Jaratkāru wanders the earth performing austerities. He is unwilling to marry, until his ancestors request him to do so for the sake of the line. Jaratkāru agrees on condition his wife also bears the same name. The snake Vāsuki then presents his sister, Jaratkāru, to the sage. Long ago, Kadrū, the mother of the snakes, cursed them to perish in Janamejaya’s sacrificial fire. The Creator, Brahmā, promised the snakes that a remnant would be saved and that Āstīka, the son of the sage, Jaratkāru, and a namesake virgin, would bring about their salvation. One evening, as the sage is sleeping with his head on his wife’s lap, it turns dusk. Worried that her husband will miss the evening ritual, she awakens him.


Incensed, the sage threatens to forsake her at once. The tortured woman begs him not to leave as she had been given to him in marriage in the hope that she would beget a son from him who would save the snakes from destruction. Then the sage says, “There is [asty eṣa] a child in you, fortunate woman…” (asty eṣa garbhaḥ subhage tava; 1.43.38). In time, the woman gives birth to a son. As his father had said of him, while he was still in his mother’s womb, “There is” (astīty uktvā), the child became known as “Āstīka” (nāmāstīketi; 1.44.20).

The reference to Saramā, the messenger of the gods, is especially significant. In the Ṛgveda,120 Saramā is sent as a messenger of the gods121 to the Paṇis, a group of anti-gods who steal the gods’ wealth (divine cattle and horses, i.e., the rays of the sun)122 every evening. Saramā journeys across the waters of the cosmic stream, Rasā, down to the Paṇis’ hiding place123 and warns them to return the cows. They refuse and in the war that follows, Indra breaks open their enclosure124 and recovers the stolen light. Saramā, who tracks down the Paṇis and aids in the recovery of light, manifests as a savior in the myth (cf. Olson 2007: 251).

As Hewitt,125 Woolsey126 et al. note, Saramā and Hermes share many functions:127 both are messengers, both are linked to the task of (hermeneutic) recovery, and both guide the soul on its afterlife journey. Duncker, drawing upon Kuhn, notes that the names Saramā and Hermes are cognate: “Hermes is no doubt derived from ὁρμή; Sanscrit sar, to flow; Zd. har, to go. The two dogs of Yama, which watch the road of the souls (vol. 3, 50), are called Sarameyas, i.e. belonging to Sarama; Kuhn has accordingly identified Sarameyas and Ἑρμείας” (Duncker 1883: 179).128

The initial episode of Janamejaya’s sacrifice and Saramā’s warning suddenly gives way to three pedagogic narratives concerning the teacher Dhaumya Āyoda and his students. These three pedagogic narratives are then followed by a further narrative dealing with Uttaṅka’s education in hermeneutics. Uttaṅka’s teacher, Veda, one day calls him and says to him: “… whenever anything is lacking in our house, I wish you to make up for it.”129 While he is away, Veda’s wife has her period and Uttaṅka is asked to inseminate her. But unlike Vyāsa, who engenders the main characters in the text, Uttaṅka refuses to inseminate his teacher’s wife.

Somewhere between blind obedience and complete randomness, lies the delicate task of interpreting a text.130

Uttaṅka reasons that Veda did not intend him to go so far in providing for what is missing. He thus interprets Veda’s words, in contrast to the first generation of students who blindly follow their teacher’s instructions. Uttaṅka’s


refusal to inseminate Veda’s wife is especially significant when seen against the background of Vyāsa’s response. Satyavatī, the queen mother, asks Vyāsa to beget sons upon Vicitravīrya’s wives in order to ensure the genealogical line’s survival. Vyāsa complies, siring Pāṇḍu and Dhṛtarāṣṭra. The epic’s creator is thus also the progenitor of its principal characters: the Pāṇḍavas and Kauravas. He creatively inseminates his narrative at several levels and triggers the main events related in the text. Uttaṅka, in contrast, maintains his distance from the narrative: he appears only at the epic’s end, following the death of its principal characters (by the time of Janamejaya’s sarpasatra, all of the Pāṇḍavas are dead as is Parikṣit). He nevertheless triggers action on another level by instigating the snake sacrifice. Whereas Vyāsa is the author-father, Uttaṅka represents the interpreter-student, appearing after the events related in the text have come to a close in order to interpret their meaning. In contrast to Vyāsa, Uttaṅka’s function is not to engender action at the textual level—to inseminate the narrative—but to engender action at the meta-textual, i.e. hermeneutic, level through a retrieval.

Veda returns home and declares Uttaṅka’s education complete: “I grant you leave to go. You will find complete success.”131 But Uttaṅka refuses to go without presenting him with his teacher’s fee (gurvartham; 1.3.97). Veda sends him to his wife, who asks for Pauṣya’s wife’s earrings.132 On his way, he encounters a man on an oversized bull. The man tells him to eat the bull’s dung. Uttaṅka hesitates, but the man says: “Eat it, Uttaṅka, do not hesitate. Your teacher himself has eaten it in his time.”133 Uttaṅka complies, and continues his journey. He finds Pauṣya and asks for the earrings. The king sends him to his wife, who gives him the earrings but warns him that Takṣaka, the king of snakes, may try to steal them. As Uttaṅka returns home, the snake indeed steals the earrings. Uttaṅka pursues him into the netherworld,134 where he sees marvelous sights: two women weaving black and white threads into a cloth, a wheel being turned around by six boys, and a handsome man. Uttaṅka praises them with verses and the man grants him a favor. Uttaṅka replies: “The Snakes shall be in my power!”135 The man tells him to blow into the horse’s anus (etam aśvam apāne dhamasveti; 1.3.156). Uttaṅka does so and smoke rushes out of its orifices. Fearful of fire, Takṣaka returns the earrings to Uttaṅka.

Uttaṅka returns to his teacher and narrates the story. He then says: “I wish to be enlightened by you, sir: what is the significance of this?”136 Veda explains the symbolism to Uttaṅka: the two women are dhātā and vidhātā, the black and white threads night and day. The wheel with twelve spokes is the day,


the six boys the seasons, the wheel itself the year. The man is Parjanya (the rain-god) and the horse Agni (the fire-god). The man on the bull was Indra, the bull itself Airāvata, the king of snakes (nāgarājaḥ; 1.3.174). Uttaṅka was able to survive the netherworld because the dung he ate was amṛta (the nectar of immortality). Following this education in the art of interpretation, the teacher then gives him leave to go. Uttaṅka then goes to Hāstinapura, where he triggers the snake sacrifice.137

If we now return to the problem of the Ādiparvan’s double beginning, we will see how the Pauṣyaparvan offers another way of entering the narrative: one that is neither cosmological nor genealogical, but hermeneutic. By embedding Vyāsa’s original narrative in the first level of sacrifice—Janamejaya’s sarpasatra—and then embedding this sacrifice in a further sacrifice, Śaunaka’s twelve-year sacrifice, the epic duplicates the outermost level of the text, thereby creating a forked structure in which is then placed the textual apparatus: contents, summary and hermeneutic and pedagogical tools. We thus have the following structure to the Ādiparvan:

Chart 2: The Forked Beginning of the Ādiparvan.138

By opening the “jaws” of the text,139 the redactors are able to place a hermeneutic apparatus in a kind of non-narrative space that is both inside and outside the text. By including the hermeneutic apparatus “inside” the text, they are able to redeem its claim to being all-encompassing;140 by placing it “outside” the text, they are able to bring the text into the reader’s horizon.


This surfeit of beginnings shows just how absurd the attempt to establish the “original” beginning on the basis of specious text-historical criteria is. Van Buitenen, for example, in his introduction to his translation of the Ādiparvan, pronounces:

While it is easy, and indeed natural, to be skeptical of many of the beginnings of the true beginning, the fact that they are there carries its own relevance. At an early enough date The Mahābhārata was conceived as standing close to the beginning of national history, so that it was only appropriate to include right at its beginning all kinds of still earlier matter. Thus The Mahābhārata became the central storehouse of Brahminic-Hindū lore; it could only have done so if it were widely considered to be what the editors of the critical edition of the text proudly claim it is: “The National Epic of India.” (van Buitenen 1973: 6)

But such a view once again privileges the idea of a single and simple beginning. A plurality of beginnings need not indicate accretion. Instead, if a plurality of beginnings violates our expectations, it reveals our expectations as linear and progressive. In a cyclical conception of narrative which mirrors the cyclical conception of eonic time, a plurality of beginnings does not violate narrative logic. In fact, only that narrative is authentic which always ends and always begins. The Mahābhārata is always ending and always beginning: already in Vyāsa’s composition and Vaiśaṁpāyana’s recitation to Janamejaya, the Kuru dynasty has ended, the great war has ended, and the great dvāpara yuga has also ended. However, in the utterance of the bard and the renewed interest of the sages of the Naimiṣa forest, the narrative begins anew.

The point here is to recognize that the epic is aware of its own structure. Thus it says, “There are brahmins who learn The Bhārata from Manu onward, others again from the tale of The Book of Āstīka onward, others again from the tale of Uparicara onward.”141 These multiple beginnings only evoke the temporality embodied in the epic: the cycle of stories mirrors the cycle of eons. Thus one has to be Janus-faced to enter and grasp the narrative logic of the epic. This condition is declared by the epic itself: “Learned men elucidate the complex erudition of this Grand Collection; there are those who are experienced in explaining it and others in retaining it.”142 Those who retain it are skilled in memory (smṛti) and those who are skilled in explaining it are skilled in hermeneutics. This is the twin task of the introductory nature of the list of contents in particular and the Ādiparvan in general. Smṛti goes backward into the past, while hermeneutics goes forward, bringing the text to us in the future. The smṛti task of memory, i.e., the philological task,143 which looks backward, is described in verses 51-94, and the hermeneutic


task or the philosophical task, which is not textual but eschatological, is described in verses 95-160. In the smṛti section, the bard provides a historical summary of the author’s origin, the epic itself, the transmission from author to bard, and the origin of the Pāṇḍavas and Kauravas and their battle. Viewing these several beginnings as so many additions to a core text diverts from a philosophical point. These several beginnings should be understood not as mechanical accretions but as the work of an author who, by providing different points of entry into the text, provides different avenues of interpretation. Thus, one could begin reading the text genealogically, cosmologically, or philosophically.

I have therefore argued for taking seriously the epic’s self-conscious imitation of the structure of cyclical time. Indeed, the text is carefully constructed to show that becoming is circular and that origins lie after ends. If, however, becoming is a closed loop with neither absolute origins nor ends, where do we begin? How do we enter the hermeneutic circle? We are presented with a textual problem as well as a cosmological problem: where to begin? The solution is in both cases the same: begin many times. Yet, how are we to understand all these beginnings? What enables us to understand a beginning if there is no absolute arché, an inceptive principle? The Mahābhārata presents two solutions:

1. The beginning can be understood in terms of a previous beginning, or2. It can be understood metaphorically.Thus the narrative beginning of 1.57, referred to as auparicarādi, defers

to the sacrificial beginning at 1.13 (āstīkādi), which in turn defers to the cosmological and genealogical beginnings at 1.1 and 1.4, jointly referred to as manvādi. The cosmological and genealogical beginnings are coeval and hence defer mutually to each other. The outermost beginning, however, is clearly the hermeneutic beginning, which, as I have argued, occurs both inside and outside of the text: the text is like an open-jawed snake, with the hermeneutic beginning reaching out to the reader. Janamejaya’s narrative, strictly speaking, exists outside the core epic narrative: Janamejaya here is an abbreviation or a code for the entire Vaiśaṁpāyana narrative and indeed for the reader herself. Saramā’s warning of an unseen danger (adṛṣṭam; 1.3.8) is meant for the reader because the Pauṣyaparvan, which occurs between the epic’s two beginnings, is ultimately the part of the text closest to the reader.


Chart 3: The Structure of the Mahābhārata.

The Ādiparvan showing how Vyāsa’s “authoritative narrative” is embedded in two sacrifices representing two narrations, one embedded within the other. The outer–most level is duplicated to embed the textual apparatus: contents, summary and hermeneutic tools.

9. Return to the Problem of “Text-Historical” “Criticism” In the previous two sections, I outlined a hermeneutic approach to the text, which takes the text’s self-understanding seriously and attempts to understand the text out of the text itself rather than applying preconceived notions to the epic. In this section, I return to Oberlies’ thesis of a three-stage historical


expansion of the epic to elucidate the inability of the “text-historical” method to account for the epic’s complexity.

Oberlies’ views typify the problems with a reductive “text-historical” approach to the Mahābhārata. Crucially, Oberlies’ analysis fails to account for the double beginning of the Ādiparvan. Rather than seeing the twin beginnings of the Ādiparvan as intentional and appreciating the careful compositional logic that guides this doubling of the outermost frame narrative, Oberlies sets out from the basic premise that the twin beginnings cannot both be equally original. Instead, he proposes that we understand the frame narratives “historically,” i.e., as evidence of separate “versions” of the text. In particular, he claims that Mahābhārata 1.4 “does not reference Janamejaya’s sarpasattra,” “this frame is not woven with the sarpasattra-frame,” and hence represents a later insertion into the text (Oberlies 2008: 91).144 In Oberlies’ reconstruction of the epic’s frame narratives, the Pauṣyaparvan (1.3) links up directly with the Āstīkaparavan (1.13) and, together, “may … have been the original frame narrative of the Mahābhārata” (ibid., 94). Further, Oberlies claims that this original frame narrative would have been “structured as a so-called neck-(saving)-narrative [Hals(lösungs)erzählung],”145 while the narrative of the snake sacrifice would have been “inserted as a switch-narrative [Schalterzählung] in the Ruru-story of the Paulomaparvan” (ibid., 94). Thus, we have the following structure to the Mahābhārata:

1st beginning: 1.1ff.

Manu-version: 1.1.50ff.

