Jump to ContentJump to Main Navigation
DharmaIts Early History in Law, Religion, and Narrative$

Alf Hiltebeitel

Print publication date: 2011

Print ISBN-13: 9780195394238

Published to Oxford Scholarship Online: September 2011

DOI: 10.1093/acprof:oso/9780195394238.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 17 September 2019

Women's Dharma

Women's Dharma

Śāstric Norms and Epic Narratives

(p.337) 8 Women's Dharma

Alf Hiltebeitel

Oxford University Press

Abstract and Keywords

This chapter explores the topics of strīdharma or womens's dharma and marriage law as the Mahābhārata portrays them through the three generations of dynastic instability that precede the generation of the epic's main heroes. This skein shows how the women marrying into the central dynastic line, beginning with the river goddess Gaṅgā, make women's dharma central to intersecting dimensions of time in which dharmic norms come under repeated scrutiny. Gaṅgā, who takes an interest in the Bhārata dynasty's “history” (itihāsa) as part of a divine plan, leaves her husband; he then marries Satyavatī, who brings her prermarital son Vyāsa, the Mahābhārata's “author”—and thus something like authorial time—into the line's genealogy. The next generation is then traced through the stories of how Gaṅgā's son Bhīṣma, ineligible to rule and sworn to celibacy, abducts the three sisters Ambā, Ambikā, and Ambālikā to marry them to Satyavatī's one remaining son, and how, once Ambikā and Ambālikā become widows, Vyāsa sires sons with them. The role of the chief queen (mahiṣī) in the Vedic horse sacrifice (Aśvamedha) is drawn into interpreting Vyāsa's unions with Ambikā and Ambālikā. In the third generation, Vyāsa's two flawed sons then marry: Pāṇḍu with Kuntī and Mādrī; Dhṛtarāṣṭra with Gāndhārī; and these three ingenious women then become mothers of the Pāṇḍavas and Kauravas.

Keywords:   strīdharma, marriage law, Gaṅgā, Bhārata dynasty, itihāsa, divine plan, Satyavatī, Vyāsa, Bhīṣma, Ambā, Ambikā, Ambālikā, Kuntī, Mādrī, Gāndhārī, mahiṣī, Āśvamedha, Pāṇḍu, Dhṛtarāṣṭra, births of Pāṇḍavas and Kauravas

As mentioned in the last chapter, toward the end of its discussion of the Yuga Purāṇa, the Mahābhārata's main story tells of a crisis of six generations, of which the two sets of vying cousins, the five Pāṇḍavas and the hundred Kauravas, make up the fourth. As I indicated, if one wants to speak of “invention,” one must do so with respect to those six generations in full, beginning with the intervention into the Mahābhārata's main dynastic line by the Goddess Gaṅgā. I also mentioned that the epic poets extract plausible historical data from the Vedic canon to trace their Bhārata itihāsa or “history of the Bhāratas” through those six generations into the “future” of the Kali yuga (see chapter 7 § A.4). As we shall now see, it is specifically through Gaṅgā that this thread of itihāsa begins to be woven into the Mahābhārata's dynastic and generational time.

This book will introduce the topic of women and dharma against that background. The obvious point is that generations require women, and the Mahābhārata's women are great, spirited, and fantastic.1 We will thus be exploring the topic of dharma over time (p.338) through the three generations of dynastic instability and crisis that precede the Pāṇḍavas’ marriage to their own fourth generation's equally if not more remarkable woman, Draupadī. As in this fourth generation, when Draupadī becomes the stake who says dharma itself is at stake at the pivotal epic dice match, the Mahābhārata women who precede her in the line can be remarkably active and loquacious about dharma, and especially when it is imperiled.

We keep ourselves largely to the Mahābhārata in this chapter, since the Rāmāyaṇa does not offer three such dynamic generations before its main story, does not place itself explicitly in a transition between yugas, does not weave itself into itihāsa or history, and—most important—does not envision dharma changing during the lifetime of Rāma, or Rāma ever changing his view of dharma. We shall come to such matters of dharma over biographical time in chapter 9.

A. Strῑdharma

Brahmanical norms for women are set forth broadly through the concept of strīdharma, “Law(s) for women” or “women's dharma.” While there is no end of śāstra-type instruction on the “dharma of women” (strī means especially but not only “wives”), and we looked briefly at some of Manu's more egregious ones in chapter 5 § E, it is too easy to suppose that it exhausts the subject.

Nonetheless, Manu encapsulates strīdharma in a famous adage that is paralleled in many of our texts:

Even in their own homes, a female—whether she is a child, a young woman, or an old lady—should never carry out any task independently (na svātantryeṇa).2 As a child, she must remain under her father's control; as a young woman, under her husband's; and when her husband is dead, under her sons’. She must never seek to live independently (na bhajet strī svatantratām). She must never want to separate herself from her father, husband, or sons; for by separating herself from them, a woman brings disgrace on both families. (M 5.147–49)

Among the dharmasūtras, the “broader” Āpastamba has no such adage.3 As Jamison indicates, Āpastamba treats women largely within “the older śrauta ritual model” of the “household pair,” in which the wife is recognized as one (p.339) of the “two masters” (svāminau; Ā 2.4.13), along with her husband (2006, 192). As was mentioned in chapter 5 (§ C at n. 60), Āpastamba also, like the Mahābhārata, allows that one may learn dharma from women and Śūdras. Baudhāyana and Vasiṣṭha, however, “quote” similar (unattributed) verses to Manu's second of these three, about a woman's passage through three phases of life—in Baudhāyana's case to make the different point that she lacks strength and cannot inherit property (B 2.3.43–45; cf. V 5.1–3). Gautama, however, keeping matters to prose, seems to have made the strictures even tighter than Manu:

A wife (strī) cannot act independently in matters related to the Law (asvatantrā dharme). She should never go against her husband and keep her speech, eyes, and actions under strict control. (G 18.1–3)

As recent studies have brought out, these texts do indirectly attribute mental agency to women (Jamison 2006), and explicitly ascribe economic and sexual agency to them (Olivelle 2005b, 248–60); but they leave us with what Jamison calls “something of a paradox—that the more woman is textually endowed with agency the more her capacity for independence is denied” and the more she needs to be kept under “guard” (2006, 201). Jamison thinks that one factor behind the increase of these strictures, and in Manu their misogynist expression, could be the growing prominence of a “new female type, the independent and religiously unorthodox woman,” the Buddhist and other heterodox nun (206).

We may expect women's “non-independence” to have narrative outlets and subversions. Even Manu, as it contracts women's worlds, may leave the tiniest room for women's dharma to breathe. On “Law Concerning Husband and Wife,” the sixteenth ground for litigation, Manu begins, “For a husband and wife who stay on the path (vartmani) pointed out by the Law, I shall declare the eternal Laws for both when they are together and when they are apart” (9.1). And, all be it that it is problematic, “even a woman … can give testimony.”4 An epic heroine's stories could be condensed into, or expanded from, these pronouncements.5 Compared to the dharmasūtrakāras, in fact, Manu does create a mini-narrative in his three famous verses. Out of Gautama's strī, which means “wife” in context, but before that “woman,” and out of the quoted adage about (p.340) a woman's three phases of life, Manu gives us a full female life cycle not only with its wants, but within the “homes” of her successive families.

B. The Law of the Mother

If Āpastamba opens the possibility of learning dharma from women within the older śrauta model of the household pair, and the epics spin that possibility out, our questions are: Under what conditions does such spinning out occur? What do women characters really desire6 … to say about dharma that would also be dharma? Let me start with this last question and propose to explore such matters chiefly under the name of the mother and call this dharma the “Law of the Mother.”7 Although a wife may also be a mother, it seems to have been easier to legislate for woman as wife (thus strīdharma) than as mother (e.g., mātṛdharma, a term one will not find in any of our classical sources). The Law of the Mother would, to begin with, be a name for something less articulated—at least as dharmaśāstra. Similarly, by the end of this chapter, we will also have discussed something we could call the “Law of the Girl,” which—as the equally unattested kanyādharma—will have proved equally resistant to proper codification.

I first used the term “Law of the Mother” in a conference presentation about two Mahābhārata cults in Tamilnadu (2004c) to address the question of whose law it is when members of a clan that has temples for Draupadī say they worship their clan deity Periyantavar, who is none other than a reincarnation of Duryodhana, because Draupadī gave permission to their ancestors to do so, but “for only a day since his laws last only a day.” In brief, Duryodhana is Draupadī's former enemy who now holds power for a day over spirits (pēy) who are exorcised mainly from women who are disrupting clan (kulam) expectations that they be dutiful wives or daughters. While Duryodhana's male priests perform the exorcisms with a clay horse that Duryodhana himself possesses, the women dispossessed are carefully attended by older women: mothers and others. My paper asked, “Is it really Duryodhana's laws that operate or Draupadi's, who sets the limit of a day to the laws of Duryodhana? Duryodhana's festival is in many ways an inversion of hers. Could one say that at a deeper level, (p.341) Duryodhana's festival takes place under the Law of the Mother?” In the discussion that followed, Mark Pizzato of the University of North Carolina, Charlotte, found the term “suggestive as a Lacanian oxymoron”—which was gratifying, since I used it bearing in mind that the term “Law of the Father” encapsulates Jacques Lacan's view that “It is the name-of-the-father that we must recognize as the symbolic function which, from the dawn of history, has identified his person with the figure of the law” (Lacan 1977, 67; Grosz 1990, 71).8 Lacan develops this terminology as a way of reading Freud's explanation of the origins of patriarchy by the Oedipus myth.

Now the type of exorcisms done in the name of Duryodhana is not done for him alone, though his are the best coordinated and probably the longest sustained. Within the same valley north of Dharmapuri, the same type of rites are also performed, at least recently, by an intermarrying clan in the name of Periyantavar-Duryodhana's wife Periyantacci, and a variation has also recently been incorporated into the festival cycle at one innovative Draupadī temple. In other words, although these rites have for some time been performed mainly in the name of a clan father (Duryodhana), they can also be performed in the name of a clan mother, while beyond that, Draupadī gives permission as a mother whose festivals make it clear that her husbands’ law, reinstituted annually by the festival that reenthrones her eldest husband Dharma (Tamil Tarumar, i.e., Yudhiṣṭhira), is at some profound level also hers (Hiltebeitel forthcoming-h). Since the Draupadī cult is a distillate of goddess worship with folk and classical Mahābhāratas, we have an opening here to think back to the Mahābhārata itself as a discursive world about which we might ask: might the Law of the Mother be a term worth considering in connection not only with Draupadī but with other Mahābhārata women whose stories intersect with hers? For the moment, let us just note that the Mahābhārata is more open to such an approach than the Rāmāyaṇa, where Rāma's dharma lies mainly in upholding the truth of his father's word and thus standing resolutely for the Law of the Father. In the Mahābhārata, on the contrary, the fatherless Pāṇḍavas follow a Law of the Mother when Kuntī, their mother, tells them by an inadvertence, thinking they have returned with alms from begging when they are really bringing home Arjuna's newly won bride, to “share it all equally” (Mbh 1.182.2), with the result that all five marry Draupadī (see chapter 10). By the end of this chapter, and further in chapter 10, I will attempt to put Kuntī's inadvertence into a deeper dharma context.

(p.342) It is beyond my competence to interpret the Mahābhārata in Lacanian terms, much less revisionist ones. When I have presented this skein to audiences without having time to go into it, I just speak of a “maternal dharma,” and would expect that for some readers of this chapter that term may be preferable. But it is striking that certain Lacanian phrases seem to be quite illuminating of the Mahābhārata's unusual treatment of fathers and mothers. Let me work from a quote that would seem to give a little room for a Law of the Mother to serve as something more than a suggestive oxymoron:

In fact, the image of the ideal Father is a phantasy of the neurotic. Beyond the Mother, the real Other of demand, whose desire (that is, her desire) one wishes she would assuage, there stands out the image of a father who would close his eyes to desires. The true function of the Father, which is fundamentally to unite (and not to set in opposition) a desire and the Law, is even more marked than revealed by this.

The neurotic's wished-for Father is clearly the dead Father. But he is also a father who can perfectly master his desire—and the same can be said of the subject. (Lacan 1977, 321)

If we start with the rival Pāṇḍava and Kaurava cousins and go back no further than their human if not always “real”9 fathers and grandfathers, it would appear that these main heroes of the Mahābhārata grow up under a near-total collapse of the paternal function. Back a generation to the “grandfathers” (having that name whether as biological parents or just generationally), one finds two sonless and dead before their time from martial (Citraṅgada) or libidinal (Vicitravīrya) excesses, and two still living, both great spokesmen of dharma: Vyāsa, usually a celibate and apparently unmarried, whose desire for dharma mixed with desires for the sons he has sired seems to provide his explanation for composing the whole poem; and Bhīṣma, whose vow of lifelong celibacy to guarantee a second marriage for his own father makes him a nonbiological grandfather who has perfectly mastered his desire, whatever the credibility of that—or the consequences.10 From Vyāsa, then, they have two fathers and a lower-rank uncle. Oldest is the Kauravas’ blind father Dhṛtarāṣṭra who cannot really rule and who can never open his eyes to the raging desires of his sons and nephews or gain wisdom on his own desires. His junior, Pāṇḍu, (p.343) after a short rule, is clearly the wished-for dead father, who has in fact died fulfilling his desires during the heroes’ childhood. Once Pāṇḍu has died, the Pāṇḍavas and Kauravas as parallel paternal cousins should accept paternal authority from Dhṛtarāṣṭra, since he and Pāṇḍu are of the same rank, having Kṣatriya mothers and the same Brahmin father, Vyāsa. But his authority is weak, vacillating, and incurably biased so long as his sons are alive. Meanwhile, not of the same rank and thus an uncle rather than a father, is Vidura, incarnation of the god Dharma, who can speak for dharma profoundly but only ineffectively since as Vyāsa's son with a Śūdra woman he cannot rule and has no authority whatsoever. All this, in Lacan's terms, would provide an ideal situation for the Kauravas and Pāṇḍavas to make rival demands for their royal patrimony.

If our heroes and villains grow up in such a world where their real or at least human fathers and grandfathers comprise such an ideally dysfunctional composite of the Symbolic father in whose name they would incarnate the transcendent Law of the Father,11 we might not be surprised to find that it is a world where a Law of the Mother could take hold, and that it will eventually take five gods to sire sons in such a family to begin to restore a paternal Law. This would not be interesting if the name of the father was always “Beyond the Mother,”12 leaving the mother, in situations of default, to be “the agent of the symbolic father, who enforces the Law-giving and lineage-maintaining power of what might otherwise be an empty paternal function” (Custodi 2005, 123). There is some of that. But what the Mahābhārata seems to be wrestling with is that when women hold the cards in speaking on dharma, and men must learn and listen, situations may emerge where the Law and indeed the name of the Mother (and when we get to it, of the Girl) may be vital to the Law's realization and at least equally as primordial as the name or Law of the Father.13 This, at least, gives us (p.344) terms with which to consider the distaff side of this family through these three generations that are rounded off, from this angle, with the five Pāṇḍavas’ legally extraordinary marriage to Draupadī.

To begin the story of this family's investment in a primordial Law of the Mother, we must take ourselves back still one generation further to the Pāṇḍavas and Kauravas’ two great grandmothers Gaṅgā and Satyavatī. The best treatments of the skein (a term of choice14) as a whole are by Biardeau (2002, 1: 212–38) and Brodbeck (2009a, 150–77), but each—for very different reasons—tends to shortchange the women. Biardeau attends mainly to their symbolic roles rather than their words, which leaves their relation to dharma rather schematic, cosmological, and abstract (see above n. 1). And Brodbeck zeroes in on the patrilineal males for whom wives are more interesting than mothers, leaving him to admit that in treating female characters, his categories “cannot do them justice” (261).15 Meanwhile, with closer attention to the women's words and moves, claims have been made for two of the women—Satyavatī (by Ghosh 2000, 33, 42) and Kuntā (by Dhand 2004)—that they are pivotal for the whole epic, which they certainly are, though neither one more than the other. Dumézil (1979, 31–45, 66–71), Doniger (1995), Jamison (1996), and Heesterman (2001, 254–59) have also made important contributions on the legal ramifications of several of the unusual marriages in the series. A methodological point here: A. K. Ramanujan had a good impulse when he criticized me for overemphasizing divine–human connections at the expense of “the architectonic complexity of the human action of the epic” (1991, 434, n. 4). But the point is valid only so long as one traces that complexity only in its (p.345) own human terms. Our epic's human action also has cosmological complexity (see Hiltebeitel 2001b, 267), as does its men's and women's words.