2nd beginning: 1.4ff. Ruru-story: 1.4.8-1.4.12 (switch-narrative)

Āstīka-version: 1.3/1.13ff. (neck-saving narrative)

Vasu-version: 1.57ff.146

Oberlies’ approach has the advantage of allowing us to map these three “versions” of the epic onto a simple chronological scheme: the “Vasu-version” would constitute the oldest epic layer, the “Āstīka-version” would have constituted a simple frame that would have allowed for this original narrative to be relayed through the use of a narrative device (i.e., a “neck-riddle”), and


the third “Manu-version” would have introduced a further level of enframing and allowed for the integration of four lists of contents147 into the epic. This smooth progression of narrative and frame elements is marred only by the insertion of the second beginning, i.e., Ugraśravas’ second arrival in the Naimiṣa forest at Mahābhārata 1.4, which interrupts the original Āstīkaparvan, otherwise “characterized by a clearly structured frame narrative.”

But in spite of its appealing simplicity, Oberlies’ model has major draw–backs:

1. It cannot account for why these later redactors felt it necessary to duplicate Ugraśravas’ arrival in the Naimiṣa forest.

2. Even if we assume, as Oberlies does, that they felt the need to introduce a “switch-narrative” in the form of the Ruru-story in the Paulomaparvan in order to enable “the link to the older Āstīkaparvan (1,13),” it is unclear why they could not just have included this one episode without duplicating the Ādiparvan’s beginning.

3. Further, it begs the question to claim that the Ruru-story was inserted in order to enable a transition between the first half of the Āstīka frame (i.e., Mahābhārata 1.3; the Pauṣyaparvan in our present epic) and the second half (Mahābhārata 1.13), when in fact these two halves would have constituted an unbroken whole prior to this insertion.

4. Alternatively, if one assumes that this continuity was interrupted by the insertion of the second beginning at 1.4, thus necessitating the insertion of a “switch-narrative” to restore the lost continuity, then one is once again confronted by the puzzle of the Mahābhārata’s twin beginnings.

5. A closer examination of Oberlies’ argument in fact shows that he, too, sees this section (i.e., the chapters from 1.4-11) as the real “insertion.” Following Ugraśravas’ second arrival at the Naimiṣa and Śaunaka’s request that he narrate the genealogy of the Bhṛgu lineage (1.4-1.5.10), “The story of the Bhārgavas... then fills up the entire remainder of the Paulomaparvan (1,5-12). This section of the Ādiparvan only relates to the section of the Mahābhārata that encloses it (1,3 and 1,13ff.) insofar as in chapter 1,11 the discussion suddenly returns to the story of Janamejaya’s Sarpasattra” (Oberlies 2008: 94).

6. In that case, however, the introduction of the concept of a “switch-narrative” does not in any way help clarify the insertion of a second beginning or the narration of the Bhṛgu genealogy from 1.4 to 1.12. Nor does it clarify why these redactors would have split apart the original Āstīkaparvan into the Pauṣyaparvan and our current Āstīkaparvan.

7. In fact, a more detailed examination of the Pauṣyaparvan demonstrates


the untenability of Oberlies’ claim. Apart from the circumstance that there is a single reference to a sarpasatra toward the end of the Pauṣyaparvan (1.3.190), the two parvans are completely dissimilar in character. The Pauṣyaparvan is mainly prose; it begins with the story of Saramā followed by three pedagogical narratives; the bulk of the book is concerned with the story of Uttaṅka, a student of the brahmin Dhaumya Āyoda in the second generation; the etiology of the snake holocaust presented in the Pauṣyaparvan is completely different from that presented in the Āstīkaparvan; and it is the brahmin Uttaṅka’s rivalry with Takṣaka and not the Parikṣit/Takṣaka conflict as in the Āstīkaparvan that provokes this first reference to the sarpasatra. There is no evidence to suggest that the Pauṣyaparvan and the Āstīkaparvan may originally have been one book.

8. If the “Vasu-version” is to be ascribed to the genre of “history,” what about the “Āstīka-version”? Is it fictional or historical? As Oberlies sets up the relation of the “Vasu-version” to the “Hals(lösungs)erzählung” of the “Āstīka-version,” this endless story whose most famous literary paradigm is The Thousand and One Nights, the latter is plainly fiction, albeit fiction created around the original historical core, i.e., the “Vasu-version.” But is the fact that Āstīka narrated it also fiction? And, if so, who is the author of this little mise-en-scène? Did Āstīka really once arrive at Janamejaya’s sarpasatra and save Takṣaka’s life, or is this also a fictional narrative?

9. If the latter, then the real “expansion” of the Mahābhārata, its meta–morphosis from history to myth, cannot have occurred through the expedient of a Hals(lösungs)erzählung, but must have occurred through the persons who introduced the Hals(lösungs)erzählung itself as a literary motif. In other words, Oberlies’ introduction of a so-called “Āstīka-version” fails to resolve the problem it was intended to be a response to, namely, of how the original epic (Oberlies’ “Vasu-version”), which Oberlies assigns to the genre of history, could evolve into the mythic account we have at present. The suggestion that Takṣaka’s Scheherazadian fate provided the occasion for a massive expansion of the epic and the interpolation of masses of mythic and didactic material into the original epic falls apart once one realizes that this requires us to assume that such a situation actually occurred. Oberlies is caught in a bind of his own making: he must attribute the expansion of the historical epic to Āstīka and then Takṣaka’s near-death encounter must have really occurred, or acknowledge that this situation is fiction and then the expansion must be attributed to the authors of the Āstīka narrative. In either case, one of the two frame narratives falls out of consideration as irrelevant: either the Āstīka narration is the historical linchpin that allows us to understand the text’s expansion and the Ugraśravas narration


is just a shell, or the Ugraśravas narration is the historical linchpin that allows us to understand the text’s expansion and the Hals(lösungs)erzählung is itself decapitated.

10. Conclusion: Establishing CriteriaThis paper has argued for a hermeneutic approach to the epic that takes the epic’s own self-understanding seriously as a text of fundamental philosophical, cosmological, and theological significance. I have argued that the two beginnings must be read in the context of this overarching program and cannot be deleted, moved, or otherwise “edited” to suit contemporary prejudices regarding texts. The analysis articulated above—of the double beginning as a meaningful component of the epic’s narrative structure—confirms the CE editors’ decision to retain the double beginning, and is indirect testimony of the rigor of their editorial praxis.

Borrowing an expression from Mahadevan’s excellent article in this volume, I may say that, just as there are “three rails” in Mahābhārata criticism, there are also “three rails” of Mahābhārata interpretation: the first is the CE which provides us with an archetype that allows us to unearth the basic narrative architecture of the epic; the second is Biardeau’s method of studying the epic within the totality of significations that Biardeau has called the “universe of bhakti,”148 a method that has been ably continued by her students Hiltebeitel, Bailey, Couture and others; and the third is the hermeneutic approach, which uses the text itself to interpret the text. Given this state of affairs, one ought to set up some minimal criteria for future Mahābhārata scholarship, especially in light of the fact that, since the CE, certain approaches such as the search for various “Ur-Bhāratas” are an intellectual embarrassment. These scholarly “theories” bring nothing new to the table: neither in terms of new ideas, nor new evidence, nor in terms of a better understanding of the epic. They are, as Hiltebeitel has argued in his contribution to the present volume, like a “hydra” that refuses to die. Merely cross-referencing other scholarship does not establish the truth of these views. Academic citation has the definite function of corroborating insights, not of accepting what is advanced as a hypothesis in one paper as established fact in another.149 With the completion of the CE, one must begin separating out the few, relatively minor corrections of the CE editors’ decisions from circular and reductio ad absurdum arguments. A prime example of this, as Hiltebeitel has shown in his article in this volume, is the notion of a Kṣatriya epic. Whatever appears not to “fit” some outmoded prejudice, be it of an Indo-European epic tradition or a hypermasculine kṣatriya saga, is excised or


attributed to “later” layers. Of this “cut-and-paste” game with layers, one can only say what Sukthankar observed of 19th-century criticism. Asking “what is the secret of this book of which India feels after nearly two thousand years that she has not yet had enough?” Sukthankar responds:

It would be a rather hazardous conjecture to suppose that such a thing might perchance happen also to the works of the critics of the Mahābhārata, for within less than half a century the lucubrations of these wiseacres have approached perilously near the limbo of oblivion, from which they are periodically snatched out by the industrious pedagogue and the curious antiquarian. (1957: 29-30)

Occam’s razor helped philosophy immensely. Likewise, in literary criticism, it is useful to assume the text as we have it as a text with some coherent meaning, unless proved otherwise. Borrowing Austin’s terminology, it is best to assume that the M0 exemplar had a basic architecture that is continued in a significant measure in all M+N moments. One must attempt to understand this architecture before excising any element as an “interpolation.” Interpolation is not explanation: it is an acknowledgement of our failure to explain the text. As an interpretive category, “interpolation” should be our absolute last choice.

As an alternative to the dogmatism of the Lassen school, I have been working since around 2007 on the idea of the Mahābhārata as a “project with a purpose,” i.e., as a self-conscious attempt to preserve Vedic knowledge and disseminate it on a pan-Indian basis, while retaining its esoteric character by concealing it among a mass of narratives. There is now general consensus that the epic is much more Vedic than previously thought, and my recent and forthcoming articles demonstrate a consistent encoding of Vedic sacrifice into the text. These overlappings can hardly be accidental, nor would it make much sense to take up an original “bardic” epic150 and insert so much Vedic/Brahmanic material into it. Regarding the thesis of Brahmanic “interpolation,” Lassen and Goldstücker are the prime culprits,151 but the myth of “Brahmanic contamination” or “Brahmanic corruption” refuses to die out: 20th century revivals include Fitzgerald (1983) and Oberlies (1995 and 2008).152 Understanding the Mahābhārata as a project with a purpose allows us to comprehend the massive scribal effort and the effort at pan-Indian dissemination behind the epic, both so ably described for us in this volume by Mahadevan, as also the presence of jointures in the text that seem to have been designed to allow for local variation and insertions.

Let me conclude with one final statement on the Mahābhārata’s double beginning. Why does the epic feel the need to place its own hermeneutic tools


within the text, in the curious space opened up between the two beginnings? The answer to this question must, once again, be sought in the epic itself. As Mahābhārata 1.56.33 and 18.5.38 demonstrate,153 the Mahābhārata takes its claim to being an all-encompassing text quite seriously. Hiltebeitel has already argued that the statement “‘Whatever is here may be found elsewhere; what is not here does not exist anywhere’ (MBh 1,56.33; 18,5.38) is not an encyclopedic slogan but an ontological claim about what counts as real, as the heterodoxies do not” (Hiltebeitel 2011a: 11). One consequence of this self-understanding, however, is that the text must contain everything within itself necessary for understanding it: the text is the sole “remainder” that is transmitted across time and space and hence must contain its own hermeneutic apparatus within itself. Ultimately, the double beginning is a consequence of this immense task: to create a kind of non-narrative space that allows for the transmission of a hermeneutic apparatus along with the text. I have further shown that such an interpretation takes the text’s own self-understanding seriously as the śeṣa which textually lives yugāntare. There is nothing outside the text, not even instructions on how to read the text. As I have shown, the first three parvans are inserted into a space in the epic, allowing them to be transmitted along with it. How are we now to understand the frames in this new interpretive architecture if not as historical? But that is another beginning.


Structure of

t he epic

Upaparvan

s O

berlies’ analysis

Oberlies’ redaction

Justification

for thesis

An

ukramaṇ

iparvan

First beginning: Mbh 1.1.1-19

List of contents: Mbh 1.1.67-94

List of contents: Mbh 1.1.95.195

The first dialogical level is that of the first frame narrative,

which narrates the context of the second narration of the M

bh.

Ugraśravas’ narration

Mbh 1.1.20-212

Second narration of the Mbh.

Parvasaṁgrah

a 1.2.1-243 List of contents M

bh 1.2.34-64 List of contents M

bh 1.2.71-234

3. Manu version beginning at 1.27ff.

“This version,

which

began w

ith the

narration of

Janamejaya’s snake sacrifice w

as, in turn, enframed w

ith the narrative of the 12 year N

aimiṣa sattra and thereby the first

dialogical level

brought in,

the title

of the

work

Mahābhārata

introduced, w

hich of

course is

found principally on this dialogical level (see p. 90-91), and this [i.e., the title] associated w

ith the name of the com

poser, V

yāsa.”

First level

of enfram

ing: the

Naim

iṣa satra

P auṣyaparvan

Mbh 1.3.1-195

2. Āstīka version beginning at 1.3ff. (lacks the outerm

ost frame and the first

dialogical level and the list of contents).

“This was fused …

with a second ‘Āstīka-version’ circulating

in Vedic circles that possessed the narrative of the snake

sacrifice of Janamejaya as a (sim

ple) frame.”

Paulomaparvan

Second beginning: M

bh 1.4.1-1.5.3

“Peculiarly, Mahābhārata 1,4 begins w

ith the same w

ords as in 1,1. But at M

bh 1,4 Janamejaya’s sarpasattra is not spoken of,

i.e., this frame narrative is not interw

oven with the sarpasattra

frame.”

Ugraśravas’ narration

Mbh 1.5.4-1.12.5

“In this second introduction, the sages gathered at the sattra request U

graśravas to wait until Śaunka com

es, and to recite that w

hich he [Śaunaka] wishes. As this one then enters (1,5.1),

he does not request Ugraśravas to recite the M

ahābhārata, but he w

ishes to hear the story [Geschichte] of the Bhārgavas (1,5.3). This then fills up the entire rem

aining Paulomaparvan

(1,5-12).”

Oberlies’ “switch-narrative” (Schalterzählung)

Ruru narrative,

concluding w

ith a

reference to Āstīka 1.8-12 “Further, this section of the Ādiparvan has to do w

ith the section of the M

ahābhārata (1,3 and 1,13ff.) enclosing it only insofar as in chapter 1,11 suddenly the discussion com

es to the story [Geschichte] of Janam

ejaya’s sarpasattra. The link to the older Āstīkaparvan (1,13) is generated w

ith the help of the Ruru narrative, a very late insertion into the M

ahābhārata.”

“That is why now

, as the M

ahābhārata is handed down,

the narration of the snake sacrifice has been inserted as a sw

itch narration [Schalterzählung] in the Ruru story [Geschichte] of the Paulom

aparvan.”