With Gaṅgā and Satyavatī on through Kuntī and the other women of her generation, we find that women's dharma is woven (sometimes literally so) into an artful story of increasing crisis among the males in the Kuru dynasty, also known as the Lunar dynasty due to its descent from the Moon god. In such circumstances, the men are no less interested in and representative of dharma than the women, but what the women have to say about it becomes increasingly decisive. Our chief attention will be on what the three generations of women who marry into this lineage before Draupadī have to say about dharma, and the moves they make regarding it.

C. Mother Gaṅgā

It all starts with a near trifle in outer space.16 “There was once a king born in the Ikṣvāku lineage” (Mbh 1.91.1) named Mahābhiṣa who, after countless sacrifices and a truthful life, became a Rājarṣi or Royal Sage in heaven. The poets are about to introduce him to the luminous celestial Gaṅgā, her robe the Milky Way, and their metaphoric range is the night sky, where Ṛṣis, royal and otherwise, are stars (Mitchiner 1982), and there are mighty winds. The celestial Gaṅgā is associated with the Parivaha wind that diffuses its waters and carries them through the sky when “agitated,” affecting the visibility of the sun and the rising moon (Mbh 12.315.46–48). “Then at some time” (tataḥ kadācit)—note this cunning narrative convention, which occurs nine times in our skein and serves as a sort of start-up and then restart mechanism—while the Gods were doing homage to Brahmā in the Royal Ṛṣis’ company, as “Gaṅgā approached the Grandfather [Brahmā], her garment, radiant as the moon, was raised by the wind” (4). Now, as Gaṅgā's garment lifts,

The host of gods then lowered their faces. But the royal Ṛṣi Mahābhiṣa looked at the river fearlessly. Mahābhiṣa was disdained by lord Brahmā, who said, “Born among mortals, you shall again gain the worlds.” (1.91.5–6)

At this point, in a fairly widespread interpolation remarked on by Biardeau, Brahmā also says that Gaṅgā will join Mahābhiṣa in this earthly destiny. According to Biardeau, the two are jointly “condamnés pour une faute légère (p.346) dans le ciel” (2002; I: 219; cf. 213). But even with the interpolation, I do not see this implication. Brahmā addresses only Mahābhiṣa:

Having stayed long among humans, you will obtain the beautiful worlds, O low-minded one whose mind is seized (hṛtamanas) by Gaṅgā. She indeed will act disagreeably in the human world. When you get angry you will then be released from the curse. (1.911*, after Mbh 1.91.6)

It seems that Brahmā only says that once Mahābhiṣa takes birth as Śaṃtanu, Gaṅgā will act disagreeably to him; not that she is “condemned” to do so. In any case, the Pune Critical Edition does well to make these lines superfluous, for it is immediately apparent that Gaṅgā's descent is voluntary and amorous: “The river, best of streams, having seen the king fallen from his firmness (dhairyāc cyutam), went away musing about him in her heart” (91.8). Brahmā can be prudish, but he does not hold her accountable for the wind blowing up her skirt. In fact, he is more often prurient than prudish, and could be punishing Mahābhiṣa for ogling a woman as something Brahmā gets a reputation for doing himself. If we look back from a Purāṇic perspective, there is an emerging irony here, since in Purāṇic myth, Brahmā is often the lascivious one disdained or punished for his gaze, which he directs most famously, but not only, at the beautiful Pārvatī when she is marrying Śiva (see e.g., Dimmitt and van Buitenen 1978, 34–35, 171; Hiltebeitel 1999c, 68–76).

Mahābhiṣa is told he can choose his father, and picks King Devāpi (91.7). He gives no reason, but in doing so he shifts from the Ikṣvāku or Solar Dynasty to the Lunar Dynasty, perhaps to avoid rebirth in his own line which I believe could be a problem (cf. Brodbeck 2009a, 153–55). Meanwhile Gaṅgā, resuming her celestial path and still musing on Mahābhiṣa, comes across the eight Vasu gods smitten with dejection, their figures bedimmed (9), which, they tell her, results from their having been cursed by Vasiṣṭha, one of the Great Ṛṣis among the Seven Sages of the Big Dipper, for having come too close to him (10–13)—all of which seems to continue to have astronomical overtones.17 The Vasus say Vasiṣṭha cursed them to be born in a womb, and, since they are unwilling to “enter an inauspicious human-female womb,” they ask Gaṅgā to be their mother (14–15). When she agrees and says they can choose their father, they pick “Pratīpa's son, King Śaṃtanu, renowned as law-abiding (dhārmikaḥ)” (16). Since Gaṅgā knows that Mahābhiṣa has chosen this same Pratīpa to be his father, the Vasus’ choice—assuming, as she must, that Mahābhiṣa will be (p.347) reborn as Śaṃtanu—is much to her satisfaction: “Such is even my mind, sinless gods, as you say. I will do his pleasure; that is your desire” (17)—and also hers. Let us note that we are poised to see for the first but not the last time a woman's mind (matam) carry prevailing force in our three-generation skein. When the Vasus insist that Gaṅgā “must throw his [Mahābhiṣa-Śaṃtanu's] newborn sons into the water so that our restoration will not take a long time,” she agrees again, but with the proviso that Śaṃtanu will keep one son. Each Vasu then imparts an eighth of his vīrya (energy/manliness/sperm) into a collective deposit for “the son you and he desire” (19–20): who will be Bhīṣma. But, add the Vasus, this son “shall not reproduce his lineage among mortals. Thus your son will be sonless, despite his possessing vīrya” (20–21). With Gaṅgā's agreement (samaya) on this further point, the “delighted” Vasus pursue their course (22). Bhīṣma's sonlessness is thus stipulated even before his father's birth, not to mention his own, by a divine compact or agreement between the Vasus and his Mother. If Mahābhiṣa cannot reenter the Solar line, he also can make only a restricted contribution to the Lunar one.

The only usage of dharma, an augmented one, in this first adhyāya of our skein is the description of the future Śaṃtanu as dhārmika, “law-abiding.” But samaya is also a legal concept. It is used early by Āpastamba to define the very “sources of Law,”18 and Gautama says, “The offspring belongs to the man who fathers it, unless there has been a compact (samaya)” (G 18.9–10)—which, with a twist, could explain why Śaṃtanu loses his first seven sons.19 At this point at least, Gaṅgā seems to have brokered this deal to leave one son to her husband, as both he and she (at least according to the Vasus) desire. And with this, she now begins to enact dharma in both gestures and words. Appearing to Pratīpa out of her own waters, she sits on his right thigh and invites him to make love to her, telling him—and implying a point of dharma20—that “abandonment of a woman in love is prohibited by the good” (92.3–5). Pratīpa knows his dharma, but seeks other ways to satisfy it and her. First, he says, “I would not go out of desire to another's woman, nor, lovely one, to one not of my own caste. Know that to be my righteous (dharmya) vow” (6). Of course he wants to know more about her, so she reassures him that she is “not unapproachable in any way,” (p.348) and moreover a virgin (kanyā) (7). Given what one soon learns about other kanyās in this skein, virginity can be renewable, and if Pratāpa's mention of “another's woman” might remind us of anything, it would be that—at least in post-epic Hinduism—Gaṅgā is the second wife of Śiva.21 Pratīpa, however, continues to demur: “I abstain from this pleasure to which you press me lest my violation of dharma would destroy what I vowed” (8). But he can still suggest a way to solve the dharma-problem her appeal seems to present: he invites her to become his daughter-in-law because she chose to sit on his right thigh, suitable for children and daughters-in-law, and avoided his left where a wife or lover (kāminī) would sit (9–11). Agreeing to what she must have herself foreseen, and thereby virtually assuring this apparently equally shrewd old king a son they both desire, Gaṅgā says,

So be it, dharma-knower. May I unite with your son. So by devotion to you will I love the famous Bhārata lineage (kulam). Whoever are the kings of the earth, you (plural: your dynasty) are their refuge. I am unable to speak the qualities (guṇas) that are renowned of your lineage in even a hundred years; its strictness is peerless (tat sādhutvam anuttamam). (12c–13)

Now telling Pratīpa he must tell his son that she must never be questioned, Gaṅgā disappears (14–16).

Even though Pratīpa and his wife are old, he “burns tapas.” And “at a certain time” Śaṃtanu is born in terms that both recall his recent celestial identity and give what seems to be a double etymology for his new name: “Mahābhiṣa became the old couple's son. He was born the continuity (saṃtāna) of a peaceful man (śāntasya); therefore he was called Śaṃtanu” (92.18). Śaṃtanu's name thus introduces a “continuity” theme that we shall see unfold,22 and rather ironically, since this heavenly migrant from the Solar Dynasty's own continuity in this line, though he does not yet know it, will go no farther than Bhīṣma. Yet as “Bhīṣma Śāṃtanava,” Bhīṣma will keep his father's name as a patronymic while affecting his family's continuity in nearly every imaginable way other than by paternity. Indeed, this double etymology would seem to have been designed less for Śaṃtanu than for Bhīṣma, who will (p.349) be preoccupied not only with his line's continuity but, at peace himself on a bed of arrows, with the “pacification” (śānti) of Yudhiṣṭhira in his vast postwar sermon on what we might call the continuity of dharma. Just to keep these ironies before us, when we come, shortly, to see that his name Bhīṣma will mean “The Terrible” for his choice not to continue his dynastic paternal line by siring sons himself, we might take his name Bhīṣma Śāṃtanava to suggest “The Terrible ‘Continuator.’”

Śaṃtanu now becomes a young man, and Pratīpa describes the beautiful divine (divyā) woman who may approach him and the conditions under which she will stay with him, saying they are to be carried out “at my appointment (man-niyogāt)” (21–23). We must note our skein's first usage of niyoga, since, even though it does not mean what it will be all about by its end—that is, the proxy siring by a man who replaces a deceased or otherwise disabled husband—the terminology seems to have been adroitly woven into the skien throughout. But for now, having made Śaṃtanu his heir-apparent, Pratīpa leaves for the forest. Soon, while hunting “along the Siddha- and Caraṇa-frequented Gaṅgā” (92.25cd), Śaṃtanu saw:

a superb woman whose figure had an intensive glowing (jājvalyamānāṃ vapuṣā) that was like the splendor of a lotus, faultless everywhere, with nice teeth, adorned with divine ornaments, wearing a subtle cloth, alone, and radiant as the calyx of a lotus …. As if drinking her with his eyes, the king wasn't satisfied (pibann iva ca netrābhyāṃ nātṛpyata narādhipaḥ). (92.26–28)

Śaṃtanu resumes the gaze that got him into trouble as Mahābhiṣa—and indeed, to drink this woman with his eyes and not be satisfied could be a reminder that he last saw her as a river of stars. Moreover, “having seen the king of great radiance going about,” she approaches him “dallying” (vilāsinī) and with “fondness come with affection,” as her dissatisfaction matched his (29).23 So he asks her to marry him. Remembering now her samaya with the Vasus, she sets her requirements: the marriage will be “at his will,” but she will do what she wants, and “I am not to be stopped or harshly addressed; … if you speak obstructively or unkindly I will no doubt leave you” (30–36). For him, it boils down to this: she is not to be questioned lest she abandon him. For her it is certainly a kind of women's law, but not one that many could impose, with a (p.350) codicil for a wife's justifiable abandonment—something opposed by Manu, who says, as we have seen, “She must never want to separate herself from her father, husband, or sons” (5.149). As their joys unfold,

by attachment to pleasure, the king, seized [as Brahmā's interpolated curse predicted] by the qualities of this foremost woman, was not aware of the many years, seasons, and months that passed. (41)

In “not so long a time” the eight Vasus are born, and Gaṅgā throws the first seven into the water, saying “I fulfill your wish.” Śaṃtanu watches, saying nothing “from fear of abandonment (tyāgād bhītaḥ)” (43–45). Here we may have the seeds of his eventual longing, once he is abandoned, for another wife, for clearly Gaṅgā, through the first seven, puts her motherly samaya before any concern for him.

At last the eighth son is born and Śaṃtanu can stand it no longer. “Yearning for a son of his own,” he speaks out while she seems to be laughing and says, “Son-killer, Stop!” to which she replies, “I will not slay your son, but this stay is now exhausted according to the samaya we made” (46–48). Gaṅgā now reveals who she is: “I am Gaṅgā, daughter of Jahnu, frequented by the hosts (gaṇas) of Great Ṛṣis; I have dwelt with you for the sake of success in accomplishing a purpose in the work of the gods (devakāryārthasiddhyartham)” (49). Although Gaṅgā emerged from her “Siddha- and Caraṇa-frequented” earthly course to sit on Pratīpa's right thigh (25), she now speaks of the heavenly “hosts of Great Ṛṣis” who frequent her heavenly course. Her “success in accomplishing a purpose in the work of the gods” is a quite precise and early indication that the “work of the gods,” which we may call the Mahābhārata's divine plan, begins to unfold with a certain complexity. That is, it involves not only the gods and Ṛṣis but this celestial goddess,24 and it will have to take in more than one generation. And it could—without it being made explicit—be thought to coincide with the yugas. The Mahābhārata's divine plan is thus far more fluid and extensive than the Rāmāyaṇa's, which concerns (at least as Vālmīki describes it) mainly just male gods and Ṛṣis, and does little to trace a course of dharma over generational time. Outside its yuga references, the Rāmāyaṇa traces dharma over time mainly through only the one generation of Rāma and his brothers, though it hints at some laxity of dharma in the long life of Daśaratha,25 and describes a lengthy course of adharma over time through the long career of Rāvaṇa.

(p.351) More immediately, we now see that what Gaṅgā has accomplished under the heading of “work of the gods” has fulfilled a samaya with the Vasus brought on by a curse by one of the most Vedic of the Great Ṛṣis, Vasiṣṭha. In fact, although Gaṅgā has just spoken of her prenuptial samaya with Śaṃtanu, there was no mention of that or any other such legal term during their premarital understanding. She had shaped that understanding, as she tells him only now, to be an extension of her samaya with the Vasus: “But this samaya was made between me and the Vasus—that as soon as they were born, I would release them from human birth,” leaving it unclear how this affects the one exception just born other than to say that she has now freed all the Vasus from Vasiṣṭha's curse (53–54). Śaṃtanu would probably be too bowled over to catch his wife on this technicality, but what good would it do him anyway? Instead he asks to learn more about Vasiṣṭha and the Vasus, and how their contretemps affected the son he must think he and he alone has just rescued from oblivion (93.1–3). What Gaṅgā tells him is this: his son is the incarnation of the god Dyaus (Sky), who was cursed by Vasiṣṭha to take birth in a womb because, as the Vasus’ ringleader, he led them, at his wife's request,26 to abduct Vasiṣṭha's divine cow (26); and that, although Vasiṣṭha shortened the terms for the other Vasus, Dyaus was cursed to “dwell in the human world for a long time by his own karma (svakarmaṇā)” (36cd; cf. 42). This would imply that Dyaus's karma will carry over into this human life, where we might have occasions to wonder whether traces of Dyaus's misadventures—foolishness done for a wife, abduction of a cow and calf—might have been left on Bhīṣma. Gaṅgā then comes to the closing words of Vasiṣṭha's curse, some of which sounds good: “… He will be a soul of dharma, conversant with all the scriptures”27 (39ab); while some is bound to be unsettling to Śaṃtanu: “The high-minded one will not reproduce among humans …” (38cd)—which Gaṅgā repeats not from Vasiṣṭha28 but from her samaya with the Vasus (91.21–22). Even more troubling, “Devoted to the welfare and his father's pleasure, he will forsake the enjoyment of women” (39cd).

This we are hearing for the first time. One cannot avoid the impression that Gaṅgā is a little free with her sources. Having finished attributing things to Vasiṣṭha, she offers a brief self-exoneration for throwing the other boys into the river for the sake of their release (mokṣārtham) from the curse, and upon that, (p.352) “the Goddess disappeared right there” (43) taking the boy with her. For Gaṅgā to vanish (antar-adhīyata)—literally, “to put herself within”—is to return to her own element, whether it be water or space (ākāśa), since she is of course the Ākāśa-Gaṅgā. In going with her, Bhīṣma's disappearance is almost like the drowning of his brothers. But Śaṃtanu knows Gaṅgā has taken him away with the promise of a long life ahead of him. Śaṃtanu goes “back to his capital afflicted with grief” (44). Having finally spoken out to keep his eighth son even though he knew it would mean losing his wife, he has suddenly lost them both.

It is now, with these events behind him, that the narrator Vaiśaṃpāyana says he will speak of Śaṃtanu's endless guṇas (93.45d) and “the splendid itihāsa called Mahābhārata” (93.46cd). The Mahābhārata's “history” begins with Gaṅgā's departure, yet also with her ongoing blessing: for thanks to her “devotion” to Śaṃtanu's father Pratīpa, she has promised that she will “love the famous Bhārata lineage” whose guṇas she is unable to recount in “even a hundred years” (92.12c–13). In effect, from a heavenly story moved down to earth, the Mahābhārata will stay largely on earth. After Śaṃtanu's “lost time” with Gaṅgā, time gets condensed into charted time along the epic's flow, beginning soon with the return of Bhīṣma. The verse on “itihāsa” ends an adhyāya. In effect, “history” begins directly after this in the next adhyāya with the story of Śaṃtanu's second marriage to Satyavatī, who is already the mother of Vyāsa, the author.29 How better to begin an “invention” of “history” than by the withdrawal of the celestial Gaṅgā, whose very intervention has resolved a crisis in the genealogy that will eventually bring forth—indeed, make possible the births of—the Kauravas and the Pāṇḍavas?30

During his wifeless rule with his son swept away, Śaṃtanu makes what he can of dharma, but with a sense of mounting insecurity. A “soul of dharma” (dharmātmā; 94.1c), he regulated the three puruṣārthas in favor of dharma (3–4); he performed his Kṣatriya svadharma like none other (5–6b); he supported the ordered relation of the four classes with Brahmins above and with each class serving the superior ones with increasing devotion down the scale (8–9; 14ab); he ruled the earth with knowledge of dharma and attained prosperity through gifting, dharma, tapas, and truth (10–11; 17); and he kept dharma and the brahman foremost in his kingdom and saw to it that no breathing creature was slain (p.353) by adharma. Moreover, “rites then began (ārabhyanta tadā kriyāḥ) for the sake of sacrifice to Gods, Ṛṣis, and Fathers” (13–15)—as if they had been interrupted. Indeed, this seeming resumption in fulfilling the “three debts” introduces a note of unease. What is this king's lineage? No doubt his rites would be for Fathers in the Lunar line he has shifted into, but what about this son who, even if he comes back, will not be able to continue them by having a son himself? All this time, for thirty-six years, Śaṃtanu “did not obtain pleasure with women”;31 still ruling, he then became a forest rover (18).

Vaiśaṃpāyana now brings the flow of time back to the river: “At a certain time (kadācit),” hunting while “following the river Gaṅgā, Śaṃtanu saw that the Bhāgīrathī had little water” (94.21). This use of the river's name Bhāgīrathī (omitted from van Buitenen and Ganguli's translations) evokes a connection between Śaṃtanu and the Solar-line king Bhagīratha, who first brought the heavenly Gaṅgā down to earth (Mbh 3.107–8; Rām 1.41–43.6). Having now also brought Gaṅgā down to earth once to marry him, Śaṃtanu will meet Bhīṣma arresting her descent. Wondering why “this best of streams no longer flows swiftly, he saw the occasion (nimittam)”: a stunning youth using a “divine weapon” to block the entire Gaṅgā with heavenly arrows (Mbh 94.22–24)!32 Śaṃtanu does not yet remember the son he had seen only at birth, who has now learned how to bewilder his father like a God or Ṛṣi by disappearing or “putting himself within” (antaradhīyata) (26–27). Suspecting this was his son, Śaṃtanu invokes Gaṅgā to appear, as she does: better dressed this time (“adorned with ornaments, attired in a dustless garment”) and holding the boy in her right hand (28–31). She tells Śaṃtanu who she is and what their son has accomplished in his time with her: he has studied the Vedas and their limbs from Vasiṣṭha; the same śāstras as Uśanas and the son of Aṅgiras (i.e., Bṛhaspati); all weapons known to Rāma Jāmadagnya; and he is “skilled in rājadharma and artha” (32–36). Though the poets do not overstate the matter, Bhīṣma has clearly been brought up by the celestial Gaṅgā and educated by Gods and celestial Ṛṣis. She has taken him up to the stars, where he could have learned Veda from a pacified Vasiṣṭha, last seen on the side of Mount Meru cursing Mahābhiṣa and Dyaus to become our father–son pair; and other celestial stalwarts have taught him the texts and topics (p.354) just mentioned, including rājadharma—about which he will expound in his first postwar anthology on that topic, where he will be more punctilious in citing his sources than his mother.33 Clearly the “work of the gods” that we have heard about first from Gaṅgā will have this carry-over on past the end of the Mahābhārata war. More immediately, though, this training has prepared her son to be the heir-apparent, which Śaṃtanu now makes him (38ef), even though he has been given the worry that he will be without issue. Meanwhile, for once there is no mention of Gaṅgā disappearing, leaving an impression that she has returned for now to being mainly a river. It is the last one sees her in this skein.

D. Mother Kālī Satyavatī

Four years pass with father and son enjoying each other's company until, “at a certain time, going into a forest along the Yamunā River, the king of the earth scented an indescribably lofty smell” (Mbh 1.94.41). Same restart mechanism, different river. Śaṃtanu is about to meet a woman introduced earlier in an account focused on her becoming the mother of Vyāsa. Outside the capital of King Vasu Uparicara of Cedi, Mount Kolāhala made love to the Śuktimatī River. King Vasu released the river with a kick, but she had become pregnant with twins, whom she gave to the king in thanks for freeing her. Vasu made the boy his marshal and married the girl, Girikā. One day Girikā lovingly told Vasu she could conceive and did the ritual bath to get a son. But this plan was strangely interrupted by Vasu's Fathers who, although they were “pleased”—presumably at the prospect of a son—told him, “Kill deer!”34 “Unable to transgress the Fathers’ appointment (niyoga), he went hunting, lovingly musing on Girikā” (38). In this reverie he ejaculated, and to keep his wife's readiness from going to waste, caught the sperm on a leaf, spoke a mantra over it, and, “aware of the subtleties of Law and Profit,” gave it to a hawk to speed it to her. Another hawk who thought it was meat attacked the carrier and the sperm fell into the Yamunā, where an Apsaras named Adrikā, cursed by Brahmā to become a fish, swallowed it. Fishermen (matsyajīvinaḥ) then caught the fish, pulled out a pair of twins from it, and told the king. Vasu kept the boy, who became King Matsya, but since the girl smelled of fish he gave her to the fisherman (dāśāya), saying, (p.355) “Let this one be yours!” Named Satyavatī, she had every virtue (guṇa), but “from dwelling amid fish-killers (matsyaghāti-), she bore for some time a fishy smell.” Out of obedience35 to her father, she plied a boat on the water (53–55), and there one day she caught the eye of the Ṛṣi Parāśara, grandson of Vasiṣṭha, who seduced her—by arranging for a fog to keep the Ṛṣis on the banks from seeing them; by promising to restore her virginity so she could return to her father; and by giving her the boon of smelling good whereby her new scent (gandha) won her the new names Gandhavatī, “the Fragrant,” and Yojanagandha, “She whom you could smell a league away” (67). She was “thrilled” with the boon, and all in all she seems quite the capable negotiator (Ghosh 2000, 35–37). That same day after Parāśara continued on his way, Vyāsa was born on an island only to depart that very day himself, his mind set on austerities, with the words, “Remembered, I will appear when things are to be done” (70cd; 1.57.31–71). Here Julia Kristeva gives us tools with which to register a useful contrast: If Gaṅgā is the “sublime mother” whose son Bhīṣma will not be able to separate from her or relate sexually to other women,36 Satyavatī is the “abject mother” whose first son separates from her instantly to practice tapas, with its overtones of gynophobia.37

With this story kept for now as a secret of her past, Satyavatī's “lofty smell” enables Śaṃtanu to trace her, and he learns from her, “I am a fisher-maid; for the sake of dharma I ply a ferry at the appointment (niyogāt) of my father, the high-souled king of fishermen” (1.94.44). But what is plying a ferry for the sake of dharma? Biardeau rightly points us ahead to descriptions of Draupadī as the boat that saves the Pāṇḍavas from the dice match by her question about dharma.38 But this ferryboat is also legal: her “father who was adjoined to dharma had a ferry” (99.6), and, it seems, appointed her its captain. And what then is a king of fisherfolk? Let us note that this adoptive father speaks in a legal manner when he uses terms like niyoga and (p.356) now dharmapatnī, “legal wife,” and samaya. Wishing to marry Satyavatā, Śaṃtanu learns that the fisher king has “a certain desire”:

If you seek her from me as your dharmapatnī, sinless one, then make a samaya with me by truth, as you are of truthful speech. I will give this daughter to you by the samaya, king, for there will surely never be another suitor like you to me. (94.47–49)

His requirement is that her son be king (51). Burning with desire, Śaṃtanu says he must think about giving him this “boon” (varam; 50, 52, 56), which Bhīṣma will later call a “bride price” (śulka; 97.14).

Bhīṣma soon sees his father morose and asks why. Śaṃtanu does “not seek vainly to marry another wife,” but still, he says, “I wish for non-destruction of continuity” (saṃtānasya–avināśāya; 94.59a–d)—reminding us, with this second mention of saṃtāna, “continuity” or “succession,” that the first usage built the term into Śaṃtanu's own name (92.18). Quoting an adage: “Those conversant about dharma say, ‘Having one son is to be sonless’” (94.59ef), he fears he will be left sonless if Bhīṣma dies fighting, as is his wont. But we know he has other good reasons to worry. What is he to make of a son who will be “devoted to the welfare and pleasure of his father” and “forsake enjoyment of women”? “Pondering” his father's demurral, Bhīṣma learns about the fishermaid from an “old councilor” of his father's and goes with some “old Kṣatriyas” to her father. The meeting “in an assembly of kings” (rājasaṃsadi, 94.68) credits the fisher king's royalty and gives further legal force to the proceedings.39 The fisher king establishes that Satyavatī, no mere ferryboat pick-up girl,40 is “the offspring of an Ārya who is your equal in qualities, from whose sperm (śukrāt) she appeared”; moreover, “Your father has been praised by him to me frequently, lad: ‘He among all kings is worthy of marrying Satyavatī’” (71–72). Bhīṣma would certainly gather that Satyavatī has a royal father who has delegated this man to arrange this very match41—and more, that the fisherman has somehow learned that Vasu was behind her birth from a fish, and has remained in touch with that king of Cedi, a kingdom located along the southern bank of the Yamunā42 where Satyavatī washed up with her original fishy smell. If we go back to her birth story, we can hardly be prepared for these intimations. Nothing was said about (p.357) one of the fishermen being a king, and Vasu showed no interest in the girl's marriage prospects, much less her origins from him. Jayatri Ghosh wishes to make the tale more plausible: “evidently, King Uparicara was attracted to a fisherwoman, had children by her, and to spare Queen Girikā's feelings, the event was transformed into a fantasy” (2000, 34). But clearly, neither Satyavatī nor her adoptive father needed to produce credentials at that time. And now these go unquestioned by Bhīṣma who, as his father's matchmaker, meets the fisherman's two demands. He renounces kingship to guarantee Satyavatī a royal son (79); and to see that he will sire no other rival, he vows celibacy even as he says he will obtain imperishable worlds in heaven (divi).43 All this is said to the celestial approbation “of the Apsaras, Gods, and the hosts of Ṛṣis (ṛṣigaṇas),” and earns him the name Bhīṣma, “the Terrible” (86–90).

Speaking on his father's behalf, Bhīṣma now says, “Ascend the chariot, mother. Let us go to our homes” (94.91). Under conditions brought about by the samaya his own mother, Gaṅgā, made with the Vasus, Bhīṣma now makes the samaya his father left pending in the name of this new mother, Satyavatī. Śaṃtanu thus enters into both marriages by restrictive samayas made on behalf of his wives as mothers-to-be; samayas that in each case introduce a supervening Law of the Mother that must rise to the occasion of a lineage whose “continuity” has been disrupted in the person of these two men cursed into being born into it: Śaṃtanu himself, cursed by Brahmā for gazing fearlessly (in his former life as Mahābhiṣa) up the skirt of mother Gaṅgā; and Bhīṣma Śāṃtanava, cursed (in his former life as the Vasu Dyaus) by Vasiṣṭha for stealing his cow. What can be said of this doubled Law of the Mother? Clearly the two women speak in some fashion together for something primordial that is deeper than kingship. Their two rivers embrace the “Mesopotamia” (doab) of “the north of Madhyadeśa,” that exemplary land of dharma.44 They speak for a primordial dharma sanctioned by the gods and celestial Ṛṣis. One, a goddess, descends from a celestial river of light; the other, born from a fish, comes from the terrestrial waters of the always darker Yamunā and is “dark” herself like her instant-Ṛṣi son, the Dark Island-born Vyāsa (Kṛṣṇa Dvaipāyana Vyāsa). One is called “child-killer,” but where she kills, it is with waters of liberation; the other is born bearing the smell of the killing of fish, suggesting something to do with the Law of the Fish (matsya nyāya) and the ongoing life–death struggle of transmigration (saṃsāra), as well as a boat across its waters.45 With Gaṅgā, the first (p.358) thing this Law of the Mother ameliorated was the punitive character of two male curses. Gaṅgā was more charmed than bothered by Mahābhiṣa's impertinence, and, given a situation where she could do no more “in person” than “love the Kuru dynasty” through two errant generations, perhaps she made up for the harshness of her first seven maternities by loving Śaṃtanu and by raising their one surviving son among Ṛṣis to be the line's highest authority on dharma. Now Satyavatā, definitely more charmed than bothered by the attentions of the Ṛṣi Parāśara, continues to speak for dharma solutions to the dynasty's disruptions (cf. Biardeau 2002, 1: 218).

Satyavatī's marriage to Śaṃtanu is over in four verses (1.95.1–4), barely enough to mention that they have two sons, and to bring closure with, “the wise king Śaṃtanu succumbed to the Law of Time (kāladharmam).”46 The next verse sets the tone and terms for much that follows. It says that when Śaṃtanu went to heaven and the oldest of the two princes, Citrāṅgada (and note: we have nothing to tell us whether Citrāṅgada ever married) was enthroned, Bhīṣma was “settled on the mind (or thought) of Satyavatī” (satyavatyā mate sthitaḥ; 5d). This half-line becomes a refrain: first, after the extravagantly martial Citrāṅgada's untimely death, when it is a question of Bhīṣma governing as regent47 while Satyavatī's second son, Vicitravīrya, is still a child (96.1d); and again, immediately after the extravagantly amorous Vicitravīrya's untimely demise, when the continuity of the now thoroughly defunct line will finally call for the extraordinary measure of proxy siring or niyoga (59d). Be it noted that Satyavatī produces sons whose rules are marked by conquest and desire, but none whose rule is marked by dharma—a pattern that will be replicated in her three grandsons. As Ghosh observes, the “reiteration of the phrase satyavatyā mate sthitaḥ underlines the acknowledgment of her decision-making role” (2000, 40). As we follow the mind of Satyavatī through these refrains and beyond, two other points bear watching: first, while the phrase “settled on the mind of Satyavatī” suggests that Bhīṣma yields to her, it also means that he and she repeatedly put their minds together; and second, that Vyāsa has told this same mother—his own real mother—that all she has to do is “remember” him (from √smṛ, which often harbors feelings of love, longing, and desire) should something need to be done. As the author, Vyāsa knows his mother's mind even from afar, and with it her desires. Perhaps by disappearing at birth with such a promise of return, he could go for a time—to requote Lacan—“beyond the Mother, the real Other of (p.359) demand, whose desire (that is, her desire) one wishes she would assuage.” But of course a mother's desires may be beginningless and unending.