“But that this w

as not the original state is suggested—

beyond the young age of the Ruru narrative and of the entire Paulom

aparvan [...]—by the follow

ing considerations: the actual continuous narration of the M

ahābhārata begins at M

bh 1,53.35: hanta te kathayiṣyāmi m

ahad ākhyānam

uttamam

, kṛṣṇadvaipāyanam

ataṁ m

ahābhāratam

āditaḥ. And one expects to ‘actually’ [‘eigentlich’] find the fram

e of the narration before this starting-point. Therew

ith the Āstīkaparvan may have

been the original frame narrative of the

Mahābhārata.”

O

berlies’ “story [told] to save someon

e’s neck” (H

als(lösungs)erzäh

lung)

Āstīkaparvan

1.13.1-53.26

“And much speaks for the fact that in this [fram

e narrative] not V

aiśaṁpāyana but Āstīka w

as the ‘first narrator’ of the M

ahābhārata and that the narration was designed as a so-

called story to save someone’s neck [Hals(lösungs)erzählung].”

“The snake king sways in the highest danger, and in order

to save his ‘neck’ [Hals], Āstīka narrates a story

[Geschichte]—the M

ahābhārata—until the life threatening

danger for Takṣaka is past. In the form of the M

ahābhārata present today, how

ever, it is a song of praise of Āstīka to Janam

ejaya that moves the latter to fulfill Āstīka’s w

ish for a suspension of the sacrifice (M

bh 1,50-51).”

Ādivaṁ

śāvataraṇaparvan

Continuation of U

graśravas/ Śaunaka “dialogical level” (1.53.27-36)

Second level of enfram

ing: Janam

ejaya’s sarpasatra

Vyāsa’s arrival at Janam

ejaya’s satra M

bh 1.54.1-24 Beginning of V

aiśaṁpāyana/

Janamejaya “dialogical level” 1.55.1 and

continued until 18.5 (except 12.327; 331; 335 and 15.42-43)

The second dialogical level is that of the second frame

narrative, which narrates the context of the second narration

of the Mbh.

Vaiśaṁ

pāyana narrative

Ādivaṁ

śāvataraṇaparvan

List of contents: M

bh 1.55 ‘Core’ M

ahābhārata 1.57.1 onward

First narration of the Mbh.

1. Vasu version beginning at 1.57ff.

(lacks the inner frame as w

ell as the outer fram

e and first four lists of contents).

“There existed an ‘unframed’ version of the M

ahābhārata—the ‘V

asu-version’—w

hich began (perhaps with a brief list of

contents [Mbh 1,55 (see p. 85)]) w

ith the introduction of V

aiśaṁpāyana as the speaker (M

bh 1,54.21, 57.75)…”

Appendix


Endnotes

1. All Mahābhārata citations refer to the Critical Edition (CE) text (Sukthankar 1933). All translations refer to the Chicago edition (cited as van Buitenen 1973).

2. Obviously, I mean the narrative plot and not some historical “core” to which was later added. The origins of the latter view can be traced back to Lassen’s attempt to reconstruct Indian history, geography, and ethnography on the basis of the epic. It becomes a staple of Germanic or Germanophilic interpretations of the epic, especially as it allows for a distinction between an “Āryan element” and an Indian element. On the history of the “core” plus “accretions” model, see Vishwa Adluri & Joydeep Bagchee, The Nay Science: A History of German Indology (n.d).

3. This aspect is usually lost in reductive analyses of the epic such as those that argue for a historical core as the “original epic” or as the “Bhārata” epic. As Austin points out in this volume, such approaches are hopelessly circular, with everything that does not fit the particular scholar’s views of the epic being excised as “late.” I discuss a recent instance of such circular reasoning in my review of A. Malinar, Bhagavadgītā: Doctrines and Contexts (Adluri 2010b).

4. On the distinction between time (kāla) and eternity as a central organ–izational principle of the epic, see my Sacrificial Ontology and Human Destiny (n.d.).

5. For a recent discussion of the relation of daivam to puruṣārtha, see Woods 2001.

6. On the Mahābhārata as a compendium on the puruṣārthas, see 1.56.21 (arthaśāstram idaṁ puṇyaṁ dharmaśāstram idaṁ param / mokṣaśāstram idaṁ proktaṁ vyāsenāmitabuddhinā //); see also 1.1.46-47 (vedayogaṁ savijñānaṁ dharmo ’rthaḥ kāma eva ca // dharmakāmārthaśāstrāṇi śāstrāṇi vividhāni ca / lokayātrāvidhānaṁ ca saṁbhūtaṁ dṛṣṭavān ṛṣiḥ //)

7. Cf. 1.1.196-197: adhyātmaṁ śrūyate yac ca pañcabhūtaguṇātmakam / avyaktādi paraṁ yac ca sa eva parigīyate // yat tad yativarā yuktā dhyānayogabalānvitāḥ / pratibimbam ivādarśe paśyanty ātmany avasthitam //

8. Scholars estimate that perhaps only 1/3 of the text is dedicated to the Kuru narrative, with the remainder containing philosophical or didactic material. Hopkins, for example, distinguishes between an “epic proper” of roughly 20,000 verses and the “moral, political, religious, and metaphysical dissertations” (1916: 325) in which this epic is embedded, but does not clarify why only the former constitutes the epic “proper.” Hopkins also famously propounded a distinction between the epic and a “pseudo-epic” (Hopkins 1906: pass.), based, perhaps, on a narrow identification of epic with a heroic narrative (Heldensage). On Hopkins’


contributions to epic chronology, see Sukthankar’s response: “I will say candidly that for all intents and purposes this pretentious table is as good as useless” (1957: 9).

9. For a recent consideration of the some of these issues, see Reich 2008.10. On the doctrine of cyclical time in the Mahābhārata, see Vassilkov 1999;

for details of how the epic replicates the cyclical structure of time, see my discussion of Uttaṅka as the link that forges the end of the narrative with its beginning (Adluri 2010b).

11. This has been a basic tenet of Western (and, especially, German) schol–arship on the epic since Oldenberg and Winternitz articulated the premise that the epic is an “ungeheuerliches Chaos” (“a monstrous chaos”; Oldenberg 1922: 1) and a “litterarisches Unding” (“a literary nonsense”; Winternitz 1909: 272). For a discussion of how these prejudices have shaped much Western scholarship, see Hiltebeitel 2001: 1-31. Hiltebeitel’s entire book is, in a sense, a rejection of these premises. See also the editors’ discussion of the Orientalist prejudice behind Oldenberg’s and Winternitz’ criticism in the respective introductions to Hiltebeitel 2011a and 2011b.

12. See Minkowski 1989 for a discussion of the principles behind such embedding.

13. Hiltebeitel has drawn attention to the presence of yet another frame situated on Mount Himavat, which he calls the Śuka-Vyāsa frame, that would in a sense be outside of even the Naimiṣa frame (see Hiltebeitel 2004, 2005). Hiltebeitel’s suggestion is appealing, as it could potentially allow us to draw together the omniscient narrator who recounts Ugraśravas’ comings and goings and the epic’s traditional author, Vyāsa, but more work needs to be done here. For my present purpose, it suffices to note that the author Vyāsa, the narrator Vaiśaṁpāyana, and the bard Ugraśravas are all contained within the narrative itself, with the reader alone, represented by Gaṇeśa, outside the epic (see Adluri 2010a for a discussion of the textual issues this raises).

14. The epic uses many words for “snake,” without noticeable difference in meaning. The term usually used for “snake” is nāga, a “serpent-demon… supposed to have a human face with serpent-like lower extremities” (Monier-Williams, sv). Nāgas are not ordinary snakes, as they frequently exhibit anthropomorphic features and are endowed with wondrous powers. Takṣaka, for example, the main snake featured in the Pauṣya- and Āstīkaparvans is a sky-going snake; his son Aśvasena is said to possess the power of māyā or illusion. In the snake underworld, Uttaṅka also encounters technological implements such as a wheel (cakraṁ; 1.3.48) and a loom (tantre; 1.3.47, 48). Although Cozad (2004)


and Vogel (1926) make a distinction between nāga and sarpa, in the epic, at least, the two terms are often used interchangeably.

15. Namely, Parikṣit and Parikṣit’s son Janamejaya. Parikṣit is descended from Arjuna via Abhimanyu.

16. Cf. Minkowski 1989: 7 and 21. According to Minkowski, the Naimiṣa forest setting provides a “situation ne plus ultra” that overcomes the danger of regress, but does not fully clarify how this is so. I find Brodbeck’s suggestion that we are, in a sense, the Naimiṣeya ṛṣis more helpful, as it takes the “reader” seriously as a component of the epic’s logic. See my Sacrificial Ontology and Human Destiny in the Mahābhārata for a discussion of the significance of the reader to the epic. See also Adluri 2010a on the addition of Gaṇeśa to the epic as its first reader for some important reflections on how reader and text work together to articulate an all-encompassing logic.

17. Things are somewhat more complex than this initial presentation suggests, of course, as there is not just one but two Ugraśravas narrations; see below. Also see Brodbeck 2009: 244-245 and Brodbeck’s contribution in the present volume.

18. Ugraśravas’ narration, of course, implicitly continues in the background, as Vaiśaṁpāyana’s entire narration is being narrated by him. I thank Simon Brodbeck for this observation.

19. Austin rightly points out that there are significant differences between the framing structure of the beginning of the epic and its close in the eighteenth major book (personal communication), a circumstance that has important bearings on my thesis of the circular structure of the epic. I would caution against expecting perfect symmetry between the Ādiparvan and the Svargārohaṇaparvan. Austin is probably right that the Ugraśravas cycle is not fused to its beginning as one might expect given the epic’s concern with maintaining a cyclical narrative architecture; it is not Ugraśravas who fuses the epic’s end to its beginning, but Uttaṅka, as I have already shown (see Adluri 2011c). Much more work needs to be done here, especially on the question of why the Uttaṅka cycle is already closed out in the fourteenth major book (the Āśvamedhikaparvan) rather than waiting till the end. Austin 2009 provides a good overview of some of the kinds of questions raised by the epic’s final major book.

20. I.e., first in relation to the reader. From a temporal perspective, however, the outer frame is the second. I retain the convention of counting the frames from the outside in throughout this article, in contrast to some scholars who count outward from the central narrative. The latter more accurately represents the temporal sequence of the two narrations, but is less evident to the reader,


who encounters the beginnings in narrative rather than chronological order.21. Credit for noticing the importance of the Naimiṣa frame must, above

all, go to Hiltebeitel (1998 and 2001).22. A number of manuscripts explicitly attribute the narration beginning

1.1.4 to the bard, inserting sautir uvāca or sūtaḥ etc. The CE leaves the identity of the narrator unresolved. I cannot agree with Tsuchida’s claim that “the Āstīkaparvan in its oldest form must have begun with the introductory passage which in the present text of the Mahābhārata occupies the initial position in the Paulomaparvan. In other words, we should suppose that the Paulomaparvan was not simply placed before the Āstīkaparvan, but inserted between the introductory passage and the main portion of the Āstīkaparvan” (Tsuchida 2008: 20). If the second beginning (which Tsuchida terms U2) was indeed the original beginning and only later extended backward to form the first beginning (Tsuchida’s U1), there would have been no reason to retain this original beginning (i.e., U2) nor, indeed, to highlight it in the Parvasaṁgraha (cf. yat tu śaunakasatre te bhāratākhyānavistaram / ākhyāsye tatra paulomam ākhyānaṁ cāditaḥ param; 1.2.29). Mehta’s suggestion of two equally original beginnings is preferable, although Mehta then defers to a specious “historical” justification; see Mehta 1973.

23. But Śaunaka is referred to as gṛhapati at 1.4.11. Van Buitenen translates both as “family chieftain,” fudging the distinction; cf. also śaunakasya kulapater at 1.1.1, which van Buitenen again renders “… family chieftain Śaunaka.”

24. lomaharṣaṇaputra ugraśravāḥ sūtaḥ paurāṇiko naimiṣāraṇye śaunakasya kulapater dvādaśavārṣike satre; 1.1.1.

25. The Bhṛgus or the Bhārgavas are a group of brahmin seers or ṛṣis thought by some to have been the redactors of the Mahābhārata. For the thesis of a “Bhārgava recension,” see Sukthankar 1936-7; see also Goldman 1977; opposed to it, see Hiltebeitel 1999; reworked in Hiltebeitel 2001: 105-118.

26. For a discussion of the ideas behind the logic of a triad of sacrifices in the Pauṣyaparvan, see Adluri 2011c.

27. See especially Tsuchida 2008 and Oberlies 2008. For Brodbeck’s justified critique of Tsuchida’s over-simplification of the Mahābhārata’s narrative structure, see his entry in this volume. I discuss Oberlies’ “onion skin” model of the epic in detail below, as it typifies the problems with so-called “text-historical” analyses.

28. See section 8 below on the significance of the first three books and, especially, of the third book, the Pauṣyaparvan, to the epic’s hermeneutic and pedagogic program.


29. manvādi bhārataṁ ke cid āstīkādi tathāpare / tathoparicarādy anye viprāḥ samyag adhīyate; 1.1.50.

30. I.e., at 1.53.10, although there is a second description of the conclusion of the sacrifice at 18.5.26-29.

31. Van Buitenen’s approach reflects a basic prejudice of modern scholarship that the epic, at its core, is concerned with a historical event—the conflict between the two rival branches of the Kuru patrilineal line (cf. van Buitenen 1973: 3)—and that philosophical and hermeneutical matter such as that contained in the Ādiparvan, the Āraṇyakaparvan, the Bhagavadgītā, and the Śanti- and Anuśāsanaparvans must be a “later” addition to what was originally in essence a simple historical narrative. Van Buitenen is especially critical of what he calls “the third perimeter,” a term that includes Ugraśravas’ narration in the Naimiṣa forest and the run-up to Janamejaya’s snake sacrifice but not the sarpasatra itself, to which van Buitenen is prepared to accord a qualified historical authenticity (cf. 1973: 3).

32. I thank Joydeep Bagchee for all translations of Oberlies.33. To further complicate matters, Oberlies footnotes this very line with

the comment “besides these, however, there must have definitely existed others, manifestly also one that—although not exclusively was nonetheless overwhelmingly—composed in Triṣṭubhs” (ibid., n. 41). But in that case, why does the hint, if it “is to be understood ‘historically’,” conceal this information from us? Does Oberlies mean this is only a partial hint? And if so, i.e., if the “hint” conceals just as much as it divulges, why ought we take it at face value? What evidence do we have for these other versions if this “hint” does not record them? Does this additional evidence corroborate this hint?