Just after we hear for the second time that Bhīṣma was “settled on the mind of Satyavatī” (1.96.1) in his regency during Vicitravīrya's childhood, we learn that as soon as the boy matured, Bhīṣma “set his mind (akaron matim) on Vicitravīrya's marriage” (2cd). Bhīṣma is minded to abduct three princess-brides for this ineffectual half-brother of his, and clearly initiates the plan—although as he sets off armed on a single chariot, it is with his (or their) “mother's consent” (anumate mātuḥ; 6). Let us mark this as the first point in a two-way concurrence, for Satyavatī will similarly ask Bhīṣma's consent (tava hy anumate; 99.17a) when she tells him her last idea on how to revive the extinct line once Vicitravīrya is dead. Etymologically we may say that their two “minds” (mati) lean “toward” (anu) each other.48 By the time Bhīṣma has completed his “superhuman feat” (karma-atimānuṣam) of abducting the three princesses and is ready to give them to Vicitravīrya, we are told that he “made the decision together with Satyavatī” (96.46). Here we come back to a point made earlier that has now become more complicated. Bhīṣma “Śāṃtanava,” who carries within his patronym the oversight of his family's “continuity,” began this project by spontaneously making two vows that complied with something fated for him by his mother Gaṅgā's samaya with the Vasus so that he could clinch the samaya by which his father could marry a second wife. Now this second mother, his father's widow, has become his handmaiden in the project of continuity, which has begun to take on the character of a kind of weave of mutually improvised dharma. Two things about this dharma: it gets rough and discordant; and Bhīṣma and Satyavatī soon find themselves co-improvising it with the author.

The abduction of the three princesses sets a rough and discordant tone from the start. The three are daughters of the king of the Kāśis (kāśipati, kāśirāja) whose city is Vāraṇāsī (Benares; 1.96.3–4), which is apparently a capital of the kingdom of Kosala.49 Although the princesses are being given a svayaṃvara (3) or “self-choice” betrothal ceremony in which they would have some choice in selecting their husbands from among the assembled kings, Bhīṣma quite literally subverts the ceremony, at least according to some wordplay by Vaiśaṃpāyana: “But when the names of the kings were proclaimed by the thousands, the lord Bhīṣma then chose50 those [maidens] himself (svayaṃ … varayām āsa)” (6)! (p.360) Lifting the women onto his chariot, Bhīṣma is quick to invoke dharma.51 But it is very selective dharma, and also opaque in the way he reaches the number eight in describing the legal options for weddings.52 Just before challenging all the other kings to stop his carrying the princesses away, he recites a blur of marriage types that are “remembered by the wise” (smṛtam budhaiḥ; 8b)—that is, probably, sanctioned by “tradition” (smṛti)—building up to the eighth:

Now know the eighth type of marriage remembered by the poets is the svayaṃvara, which princes praise and observe. But dharma experts call the bride best who is carried off forcibly. So, kings, I will carry these off by force! (10c–12b)

It would seem that Bhīṣma is interpreting this “eighth” form as susceptible to the rules of his sixth, which he has just described simply as “taking leave by force” (9c), implying the abduction of the bride according to the so-called Rākṣasa mode, which is implicitly what he turns the Kāśi princesses’ svayaṃvara into. Bhāṣma is clearly familiar with there being eight (he numbers only the eighth) acceptable forms of marriage, such as finds its way into some dharmasūtras and Manu.53 But the svayaṃvara is not among the eight these texts mention, which fits our sense that Manu speaks from a tradition that avoids role modeling by princely women.54 The Mahābhārata, in fact, accurately recounts the same eight types in the same order as Manu when King Duṣyanta cites “Manu Svāyambhuva” on their descending order of preference by varṇa (Mbh 1.67.8c–9b; cf. M 3.21) before he praises the Gāndharva mode or “love match,” by which he is trying to seduce Śakuntalā.55 It would thus seem that Bhīṣma, like Duṣyanta, knows what he is doing. Taking advantage of urgency to make his list cursory, which may allow him to be vague about the Prājāpatya mode, he bends the Law by making double-talk about the (p.361) three types of marriage most suitable for Kṣatriyas, elevating the Rākṣasa mode while suppressing the Gāndharva mode to make room for the svayaṃvara.56 Among these three, the Gāndharva mode is similar to the svayaṃvara in that in both the woman participates in choosing: in the Gāndharva mode privately, in the svayaṃvara publicly.57 And the svayaṃvara and the Rākṣasa modes share feats of valor that allow Bhīṣma to justify his intrusion as dharma.58 Just to confirm, Vaiśaṃpāyana calls it kṣatradharma as Bhāṣma fights off the challenge of the last determined suitor (29).

Nonetheless, although this may all be Bhīṣma's derring-do, it is important not to miss that it has Satyavatī's “consent,” both in the planning and the result.59 Satyavatī, who would certainly know what it is to be well protected in marital arrangements, seems to have left these maidens to something close to the law of the fishes. It is hard to imagine what the three princesses would think while being carted off on a chariot under attack by thousands of kings: off on a wild adventure?60 spellbound into inaction?61 traumatized?62 forced to calculate between wishing for the good or bad aim of their suitors in missing them? wishing someone would kill Bhīṣma? hoping that Bhīṣma's arrows protect them?63 And what would they think, once he has defeated the last challenger, when he treats them “like daughters-in-law, younger sisters, or daughters” as they cross the lovely landscape on the way back to Hāstinapura (96.42–44)? They do not appear to learn this “Terrible” man's purpose until this “dharma-knower” has given (pradadau) them to his younger brother and (p.362) begun preparations for their wedding “according to the Law of the good” (satāṃ dharmeṇa dharmajñaḥ), as he had decided with Satyavatī (96.45–46).

But now the eldest of these still unnamed maidens has a problem. While Bhīṣma is making wedding preparations, she reveals that she had prearranged with King Śālva of Saubha to choose (vṛ) him at the svayaṃvara, having already done so in her heart (manasā), as indeed Śālva had also formerly chosen her. Moreover, this was also the desire of her father (1.96.48–49). In other words, beneath this svayaṃvara there were some elements of a “love-match,”64 the very thing Bhīṣma suppressed in his patchwork recital of the eight modes of marriage. Śālva was of course the last challenger, and the eldest abductee who had chosen him could well have wished for Bhīṣma's death—as she will soon do for the rest of her life, and indeed her next one. Her choices are thus important. On one level, we may be back in the divine plan, for as Biardeau remarks, rather than taking on the role of dharmapatnī, a wife in accord with dharma, that Bhāṣma offers her, she seems to have made the error of choosing for herself an Asura demon who had incarnated on earth.65 And while that would not tell her story as she lives it, she is portrayed as defiant. Up to now she seems to be speaking only to Bhīṣma, but they are actually “in an assembly of Brahmins” (viprasaṃsadi; 96.50b). There she calls him “dharma-knower” and challenges him, now that he knows of her prior choice, to conduct himself in accord with dharma (dharmam ācara) (50)! Bhīṣma, who has subverted the svayaṃvara into a Rākṣasa-mode abduction and explained it as dharma by suppressing mention of a Gāndharva match that could be latent within it is suddenly challenged to rule on the two very things he had ruled out: a woman's choice and an apparent love match. Moreover, he is challenged to find a “courtroom” solution to a dharma predicament, as he will be again during Draupadī's humiliation at the dice match. But unlike his equivocation on Draupadī's question, this time “the dharma-knowing hero” and the Brahmins decide, with no reported discussion and without consulting Satyavatī, to give the oldest sister leave to go her own way. In these verses we learn that her name is Ambā, “Mother,” and in the next verse that her two sisters, as they marry Vicitravīrya, are called Ambikā and Ambālikā (52), both diminutives for “Mother.” The isolated “Mother” Ambā gets no (p.363) help from these two “little mothers” or her mother-in-law Satyavatī, who seems to have no voice in this assembly of Brahmins; and her “father's former desire” to see her married to Śālva is a matter that has no legal force. Although she will remain pertinent to this skein, this is the last one hears of her in it. Bhīṣma will bring her story up to date shortly before the war to explain why he will not fight a brother of Draupadī named Śikhaṇḍin, who is Ambā reborn after she had been rejected not only by Śālva but by Bhīṣma himself and has died vowing Bhīṣma's death in her next birth.

Neither Ambikā nor Ambālikā are ever ascribed a word. “Tall and dark (bṛhatī śyāme) and with blue-black curling hair, red-pointed nails, and buxom breasts and buttocks,” they turn Vicitravīrya from being dharma-minded to desire-minded (dharmātmā kāmātmā samapadyata) for seven years until he dies from “consumption” (1.96.53–58). Once again Bhīṣma is “settled on the mind of Satyavatī” (59d), who is now not only a widow but a grieving childless mother:

Then Satyavatī, distressed, wretched, eagerly longing for sons, having done the funeral rites for her son with her daughters-in-law, Bhārata, a thinking [or proud or eminent] woman (māninī), having reflected upon dharma and the Fathers’ lineage and the Mothers’ lineage (dharmaṃ ca pitṛvaṃśaṃ ca mātṛvaṃśaṃ ca), spoke to Gaṅgeya: “The ancestral offering (piṇḍa), fame, and continuity (saṃtānam) of the famous Kauravya Śaṃtanu who was always a man of dharma are established in you.” (97.1–3)

When Satyavatī mentions the Mothers’ lineage along with that of the Fathers, one might think she is referring to her own birth line (or lines … ). But the ancestral piṇḍa—a rite where rice balls are offered up on behalf of both paternal and maternal grandparents in a son's patriline—tells us otherwise. She is thinking of all the men (notably her deceased husband and sons) and all the women (dead and especially living, including herself, Ambikā, and Ambālikā) married into the one dynastic line whose “continuity” she says she seeks now to secure through Bhīṣma. I would not rule out that she could also be thinking of her husband's first wife Gaṅgā, now back in heaven, but probably not as the recipient of a piṇḍa, since Gaṅgā is immortal. Both Gaṅgā and Satyavatī link the dynasty's “continuity” through Bhīṣma with dharma: the one through history (itihāsa), the other through lineage (vaṃśa, kula). With Satyavatī, we recall that the ancestral offering is one of the three debts and one of the five mahāyajñas that are done for the benefit of both male and female “fathers” as joint sacrificers.

Clearly, however, even if Satyavatī considers Bhīṣma the “best of the upholders of dharma” (7b), she can hardly expect that he will do what she (p.364) now asks of him, and we know what else she could have in the back of her mind even if she has apparently not thought of it yet:

I shall enjoin (viniyokṣyāmi) you in a task. When you have heard it, you can do it …. Your brother's two chief queens (mahiṣyau), auspicious daughters of the Kāśi king, both lovely and in the bloom of youth, are desirous of sons, Bhārata. Beget children on them for the continuity of our line (saṃtānāya kulasya naḥ) by my appointment (man-niyogāt), fortunate one. You are able to do dharma here. Be consecrated as king of the realm and rule the Bhāratas. Take a wife by the Law (dharmeṇa). Do not drown (your) grandfathers. (97.7cd, 9–11)

Satyavatī mentions dharma at every turn (from 97.2–7 she refers to it seven times), but when it comes to being able to do dharma in two contradictory ways, Bhīṣma has established his preference for celibacy. No doubt Ghosh has a point that Satyavatī is taking a courageous risk, as she seems to leave a possibility that Bhīṣma could have a son who might succeed him as king with a new wife (2000, 41–42), although Satyavatī has certainly made her two widowed daughters-in-law his first “appointed” priority. But rather than “renouncing her self-interest irreversibly,”66 I think Satyavatī is urgently throwing different options on the table more as a hastily calculated risk. Though jumbled, and whether or not she has her own next step in the back of her mind, she drives straight to the point that she knows Bhīṣma must reject. At first, she seems to be talking about niyoga as the appointment to a proxy siring, which perhaps by some wild stretch of the imagination she might think Bhīṣma would think of a way to do. But in short compass she asks him to do dharma by breaking both of his vows in order both to rule and to marry. Let us also note, and hold for further discussion, the incongruity of her mentioning two chief queens (mahiṣīs).

Bhīṣma reminds Satyavatī of both of his vows. Restating them, he says that even though her words are the highest dharma, nothing will make him forsake his truth even if “the Dharma King (Yama) should abandon dharma” (97.13–18). Satyavatī now admits, “I know the truth which you spoke for my sake,” and she knows that Bhīṣma stands utterly upon it (20–21b); but she persists with one more appeal:

Look to the Law of Distress. Carry out the ancestral yoke. Act, scourge of your foes, so the family thread (kulatantu) and dharma will not be lost and your wellwishers may rejoice. (97.21c–22)

(p.365) According to the narrator Vaiśaṃpāyana, Satyavatī, still “wretched, eagerly longing for sons,” is now reduced to “babbling and speaking astray from dharma (dharmād apetam)” (23). It seems that she is left to hold dharma and what she calls the bare “family thread” or “thread of the line” together all by herself. As she perceives, there seems to be no dharma solution outside the Law of Distress, for which Bhīṣma will reveal some uneasiness when he approaches that topic in the Āpaddharmaparvan of Book 12 (Bowles 2007, 192–94). But let us note that one imaginable solution goes unimagined: Satyavatī, now certainly no less a widowed queen than her daughters-in-law, would be available for a proxy siring herself. Like her two daughters-in-law, whom she will soon describe as “desirous of sons by the Law” (putrakāme ca dharmataḥ; 99.43), she too, as just mentioned, is “eagerly longing for sons” (putragṛddhinī). In these epics it would be no excuse to say she is too old (we have met one comparable figure: Bhīṣma's paternal grandmother, the wife of Devāpi; another is Kausalyā in the Rāmāyaṇa). As we shall see, this unmentionable (certainly Bhāṣma could not recommend it to his “mother”) has some interest as a structural possibility. But back to reality, Bhīṣma does have an āpaddharma solution to offer that he can leave her and himself biologically out of, which he introduces as kṣātra dharma:

I shall tell you, queen, the eternal Kṣatriya dharma by which even the continuity of Śaṃtanu (śaṃtanor api saṃtānam) may be imperishable (akṣayam) on earth. Having heard it, and having looked to the loom of the world (lokatantram), carry it out with wise household priests who are skilled in the aims of āpaddharma (97.25–26).

Once she listens, Satyavatī will not be left holding the last thread (tantu) of the line. For in his tightly packed response that “picks up” twice on the verbal root√tan in tantu, “thread,” Bhīṣma offers her as dharma a way that she can work that thread back into the “continuity” (saṃtāna) of Śaṃtanu, having looked to the “loom” or “course” (tantra) of the world! It will be Satyavatī's—and our—task to unravel this teaser, which punctuates the end of an adhyāya. But first she must hear more from Bhīṣma, who has yet to know all that lies potent in his own words.

Without mentioning niyoga (or any derivative), which calls for the intervention of a (preferably older) brother of a deceased or otherwise incapacitated husband, Bhīṣma recommends—as a kind of alternative, or as Doniger puts it, as “pseudo-niyoga” (1995, 174)—that a Brahmin do the job, and provides two precedents. First, placing his proposal under the heading of Kṣatriya dharma, he invokes the story of Rāma Jāmadagnya: after he killed all the Kṣatriyas twenty-one times, their widows regenerated that population by uniting with (p.366) Brahmins (1.98.1–5).67 Then he works back to that point rather obliquely by telling of the blind Ṛṣi Dīrghatamas, who was cursed in the womb to (as his name indicates) “long darkness” (6–19). When an apparently impotent king Bali engages him to sire sons with his wife, the Ṛṣi does this, but not before the queen rejects him for his blindness and age and refers him to her Śūdra nurse, with whom he first sires eleven of “his own” sons (20–33). As several (Doniger 1995, 174–75; Biardeau 2002, 1: 216 n. 21; Dhand 2004, 40, 42) have noted, there are foreshadowings here—blindness from a curse, rejected proxy, children with the servant—of what will soon be replicated in the Kuru line. But in brief, citing also a “Vedic” adage,68 Bhīṣma thinks all that is necessary is to find a Brahmin “of stern spirit” (niyātātmā; 4c) and for the women in question to “keep in mind the Law” (dharmam manasi saṃsthāpya; 4–5). Satyavatā herself should thus invite a suitable Brahmin of qualities (guṇavān) to prosper the continuity (saṃtāna) of the lineage “in the fields of Vicitravārya” (99.1–2).