34. Hiltebeitel (2001) has already shown that the epic poets make use of “orality” as a literary trope, i.e., that the epic’s literary form is intentionally structured so as to give the impression of being an oral epic or one with an oral background to it. Perhaps the epic poets are being just as playful when they allow us to discern “hints” of history in the epic. In any case, we must take these hints with caution. If we can speak of “history” at all here, it is of the epic’s fiction of history. It is almost as the epic, somehow anticipating 19th century historicism, already sets up its project to defeat straightforward historicist readings. In fact, the one “hint” the epic really does gives us is when it calls itself itihāsapurāṇa (1.1.204). As Greg Bailey points out (personal communication), traditional etymology derives purāṇa(va) from purā and nava (= the narrative of the past renewing itself). The term itihāsapurāṇa would then be a designation for the continuous transformation of history (itihāsa) into myth (purāṇa), i.e., of


the renewal of time itself. The epic, as I have argued throughout this paper, is aware of the irrelevance that threatens all history; through its complex narrative strategies (framing, double beginnings, beginning at the end), it is constantly attempting to overcome this irrelevance, and defeat historicism.

35. Brodbeck makes the point especially well in his recent book; see Brodbeck 2011.

36. This is, of course, simply untrue, since history is equally metaphysical in that it elevates the individual empirical object to a well-determined thing (wohlbestimmtes Ding), which not only contains within itself the ground of all our cognition but also more than a finite intellect can ever come to know of it. In other words, history does not do away with transcendence; it simply displaces the idea of transcendence into material reality. For a penetrating analysis of the epistemological and metaphysical prejudices at the root of the modern faith in the empirical individual, see Schmitt 2003; see also Gadamer (2004) for an excellent overview of the rise of historicism.

37. Conflict between Arjuna’s and Takṣaka’s respective lineages is a recurrent motif in the Ādiparvan. Arjuna, from whom Janamejaya is descended via Parikṣit, destroys Takṣaka’s home and kills his wife, “herself a daughter of the Snakes” (bhujagātmajā; 1.218.6). Takṣaka himself escapes because, at the time of the burning of the Khāṇḍava forest, he was away “in the Field of the Kurus” (kurukṣetre ’bhavat tadā; 1.218.4). His son Aśvasena also succeeds in escaping through his terrible power of illusion (māyāṁ ... ghorāṁ; 1.218.10), but Arjuna curses him to never find shelter (apratiṣṭho bhaved iti; 1.218.11). The conflict between Arjuna and the snakes also recurs in the central battle-books, specifically in the Karṇaparvan. Karṇa is said to aim a “serpent-mouthed arrow” (sarpamukhaṁ; 8.66.5) of the line of Airāvata (airāvatavaṁśasaṁbhavaṁ; 8.66.6) at Arjuna; some mainly northern manuscripts go even further: the snake is identified as Aśvasena, who, having escaped the destruction of the burning Khāṇḍava forest, now enters Karṇa’s arrow seeking revenge (cf. app. 1, no. 40). On snake-related imagery in relation to Karṇa, see Hiltebeitel 2007.

38. Snake and human genealogies seem to be particularly susceptible to such “cross-overs.” Nahuṣa, an early ancestor of the Kuru line, is cursed to become a snake when he strikes sage Agastya either with his foot or feet (versions vary; cf. 12.329.38; 5.17.11; and 13.103.20). Dhṛtarāṣṭra is often referred to as a snake in the epic (cf. 1.3.142; 1.31.13; 1.52.13; 2.9.9; 4.2.14; 5.101.15; 8.24.72; 16.5.14) and Janamejaya himself occurs once as a snake’s name in the Mahābhārata (2.9.10; cf. also Pañcaviṁśā Brāhmaṇa, Baudhāyana Śrauta Sūtra, Baudhāyana Gṛhya Sūtra which list Janamejaya as one of the snakes to officiate at the first sarpasatra; for the full


citations, see Minkowski 1989). Aśvasena is also the name of one of Dhṛtarāṣṭra’s sons (1.108.9). Especially significant in this context is the close assocation between Arjuna and the snakes: some of the most prominent snakes in the epic (Karkoṭaka, Śeṣa, Vāsuki, and, especially noteworthy, Takṣaka; cf. 1.114.60) are said to be in attendance at his birth; and Arjuna himself is married to a snake, the nāgā princess Ulūpī, who drags him underwater and will not release him until he consents to father a son on her (1.206.12-32; cf. also 1.2.91). Irāvān, Arjuna’s son of Ulūpī, later plays a minor but crucial role in the Kurukṣetra battle, sacrificing his life in order that his father and paternal uncles may win the war. On Nahuṣa’s fall, see Hiltebeitel 1977.

39. “Peculiarly, Mahābhārata 1,4 begins with the same words as in 1,1. But at Mbh 1,4 Janamejaya’s sarpasattra is not spoken of, i.e., this frame narrative is not interwoven with the sarpasattra frame. In this second introduction, the sages gathered at the sattra request Ugraśravas to wait until Śaunaka comes, and to recite that which he [Śaunaka] wishes. As this one then enters (1,5.1), he does not request Ugraśravas to recite the Mahābhārata, but he wishes to hear the story [Geschichte; perhaps “history”?] of the Bhārgavas (1,5.3). This then fills up the entire remaining Paulomaparvan (1,5-12). Further, this section of the Ādiparvan has to do with the section of the Mahābhārata (1,3 and 1,13ff.) enclosing it only insofar as in chapter 1,11 suddenly the discussion comes to the story [Geschichte] of Janamejaya’s sarpasattra. The link to the older Āstīkaparvan (1,13) is generated with the help of the Ruru narrative, a very late insertion into the Mahābhārata. That is why now, as the Mahābhārata is handed down, the narration of the snake sacrifice has been inserted as a switch narration [Schalterzählung] into the Ruru story [Geschichte] of the Paulomaparvan” (Oberlies 2008: 94).

40. On the absence of justification for claiming that the Pauṣyaparvan was originally part of the Āstīkaparvan, see section 9 below.

41. I say more about the problems Oberlies is contending with in section 9 below.

42. On the concept of “neck-riddles” or “capital-riddles” (Halslösungsrätsel) see Elias 1995; see also Meyer 1967. The neologism Hals(lösungs)erzählung appears to be Oberlies’ invention, created in analogy to the German Halslösungsrätsel but with an extended narration replacing the traditional riddle, Rumpelstiltskin providing, of course, the most famous example of a Halslösungsrätsel and The Thousand and One Nights of a Hals(lösungs)erzählung. In both cases, the objective is the same: to save either one’s own or another’s “neck.” How exactly this applies to Takṣaka Oberlies does not clarify. It is unlikely that he means to


suggest that there is some “riddle” to be solved at the end of Āstīka’s narration as is the case in Rumpelstiltskin, but I also doubt he means that king Janamejaya falls in love with Āstīka as the Scheherazade parallel would suggest. Brodbeck notes that Oberlies’ point makes sense if we accept the equation of Āstīka with Vaiśaṁpāyana (personal communcation, see also Brodbeck 2009, esp. 173-174); on the problems with this equation, see below.

43. Oberlies 2008: 94-95. Contrary to popular belief, snakes do possess necks, although they are highly atrophied (the majority of a snake’s body is comprised of an extremely extended thorax). But I take it that Oberlies does not mean us to take the “Hals(lösungs)”-aspect literally.

44. I.e., its core “historical” narrative of the “Vasu-version.”45. “King Vasu is present from Mbh 1,57—i.e., the immediate beginning

of the ‘Vasu version’—as a (typical) cultural founder [Kulturstifter] and the time in which he rules as a golden age [Goldenes Zeitalter] (on this see in detail in another passage). This will (ultimately) be the reason why this version of the epic began with his story [Geschichte]” (Oberlies 2008: 95, n. 52).

46. Oberlies argues for the priority of the “Vasu-version” on the grounds that “[t]his will (ultimately) be the reason why this version of the epic began with his story [Geschichte]” (95, n. 52), but the argument is a non sequitur. The claim is only tenable, if one already assumes in advance that every epic is necessarily built up on or around some historical narrative, as Oberlies, in fact, does; cf. especially the claim in Oberlies 1998 that “in the Mahābhārata—as in so many heroic epics [Heldenepen]—historical and quasi-historical facts are transformed through a narrative template” (139, n.51).

47. Oberlies’ reconstruction does answer some questions, in particular, the question of how the original “historical” “Vasu-version” could give rise to the highly mythologized and theological epic we have at present (cf. van Buitenen on “inept mythification”; 1973: xx), but these are concerns endemic to German Lutheranism rather than being reflective of genuine issues in the epic; see my discussion of Oberlies’ motivations below.

48. Oberlies, in fact, concedes that the epic is carefully redacted (cf. Oberlies 2008: 75-78), but then defers to specious text-historical justifications for this seeming order. Here one again sees a massive historicist prejudice at work that has so marred German epic scholarship.

49. For a useful discussion of hermeneutics in recent philosophy, see Malbon 1983. The term “hermeneutics” shares “a linguistic root with the name of the Greek god Hermes, the messenger of the gods and the inventor or discoverer of language and writing. The three basic meanings of hermeneuein are: (1) to


speak (or express or say), (2) to explain (or interpret or comment upon), (3) to translate.” The “foundational Hermes process” at work in all three cases is thus the same: “in all three cases, something foreign, strange, separated in time, space, or experience is made familiar, present, comprehensible; something requiring representation, explanation, or translation is somehow ‘brought to understanding’ is ‘interpreted’” (quoting and heavily paraphrasing Malbon 1983: 212)

50. Paraphrasing Wittgenstein, one could perhaps formulate a principle of the form: Whereof there is no manuscript evidence, thereof one must remain silent. In Wittgenstein’s Tractatus-Logico-Philosophicus, where the principle originally appears (“Whereof one cannot speak, thereof one must be silent,” 7; Ogden trans.), the principle functions to separate meaningful combinations of signs from nonsensical combinations. In our present case, its function would be to separate scholarship from unverifiable speculation.

51. The prejudice appears to be innate to modernity; on Annas’ criticisms of Plato, see Annas 1981, and, for my rejoinder, see Adluri 2010c.

52. See Mehta 1973, and, more recently, Tsuchida 2008 and Oberlies 2008. Tsuchida’s work is riddled with logical fallacies such as petitio principii, ignorantia elenchii, and at least one ad hominem attack (cf. Tsuchida 2008: 6); I cannot discuss Tsuchida in the present paper, but I take up Oberlies’ work, which is based mainly on Tsuchida’s, in detail in section 9.

53. On Nīlakaṇṭha’s editorial principles, cf. Sukthankar 1933: lxv-lxx. In Sukthankar’s words, “Nīlakaṇṭha’s guiding principle, on his admission, was to make the Mahābhārata a thesaurus of all excellences (culled no matter from what source” (ibid., lxvii; emphasis in original). For a useful introduction to Nīlakaṇṭha’s work, see Minkowski 2010.

54. Sukthankar 1933: lxxxvi; emphasis in original. Even though Sukthankar hesitates to acknowledge both beginnings as genuine, he remains true to the principles established for the reconstruction of the CE text.

55. Ibid., v. The criterion of “appropriateness,” unfortunately admits of no objective quantification or evaluation; as a subjective category, it has been utterly abused in “text-historical” criticism to make the text conform to specific ideologies or perceptions of what an ideal “Ur-Bhārata” would have looked like.

56. Cf. Sukthankar 1933: 93. “Before loma°, K3 Ñ V1 B (except B3; B1 in brackets) Dn3 D1.4.6 Nīlp ins. sautiruvāca (cf. Nīl. comm.); T (T2 with prefixed śrī) G4-6 M2.4 (with prefixed śrī).5 sūtaḥ.” With the exception of Ś, manuscripts from all major groups are present in this list.


57. See Mahadevan’s contribution in this volume for evidence of the antiquity of the text as recovered in the CE; based on the “three rails” of criticism, Mahadevan suggests a date of 3rd to 2nd century BCE for the CE text, putting it substantially further back than the four centuries of evidence of the manuscripts. If the two beginnings are not original, they must have been added prior to the date supplied by Mahadevan, and that is to say, prior to the CE, which is to enter the realm of Austin’s hypothetical M-N text(s). Needless to say, that puts us in the realm of unverifiable speculation, based on nothing more than some scholars’ intuitions of some kind of original “heroic epic” (Heldenepos), a prejudice that can be traced back to 19th century German longings for “Āryan” or “Indo-Germanic” origins; on the origins of the Indogermanisches Urepos hypothesis, see Vishwa Adluri & Joydeep Bagchee, The Nay Science: A History of German Indology (n.d.).

58. By this I mean, of course, that the critic must provide manuscript evidence, rather than speculate about a text that “could have been, should have been.” Such speculations do extraordinary harm to the discipline of philology, as they completely undermine all scholarly canons: it would seem that, misled by the maxim of text-critical reconstruction brevior lectio praeferenda est (the shortest text is to be preferred), some scholars have taken it to the extreme of hypostatizing shorter and shorter texts to the point of volatilizing the text as we have it entirely. Our problem, then, is not so much one of “Einschaltungen” in the epic as the irrational “Ausschaltung” of the manuscript evidence in search of a hypothetical “text that is no text.” Fitzgerald’s hypothesis of a “Pāṇḍava-Bhārata” perfectly illustrates the problems with this approach (see Fitzgerald 1983 and 2010).

59. See Hiltebeitel’s contribution in this volume.60. I owe the idea of “reading communities” to Alf Hiltebeitel, who has

been especially instrumental in redirecting our attention to the reception of the text among its traditional reading communities as a crucial element in understanding the text (see Hiltebeitel 2005 and 2006; on the significance of Hiltebeitel’s turn to a consideration of the reception of the epic, see the respective introductions to Reading the Fifth Veda (Hiltebeitel 2011a) and When the Goddess was a Woman (Hiltebeitel 2011b).

61. Cf. van Buitenen for the suggestion that if it “pleased” the manuscript owner he could “insert in his loose-leaf book a couple of leaves containing a variant version of one of those stories” “without compunction” (1973: xxix).

62. In fact, even if we were to assume a redactor as Tsuchida, Oberlies et al. do, there is no reason why the redactor, who undertook to collate, edit,


and unify an entire library of stories and retellings, could not have simply eliminated one beginning. Tsuchida and Mehta both proffer putative reasons for why he could not or did not wish to do so, but this is to enter the realm of psychobiography. Tsuchida proceeds ex hypothesi throughout; the following passage provides some indication of the circular and self-referential nature of his arguments: “It goes almost without saying that these two sub-Parvans attained their present shape only as the result of a gradual and intricate process of enlargement. One can, nevertheless, fully grasp what the original compiler of the prologue intended to present with his compilation. His main purpose was to narrate Janamejaya’s celebration of the sarpasatra and other events which finally converge on the start of Vaiśaṁpāyana’s recital of the epic. Most probably it was this very compiler of the prologue who elaborated framework U-2 in order to put his own genesis of the Mbh into the mouth of some authoritative narrator. To the question of whether this compiler ever consulted any other independent version U, now lost, or whether he simply followed the current tradition of Ugraśravas’ narratorship of the epic and purāṇic texts one cannot give any exact answer, though the latter supposition seems more plausible than the former. The compilers, indeed, who laid out frameworks U-1 and U-2 must have been still quite well-acquainted with the ancient bardic tradition. But there is no need at all for us to think that they also belonged to the same class of sūta as Ugraśravas, Lomaharṣaṇa and Saṃjaya. Most probably the epic texts they handled in their compilatory activities had already been transmitted in written form. It would thus be futile to look for any direct vestige of the oral tradition in the frame-construction of the present Mbh” (Tsuchida 2008: 9; italics mine).

63. Text-historical “reconstructions” of the process by which the double beginning arose are entirely conjectural and subjective. M. Biardeau has rightly drawn attention to brahmins’ memories (Biardeau 2002, 2: 747-749). While I cannot agree with her on the issue of orality, one must take her suggestion seriously: it is hardly likely that brahmins capable of reciting the entire epic from memory would not be aware of the “dissonance” among its first four opening parvans. Nor is the stereotype of brahmins committed to an unchanging textual corpus realistic, as Nīlakaṇṭha’s edition demonstrates. If the double beginning was added to the text at a certain stage in the Mahābhārata’s history, then it was as part of a highly sophisticated redactorial and hermeneutic program—an alternative, however, I reject, as the “archetypal redactor” is an invention of 19th century text-historicism based on models of textual transmission and growth that are simply inadequate to the reality of the Indian context.


64. For a more detailed analysis of the significance of the number eighteen, see Sarin 2004.

65. dvādaśa varṣāṇi; 1.204.8 and 1.205.30. Van Buitenen translates: “twelve months.”

66. Oberlies 2008: 75; but see Oberlies 2005 on the issue of whether 12 months or 12 years are intended.

67. On Ruru’s doubling as a metaphor for self-reflection leading to the arrival of being (Āstīka), see Adluri & Bagchee (n.d.).

68. On doubling as a feature of being itself, see my recent book (Adluri 2011a).

69. Austin rightly points out that many of these features could also occur in a narrative based on linear conception of time (personal communication). A detailed analysis of doubling as a feature of being itself would take me far afield of the present paper (see, however, Adluri 2011a), so I will just note that the Mahābhārata’s use of doubling is not restricted to the observation that things recur in time but has the function of showing that the universe is a copy without an original or, in Plato’s words, the universe is a copy of a copy. It is this philosophical aspect to doubling that is important here.

70. I gratefully acknowledge Simon Brodbeck’s hint that the idea of “expansion” is the epic’s own fiction; in retrospect, my previous draft (which had the words “Acknowledging that…”) conceded too much.

71. On the problem of positing absolute origins, see section 8 below. Every serious attempt at understanding the epic must take the hermeneutic framework provided by the epic itself seriously; without taking seriously indications such as the separation of being and becoming (1.1.187-190 and 1.1.191, 1.1.193-195), the separation of eonic time from the eternity of Brahman (1.1.37-38 and 1.1.27-28), and the difficulty of making a beginning in a narrative conception that does not allow for absolute beginnings in time, there can be no meaningful interpretation of the epic, in spite of the masses of “critical” scholarship produced on it.

72. Hiltebeitel 2001; see esp. the chapter “Conventions of the Naimiṣa Forest” for its rewarding discussion of the importance of the Naimiṣa frame, long neglected by the “analytic” school, to the epic.

73. janamejayasya rājarṣeḥ sarpasatre mahātmanaḥ / samīpe pārthivendrasya samyak pārikṣitasya ca // kṛṣṇadvaipāyanaproktāḥ supuṇyā vividhāḥ kathāḥ / kathitāś cāpi vidhivad yā vaiśaṁpāyanena vai; 1.1.8-9.

74. śrutvāhaṁ tā vicitrārthā mahābhāratasaṁśritāḥ; 1.1.10. There is also a suggestion in the second beginning that Ugraśravas is not as yet fully proficient


in the narrative: Śaunaka asks whether he has “learned” (adhīṣe; 1.5.1) whatever had been learned long ago by his father, and seems to be testing either his memory or his understanding at select intervals in the narrative (cf. 1.14.1-3; 1.31.1-3; 1.36.1-2). In contrast, the Naimiṣa sages treat Ugraśravas much more deferentially, as someone well-versed in the narrative, and Ugraśravas’ response reflects a quiet confidence (cf. “it is the delight of the learned”; 1.1.26) that is perhaps absent in the second narration.

75. dvaipāyanena yat proktaṁ purāṇaṁ paramarṣiṇā / surair brahmarṣibhiś caiva śrutvā yad abhipūjitam // tasyākhyānavariṣṭhasya vicitrapadaparvaṇaḥ / sūkṣmārthanyāyayuktasya vedārthair bhūṣitasya ca // bhāratasyetihāsasya puṇyāṁ granthārthasaṁyutām / saṁskāropagatāṁ brāhmīṁ nānāśāstropabṛṁhitām // janamejayasya yāṁ rājño vaiśaṁpāyana uktavān / yathāvat sa ṛṣis tuṣṭyā satre dvaipāyanājñayā // vedaiś caturbhiḥ samitāṁ vyāsasyādbhutakarmaṇaḥ / saṁhitāṁ śrotum icchāmo dharmyāṁ pāpabhayāpahām; 1.1.15-19.

76. On the epic’s popular title as a strī-śūdra-veda or a “Veda for women and śūdras,” I find Black’s discussion in a forthcoming article especially useful. Black notes that “Although the Critical Edition does not contain the well known description of the epic as a text ‘for women and śūdras’, the Mahābhārata does seem to regard itself as delivering a universal message. In addition to the numerous phalaśrutis throughout the text that address audiences beyond those who are male and of the twiceborn classes, Vyāsa himself, in the Śāntiparvan, instructs his disciples to teach his story to members of all four varṇas (12.314.45). In light of the author’s own instruction to his students, what better way to reach a diverse and inclusive audience than to have Brahmanical knowledge communicated by someone of lower birth. Indeed, without making any claims about the ‘real’ history of the text, this scenario seems to be the one that the Mahābhārata tells about its own transmission: originating among brahmins, but learned by sūtas such as Ugraśravas who, implicitly, share such tales and legends with a wide audience, particularly when they frequent popular pilgrimage sites, such as the ones Ugraśravas visited before arriving in the Naimiṣa Forest” (Black, forthcoming: 11). On the term strī-śūdra-veda itself, Black notes that the description “appears in the Bhāgavata Purāṇa (1.4.25), which says that Vyāsa composed his story out of compassion for women, śūdras, and uneducated twice-borns. Nonetheless, there are a number of individual phalaśrutis throughout the text that offer rewards for śūdras and women” (ibid., n. 16).

77. See Brodbeck 2009: 244-245, esp. n. 40. See also Brodbeck in this volume: “there were two Naimiṣa Forest recitations: one by Ugraśravas to Śaunaka, and a later one by Ugraśravas to the ṛṣis. This is clear when Ugraśravas, addressing the


ṛṣis, refers to his dialogue with Śaunaka as having already happened.”78. yat tu śaunakasatre te bhāratākhyānavistaram / ākhyāsye tatra paulomam

ākhyānaṁ cāditaḥ param; 1.2.29.79. Cf. van Buitenen 1973: 439. “Told at Śaunaka’s session: but he is yet to do

it; below, 1.4-5; 13.” 80. A study of the terms vistareṇa and punaḥ suggest that duplications

in the narrative are due to a desire for poetic retelling rather than being insertions of different versions of stories. Śaunaka, for example, frequently requests longer, more detailed accounts (vistareṇa) from the bard, who then follows up his initial narration with a longer version. The word vistaram is used at 1.2.29 to express the relation of the sections of the text that precede the Paulomaparvan to the Mahābhārata from this book onward. It reappears at 1.2.126 and clarifies the entire sub-narrative of the “Rāmāyaṇa” as an elaboration. In the Paulomaparvan, the bard provides a summary genealogy of the Bhṛgus. Śaunaka then asks for further details of Bhṛgu’s son, Cyavana. Ugraśravas complies and later says that he will present a full account (vistareṇa pravakṣyāmi; 1.8.3) of all of Bhṛgu’s descendants, including Ruru. Śaunaka later explicitly requests a fuller account of Āstīka of Ugraśravas (saute kathaya tām etāṁ vistareṇa kathāṁ punaḥ; 1.14.1). This pattern is repeated in the Āstīkaparvan (vistareṇa punar vada; 1.45.1 and sarvaṁ vistaratas; 1.48.3) and in the Ādivaṁśāvataraṇaparvan (vistaraśravaṇe jātaṁ kautūhalam atīva me; 1.56.2 and sa bhavān vistareṇemāṁ punar ākhyātum arhati; 1.56.3). The next time the word vistareṇa appears, it is within the inner narrative: Janamejaya rather than Śaunaka addresses the dvijottama Vaiśaṁpāyana (etad icchāmy ahaṁ śrotuṁ vistareṇa dvijottama; 1.71.2).

81. etat parvaśataṁ pūrṇaṁ vyāsenoktaṁ mahātmanā / yathāvat sūtaputreṇa lomaharṣaṇinā punaḥ // kathitaṁ naimiṣāraṇye parvāṇy aṣṭādaśaiva tu; 1.2.70-71.

82. Brodbeck 2009: 245, n. 40.83. “The latter/later satra is iterable because although it only happens once in

the text, it can happen each time the text is presented—be this to many listeners of a reciter at once, or to one silent reader at once” (personal communication).

84. Without the addition of the reader, who represents consciousness and life, the text remains dead, a corpus suitable only for autopsy (Greek autopsein = “to see for oneself”). Approaches that seek to reduce the text to nothing more than a corpus available for “text-historical” dissections are thus incapable of appreciating the sublime logic of this great epic, as Sukthankar already demonstrated (see Sukthankar 1957). Sukthankar’s On the Meaning of the Mahābhārata should be made required reading for anyone attempting his or her hand at “Mahābhārata criticism.”


85. This is a practice attested to in the recitation of mantras; in the Gāyatrī mantra, for example, it is customary to recall the god (Savitṛ), then the ṛṣi or sage to whom the mantra is traditionally attributed (Viśvamitra), and then the meter (gāyatrī). Its occurrence here in the Anukramaṇiparvan hints at the epic’s self-conscious understanding of itself as a mantra; on the Mahābhārata as a mantra, see my Sacrificial Ontology and Human Destiny in the Mahābhārata.

86. niṣprabhe ’smin nirāloke sarvatas tamasāvṛte / bṛhad aṇḍam abhūd ekaṁ prajānāṁ bījam akṣayam // yugasyādau nimittaṁ tan mahad divyaṁ pracakṣate / yasmiṁs tac chrūyate satyaṁ jyotir brahma sanātanam; 1.1.27-28.

87. The term bhavagītā is mine. Although Dhṛtarāṣṭra’s bhavagītā precedes the Bhagavadgītā from the perspective of textual order, from a chronological perspective it is actually later, occurring after Bhīṣma’s fall and after the end of the Kurukṣetra battle. Dhṛtarāṣṭra has already heard the Bhagavadgītā’s soteriological philosophy, but appears to have gained nothing from it, as his lament and Saṁjaya’s subsequent reprimand demonstrate.

88. kaṣṭaṁ yuddhe daśa śeṣāḥ śrutā me; trayo ’smākaṁ pāṇḍavānāṁ ca sapta / dvyūnā viṁśatir āhatākṣauhiṇīnāṁ; tasmin saṁgrāme vigrahe kṣatriyāṇām // tamasā tv abhyavastīrṇo moha āviśatīva mām / saṁjñāṁ nopalabhe sūta mano vihvalatīva me; 1.1.158-159.

89. kālamūlam idaṁ sarvaṁ bhāvābhāvau sukhāsukhe // kālaḥ pacati bhūtāni kālaḥ saṁharati prajāḥ / nirdahantaṁ prajāḥ kālaṁ kālaḥ śamayate punaḥ // kālo vikurute bhāvān sarvāml loke śubhāśubhān / kālaḥ saṁkṣipate sarvāḥ prajā visṛjate punaḥ / kālaḥ sarveṣu bhūteṣu caraty avidhṛtaḥ samaḥ // atītānāgatā bhāvā ye ca vartanti sāṁpratamtān / kālanirmitān buddhvā na saṁjñāṁ hātum arhasi; 1.1.187-190.

90. atropaniṣadaṁ puṇyāṁ kṛṣṇadvaipāyano ‘bravīt /… bhagavān vāsudevaś ca kīrtyate ‘tra sanātanaḥ / sa hi satyam ṛtaṁ caiva pavitraṁ puṇyam eva ca // śāśvataṁ brahma paramaṁ dhruvaṁ jyotiḥ sanātanam / yasya divyāni karmāṇi kathayanti manīṣiṇaḥ // asat sat sad asac caiva yasmād devāt pravartate / saṁtatiś ca pravṛttiś ca janma mṛtyuḥ punarbhavaḥ; 1.1.191, 193-195.