So it is, with no hint that she has thought of it only suddenly, but rather because she knows Bhīṣma speaks the truth and because she trusts him, and for the continuity (saṃtāna) of the line, that Satyavatā now speaks of her premarital affair “with a faltering voice and as if smiling shyly” (1.99.3–4). She begins with how lawful everyone is and was: to Bhīṣma, “In our family you are dharma, you are the truth, you are the ultimate resort”; her father was legally engaged (dharmayukta) in his ferryboating; and “the supreme Ṛṣi Parāśara was the best of dharma's upholders” (5–7). She tells how she was seduced by Parāśara's boons of smelling beautiful and restored virginity, how once born Vyāsa divided the Veda, etc. And she concludes that although she can bring Vyāsa there, they must co-invite him:

He surely, when appointed (niyuktaḥ) by me and by you, … will beget beautiful offspring on your brother's fields. He has said, “You may remember me when a task need be done.” I shall now recall him, strong-armed Bhīṣma, if you wish. Surely, Bhīṣma, with your consent (anumate) the great ascetic will beget sons in the fields of Vicitravīrya. (15–17)

Note that it is Satyavatī who introduces the terminology of niyoga in this irregular manner of appointing a Brahmin—but not without regularity, since Vyāsa (on the mother's side), like Bhīṣma (on the father's side), is an older (half-) brother of Vicitravīrya.

(p.367) Bhīṣma folds his hands in homage. Somehow knowing something of Vyāsa's reputation, he praises Vyāsa's insight into the three puruṣārthas, and approves. Here, for the first and only time in this skein, Satyavatī is given a provocative name:

Then, when Bhīṣma had given his promise, O scion of the Kurus, Kālī directed her thought to the Muni Kṛṣṇa Dvaipāyana (21).

As far as I can see, Satyavatī's name Kālī has not been sufficiently considered.69 Only Biardeau seems to have grappled with it, but she loses its singularity in her mythographic categories: Satyavatī's name, “la Noire,” occurs in our present passage after Satyavatī mentions Vyāsa's name Kṛṣṇa, “black” (2002, 1: 216); and when Vaiśaṃpāyana mentions it, he “goes so far as to call [her] Kālī, ‘la Noire,’ from the name of the goddess under the cruel form of slayer of the Buffalo asura” (220). Biardeau offers no explanation of what Satyavatī has to do with slaying the Buffalo Demon, and would seem to overparticularize70 where Kālī is something more general: the name of a goddess recognized for being above all, like Śaṃtanu's other wife Gaṅgā, a very famous Mother.

This is not an easy point to make, since, as has been recognized for some time (see Kinsley 1975, 89), there seems to be only one place in the Mahābhārata, at least in the Critical Edition, which could be used to support the idea that the epic poets even know this “dread goddess.”71 But it is not a negligible usage, and despite what I said some years ago, it is not likely to be an “intrusion.”72Kālī” is used either as an adjective or a proper name to introduce the goddess Kālarātrī, the “Night of Time,” also calling her a Kṛtyā, an apparitional goddess of black magic.73 It describes Kālarātrῑ as she is seen by the warriors about to be slaughtered by Droṇa's son Aśvatthāman as he is possessed by Śiva during the night massacre that brings the Mahābhārata war to its final cataclysm, and as she was seen by those warriors in dreams all during the war (10.8.64–67). As Jacques Scheuer cleverly puts it, Kālarātrī “is presented with the features of a black woman (kālī, the first word of the description, is nearly a proper name) … ” (1982, 316). And indeed, Biardeau takes it as a proper (p.368) name.74 In either case, I think it is plausible to think of it as a reference to Kālī (cf. Johnson 1998, 115). But even if this Kṛtyā Kālarātrī is just another “dark woman,” we still have the case of Satyavatī, which could stand alone in the epic as a reference to Kālī. What then can we learn from the uses of this name for her?

First of all, as regards this one reference to Satyavatī as Kālī in this skein, I believe Biardeau has the right impulse to note that it occurs after Satyavatī mentions Vyāsa's name Kṛṣṇa. Biardeau also links Satyavatī–Kālī with the “epic constant” of darkness: as with the dark Yamunā from which she is born, both she and her son are among the dark “personages connected with saving the dharma of the world of transmigration” (2002, 1: 201). Contextually, however, when Vaiśaṃpāyana invokes this name, he reminds us, just as Satyavatī recalls Vyāsa's birth, indeed, as she brings him back into her world by remembering him, that she was “Kālī” when he was born. This is also what the higher order narrator, Ugraśravas, has said earlier when he introduced Vyāsa upon his arrival at Janamejaya's Snake Sacrifice to hear his epic composition recited for the first time. Vyāsa is:

he whom Kālī had birthed from Parāśara, son of Śakti, even as a virgin on a Yamunā island, the grandfather of the Pāṇḍavas who, the same day he was born from his mother, matured his body by will. (1.54.2–3b)

The case is also similar in the prose genealogy of the whole dynasty that Vaiśaṃpāyana delivers just before he spins out our present skein from it. When he comes to Satyavatī she is called Gandhakālī (90.51), “Kālī of the Smell”; and although the two names are used together to describe everything she does, Gandhakālī is used more tightly with reference to her mothering and premarital birthing of Vyāsa:

Bhīṣma, to do his father a favor, brought him Satyavatī, the mother whom they called Gandhakālī, whose premarital son (kānīna) Dvaipāyana was a child from Pārāśara …. (1.90.51–52; cf. M 9.160, 172)

Whereas “Kālī of the Smell” would presumably have the smell of dead fish, which Satyavatī loses while keeping her virginity thanks to Parāśara's boons, she is not called Gandhavatī75 or Yojanagandhā (1.57.67) until she has started (p.369) to smell like perfume. I think the poets describe Satyavatī as having undergone a transformation similar to ones undergone by other goddesses, but especially by Kālī, from dark to beautiful, usually golden or light (gaurī), to make her marriageable to Śiva.76

Although it is not the case in our present skein, Satyavatī also carries her name Kālī into her marriage to Śaṃtanu. There are some striking instances when Bhīṣma tells Duryodhana the story of Ambā (in the Ambopākhyāna) to explain why he will not fight Śikhaṇḍin, and again, it is usually a question of her being a mother.77 For instance, Bhīṣma says that when Śaṃtanu was worried that having one son was like having none, “Knowing his desire I brought him the mother Kālī (kālīṃ mātaram āvaham)” (5.145.18cd). But most fascinating are Bhīṣma's usages when he describes the dire condition of the heirless realm after the death of Vicitravīrya:

When Indra no longer rained on the kingless kingdom, the subjects hastened to me, oppressed by hunger and fear. “All the subjects are dwindling! Be our king and revive us! Be blessed, drive away the plagues (ītayo nuda). All your subjects are oppressed by the most terrible diseases (vyādhibhir) and but few remain, Gāṅgeya. You can rescue us! Dispel the diseases, hero, protect the subjects by dharma lest the kingdom fall to ruin while you are alive.” The subjects wailed, yet my mind was not shaken. Remembering the conduct of the good, I kept to my oath, great king. The cityfolk and my auspicious mother Kālī (mātākālī ca me śubhā), the retainers, the house priests and preceptors, and the learned Brahmins, much scorched, told me, “Be the king!” continuously …. At their words, son, I folded my hands, greatly distressed and unhappy, and told them again and again the oath I had made out of deference to my father, that for the sake of the lineage I would keep my seed drawn up and not be king. Then, having folded my hands, I placated my mother again and again, O king: “Mother (amba), though I am born from Śaṃtanu and carry the lineage of Kuru, I cannot belie my oath, withal for your own sake. Do not appoint (mā … niyojaya) the yoke to me. I am your servant and slave, O mother (amba) who loves her children!” (5.145.24–29, 31–33)

It is fascinating to see Bhīṣma telling Duryodhana that he called Satyavatī ambā while telling him the story of Ambā—as if all pestilential mothers were one, at (p.370) least to him.78 The conditions that induce Bhīṣma and Satyavatī to call in Vyāsa, which as Bhīṣma says was the next thing they did (5.145.34–35), are thus ones that one could associate with Kālī as a mother of ills and destruction—though again, this is not to say Satyavatī is that goddess, only that she is being called Kālī in a way that seems to recall such a goddess. The name deepens Satyavatī's primordiality to put it on a par with Gaṅgā's, with each equally intertwined into the Kuru dynasty's lineage problems and solutions. As Bhīṣma tells it, the public thinks the restoration of dharma is up to him, but he knows that it must also be up to her. In our skein, these dire conditions will soon be described much more minimally by Satyavatī to Vyāsa.

But first, even while Vyāsa was propounding the Vedas, he appeared “instantly, having discerned his mother's thought” (1.99.22), and without her having said a word. As she embraces him with tears, she is once again the fisherwoman (dāśeyī), and he sprinkles his “distressed mother” with water, greeting her too before saying, “I have come to do what you intend. Command me, you who know the true nature (tattva) of dharma, and I will do what pleases you” (25). If there was any doubt when she veered from dharma or deferred to Bhīṣma about it, here we have the word of the author that she knows its true nature. Vyāsa even seems eager to act on her command, as if he equates the Law with whatever pleases his mother. Satyavatī begins with an assertion of the parity of father and mother in parenting sons: “Sons are born the common property of the mother and father, O poet (kavi). As the father is their owner (svāmī), so is the mother, no doubt” (28). Drawing together the brotherhoods of Vyāsa and Vicitravīrya on the mother's side and Bhīṣma on the father's, and mentioning Bhīṣma's disinclination to rule or have children, she now proposes this:

Out of esteem for your brother, for the continuity (saṃtāna) of the line, at Bhīṣma's word and by my own appointment (niyogāt), blameless one, out of compassion (anukrośāt) for beings and for the protection of everyone, with non-cruelty (ānṛśaṃsyena) you must do what I am proposing. (32–33)

Note, again, that it is she who must make the appointment. Vyāsa is to beget sons on the two lovely wives of Vicitravīrya, who are “desirous of sons in accordance with dharma” (34). Vyāsa's response is to confirm Satyavatī's profundity when it comes to dharma:

(p.371) You know dharma, Satyavatī, both the higher and the lower (paraṃ cāparam eva ca). And since, knower of dharma, your mind is set on dharma, therefore by your appointment (niyogāt), having pointed out what is needed with respect to dharma, I will do what you desire as it is found to be ancient practice. (36–37)

Indeed, Vyāsa says that Satyavatī speaks both for the higher and lower dharma, which, in the words of the passage that introduces him at Janamejaya's Snake Sacrifice, he can certainly do: “He was a Brahmarṣi who knew the high and the low (parāvarajñaḥ), a poet, a man of true vows, and pure” (1.54.5cd).

But exactly what would the high and low be in the current situation? Vyāsa may be preparing Satyavatī for a little problem he has just mentioned: that he will now indicate what is needed with respect to dharma before he can undertake “her appointment.” Vyāsa says he will give his brother sons the likes of Mitra and Varuṇa, but that the two widows must submit to a year-long vow before he lies with them (1.99.38–39). That may be the higher law, and, if so, Satyavatī is urgent for a lower one: “In kingless kingdoms there is no rain, no gods. How, lord, can a kingdom be preserved that has no king!” (99.40c–41b). In that case, says Vyāsa, to bear a superior child the minimal vow will be that the widows must bear his ugliness: “my smell (gandha), my looks, my garb, and my body” (43ab).

Meeting Ambikā in secret, Satyavatī begins, “Kausalyā, what I say to you is the loom of the law (dharmatantram). Listen to me” (99.45ab). Going on to say that the appointment is in accord with Bhīṣma's view (buddhi) and yet still up to her own view (buddhi), “somehow (kathaṃcid) she persuaded that dharma-farer [Ambikā] by appealing to dharma” (1.99.46–47, 49). As the “somehow” suggests, those who are “appointed” are not always so accepting of this “ancient practice.” Whatever pestilential conditions may have already beset this kingless kingdom, Satyavatī's urgency looks like it is the first thing to deprive the Kurus of kings the likes of the Vedic sovereigns Mitra and Varuṇa. And the second is that even though Vyāsa has just said that without the year's vow it will depend on the deportment of the widows whether they get superior sons, Satyavatī does not sufficiently warn at least the first of them, Ambikā, and actually seems to mislead her, telling her to expect a “brother-in-law” (devara), which makes Ambikā think she will be lying with “Bhīṣma or another of the Kuru bulls” (100.2–3; see Biardeau 2002, 1: 217; Dhand 2004, 35). The result is that when Ambikā shuts her eyes at Vyāsa's ugliness, Vyāsa curses her son to be born blind “because of his mother's defect of virtue” (matuḥ … vaiguṇyād; 10c).79 (p.372) Satyavatī quickly recognizes that no blind man can be king of the Kurus and seems to correct what may have been an oversight in her first appointment. This time she asks Vyāsa to “grant a protector of the lineage of affines (jñāti) and an increaser of the lineage of Fathers—a second king of the Kuru lineage” (12). Note her three differentiable usages of “lineage” (vaṃśa) in this one verse. The lineage of affines would include Satyavatī and Vyāsa, who would belong to the Kuru vaṃśa on her side as her son, but now also as a father. Vyāsa meets Satyavatī's new request, but minimally: because Ambālikā pales, he curses her son Pāṇḍu to be born pale (14–21).

Finally, Satyavatī gives the first widow a second chance, but when Ambikā recalls Vyāsa's smell (gandha) and appearance (obviously, smell stands out in her recollections) and dresses a beautiful Śūdra servant to replace her in bed, Vyāsa and the Śūdra woman produce the “half-breed” (karaṇa) Vidura, an incarnation of the god Dharma and skilled in both dharma and artha (100.22–28, 101.27–28).80 Whereas the Law of Mother Gaṅgā shortchanged the Kuru dynasty, giving it kings well-versed in dharma but only for one generation, the Law of Mother Kālī continues to beset it with too many flawed successors: two dead, and three now each with a defect as to being king, with only Pāṇḍu's blemish leaving him eligible.

Yet if Satyavatī seems to leave nothing but loose ends, she also ties together the continuity theme with a curious term: what she speaks, she says, is “the loom of the Law” (dharmatantram). Van Buitenen translates dharmatantram as “under the Law” (1973, 235); Ganguli as “It is consistent with virtue” ([1884–96] 1970, 1: 246). Both are apt but neither captures the verbal play that makes this usage nodal and justifies such a literal translation once again. As mentioned, Satyavatī seemed to be left holding the bare “family thread” or “thread of the line” (kulatantu) alone (97.22) until Bhīṣma offered his recommendation that she look to the “loom” or “course of the world” (lokatantram) and consult with household priests conversant with āpaddharma (26). With the concurrence of Vyāsa, who will serve the Pāṇḍavas as a priest in their great rituals, though not as their purohita (see Sullivan 1990a, 31–34), (p.373) she now urges as dharmatantra the first of the three unions that will produce what Bhīṣma will soon enough, after a time of onrolling dharma under his regency,81 confirm are the needed threads to continue the lineage into the next generation. As he says to Vyāsa's three sons, and Vidura in particular, when he begins to plot their weddings:

Our family, protected by the great-spirited dharma-knowing kings of old, has never come to ruin here; and through me, Satyavatī, and the great-spirited Kṛṣṇa [Vyāsa], it has been firmly established in yourselves, the threads of the line (yusmāsu kulatantuṣu). (103.2–3)

Indeed, well before this in the text (or after it in the sequence of frames), the higher order narrator Ugraśravas prepared his audience, the Ṛṣis of the Naimiṣa Forest, to anticipate such “extended” verbal play, saying of Vyāsa when he comes to Janamejaya's Snake Sacrifice:

It was he who with holy renown and great fame begot Dhṛtarāṣṭra, Pāṇḍu, and Vidura, extending (or stringing along) the continuation (or stretching over) of Śaṃtanu (śaṃtanoḥ saṃtatiṃ tanvan). (1.54.6)

We see then how Mother Kālī Satyavatī has reset the shuttles of the divine plan along with Bhīṣma and Vyāsa. But let us not forget that Śaṃtanu himself and above all his first wife, Gaṅgā, set the first threads to this tapestry. Indeed, still within the higher order narration of the Bard Ugraśravas, I believe we are taken back to one of the riddles that provide an allegorical backcloth to the entire Mahābhārata in the story of Uttaṅka (see n. 14). These two Mothers—the bright woman of the night and the dark woman brought ashore by day—give genealogical shape to the Uttaṅka story's unique identification of Dhātṛ and Vidhātṛ, the Placer and Ordainer, respectively, as goddesses, weavers on the loom (tantre) of Time whose black and white threads (tantavaḥ) are woven into nights and days (1.3.167, 172), just as the children and descendants of these two mothers are threads woven into the nights and days of this text.