91. This is not to deny the richness of the epic’s contents, but to underscore the problem of mortality as the epic’s enduring and central concern. Austin rightly points out that the epic is also concerned with “themes of the problem of dharma, necessity of violence, renunciation, the divinity of Kṛṣṇa etc., none of which can simply be reduced to the other” (personal communication). My point, however, is not so much that there is one single problem in the epic as that there is a basic concern with being in time expressed in the form of the question “how now to live?” that provides the motive force for many of the issues Austin rightly cites. Questions of dharma, of violence, renunciation, etc., only make


sense against the background of this basic problem of being in time; which, of course, is not to reduce them to this problem but to acknowledge the infinite richness of the problem.

92. catvāra ekato vedā bhāratam ca ekam ekatah / samāgataih surarṣibhis tulām āropitam purā / mahattve ca gurutve ca dhriyamāṇam tato adhikam // mahattvād bhāravattvāc ca mahā bhāratam ucyate / niruktam asya yo veda sarva pāpaih pramucyate; 1.1.208-209.

93. Samantapañcaka is another name for Kurukṣetra or the “Field of the Kurus” where the battle between the Kauravas and Pāṇḍavas was fought. It gets its name “five lakes” from the legend that Rāma Jāmadagnya filled five lakes with the blood of the kṣatriyas. Curiously, throughout the Anukramaṇi- and Parvasaṁgrahaparvans, this site is referred to by its older name rather than as Kurukṣetra: Ugraśravas, asked to narrate his doings, recounts that he was present at Janamejaya’s sarpasatra where Vaiśaṁpāyana recited the Mahābhārata, after which he journeyed to Samantapañcaka (1.1.11; cf. also 1.2.1) the site of the great battle between the Kauravas and Pāṇḍavas. In fact, the name Kurukṣetra never occurs within the Ugraśravas narration proper (i.e., excluding the Vaiśaṁpāyana narration, which is embedded in the Ugraśravas narration), with the exception of the third minor book, the Pauṣyaparvan, which is embedded between the two beginnings (1.3.1, 144, 145). It is almost as though Ugraśravas’ narration, which is soon to explode into the massive account of the violent conflict between the Kuru princes, is looking back more to its mythic and sacrificial predecessor, Rāma Jāmadagnya’s slaughter of the kṣatra, than forward to the upcoming events of the Kuru bheda narrative.

94. Credit for noticing the significance of the yuga scheme, and its relation to the avatāra myth, must go to Biardeau (1976). See also Couture’s recent essay on avatāra, continuing Biardeau’s excellent work, as relating to the concept of divine “play” (Couture 2001). As though anticipating that the main weakness of Western “critical” scholarship would turn on its inability to appreciate the centrality the Mahābhārata as divine play, see Sukthankar (1957: 25): “How could European savants, lacking as they do in their intellectual make-up the millenia old back-ground of Indian culture, ever hope to penetrate this inscrutable mask of the Unknowable pulling faces at them, befooling them and enjoying their antics?”

95. purāṇam akhilaṁ tāta pitā te ’dhītavān purā / kaccit tvam api tat sarvam adhīṣe lomaharṣaṇe // purāṇe hi kathā divyā ādivaṁśāś ca dhīmatām / kathyante tāḥ purāsmābhiḥ śrutāḥ pūrvaṁ pitus tava // tatra vaṁśam ahaṁ pūrvaṁ śrotum icchāmi bhārgavam / kathayasva kathām etāṁ kalyāḥ sma śravaṇe tava; 1.5.1-3.

96. yad adhītaṁ purā samyag dvijaśreṣṭha mahātmabhiḥ / vaiśaṁpāyanaviprādyais


taiś cāpi kathitaṁ purā // yad adhītaṁ ca pitrā me samyak caiva tato mayā; 1.5.4-5.97. Cyavana means “moving, the being deprived of, falling from any divine

existence for being re-born as a man” (Monier-Williams, sv). It derives from √cyu which means “to come forth from, to drop from, to fall down, to die, to be deprived of, to perish” (Monier-Williams, sv). The Mahābhārata itself high–lights this etymology in elucidating Cyavana’s name: “And the child she bore alive in her womb, O descendant of the Bhṛgus, angrily fell [cyutaḥ] from his mother’s womb and thus became known as Cyavana” (tataḥ sa garbho nivasan kukṣau bhṛgukulodvaha / roṣān mātuś cyutaḥ kukṣeś cyavanas tena so ’bhavat; 1.6.2). Etymologies are a crucial element in understanding the epic, as Śaunaka’s question to the bard in the Āstīkāparvan demonstrates. Śaunaka asks the bard: “This I wish to hear. Pray tell me the etymology of jaratkāru.” When the bard resolves the word into jarā and kāru and interprets the word as “monstrous destruction” (cf. jaratkāru niruktam tvam yathāvad vaktum arhasi // jarā iti kṣayam āhur vai dāruṇam kāru samjnitam; 1.36.2-3), demonstrating that he has grasped the soteriological ontology implicit in the Jaratkāru narrative, Śaunaka laughs out loud and says “That fits!” (upapannam iti; 1.36.5). One must read this passage in the full light of a previous occurrence of etymology in the text, in which the word “Mahābhārata” is etymologically related to salvation: “Once the divine seers foregathered, and on one scale they hung the four Vedas in the balance, and on the other scale The Bhārata; and both in size and weight it was the heavier. Therefore, because of its size and weight, it is called The Mahābhārata—he who knows this etymology is freed from all sins” (catvāra ekato vedā bhāratam ca ekam ekatah / samāgataih surarṣibhis tulām āropitam purā / mahattve ca gurutve ca dhriyamāṇam tato adhikam // mahattvād bhāravattvāc ca mahā bhāratam ucyate / niruktam asya yo veda sarva pāpaih pramucyate; 1.1.208-209). For a fuller discussion of how the etymologies of the names in the Paulomaparvan encode a philosophical message, see Adluri & Bagchee (n.d.). I thank Christopher Austin for pointing out that tying playful etymologies to salvation is routine procedure in the Brāhmaṇas (personal communication).

98. For an analysis of the Ruru narrative, especially in its soteriological aspect, see Adluri & Bagchee (n.d.).

99. asty eṣa garbhaḥ subhage tava; 1.43.38,100. astīty uktvā gato yasmāt pitā garbhastham eva tam / vanaṁ tasmād idaṁ

tasya nāmāstīketi viśrutam; 1.44.20. I thank Gregory Bailey for this translation (personal communication).

101. I discuss both these narratives in greater depth in my Sacrificial Ontology and Human Destiny in the Mahābhārata.


102. For a discussion of the significance of Āstīka, who represents the soteriological power of being that is born through hermeneutics, see n. 119 below. Āstīka is key to understanding why the first narration concludes with the Āstīkaparvan and a new beginning is made from 1.57 onward, i.e., in the Ādivaṁśāvataraṇaparvan. For a more detailed discussion of how the epic’s opening books articulate a comprehensive soteriological program, see my Sacrificial Ontology and Human Destiny in the Mahābhārata.

103. By “becoming,” I understand the experience of change in its twofold aspect of coming-to-be (Greek genēsis) and perishing (phthora). I use the overarching term “becoming” to convey the sense of a range of terms such as bhavābhavau (becoming, literally “being—non-being”), vṛtti (disturbance), saṁsāra (eternal recurrence), jayājayau (victory and defeat), lābhālābhau (gain and loss) and sukhaduḥkha (pleasure and pain). I translate these with the term “becoming.” The epic also often uses the word kāla (time) in place of bhavābhavau; in these cases as well, I translate with “becoming.” For references to bhavābhavau, see 3.148.9 (in relation to the yugas and the puruṣārthas), 3.279.10 (in relation to pleasure and pain), 5.36.45 and 12.26.31 (on self-control and salvation), 5.39.1 (in relation to finitude and fate); see also 12.137.51, 12.221.94 and 12.233.11.

104. For a discussion of the four “genera of becoming” (sacrifice, cosmology, genealogy, and war or agōn), see my Sacrificial Ontology and Human Destiny in the Mahābhārata. The four genera are crucial to understanding the Mahābhārata as the entire epic is articulated in terms of these four genera. The four genera are also key to understanding the epic’s dual claim to being both an itihāsa (cf. itihāsam; 1.1.24, 52, 204) and a Veda (cf. kārṣṇaṁ vedam; 1.1.205) as they provide a way of speaking about “becoming” without reifying it, as is the case in history.

105. The obsession with a historical kingdom has been a characteristic feature of German epic scholarship since C. Lassen (1837) who sought to reconstruct Indian ethnology and prehistory on the basis of the epic, the Kauravas and Pāṇḍavas being identified with “white Aryans [weisse Arier]” and “black aborigines [schwarzen Urbewohner]” respectively and the epic as a whole being interpreted as the record of a historical conflict for white supremacy. Lassen’s racial hypothesis lays the fundament for over two centuries of German epic studies, beginning with Holtzmann, Sr. (1854), author of the infamous “inversion hypothesis,” according to which the Kauravas were the heroes of the “original” epic and were later denigrated by scurrilous “Brahmanic” redactors. Goldstücker (1879) sought to anchor Holtzmann’s thesis in the text by distinguishing a Bhārata of 24,000 verses from the Mahābhārata of 100,000


verses. Wecker (1888) continues this strain of “Indo-Germanic” thinking by setting up explicit comparisons between the Mahābhārata and the Nibelungenlied (Dhritaraschtra = Armenrich, Bhischma = Rüdiger, Karna = Siegferd, Arjuna = Iring, Krischna Kesava and Krischna Draupadi = Kriemhilde, etc.). Finally, Holtzmann, Jr. (1892) saw in the epic evidence of a “Brahmanic Counter-Reformation [Gegenreformation]” against a supposed Buddhist Enlightenment, explicitly describing Aśoka as a mixture of Frederick the Great and Lessing! The obsession with projecting Lutheran-Protestant values onto Buddhism continues in some more recent literature; a case in point being Bronkhorst 2011.

106. Not shown on this chart is the shared opening line—lomaharṣaṇaputra ugraśravāḥ sūtaḥ paurāṇiko naimiṣāraṇye śaunakasya kulapater dvādaśavārṣike satre (1.1.1 and 1.4.1)—which holds the two beginnings together.

107. The imagery is Vedic, but I argue one can find similar descriptions in the epic as well; most notably, in the description of the churning of the ocean (1.16.1-40).

108. For raṇayajña, see 2.20.15 (svargaṁ hy eva samāsthāya raṇayajñeṣu dīkṣitāḥ / yajante kṣatriyā lokāṁs tad viddhi magadhādhipa), 5.57.12 (ahaṁ ca tāta karṇaś ca raṇayajñaṁ vitatya vai / yudhiṣṭhiraṁ paśuṁ kṛtvā dīkṣitau bharatarṣabha), 5.154.4 (raṇayajñe pratibhaye svābhīle lomaharṣaṇe / dīkṣitaṁ cirarātrāya śrutvā rājā yudhiṣṭhiraḥ), 9.59.25 (yuddhadīkṣāṁ praviśyājau raṇayajñaṁ vitatya ca / hutvātmānam amitrāgnau prāpa cāvabhṛthaṁ yaśaḥ).

109. Sacrificial remainders can be of multiple kinds, as I have already argued in my book Sacrificial Ontology and Human Destiny in the Mahābhārata. Parikṣit is the “genealogical remainder,” but there is also a “textual remainder,” namely, the Mahābhārata itself, which survives the previous cosmological cycle (dvāpara yuga) to enter the next (kali yuga).

110. Cf. Bhagavadgītā 3.13. yajñaśiṣṭāśinaḥ santo mucyante sarvakilbiṣaiḥ / bhuñjate te tv aghaṁ pāpā ye pacanty ātmakāraṇāt //

111. Of course, immortality is of two types: the limited immortality offered by historical fame and the transcendence of being. Janamejaya, I argue, attains the higher immortality of being through the sarpasatra. For an articulation of my thesis of “double transcendence” in Plato, see Adluri 2011d.

112. Oldenberg does not tell us which verse except to note that it is from the Atharvaveda: “A royal father and a royal son, both surrounded by great splendor. The poet of a poem preserved in the Atharvaveda tells us where they ruled: over the Kurus, precisely the tribe that stands in the center of the later Vedic period and the Mahābhārata” (Oldenberg 1922: 8).

113. On Āstīka as being, see my comments below.


114. “Original,” of course, not in the sense of being temporally prior, but in the sense of having a logical and hermeneutic priority over the other beginnings. As Heidegger emphasizes over and over, human existence always already finds itself within a certain hermeneutic situation, a “state of having been interpreted” (Ausgelegtheit) that is the condition of its being able to carry out an explicit inquiry or discussion. It is in this sense that the Pauṣya narrative has a certain priority over the others: as the hermeneutic beginning it refers to the fact that we always already find ourself within a certain situation of interpretation and understanding that first enables explicit reflection on origins or causes, such as the reflection on cosmological or genealogical beginnings found in the Anukramṇiparvan and the Paulomaparvan respectively. For a concise treatment of the concept of the “hermeneutic situation,” see now Heidegger 2005 (Anhang III).

115. The exceptions are Upamanyu’s verses in praise of the Aśvins (1.3.60-70), Uttaṅka’s verses in praise of the snakes (1.3.38-46), Uttaṅka’s verse in praise of the wondrous sights he sees in the underworld (1.3.150-153), and king Janamejaya’s response to Uttaṅka (1.3.183-184).

116. yat tu śaunakasatre te bhāratākhyānavistaram / ākhyāsye tatra paulomam ākhyānaṁ cāditaḥ param; 1.2.29.

117. janamejayaḥ pārikṣitaḥ saha bhrātṛbhiḥ kurukṣetre dīrghasatram upāste / tasya bhrātaras trayaḥ śrutasena ugraseno bhīmasena iti; 1.3.1

118. On initiatory motifs in the Pauṣyaparvan, see Adluri 2010a; on sacrificial and pedagogic aspects, see Adluri 2011c.

119. As Janamejaya’s snake sacrifice unfolds, Āstīka arrives at the sacrificial grounds and praises the king. The king offers him a boon and the brahmin asks that the sacrifice be stopped. The king implores him to ask for any other boon, but not to demand the sacrifice’s interruption. Takṣaka, the intended victim of the sacrifice, has begun his downward fall out of the sky into the sacrificial fire, when Āstīka says to him three times “Stay! Stay!” (tiṣṭha tiṣṭheti; 1.53.5) and arrests the frightened snake’s fall. The king then assents to Āstīka’s wish and ends the sacrifice.

120. On the Ṛgvedic background to the Mahābhārata, see Feller 2004. Feller demonstrates that the Pauṣyaparvan is especially rich in Ṛgvedic resonances, but does not discuss the Saramā myth. See also Adluri 2009.

121. Cf. Bloomfield 1896: 425; Hopkins 1908: 505; Srinivasan 1973: 45.122. ṚV 10.108.7 identifies this wealth as gó, aśva, vásu (7.90 adds híraṇya

to the list). However, cows and horses seem to be metaphorically identified with the dawn in this hymn, and the entire myth seems to be an aetiological


description of the nightly disappearance of the sun and its reappearance with the rays of the dawn. For horses see Macdonell 1897: 31, 47. For cows see Srinivasan 1973: 53-4. Srinivasan notes: “It is well known that terms meaning ‘cow’ may be used metaphorically for ‘rays of light, rays of dawn, dawns.’ The Indian lexical material is the earliest source to direct our attention to this figure of speech. Nighaṇṭu 1.5 lists gāvaḥ and usrāḥ as raśmi terms, and, Nirukta 2.6 reiterates this: sarve ‘pi raśmayo gāva ucyante. In a list of animals associated with different gods in the capacity of vahanas, Nighaṇṭu 1.15 mentions aruṇyo gāva uṣasām.” Srinivasan continues: “How do we know that the Paṇis withhold the rays of the Dawn? The frequent use of usrā and usriyā in this context is our first indication. Both terms are derivatives of vas ‘to shine’… In the myth of the Paṇis, usríyā = the cow as ‘light’ (e.g. 7.57.7; 7.81.2) as well as the oblation-giving cow (4.50.5)” (ibid.).

123. Although the Paṇis are identified as dásyus, suggesting a netherworld home, the myth does not explicitly support this interpretation. However, scholars (e.g., Srinivasan) have attempted to locate the Paṇis’ home “in the western region of a lower world,” a description that would also accord with the experience of the sun’s passage across the sky and nightly setting in the west.

124. Indra cuts through the enclosure (valá; 10.67.6) where the cows have been hidden; he makes a path to drive out the cows (3.30.10; cf. also 2.14.3). Other passages, however, refer to Bṛhaspati, the Aṅgirases, the Navagvas and Daśagvas, as well as other priests in this context. On the identity of the Paṇis’ foes, see Srinivasan 1973: 49-52.

125. “Throughout the Rigveda and Brāhmaṇas the dog and Agni are both regarded as messengers of the gods. As Sāramēya, the Greek Hermes, he is both messenger and watch-dog, and both chronologically and mythologically he and Sarama, the dawn, stand, as Max Müller says, ‘on the threshold that separates the gods of light from the gods of darkness’” (Hewitt 1890: 441-2).

126. “In the Vedic religion a dog was sent by Yama to accompany the soul on its journey after death and two four-eyed dogs guard the road that leads to the abode of Yama. And the dogs of Yama were called Sarameyas, which in Greek form, according to Dr. Kuhn, became Hermeias or Hermes, death’s messenger, who was an infernal god, and conducted souls in their exit” (Woolsey 1993: 219).

127. West notes the connection between the Greek god Pan and the Vedic god Pūṣan as well as Pan and Hermes. Interestingly, the same complex of associations (cattle, conducting the souls into the netherworld, making things visible, guiding) applies to all three gods as well as to Saramā. “Some of


Pūṣan’s functions parallel those for which Hermes is noted rather than Pan. Hermes too is a good lookout…, a god of roads… and a guardian of flocks and herds. As [psuchopompos] he guides the dead on the path that they must go, and similarly Pūṣan conducts the dead to join their ancestors (RV 10. 17. 3-6; AV 18. 2. 53-55, cf. 16. 9. 2)… With his knowledge of ways and byways, Hermes can spirit away cattle or other property; he is the patron god of the sneak-thief. But by the same token he is good at finding things that are hidden, he knows where animals have strayed, and he gets the credit if someone makes a lucky discovery…As [mastérios] (Aesch. Supp. 920) he helps people track down their stolen property. Pūṣan for his part is the patron of professional trackers, and can bring lost, hidden, or stolen goods to light [RV 1. 23. 13; 6. 48. 15, 54. 1-2, 8, 10; AV 7. 9. 4], and the same can be said of Hermes. So the Arcadian Pan and the Panhellenic Hermes overlap, and both have many features in common with Pūṣan. Pan was held to be Hermes’ son. It seems likely that originally they were the same” (2007: 282-3). West’s suggestion is especially significant given that Śaunaka, the interlocutor of the Mahābhārata, adopted a Bhāradvāja, seers who were especially known for their worship of Pūṣan (cf. Sarmah 1991: 197).

128. See also Kramrisch 1975: 236. “Although not definitely proved, her name seems to derive from sar, to speed.”

129. … yat kiṁ cid asmadgṛhe parihīyate tad icchāmy aham aparihīṇaṁ bhavatā kriyamāṇam iti; 1.3.86.

130. Ugraśravas and Uttaṅka thus respectively embody the double function of redactorial activity: 1. Preserving and transmitting the text. 2. Explaining the text. The Mahābhārata itself notes these dual functions: “Having expiated upon this great erudition, the seer thereupon made a summary thereof; for the wise wish to retain it for this world, in its parts and its entirety. There are brahmins who learn The Bhārata from Manu onward, others again from the tale of The Book of Āstīka onward, others again from The Tale of Uparicara onward. Learned men elucidate the complex erudition in this Grand Collection; there are those who are experienced in explaining it, others in retaining it” (vistīryaitan mahaj jñānam ṛṣiḥ saṁkṣepam abravīt / iṣṭaṁ hi viduṣāṁ loke samāsavyāsadhāraṇam // manvādi bhārataṁ ke cid āstīkādi tathāpare / tathoparicarādy anye viprāḥ samyag adhīyate // vividhaṁ saṁhitājñānaṁ dīpayanti manīṣiṇaḥ / vyākhyātuṁ kuśalāḥ ke cid granthaṁ dhārayituṁ pare; 1.1.49-51).

131. tad anujāne bhavantam / sarvām eva siddhiṁ prāpsyasi; 1.3.92.132. The link between earrings and hearing (śravanam) is suggestive. Fur–

ther, we may recall that the primary Indian characterization of scriptural or


authoritative texts is śruti (lit. “that which is heard”), suggesting that Uttaṅka’s quest is symbolic of the recovery of sacred meaning or insight. There is another aspect to the story that strengthens the association of earrings with śruti. Staal notes that the “Vedic text, the Aitareya Āraṇyaka (5.5.3) states that a pupil should not recite the Veda after he has eaten meat, seen blood or a dead boy, had intercourse or engaged in writing” (1979: 122-3). In the Uttaṅka narrative, Uttaṅka is unable to see Pauṣya’s wife because he has eaten food previously and is therefore in a state of pollution. Besides relating the earrings to śruti, there are other alternatives: one is to look at the story of Karṇa’s earrings, but that is another story.

133. bhakṣayasvottaṅka / mā vicāraya / upādhyāyenāpi te bhakṣitaṁ pūrvam iti; 1.3.104.

134. Snakes in the Mahābhārata are a symbol of hermeneutics: 1. The image of a snake looping back on itself, although not explicitly provided by the epic but familiar from Gnostic imagery, provides the most impressive description of the text’s self-reflexive character. 2. The snake cycle in the Mahābhārata functions as a hermeneutic laid over the core narrative. Understanding the fate of the snakes lets us understand the fate of the Kuru dynasty. The snake genocide is a foil for the upcoming human genocide that reveals its underlying logic: although most of the snakes are destroyed, a remnant escapes: Takṣaka, who is the sacrificial remainder. 3. The snake realm is the realm of hermeneutics where the inner workings of the universe (fate, time, space and destiny) become visible. This realm requires interpretation in order for its real meaning to become visible. Interestingly, Nāgārjuna, a Buddhist from South India born into a brahmin family, is said to have descended into the snake netherworld to obtain The Hundred Thousand Verse Prajñāpāramitā Sūtra, following which he acquired the name Nāgārjuna. The repetition of a descent into the snake realm in pursuit of wisdom is worth noting, as is the fact that the Mahābhārata is also traditionally said to be comprised of 100,000 verses.

135. nāgā me vaśam īyur iti; 1.3.155.136. tad icchāmi bhavatopadiṣṭaṁ kiṁ tad iti; 1.3.171.137. The great sarpasatra is thus not an actual sacrifice of snakes, but nor is it

a parable for the defeat of snake-worshipping peoples (see Cozad 2004, Kosambi 1964, Pargiter 1913). Instead, it should be read as a parable for interpretation, through which activity the soteriological power of being becomes manifest. It replicates the Kurukṣetra battle at a meta-textual level, allowing us to understand the cosmological and eschatological significance of this battle.

138. Not shown on this chart is the shared opening line—lomaharṣaṇaputra ugraśravāḥ sūtaḥ paurāṇiko naimiṣāraṇye śaunakasya kulapater dvādaśavārṣike satre


(1.1.1 and 1.4.1)—which holds the two beginnings together, or the fact that the Uttaṅka narrative connects the epic’s hermeneutic “beginning” with its end through his two biographies, one in the Pauṣyaparvan and the other in the Āśvamedhikaparvan; for a fuller discussion see Adluri 2011c. The epic is thus not as open-jawed as in this schematic presentation: it is circular, with its end forged to its beginning, as shown in the next chart.

139. Cf. Garuḍa’s theft of the somā, which he “envelope[s] … without drinking it” (apītvaivāmṛtaṁ pakṣī parigṛhyāśu vīryavān; 1.29.11). The basic idea is the same: the hermeneutic key to the text, the secret of immortality, is contained in the text’s “mouth.” The fact that Garuḍa rejects amṛta and asks that the snakes be his food instead is also suggestive. Cf. now my comments on Janamejaya above.

140. Cf. 1.1.48. “… everything has been entered here, and this describes this Book” (iha sarvam anukrāntam uktaṁ granthasya lakṣaṇam).

141. manvādi bhārataṁ ke cid āstīkādi tathāpare / tathoparicarādy anye viprāḥ samyag adhīyate; 1.1.50.

142. vividhaṁ saṁhitājñānaṁ dīpayanti manīṣiṇaḥ / vyākhyātuṁ kuśalāḥ ke cid granthaṁ dhārayituṁ pare; 1.1.51.

143. Obviously, by “philology” I do not mean what is understood as “scientific” or “critical” philology. Contemporary philology is as far-removed from the meaning of this word as it is from the genuine smārta tradition; on the meaning of philologia (“love of the logos”) and its essential connection with philanthrōpia (“love of humanity”), see Plato’s Phaedo 89d-e. Plato, who is the first to use the term philologia, links philologia to the argument for the immortality of the soul and sets apart the philologos (“the lover of speech,” of which Socrates is, of course, the paradigmatic example for Plato) from the misologos (“the hater of discourse”) who is characterized by misanthrōpia.

144. “Die Kapitel 1,4 bis 1,12 sind ein Einschub in eine bestehende textliche Umgebung.”

145. “The snake king Takṣaka sways in the highest danger, and in order to save his ‘neck’, Āstīka narrates a story—the Mahābhārata—until the life-threatening danger for Takṣaka is past.” Ibid., 94-95.

146. I have focused here on the main elements of Oberlies’ reconstruction, but his actual scheme is far more complex. I provide an overview of his complete thesis in the appendix.

147. I.e., at Mahābhārata 1.1.67-94, 1.1.95-159, 1.2.34-69, and 1.2.71-234.148. The Mahābhārata is, as Biardeau so elegantly puts it, “le monument

principal, et sans doute le plus ancien, de la bhakti” (1981: 78, n. 1.), and it


therefore must be read, as I have argued in this paper, in the context this self-understanding rather than being forced to conform to arbitrary and extraneous models.

149. I have in mind especially Oberlies’ reliance on Tsuchida, even though the latter’s work is riddled with logical flaws (including an aad hominem attack against Mehta). In many ways, it would have been preferable if Oberlies had cited his own work, or attempted to arrive independently at the case he wants to argue rather than deferring to Tsuchida; cf. Oberlies 2008. For earlier criticisms of Oberlies’ citational praxis, see Grünendahl 2002.

150. See Hopkins, Fitzgerald, and other defenders of the oral composition hypothesis.

151. For a discussion of Lassen and Goldstücker, see the editors’ introduction to Hiltebeitel 2011b: x, n. 15, see also xxvii, n. 72.

152. The thesis is hardly new as it was first advanced by Lassen in 1867, followed by Goldstücker in 1879, Holtzmann in 1892, and made into a cottage industry by Hopkins (1899 and 1906). It has continued since then in some form or another as a fundament of Germanophonic epic scholarship.

153. dharme cārthe ca kāme ca mokṣe ca bharatarṣabha / yad ihāsti tad anyatra yan nehāsti na tat kva cit //

Bibliography

Vishwa Adluri. 2011a. Parmenides, Plato and Mortal Philosophy. London: Continuum Publishing.

———. 2011b. “Hermeneutics and Criticism.” In Ways and Reasons for Thinking about the Mahābhārata as a Whole. Ed. Vishwa Adluri. Pune: Bhandarkar Oriental Institute.

———. 2011c. “Hermeneutics and Narrative Architecture in the Mahābhārata.” In Ways and Reasons for Thinking about the Mahābhārata as a Whole. Ed. Vishwa Adluri. Pune: Bhandarkar Oriental Institute.

———. 2011d. “Art-Critic and Warrior: Book X of the Republic on Becoming a Platonic Philosopher.” Paper presented at Xavier University.

———. 2010a. “The Perils of Textual Transmission: Decapitation and Recapitulation.” Seminar, no. 608, special issue titled The Enduring Epic: A Symposium on Some Concerns Raised in the Mahābhārata: 48-54; translated into French as “Les dangers de la transmission textuelle: décapitation et récapitulation” by G. Schaufelberger.


———. 2010b. Review of The Bhagavadgītā: Doctrines and Context, by Angelika Malinar. History of Religions 50,1: 102–107.

———. 2010c. “The Importance of Being Er.” Proceedings of the IX Symposium Platonicum; Plato’s Politeia, vol. 1. International Plato Society: 144-150.