E. The Transitional “Three Mothers”

As we move from the generation of Satyavatī's sons to that of her grandsons, Bhīṣma once again brings in three women to be queens of the royal line: (p.374) Gāndhārī, to marry Dhṛtarāṣṭra; and Kuntī and Mādrī to marry Pāṇḍu (103–5). He also secures an unnamed bastard daughter of a king82 to be Vidura's wife, with whom Vidura has many virtuous sons who are unnamed like their mother (106.12–14). Presumably Bhīṣma follows custom in finding brides for brothers in their order of seniority. We must soon bring closure on this headline role of Bhīṣma, by which Dumézil dubs him “Bhīṣma marieur,” Bhīṣma the matchmaker (1979, 66–71). Dumézil covers Bhīṣma's part in arranging the marriages of Satyavatī to his father Śaṃtanu; of two of the three sisters he abducted for his half-brother Vicitravīrya; and now of these three wives for Dhṛtarāṣṭra and Pāṇḍu (66–67). Indeed, not only can we add to this his arrangement of Vidura's marriage, we could also suspect that his matchmaking role was set for him in his previous life as Dyaus when he was the chief among the eight Vasus to get the samaya (91.21–22) or boon (varam, 93.40–42) of Gaṅgā by which she married Mahābhiṣa-Śaṃtanu and became Bhīṣma's own mother.

In any case, the first thing we may note about “Bhīṣma marieur” is that, once he has accomplished these three generations of matchmaking, he seems to have finished all work of this kind. When it comes to the marriage of Draupadī, such family business has been handed over to his logical counterpart and co-“grandfather” Vyāsa, who, from Draupadī's wedding on to the revival of the Pāṇḍavas’ grandson Parikṣit, gets help from Kṛṣṇa in carrying forward the ongoing wonders of the line's continuity. From such hindsight, Bhīṣma's matchmaking seems to have been doomed to repeat certain types of failures through the two transitions under his watch—failures that he was hardly the one to correct.

In the first transition, Bhīṣma seems bent on repeating a fateful fascination with primal Mothers, with the first of whom, Gaṅgā, he in effect—as Dyaus—arranged his own conception. If I have invoked the primordial renown of Gaṅgā and Kālī Satyavatī as Mothers, I passed too quickly over an evident continuity with the names of their three successors: the three Kāśi princesses Bhīṣma abducts to marry Vicitravīrya. Bhīṣma's and Satyavatī's intention is to bring these three “Mothers” into the lineage before they ever are such, in this case mothers less by reputation than by their names, which mean something like Mama (Ambā), Mamita (Ambikā), and Mamacita (Ambālikā), or, if one prefers, “Mummy, Mummikins, little Mummy” (Jamison 1996, 67, 69, 79). They are brought in to be mothers like Gaṅgā and Satyavatī, but unlike them they are brought in violently; and their manner of becoming mothers will not be to their liking. Moreover, as both Jamison (p.375) and Biardeau have recognized (without, it seems, anyone else having waded in on it), these three names in combination cannot be disentangled from their use in a celebrated—if often called “notorious”—scene in the great royal horse sacrifice, the Aśvamedha. This usage occurs when the three names are invoked in a formula special to the Aśvamedha while the chief queen or mahiṣī is having undercover ritual intercourse with the sacrificed horse. The formula has what Jamison calls “clean and dirty versions” (79). The first and most widely mentioned in the texts is used when the mahiṣī is led up to the slain horse accompanied by her cowives (each of whom may have a hundred attendants). Here the mantra is:

O Ambā, Ambālī, Ambikā [var. Ambā, Ambikā, Ambālikā]83

No one is leading (nayati) me.

The horsikins is sleeping. (Jamison 1996, 67, trans. TS

“Leading,” according to Jamison (67, 274 n. 108), probably introduces a conventional sense of √ to suggest that the mahiṣī is being led, as it were, into matrimony. Then, once the mahiṣī has settled herself with the horse beneath a blanket84 and either the Adhvaryu priest or the king (her husband as Sacrificer) has uttered the most straightforwardly erotic of all the rite's mantras, focusing on her sexual pleasure, the mahiṣī, having “manipulate[d] the dead horse into some sort of copulatory position,” in some texts modifies her mantra to say:

O Ambā, Ambālī, Ambikā85

No one is fucking (yabhati) me.

The horsikins is sleeping. (Jamison 1996, 69, trans. TS

With or without these words, which “mock” or “scold” the horse, the rite clearly invites the dead horse to regain its sexual stamina,86 for one of the Aśvamedha's purposes is to induce fertility, and specifically to obtain royal offspring (66, 76; cf. 242). Meanwhile, the cowives and their attendants circulate back and forth (p.376) around the horse and the mahiṣī slapping their thighs and fanning with their hems, and then exchange “slangy and crude” riddling mantras with the priests (65–66, 69–70).

The use of this formula to overcome childlessness gives us, then, an obvious point in common with what Bhīṣma and Satyavatī initially wish from the three Kāśi sisters, and continue to wish from the two younger of them even after they have become widows. But as Jamison and Biardeau recognize, the connection is anything but obvious, in no small part because there is no Aśvamedha at this point in the Mahābhārata (indeed, with Bhīṣma as regent, there is no king to perform one). Jamison, looking back to Vedic sources, proposes that the two usages can be illumined by a third: a “husband-finding” (pati-vedana) ceremony in its form as an “addendum” to the third of the Four-monthly (Caturmāsyāni) seasonal rites called the Sākamedha, in which Rudra is worshiped with a Traiyambaka Homa that includes Aśvamedha features: especially (along with more back and forth thigh-slapping87) the names of the “three mothers” formula condensed into “Traiyambaka” as a name for Rudra, who is worshiped both to remove the young woman's lack of marital success and to secure her first unborn descendant (242–43, 245–46). Biardeau, on the other hand, looks toward Purāṇic usages and finds the “link … between the rite and its epic role” (if I am unpacking her brief discussion correctly) in the proposition that the triple name in the Aśvamedha would evoke one person, the mahiṣī, “transformed into three women” as the Kāśi sisters, and that the one woman chiefly in question would thus be Ambikā—whom she calls here “the mother of Pāṇḍu,” but must mean the mother of Dhṛtarāṣṭra—not only as the mahiṣī, but in anticipation of her name becoming “one of the most frequent names of the Goddess” (2002, 1: 220).

I believe that both these authors have found pieces to a puzzle that the Mahābhārata poets have intentionally left incomplete. And if that is so, then so it must remain. But I think there are more pieces to be found if we allow for two things: first, that the Mahābhārata episode alludes knowingly to this Aśvamedha mantra and its ritual scene without being willing to make it obvious. For I think what it wants to do is pull off an innuendo that beneath the already potentially disagreeable submission to proxy siring (niyoga) to which the two widows are asked to submit, there is the further suggestion that the smelly and misshapen author Vyāsa is cryptically taking on the role of the revived sacrificial horse. Second, I would urge that if we dig into the contextual site of this (p.377) episode while thinking more or less synchronically across the two Sanskrit epics rather than reaching back to supposedly older versions of the epic story, which I consider unlikely,88 or resting the case on anticipatory epic outcroppings of postepic goddess mythology, we can see a few more pieces of the puzzle crop up from the text itself.

The first move, then, is to the Rāmāyaṇa, where a parallel scheme contextualizes this interpretation. It is a case of the Rāmāyaṇa having two Aśvamedhas where the Mahābhārata seems—until one sees the double parallel—to have only one. For whereas the Mahābhārata slots this present story of the rescue of the Kuru line amid (and, I argue, colored by) the other rescues of the line that come before and after it in the same skein, and has an actual Aśvamedha in its main narrative only later as a sin-cleansing rite after the Mahābhārata war (14.70.15–16; Hiltebeitel [1976] 1990, 292), the Rāmāyaṇa has two Aśvamedhas in its main narrative: the first slotted at the only parallel point where it is a question of rescuing the Ikṣvāku line by the birth of Daśaratha's four sons headed by Rāma,89 and the second, as in the Mahābhārata, a postwar sin-cleansing90 rite of realm consolidation. Something of this complexity calls for a fuller exposition elsewhere.91 But let us not miss the ways that both epics keep the Aśvamedha's reputation for replenishing the continuity of royal lineages even in their second Aśvamedhas. In the Mahābhārata, it is during the postwar Aśvamedha that Kṛṣṇa revives the stillborn Parikṣit, as Vyāsa had promised. And in the Rāmāyaṇa, it is during the postwar Aśvamedha that Rāma discovers his sons Kuśa and Lava, born to Sītā during her banishment in Vālmīki's hermitage, who will now be Rāma's recovered heirs.92

(p.378) What then are the additional puzzle pieces that crop up from the site of Vyāsa's “appointment” to sire sons on the two Kāśi widows? The first two are his insistence that what is needed with respect to dharma is that the two widows undergo a year-long vow before he lies with them, and that should they do this he will give his deceased brother sons the likes of Mitra and Varuṇa (1.99.38–39). However general it may appear, the stipulation of a year-long vow is an exact counterpart to the Aśvamedha requirement that the king and his queens must remain sexually abstinent during the full year that the horse wanders free.93 Moreover, Vyāsa's precision regarding the pair Mitra and Varuṇa is unusual in the epic and certainly has a Vedic ring.94 It is important to remember who is speaking here and what he has been doing. The last we heard of Vyāsa before Satyavatī recalled him as needed was that he had gone off to the Himalayas to divide the Vedas and impart the Mahābhārata as the fifth Veda to his five disciples, who were to proclaim it as the Bhārata (1.57.73–75). These precisions about a year-long vow, night-visitation, and Vedic-issue sons thus come from a Veda-knowing author who could be alluding to what we learned in chapter 3 § D: that the Aśvamedha variously identifies the king with dharma, and that the Rājasūya, the other great Vedic royal ritual, invokes Mitra as “lord of truth” and Varuṇa as “lord of dharma” in announcing the newly consecrated Bhārata king (MS 2.6.6; TS–2). Such a momentary fusion of elements from of two great royal rituals would fit our scene, since Vyāsa would be engendering just such a king—or two. It is also in this domain of Vedic allusion that I would propose some explanation of how and why Vyāsa would be encrypting himself as a revived sacrificial horse.

First of all, the Mahābhārata is certainly in step with the dharmasūtra–dharmaśāstra literature and the Māmāṃsā in considering Veda as a source of dharma. But more than these surrounding literatures, and also more pervasively than the Rāmāyaṇa, it is disposed to give new and often enigmatic shape to Vedic usages and practices in its main and ancillary narratives. What we are finding in our current skein is that this can be particularly the case where the Vedic practices themselves—as with this kingless Aśvamedha scene, and also with niyoga—are already enigmatic and of dark and doubtful dharma from the standpoint of post-Vedic Brahmanical culture. Their cachet lies in their being (p.379) reminders and remainders of dharma changing over time. It is thus telling that it would be Vedavyāsa who confirms these practices as dharma. Secondly, Vyāsa makes a singular appearance in the Nārāyaṇāya, which, I have argued (Hiltebeitel 2006a, 249–50), offers a bhakti decryption of a number of the epic's Vedic and purāṇic allusions. As mentoned in chapter 6, the Nārāyaṇīya comes to its deepest level of disclosure when Vyāsa tells Janamejaya, who has just been instructed to perform a Horse Sacrifice on top of his Snake Sacrifice (12.334.8–9), about the manifestation of Nārāyaṇa as the Horse's Head (Hayaśiras) while it is further disclosed that Vyāsa is a “portion” of Nārāyaṇa as well.95 These Nārāyaṇīya enigmas give a glimpse of Vyāsa's horseplay that might not be limited to speaking from the horse's mouth. The interpretation is in any case proposed as similar to the one Jamison offers in interpreting Ṛgveda 10.86 about Indra's monkey-companion Vṛṣākapi “as a veiled Horse Sacrifice” that describes, among the benefits brought by “Indraṇī's mating with Vṛṣākapi,” the restoration of Indra's worship with bulls and soma, the reaffirmation of his power, his recovery of good erections, and his attainment of sons (1996, 74–88 quoting 81, 82–83).

Now Vyāsa only says he will sire sons the likes of Mitra and Varuṇa (Mbh 1.99.38b), not kings. But we know that is what is at stake even if we cannot derive how this would have worked out. Dumézil seems to ignore this verse, perhaps because it does not help his case that an original set of incarnations has been effaced. For Dumézil, Pāṇḍu would originally have incarnated Varuṇa; Dhṛtarāṣṭra and Vidura likewise the two “minor sovereigns” Bhaga (god of destinies) and Aryaman (god of Ārya clans); and Yudhiṣthira would have been sired by Mitra were it not that a “clumsy retouch” had replaced Mitra by Dharma (1968, 146–48, 152, 159–60, 170–74). Going well beyond Jamison's assumption “that some version of the epic story of Ambā and her sisters already existed in early Vedic” (1996, 304 n. 94; see n. 95 above), Dumézil was convinced that he could recover a whole Mahābhārata whose “primary form [was] contemporary with the oldest Vedic times, or anterior” (172). That idea seems now to be a mirage.96 But the real point here is that Vyāsa does not get to sire sons who would incarnate dharma in the likes and names of Mitra and Varuṇa because (p.380) Mother Satyavatī says matters are too pressing to give the widows a year for such a vow. Instead of being a case of lost Vedic symmetries, it is a matter of something willfully set ajar.

It is thus not only on male initiative that we find Aśvamedha pieces set askew. Before Satyavatī gets to Vyāsa, she first tries to enjoin or appoint (vi-ni-√yuj; niyoga) Bhīṣma to sire sons with his brother's “two mahiṣīs.” Can there be two chief queens (mahiṣīs)? This is the only dual of mahiṣī (mahiṣyau; 1.97.9a) in either epic. The Rāmāyaṇa's first Aśvamedha incongruously requires more than one queen to lie with the horse, but on unequal terms: only Daśaratha's aged queen Kausalyā passes the night with the horse out of desire for dharma (dharmakāmyayā) as his mahiṣī, whereas his other two queens, Kaikeyā and Sumitrā, cast in the well-recognized Aśvamedha roles of the favorite (vāvātā) and discarded (parivṛktī) wife, respectively, are said only to have also united with the horse, apparently more briefly (Rām 1.13.27–28; cf. Jamison 1996, 66, 87, 274 n. 104). Just as the Rāmāyaṇa has intelligible narrative reasons to innovate in giving each of Daśaratha's queens some time with the horse, the Mahābhārata would have its reasons to be inventive in having two mahiṣīs.97 Yet the reasons are immediately curious, since neither Ambikā nor Ambālikā would be a mahiṣī if Ambā were still around. But of course this is Satyavatī speaking, not Vyāsa. It is never explicitly two mahiṣīs when Vyāsa lies with Ambikā and Ambālikā. Satyavatī's usage at this point in speaking to Bhīṣma would seem to be a reminder that Ambā's unavailability as a Mother complements Bhīṣma's as a Father, and an intimation that despite the legal incongruity, to speak of Ambikā and Ambālikā as two mahiṣīs gives them both an equal chance to become mother of the one desired royal heir.