———. 2009. Review of The Sanskrit Epics’ Representation of Vedic Myths, by Danielle Feller. Asiatische Studien/ Études Asiatiques A 63, 2: 480-3.

———. n.d. Sacrificial Ontology and Human Destiny in the Mahābhārata. Unpublished manuscript.

——— & Joydeep Bagchee. n.d. “From Poetic Immortality to Salvation: Ruru and Orpheus in Indic and Greek Myth.” Forthcoming in History of Religions.

——— & Joydeep Bagchee. n.d. The Nay Science: A History of German Indology. Annas, Julia. 1981. An Introduction to Plato’s Republic. Oxford: Oxford University

Press. Austin, Christopher. 2009. “Janamejaya’s Last Question.” Journal of Indian

Philosophy 37: 597–625.——-. 2011. “Evaluating the Critical Edition of the Mahābhārata: Inferential

Mileage and the Apparatus Materials.” Journal of Vaishnava Studies 19,2 (Spring) 71-86.

Bagchee, Joydeep. 2011. “Inversion, Kṛṣṇafication, Brahmanization: The Explanatory Force of Some Extraordinary Figures of Speech.” Journal of Vaishnava Studies 19,2 (Spring): 127-141.

Biardeau, M. 2002. Le Mahābhārata: Un récit fondateur du brahmanisme et son interprétation. 2 vols. Paris: Seuil.

———. 1981. L’Hindouisme: Anthropologie d’une civilisation. Paris: Flammarion.———. 1976. “Études de mythologie hindou, IV. Bhakti et avatāra.” Bulletin

d’École Française d’Extrême Orient 63: 111-263.Black, Brian. Forthcoming. Forthcoming.“Transitions and Transmissions

in the Mahābhārata: Revisiting the Ugraśravas/ Śaunaka Frame Dialogue.” In Revisiting Transitions in Indian History. Ed. Ranabir Chakravarti & Kumkum Roy. Delhi: Oxford University Press.

Bloomfield, Maurice. 1896. “Contributions to the Interpretation of the Veda.” The American Journal of Philology 17,4: 399-437.

Brodbeck, Simon. 2009. The Mahābhārata Patriline: Gender, Culture and the Royal Hereditary. Farnham, Surrey: Ashgate.

——-. 2011. “Analytic and Synthetic Approaches in the Light of the Poona Edition of the Mahābhārata and Harivamśa.” Journal of Vaishnava Studies 19,2 (Spring) 223-250.


Bronkhorst, J. 2011. Buddhism in the Shadow of Brahmanism. Leiden: Brill Publishing.

Couture, André. 2001. “From Viṣṇu’s Deeds to Viṣṇu’s Play, Or Observations on the Word Avatāra as a Designation for the Manifestations of Viṣṇu.” Journal of Indian Philosophy 29: 313–326.

Cozad, Laurie. 2004. Sacred Snakes: Orthodox Images of Indian Snake Worship. Aurora, CO: The Davies Group.

Duncker, Max. 1883. History of Greece: From the Earliest Times to the End of the Persian War, vol. 1. London: R. Bentley & Son.

Elias, Michael. 1995. “Neck-Riddles in Mimetic Theory.” Contagion: Journal of Violence, Mimesis, and Culture 2: 189-202.

Feller, Danielle. 2004. The Sanskrit Epics’ Representation of Vedic Myths. Delhi: Motilal Banarsidass.

Fitzgerald, James. 1983. “The Great Epic of India as Religious Rhetoric: A Fresh Look at the Mahābhārata.” Journal of the American Academy of Religion, 51,4: 611-630.

———. 2010. “No Contest between Memory and Invention: The Invention of the Pāṇḍava Heroes of the Mahābhārata.” In Epic and History. Ed. David Konstan & Kurt A. Raaflaub. London: Blackwell.

Gadamer, Hans-Georg. 2004. Truth and Method. London: Continuum.Goldman, R. P. 1977. Gods, Priests, and Warriors: The Bhṛgus of the Mahābhārata.

New York: Columbia University Press. Goldstücker, Theodore. 1879. Literary Remains of the Late Professor Goldstücker,

vol. 2. London: W. H. Allen.Grünendahl, Reinhold. 2002. “On the Frame Structure and ‘Sacrifice Concept’

in the Nārāyaṇīya and Tīrthayātrā Sections of the Mahābhārata, and the Craft of Citation.” Zeitschrift der Deutschen Morgenländischen Gesellschaft 152, 1: 309-340.

Heidegger, Martin. 2005. “Phänomenologische Interpretationen zu Aristoteles. Anzeige der hermeneutischen Situation.” In Martin Heidegger, Phänomenologische Interpretationen ausgewählter Abhandlungen des Aristoteles zur Ontologie und Logik. Frankfurt am Main: Vittorio Klostermann.

Hewitt, James. 1890. “Notes on the Early History of Northern India. Part IV. Essay on the Pre-Vedic History of India and the Identity of the Early Mythologies of Europe and Asia, Founded on a Study of the Brāhmaṇas and of Sacrificial Observances.” The Journal of the Royal Asiatic Society of Great Britain and Ireland (Apr.): 319-481.


Hiltebeitel, Alf. 2011a. Reading the Fifth Veda: Studies on the Mahābhārata. Essays by Alf Hiltebeitel, vol. 1. Ed. Vishwa Adluri & Joydeep Bagchee. Leiden: Brill.

———. 2011b. When the Goddess was a Woman: Mahābhārata Ethnographies. Essays by Alf Hiltebeitel, vol. 2. Ed. Vishwa Adluri & Joydeep Bagchee. Leiden: Brill.

———. 2011c. “On Sukthankar’s “S” and Some Shortsighted Assessments and Uses of the Pune Critical Edition.” Journal of Vaishnava Studies 19,2 (Spring): 87-126.

———. 2007. “Krishna in the Mahābhārata: The Death of Karṇa.” In Krishna: A Source–book. Ed. Edwin F. Bryant. 23–76. New York: Oxford University Press.

———. 2006. “The Nārāyaṇīya and the Early Reading Communities of the Mahābhārata.” In Between the Empires: Society in India 300 B.C.E. to 400 C.E. Ed. Patrick Olivelle. 227–255. New York: Oxford University Press.

———. 2005. “Weighting Orality and Writing in the Sanskrit Epics.” In Epics, Khilas, and Purāṇas: Continuities and Ruptures. Proceedings of the 3rd Dubrovnik International Conference on the Sanskrit Epics and Purāṇas, Sept. 2002. Ed. Petteri Koskikallio. 81–111. Zagreb: Croatian Academy of Sciences and Arts.

———. 2004. “More Rethinking the Mahābhārata: Toward a Politics of Bhakti.” Indo-Iranian Journal 47: 203-227.

———. 2001. Rethinking the Mahābhārata: A Reader’s Guide to the Education of the Dharma King. Chicago: University of Chicago Press.

———. 1998. “Conventions of the Naimiṣa Forest.” Journal of Indian Philosophy 26: 161-171.

“Reconsidering Bhṛguization.” In Composing A Tradition: Concepts, Techniques and Relationships. Proceedings of the 1st Dubrovnik International Conference on the Sanskrit Epics and Purāṇas, Aug. 1997. Ed. Mark Brockington & Peter Schreiner. 155-168. Zagreb: Croatian Academy of Sciences and Arts.

———. 1977. “Nahuṣa in the Skies: A Human King of Heaven.” History of Religions 16,4: 329-350.

Hopkins, E. W. 1916. “Mahābhārata.” In Encyclopedia of Religion and Ethics, vol. 8. Ed. James Hastings. New York: Charles Scribner’s Sons.

———. 1908. “The Mediatorial Office of the Vedic Fire-God.” Harvard Theological Review 1,4: 507-512.

———. 1901. The Great Epic of India: Its Character and Origin. New York: Charles Scribner’s Sons.

Klostermaier, Klaus. 2007. A Survey of Hinduism. Albany: SUNY Press.


Kosambi, D. D. 1964. “The Autochthonous Element in the Mahābhārata.” Journal of the American Oriental Society 84,1: 31-44.

Kramrisch, Stella. 1975. “The Indian Great Goddess.” History of Religions 14,4: 235-265.

Mahadevan, T. P. 2011. “Three Rails of the Mahābhārata Textual Tradition.” Journal of Vaishnava Studies 9,2 (Spring): 23-69.

Malbon, Elizabeth Struthers. 1983. “Structuralism, Hermeneutics, and Con–textual Meaning.” Journal of the American Academy of Religion 51,2: 207-230.

Mehta, Mahesh. 1973. “The Problem of the Double Introduction to the Mahābhārata.” Journal of the American Oriental Society 93,4: 547-550.

Meyer, Hansjörg. 1967. Das Halslösungsrätsel. Ph.D. Dissertation, University of Würzburg.

Minkowski, C. Z. 2010. “Nilakantha’s Mahabharata.” Seminar, no. 608, special issue titled The Enduring Epic: A Symposium on Some Concerns Raised in the Mahābhārata: 32-38.

———. 1989. “Janamejaya’s Sattra and Ritual Structure.” Journal of the American Oriental Society 9,3: 401- 420.

Oberlies, Thomas. 2008. “(Un)ordnung im Mahābhārata: Rahmenerzählungen, Gesprächsebenen und Inhaltsangaben.” Studien zur Indologie und Iranistik 25: 73-102.

———. 2005. “12 Monate oder 12 Jahre der Verbannung? Zur Komposition von Mbh 3,90-140.” Bulletin d’Études Indiennes 22-23: 273-285.

———. 1998. “Die Ratschläge des Sehers Nārada: Ritual an und unter der Oberfläche des Mahābhārata.” New Methods in the Research of Epic. Ed. Hildegard Tristram. Tübingen: Gunter Narr Verlag.

Oldenberg, Hermann. 1922. Das Mahābhārata: Sein Inhalt, Seine Entstehung, seine Form. Göttingen: Vandenhoeck & Ruprecht.

Olson, Carl. 2007. The Many Colors of Hinduism: A Thematic-Historical Introduction. New Brunswick, NJ: Rutgers University Press.

Pargiter, F. E. 1913. The Purāṇa Text of the Dynastics of the Kali Age. Oxford: Oxford University Press.

Reich, Tamar C. 2008. “Sacrificial Violence and Textual Battles: Inner Textual Interpretation in the Sanskrit Mahābhārata.” History of Religions 41,2: 142-169.

Sarin, San. 2004. “Mahābhārata: The Numbers 18 and 108 through Akṣauhiṇī.” Indologica Taurinensia 30: 237-252.

Sarmah, Thaneswar. 1991. The Bhāradvājas in Ancient India. Delhi: Motilal Banarsidass.


Schlingloff, Dieter. 1969. “The Oldest Extant Parvan-List of the Mahābhārata.” Journal of the American Oriental Society 89,2: 334-338.

Schmitt, Arbogast. 2003. Die Moderne und Platon. Stuttgart: J. B. Metzler. Slaveva-Griffin, Svetla. 2003. “Of Gods, Philosophers, and Charioteers: Content

and Form in Parmenides’ Proem and Plato’s Phaedrus.” Transactions of the American Philological Association (1974-) 133,2: 227-253.

Srinivasan, Doris. 1973. “The Myth of the Panis in the Rig Veda.” Journal of the American Oriental Society 93,1: 44-57.

Staal, J. F. 1979. “The Concept of Scripture in the Indian Tradition.” In Sikh Studies: Comparative Perspectives on a Changing Tradition. Ed. M. Juergensmeyer & N. Gerald Barrier. 121-24. Berkeley: Berkeley Religious Studies Series.

Sukthankar, V. S., ed. 1933. “Prolegomena.” The Ādiparvan for the First Time Critically Edited. Pune: Bhandarkar Oriental Research Institute.

———. 1930. “Epic Studies II: Further Text-Critical Notes.” Annals of the Bhandarkar Oriental Institute 11,2: 165-191.

———. 1930. “Epic Studies III: Dr. Ruben on the Critical Edition of the Mahābhārata.” Annals of the Bhandarkar Oriental Institute 11,3: 259-283.

———. 1935-36. “Epic Studies IV: More Text-Critical Notes.” Annals of the Bhandarkar Oriental Institute 16,1: 90-113.

———. 1936-7. “Epic Studies VI: The Bhṛgus and the Bhārata: A Text-Historical Study.” Annals of the Bhandarkar Oriental Institute 18,1: 1-76.

———. 1938-39. “Epic Studies VII: The Oldest Extant Manuscript of the Ādiparvan.” Annals of the Bhandarkar Oriental Institute 19: 201-262.

———. 1957. On the Meaning of the Mahābhārata. Bombay: Asiatic Society. Tsuchida, Ryutaro. 2009. “Some Reflections on the Chronological Problems of

the Mahābhārata.” Studies in Indian Philosophy and Buddhism 16: 1-24. ———. 2008. “Considerations on the Narrative Structure of the Mahābhārata.”

Studies in Indian Philosophy and Buddhism 15: 1-26.———. 2006. “The Formation of the Anukramaṇi- and the Parvasaṁgrahaparvan

of the Mahābhārata.” Studies in Indian Philosophy and Buddhism 13. Vassilkov, Yaroslav. 1999. “Kālavāda (the doctrine of Cyclical Time) in the

Mahābhārata and the concept of Heroic Didactics.” In Composing A Tradition: Concepts, Techniques and Relationships. Proceedings of the 1st Dubrovnik International Conference on the Sanskrit Epics and Purāṇas, Aug. 1997. Ed. Mark Brockington & Peter Schreiner. 17-33. Zagreb: Croatian Academy of Sciences and Arts.

Vogel, J. Ph. 1926. The Indian Serpent-Lore. London: Probsthain & Co.


West, M. L. 2007. Indo- European Poetry and Myth. Oxford: Oxford University Press.

Winternitz, Moriz. 1909. Geschichte der indischen Literatur, vol. 1. Leipzig: Amelang.

Woods Julian. 2001. Destiny and Human Initiative in the Mahābhārata. Albany: SUNY Press.

Woolsey, John Martin. 1993. Symbolic Mythology and Translation of a Lost and Forgotten Language. Whitefish, MT: Kessinger Publishing.

5.-Vishwa1

Documents

Transcript of 5.-Vishwa1