Here some pieces noted by Biardeau and Jamison also fall into new places. The older of the two widows, Ambikā, should be—now already, but especially next with Vyāsa—not only (as per Biardeau) the single focus of the three names of the Aśvamedha mantra; she should be the single mahiṣī. And further (as per Biardeau), not only does Ambikā become a frequent name for the goddess; as Jamison shows, hers is the predominant name among the “three Ambikās” (p.381) referred to in Rudra's epithet Tryambaka, which refers to Rudra as “possessing three Ambikās.”98 Indeed, although the Aśvamedha mantra's “three mothers” are enough to call for the question of the epic's use of these ritual names, it is worth pressing Jamison's extension to the discussion of the Traiyambaka Homa further to ask whether the epic poets would have been alluding to this rite as well, as Jamison wants to suggest.99 It would seem that the Traiyambaka Homa's “‘husband-finding’ spell”—a “somewhat sinister ritual performed outside consecrated ground in an inauspicious, indeed dangerous place,” a crossroads—would be an addendum to a year's-end “Four-monthly” rite for a girl who is running out of time in finding a mate, and who, with her parents and no doubt at their urging, is calling on Rudra and his inauspicious sister to apotropaically remove her unfavorable condition (Jamison 1996, 242–44; 303 n. 80). For Ambikā as Rudra's sister is linked with autumn, which in some texts is Rudra's season of “special murderousness” due to her influence (cf. 241, 245, 304 n. 94).100 All this is fitting for an epic plot in which Ambikā, as a result of her larger failure in meeting the reduced vow demanded by Vyāsa of bearing “my smell, my looks, my garb, and my body” (99.43ab), would be doubled by a second mahiṣī, Ambālikā, because Ambikā's son will be blind and unfit to rule.

Further, setting aside her assumption (to my mind unconvincing) that this ritual, like the Aśvamedha, would have complemented an older version of the epic story, Jamison also relates the Traiyambaka Homa to Bhīṣma's abduction of the three Kāśi sisters on two interesting points. First, remarking that “the connection seems hard to gainsay, especially because the sisters also occur in a marriage context,” she suggests that “the epic maidens would provide bad role models for the husband-seeking girl of the Sākamedha” (245). Rather than assuming an old para-Vedic story, however, I think it preferable to ask what the epic poets might have made of the model of the Vedic rite. From this angle, the “husband-finder” of the Traiyambakahoma would be a good role model, given her bad situation, for something untoward to go further haywire in an epic (p.382) series of turnabouts, which would include Bhīṣma's turning of a husband-finding ritual into a wife-finding one—something even more basic than his turning a woman's-choice svayaṃvara into a man's-choice svayaṃvara, which, as Jamison observes (1996, 299 n. 38, cited above), he does with verbal precision. Again it would be a matter of inversion via allusion, although in this case less explicitly. Second, Jamison goes on to say that while Ambā became a murderous avenger after the abduction causes her to lose her husband, “[e]ven Ambikā and Ambālikā, though they settled happily enough into their married life after their unexpected abduction … , might not have chosen this particular method of pati-vedana [husband-finding] if they had their druthers” (245). But we know they did have their druthers. They were going to have a svayaṃvara. It would seem that as personifications of “the tryambaka,” the three Kāśi sisters would be embodiments of a pretext-pati-vedana or “husband-finding” by svayaṃvara that goes awry, at least for them, because Bhīṣma carts them off Rākṣasa style to find their rather limited un-chosen husband Vicitravīrya, whom Ambā in fact rejects.

Of course here we have a way to suggest that if the epic poets recall both Vedic rites, it is separately: the “husband-finding” ritual would underlie the beginning of the three sisters’ adventure into the Kuru dynasty. And the Aśvamedha invocation of Ambā, Ambikā, and Ambālikā would underlie the point where Ambikā and Ambālikā—and then Ambikā's Śūdra servant—lie with the smelly misshapen Vyāsa. Once again, there is something askew with the Śūdra servant in place of the oldest Mother Ambā, something that carries forward into the unfolding plot and allows us to move from this one threesome to the next: that is, to the transition from the three Kāśi sisters to Gāndhārī, Kuntī, and Mādrī.

As we approach our skein's last segment, let us take note of three continuities. First, as is apparent by now, in moving from generation one to two through an intensification of Vedic allusions centered on the Aśvamedha, we saw no diminution in allusive references to Mother Goddesses or to Rudra–Śiva. In both cases, we would seem to have cumulative evidence that the Mahābhārata either knows or anticipates more “purāṇic” mythology about these deities than it tells us. Is Gaṅgā already married to Śiva when she marries Śaṃtanu and becomes the mother of Bhīṣma? Is there already a Kālī behind Kālī–Satyavatī? Do the three Kāśi sisters, probably from Vārāṇasī, not only recall Vedic affinities with Rudra but already come from a city associated with Śiva?101 It is unnecessary to push these associations to (p.383) a positive answer on each or any count. The important point is that references to Śiva and possible allusions to “later” goddesses also remain prominent in the stories of Gāndhārī and Kuntī. If nothing else, the Mahābhārata provides a semiotics for later goddesses to find their syntax with Śiva. Second, we will notice in the movement from generation two to three a certain consistency, along with a downward vector, that one might seek to explain as an incomplete fulfillment of the three goals of human life (the trivarga), perhaps complemented by a Dumézilian analysis of incomplete tripartition. While in each generation there are impressive martial (artha or second function) feats by Citrāṅgada and Pāṇḍu and excessive amatory (kāma or third function) ones by Vicitravīrya and Dhṛtarāṣṭra, righteous pursuits (dharma or first function) are compromised, first at the top by Bhīṣma and then at the bottom by Vidura, where dharma is derailed from actual rule. Third, just as we may assume it would not have been easy for Bhīṣma and Satyavatī to find a bride for Vicitravīrya, we may assume it will not be so easy for them to find one for Dhṛtarāṣṭra. As we move from the three Mothers of generation two to the three Mothers of generation three, there will be in each case a seniormost mother (Ambā, Gāndhārī) connected with Śiva who either rejects marriage into the line (Ambā) or marries into its most unpromising senior branch after the line itself has split (Gāndhārī).

F. Mothers Kuntī and Gāndhārī

So Bhīṣma once again brings in three women: Gāndhārī (103.9–17), Kuntī, and Mādrī (105.1–6). As these accounts unfold, two things become apparent. First, it quickly emerges that our skein makes Kuntī its new rising star,102 for between Gāndhārī's marriage and her own, a whole adhyāya (104) is dedicated to Kuntā's childhood. She is the only woman marrying into the chief royal line of either epic whom we get to know well as a girl, and not only now but through her own adult recollections. Second, this is the first generation in this skein in which there is rivalry between “cowives.” Although there is rivalry between cowives in the lineage before our skein (see Dumézil 1973, 16–18; Defourny 1978, 107–37; Brodbeck 2009a, 128–29), the present case takes us back above all—as Gaṅgā and Kālī–Satyavatī did to the female weavers Dhātṛ and Vidhātṛ—to a prototype from the early prolegomenal matter in Book 1: in this case, the rivalry between Kadrū and (p.384) Vinatā, the cowives of the Ṛṣi Kaśyapa and the mothers of snakes and birds.103 Gāndhārī and Kuntī compete over who will be the first to give their husbands an heir, and then Kuntī and Madrī show some rivalry. If one recalls that Kuntī is Kṛṣṇa's paternal aunt, one has one of the reasons why these rivalries revolve around her, and why she is pivotal to the change, noticed in the previous section, to Vyāsa and Kṛṣṇa in the handling of the family's dynastic fortunes. Yet even as Kuntī becomes salient, our skein—right down to its well-marked end—keeps all three generations of women (Gaṅgā, on this point, excepted) in the living picture.

Bhīṣma prefaces his renewed matchmaking in a passage we have already noticed. Addressing Vidura, he mentions the continuity (saṃtāna) now established in the persons of Dhṛtarāṣṭra, Pāṇḍu, and Vidura as “threads of the line” (103.3). But although Bhīṣma mentions Vidura as one of the “threads,” he rather tries to draw him into being one of the threaders: “For this lineage to grow like the ocean, son, I, but especially you, must take care …” (103.4). Bhīṣma speaks favorably of three royal lines he considers to have produced suitable brides for Dhṛtarāṣṭra and Pāṇḍu, and, saying he thinks these prospects should be wooed, asks Vidura's opinion. But Vidura demurs: “You are our father, our mother, and our supreme guru”; Bhīṣma should decide and act on these matters himself (5–8). Calling Bhīṣma “mother” here could suggest that Satyavatī, Ambikā, and Ambālikā are not being consulted.

Biardeau offers an incisive key to interpreting Bhīṣma's choices: Kuntī's link with the earth, most notably through her name Pṛthā, is with the earth totally, even though her Yādava people are not associated with any land, whereas the names of Gāndhārī (Woman of Gāndhāra) and Mādrī (Woman of Madra) identify them only by their lands, and with neighboring northwestern lands ruled by or associated with incarnate demons in an area propagated by Buddhism (2002, 1: 231–33). Indeed, Madrī is even called Bāhlīkī (Woman of Bactria)104 once (1.116.21a), when Kuntī says how lucky Mādrī was to have been the last to see Pāṇḍu's face making love.

(p.385) Dumézil, however, appears to have made the most cogent attempt to interpret these three marriages as an ensemble, taking Bhīṣma to have implemented them in trifunctional order as “un théorie en action des modes archaϯques de marriage” (1979, 71). Dhṛtarāṣṭra's marriage with Gāndhārī, who is “given” to him by her father, mother, and brother (Mbh 1.103.11e, 12c, 15a) with a large dowry (14c), has earmarks of a Brāhma-mode marriage; Kuntī's “self-choice” (svayaṃvara) of Pāṇḍu comes next; and Pāṇḍu's marriage to Mādrī by purchasing her would be an Āsura-mode wedding (69–71). Dumézil thinks the operative theory would be older than the classical enumeration of eight types of marriage in which the Brāhma mode can be deemed suitable for Brahmins; the svayaṃvara—discounted in the dharmaśāstra lists—special, as we have seen in § D above, in the epics, at least, for Kṣatriyas; and the Āsura mode for Vaiśyas and Śūdras (33). But again, while the theory may be old and parallelled by marriage sequences in other Indo-European heroic traditions, this does not give the “entire episode” a special “antiquity” (70). Rather than reflecting an older account of an archaic trifunctional order translated and updated into terms of classical dharma, the epic continues to have Bhīṣma arrange marriages in which dharma remains askew. Kuntī's marriage to Pāṇḍu does not seem to present any immediate dharma problems, but considering that Kuntī's Yādava origins connect her only loosely with a royal lineage,105 it is not clear why she would be having a svayaṃvara at which to have rather miraculously “found (avindata)”106 Pāṇḍu “in the midst of a thousand kings” (105.2cd)! Mādrī's marriage in the Āsura mode can also be complicated, as Dumézil acknowledges elsewhere (1968, 75–76), since her brother Śalya is an incarnate Asura or demon.107 But it is especially in finding a bride for Dhṛtarāṣṭra that Bhīṣma sets things askew.

Here, where Dumézil sees the most “august” of the Indo-European “procedures ‘civilisés’” reserved for Dhṛtarāṣṭra, his handling shows the strain of his compartmentalization. Dumézil presents as lofty a picture as he can: “had it not been for his infirmity,” Dhṛtarāṣṭra “would have been king,” and his blindness is a kind of insight (1979, 70). But he fails to appreciate that in arranging for Dhṛtarāṣṭra to receive a bride as a gift, Bhāṣma enables a violation of the Kṣatriya ethos of not accepting gifts that he himself had so ably put on (p.386) record when he took the three Kāśi sisters ungiven yet gave them to Vicitravīrya.108 This time, when Bhīṣma goes to extraordinary measures to get someone else to give a bride to a problem child, it is not in an “august” remnant of a more civilized time but in the sad light of Dhṛtarāṣṭra's blindness. Indeed, when King Subala of Gāndhāra receives Bhīṣma's messenger, he agrees in mind (buddhyā) about the groom's lineage, fame, and conduct, but only after considering (prasamīkṣya) that Dhṛtarāṣṭra is blind (103.11; cf Brodbeck 2009a, 168). But it is when Bhīṣma first hears about Gāndhārī herself that we might wonder how desperate, or perhaps heedless, he is:

Then Bhīṣma heard from Brahmins that Subala's daughter Gāndhārī had propitiated the boon-granting god who took the eyes of Bhaga, Hara (bhaganetraharaṃ haram), and that the auspicious Gāndhārī, it was said, had obtained the boon of a hundred sons. (103.9)

The associations with Śiva of course continue, but their ominousness is now rather evident. As we descend into blindness along with this bride who will be given, Śiva is doubly Hara, the god who “takes” or “removes,” and what he removes is sight itself from the god of destinies—something that might complement what we have understood so far: that Dhṛtarāṣṭra's blindness resulted from Little Mother Ambikā's shock upon seeing Vyāsa. If Bhīṣma is “taking care” in selecting this bride, it would seem to have been bad judgment on his part, at least knowing of this boon from Śiva, for thinking a hundred sons was a good idea,109 and on Vidura's part for giving Bhīṣma carte blanche in all the matchmaking.

Gāndhārī, a woman who fares in dharma—a dharmacāriṇī (1.103.11ef)—has already been given by her father when we meet her. Realizing that her mother and father wish to give her to a blind man (12), she blindfolds her eyes and says, at least literally, “I would not eat before my husband” (nātyaśnīyāṃ patim aham; 13e). In saying this, she is twice said to be “vowing utter fidelity to her husband” (13d; 17b); but taken along with Vaiśaṃpāyana's adhyāya-closing statement that she never again mentioned “other men” (17cd), her first curt and haunting words could suggest that the bandaging of her eyes is not only an act of spousal consolidarity and deference (not eating before one's husband is (p.387) just ordinary good behavior) but a curbing of her desire—about which there is more that can be said. Mahābhārata folklore of northwest Tamilnadu knows an irresistibly pertinent story. Gāndhārī once wanted to make Duryodhana invulnerable by the power amassed from bandaging her eyes. She would lift her bandage to see him, and told him to come before her naked. But Kṛṣṇa got Duryodhana's sister Duḥśalā to appear, and Duryodhana, feeling shy, covered his genitals and thighs with a banana leaf, leaving that area vulnerable when Gāndhārī looked. The subtext of desire is evident: Gāndhārī wanted to see her son naked. And the banana leaf could suggest that Duryodhana sensed that she wanted not only to see but eat, which could recall what she says in Sanskrit about not eating before her husband.110 For now, however, we must appreciate that Dhṛtarāṣṭra's Little Mother Ambikā and his fiancée Gāndhārī have both closed their eyes around their desires. But whereas Ambikā did so for only a fairly short but fateful time—maybe just upon first seeing Vyāsa, probably for the duration of their unpleasant encounter—Gāndhārī has vowed to close them for a lifetime.

Before the Critical Edition gets to Vaiśaṃpāyana's briefer descriptions of Bhīṣma's arrangements for Kuntī and Mādrī to marry Pāṇḍu, it now pauses for an adhyāya, as noted, over Kuntī's childhood to tell how she became a mother even while remaining a kanyā, and to explain her secrecy about Karṇa, the son she bore and abandoned “to hide her misconduct and out of fear of her relations” (1.104.13cd). Like Gaṅgā, Kuntī puts her son in a river,111 and like Satyavatā she gets to remain a kanyā and hides the youthful affair from her adoptive relatives.112 But in Kuntī's case, compared with Satyavatī's, being a kanyā has more the surcharge of her being “just a girl.”113 Unlike Satyavatī, Kuntī does not bargain to remain a virgin, she just remains a girl: indeed, in a fuller version of her story, “a free female” (svatantrā). In that elaboration, Sūrya offers gratis that she will remain a kanyā even after their union, clarifying with a contrived but surprising etymology: “O sweet-smiler, neither your father, mother, nor elders prevail; hail to you of choice hips. Hear my word. A free female (svatantrā), (p.388) since she desires all (sarvān kāmayate yasmāt), is called kanyā from the root kan” (3.291.12–13). Sūrya says, I believe, that as a free female, a girl is free to desire the world (or all things) and to make the world (or all things) desirable.114 Known from birth as Pṛthā—“the Wide,” evoking the Broad Earth Pṛthivī and her “girlish wide eyes” (kanyām … pṛthām pṛthulalocanām; 3.287.12cd)—she is the eldest child of the Yādava chief Śūra and older sister of Vasudeva (who is perhaps still yet to become Kṛṣṇa's father), and she gets her name Kuntī from her adoptive father Kuntibhoja, the otherwise childless son of Śūra's father's sister, because Śūra had promised and given Kuntibhoja his firstborn child as “a friend to a great-souled friend” (sakhā sakhye mahātmane; 1.104.1–3). This exchange is not one that Kuntī will later recall happily. As she tells Kṛṣṇa during his prewar embassy to the Kauravas,

I censure not myself nor Suyodhana,115 but my father by whom I was transferred to Kuntibhoja as wealth is by rogues. I was a child playing with a ball in my hand when your grandfather gave me to Kuntibhoja as a friend to a great-souled friend! I was humiliated (nikṛtā) by my father and maternal uncles,116 Foe-scorcher. (5.88.61–63b)

Yet from this disquieting friendship unfolds the friendship of Kṛṣṇa and Arjuna and a hidden friendship between Kṛṣṇa and Karṇa that we will return to in chapter 12 (see Hiltebeitel 2007b, 30–31). For now, however, the focus is on how Kuntī got the mantra that made her a special kind of mother:

In her father's [Kuntibhoja's] house she was appointed (niyuktā) to the honoring of Gods and guests; and so she came once to serve that fierce and terrible Brahmin of strict vows whose design in dharma is hidden (nigūḍhaniścayaṃ dharme), whom they know as Durvāsas. This fierce man of honed spirit she satisfied with all her efforts, and the Muni, with foresight into the Law of Distress (āpaddharma-anvavekṣayā), gave her a mantra combined with sorcery (mantram … abhicāra-abhisaṃyuktam) and said to her: “Whichever God you call up with this mantra, by this or that one's grace there will be a son for you.” When that Brahmin had said this, from curiosity, and being just a girl (kanyā satī), the famous one then invoked the Sun God …. (1.104.4–8)

(p.389) From this momentous passage, whose essentials Kuntī herself will soon rephrase,117 we need only know that Durvāsas is an irascible and inveterately hungry Ṛṣi who incarnates Śiva, to sense all that would be at work when, with “foresight in āpaddharma” and as “one whose design in dharma is hidden,” he gives this sorcerous Mother-making mantra to this wide-eyed girl.

In the fullest version of this story, Kuntibhoja charges Kuntī to give to Durvāsas ungrudgingly or disinterestedly (amatsarāt; 3.287.15d), to which she replies, “It is by my own nature (svabhāva) that I would honor the twiceborn” (288.2ab)—powerful words from a girl. Here no sorcery is mentioned, but she seems to paralyze the god (289.17–18), and Sūrya calls Kuntī mattakāśinī (291.21), which could be “bewitching one.” Much later, when she asks Vyāsa to raise the slain Karṇa from the Gaṅgā so she can see him while Dhṛtarāṣṭra and Gāndhārī also see the slain Kauravas, she has more rueful memories: that Durvāsas and Sūrya both threatened to curse her, and that she again became a kanyā by the god's grace as Durvāsas had foretold (15.38.1–17). Here she also recalls that Durvāsas told her she would become the mother of Dharma (dharmasya jananī; 6a)! In any case, as we shall soon see, not only the gods but all the Ṛṣis must know that Kuntī keeps her secret sorcerous knowledge into her marriage to Pāṇḍu as part of her “innate” service to the twiceborn, and thus to the Ṛṣis themselves. This makes it part of the divine plan. But more than this, again something we could call primordial is drawn from the distaff side into the Kuru line: this time in a secret that the gods and Ṛṣis keep with a “free female” girl made old (as a mother) before her time. I would suggest that some such sense lies behind an understanding of Pṛthā–Kuntī in the north Indian Mahābhārata folklore of sub-Himalayan Garhwal, where she is renowned for her great old age in the Pāṇḍav Līlās (dance-dramas about the Pāṇḍavas and company) and known as the elder sister of Bhūmī, Mother Earth (Sax 2002, 71–74)! In one story, Arjuna follows her and Draupadī—his mother and his wife—only to find out who they really are: the chief “hags” among the Sixty-Four Yoginīs, with Kuntī the eldest among them, who appear before him as vultures to forecast the Mahābhārata war (Sax 2002, 153–55; cf. 144). “Hags” is Sax's translation of Garhwali pañcāli, “bird” (Hindi pakṣī), evoking Draupadā's name Pāñcālī. Note that Kuntī and Draupadī are birds, of whom one is soon reminded—along with the archetypal story of Kadrū and Vinatā—by the story of Gāndhārī's pregnancy.

(p.390) As Gāndhārī's story is told, it mirrors Kuntī's in ways that deepen their differences around themes we have been discussing.

Gāndhārī once satisfied Dvaipāyana [Vyāsa] when he had arrived exhausted with hunger and fatigue. Vyāsa granted her a boon. (Mbh 1.107.7)

Vyāsa's hunger, elsewhere unheard of, is here to match that of Durvāsas, which in Kuntī's story is only implicit yet there by his ornery reputation. Each Ṛṣi is “satisfied” using the same verbal root tuṣ: just as Kuntī “satisfied” (atoṣayat; 104.5c) Durvāsas, an incarnation of Śiva, and Gāndhārī “satisfied” (toṣayām āsa; 107.7c) Vyāsa, an incarnation of Viṣṇu–Nārāyaṇa. Their involvements thus reflect the intertwined workings of the two major gods, and not just on the sides where one would expect them. Yet if these two Ṛṣis are similarly satisfied, they reward their hostesses differently. Durvāsas gives an unsolicited mantra to a curious unmarried wide-eyed girl that opens the world ahead of her. Vyāsa gives a boon or choice (varam) to a married woman who has closed her eyes to the world, who now chooses—and there is no saying what else she could have chosen—only to go deeper into her own darkness by fine-tuning the troublesome boon that she has already gotten from Śiva: “She chose that her hundred sons would resemble her husband (sadṛśam bhartuḥ)” (107.8ab)—a man whom she had never seen, and how they should resemble him she does not say (it is tempting to take sadṛśam, “resembling, similar to, with the appearance or look of,” as a pun on Dhṛtarāṣṭra's blindness).118 Indeed, Vyāsa had already predicted at Dhṛtarāṣṭra's conception that this son of his would have a hundred powerful (mahābalāḥ) sons (100.10ab), so if Gāndhārī adds anything, it is not to their strength. And while it is not clear whether Vyāsa adds anything at this point either, the outcome is that Gāndhārī will have all her sons, plus an unasked-for daughter gratis,119 the aforementioned Duḥśalā, all in one horrendous two-year pregnancy that she interrupts only to abort it when she hears that Kuntī has had a son “of splendor like the (p.391) morning sun (bālārkasamatejasam)” (107.10ab)—presumably Yudhiṣṭhira but sounding (no doubt intentionally) more like Karṇa, as is also the case when Gāndhārī tells Vyāsa just after this that she aborted her belly because she had “heard that the eldest son of Kuntī (jyeṣṭhaṃ kuntīsutam) was born of splendor like the sun (ravisamaprabham)” (107.15).120 It would seem that we are entitled to fantasize. Although jyeṣṭha could mean “most excellent” rather than “eldest,” the latter very common meaning, when referring to siblings, would suggest that Gāndhārī knows that Kuntī now has more than one son! Yet it would still take a lot to explain why Gāndhārī would be jealous about a déclassé son born years ago out of wedlock, or why she would be referring to Karṇa while really worrying about Yudhiṣṭhira. The ambiguity would seem to foreshadow the emerging affinities, despite their opposition, between Karṇa and Yudhiṣṭhira as the two most legitimate heirs to the throne. In any case, Gāndhārī's miscarriage requires Vyāsa himself to save the day, as it were, by dividing the mass of flesh into a hundred and one pieces while giving directions on how to let them continue to gestate in pots.121

Even though we cannot be sure whose birth Gāndhārī is talking about, much less how she could have heard about the birth of either Karṇa or Yudhiṣṭhira at this point, her hearing about Kuntī's sunlike firstborn son makes their rivalry the turning point in the Kauravas’ birthing. This simultaneity was important enough to have required this seeming mention of Yudhiṣṭhira's birth before our skein actually gets to it, and to interrupt the narrative of Pāṇḍu's strange reign—at least in the Critical Edition, which again, I believe, shows the soundness of defaulting to the Northern Recension where the Southern Recension gives the story differently.122 In the Critical Edition, Pāṇḍu's rule is narrated in two stages that cover three phases of his career, with the middle phase interrupted by the story of Gāndhārī's pregnancy. These three phases (p.392) clearly follow the one-sided role models set for him by his three or four “fathers.”123 The first (105.6–106.5) begins when he has married Mādrī:

When he had wed her, Pāṇḍu, who was endowed with both strength and enterprise, desired to conquer the earth and went at his enemies in all their multitudes. (105.7)

In this expansionist124 martial phase, he is like Citrāṅgada and/or Bhīṣma, but his conquests—most notably in Magadha as part of a drive to the east (105. 10–12)—now take place in a “real” political geography that will also challenge his Pāṇḍava sons (see Biardeau 2002, 1: 225). Note that when Pāṇḍu returns and greets the feet of his “father” (105.25), the narrator is referring to “father” Bhīṣma just as Pāṇḍu is about to deliver the war booty to him, Satyavatī, and Ambālikā, with something too for Vidura (106.1–2). Second is a phase of hunting life, in two parts. First, before we hear about Gāndhārī's pregnancy, we learn that Pāṇḍu was provisioned in the forests by his brother Dhṛtarāṣṭra and joined there by his two wives in an amorous setting.125 Then after the narration of Gāndhārī's boon and the births of the hundred Kauravas (107–8), Pāṇḍu shoots a Ṛṣi named Kiṃdama,126 who is mating while disguised as a deer (109.5–31). According to the dying Kiṃdama, Pāṇḍu is now tinged with lust and greed (109.11cd), which Pāṇḍu then links back to the lustfulness of Vicitravīrya as a strong vice he must get rid of (110.2–6). Note that Kiṃdama curses Pāṇḍu because of his cruelty in shooting a mating creature, which was “most unrighteous” (adharmiṣṭham; 21d) because it “frustrated a cherished fruit of the puruṣārthas” (19e; cf. 23d)—that is, the trivarga—which, as a king, Pāṇḍu was supposed to protect. Finally, Pāṇḍu's third phase (110.3–116.12) occurs under Kiṃdama's curse that he too will die should he engage in the act of love. Even though he manages to have sons before he dies under these conditions, “his last years pass without his exercising his royal duties in a normal manner” (Biardeau 2002, 1: 237). Here Pāṇḍu's turn to sexual continence puts him again in the same league as Bhīṣma. But more than this, what motivates him to practice tapas is the example of his “real” father Vyāsa (110.3–6, 29); and if Vyāsa is his model, it could also motivate Pāṇḍu's eventual incontinence.127 (p.393) Throughout all this, poor Pāṇḍu seems to have only increasing difficulties with this fractured paternal imago, and gets very little from his silent mothers. What good fortune he has must come from his wives, who love him.

Still following the Critical Edition and the Northern Recension, we now come to the turning point in the life of Kuntī, Mādrī, and Pāṇḍu. In four verses that the Southern Recension discards (109.1–4; see n. 122 above), King Janamejaya, having heard about Gāndhārī's pregnancy and the news she gets of Kuntī's sunlike child, now asks Vaiśaṃpāyana to tell him in full about the Pāṇḍavas’ births as partial incarnations of gods. It is these verses that prompt Vaiśaṃpāyana to relate the story of Pāṇḍu's deer hunt and Kiṃdama's curse, which occurs, of course, in both recensions: Pāṇḍu will die “when you are lying with a woman you love, overcome by love, blinded by your passion,” moving in that happiness not only himself but bringing about the death of the beloved he lies with, who will follow him out of devotion (bhakti), coming under the sway of the King of Ghosts.128 It could be either wife, and it is clearly bad news for both of them even if it is not clear what they learn about the curse. There is nothing to indicate that they heard it uttered, or what Pāṇḍu told them about it. All we know is this: when the deer-Ṛṣi died after speaking, Pāṇḍu was overwhelmed with grief (śokārtaḥ); then together with his wives he was stricken with grief and sorrow (śokaduḥkhārtaḥ) as they lamented the deer like a kinsman (109.31–110.1)! As Pāṇḍu turns his mind to asceticism, he speaks only of how all this affects him. And with that we embark on the third and last phase of his career—modeled, as he says, after Vyāsa—and the point toward which this chapter (and indeed, our whole skein) has been driving: his dialogue with Kuntī. But we cannot get there before another intervention by the Ṛṣis.

G. Kuntī, Mādrī, and Pāṇḍu among the Hundred Peak Mountain Ṛṣis

Pāṇḍu now utters one of the Mahābhārata's moving evocations of the renunciant life, holding it up as a sad but also beautiful ideal that the epic, however, will not allow its householder kings to live. Indeed, it will be like father like son all over again. Just after Pāṇḍu says he will emulate his father Vyāsa by yoking himself to austerities (110.6), he expresses himself in a train of thought (110.7–18) that his son Yudhiṣṭhira will also follow, quite precisely from point to point and using (p.394) many of the exact same words, phrases, and lines,129 when he says he wants to renounce the world after the war upon Kuntī's revelation that Karṇa was his elder brother. No matter how poignantly such a royal householder expresses this ideal, it moves his household to speak against it, and Pāṇḍu's entire household at this time and place is Kuntī and Mādrī, who now protest when he finishes this speech, which seems to ask them, while he remains behind to begin a life of solitude, to accompany Pāṇḍu's retinue (there are unnamed Brahmins, followers, and servants with them [110.37, 39]) to Hāstinapura, the City of the Elephant, to bring the court there word of his decision (110.22–24). Saying the same words jointly (25), Kuntī and Mādrī reply,

There are other stages of life (anye … āśramāḥ) that you can undertake together with us, your wives by the Law (dharmapatnīs), O Bharata Bull, and still do great austerities. And you surely and without fail will find heaven too. We shall abjure all our senses and, devoted to our husband's world and forsaking the pleasures of love, we too shall undertake severe austerities. If you desert us, wise king who are the lord of your people, then of a certainty we shall give up our lives this very day. (110.26–28)

Kuntī and Mādrī thus speak jointly as dharmapatnīs (not as co-mahiṣīs) to redirect Pāṇḍu toward an āśrama or “life-stage” that allows their participation.130 If that is their will, he says, he will be a vānaprastha with them and follow the “harsh and ever harsher rule (vidhi) of the Forest Treatises” (110.34–35). Although only Pāṇḍu mentions texts, his two dharma-wives would seem to know the gist of what is in them. The three now remove their royal paraphernalia for their retinue to take back to Hāstinapura along with word of his decision (36–40); and when Dhṛtarāṣṭra hears “from them as to all that had happened in the great forest, he mourned after Pāṇḍu” (41). Although Vaiśaṃpāyana has by now (in the Critical Edition) told Janamejaya about Dhṛtarāṣṭra's hundred sons, Gāndhārī has not yet had them. But the race is on and no one could now expect Kuntī to become a mother.

(p.395) The three now start off into the Himalayas, crossing over certain mountains—Himavat, Gandhamādana (220.42–43)—that their sons and Draupadī's travels will make more familiar.131 But their further journey, on which they begin to be “protected by Great Beings (mahābhūtas, possibly the five elements), Perfected Ones (Siddhas), and the Supreme Ṛṣis (Paramarṣis),” takes them to Mount Śataśṛṅga, the Hundred-Peak Mountain (110.44–45), a destination that is otherwise, it seems, virtually uncharted.132 But I believe it is invoked by Damayantī when she is estranged from her husband Nala and addresses a mountain somewhere in central India:

O best of mountains, have you with your hundred peaks (śṛṅgaśatair) that scrape the sky perchance seen king Nala in this terrible forest? (3.61.50)

What is most pertinent about this turning point in the famous love story of Nala and Damayantī, a “mirror story” to the Pāṇḍavas and Draupadī's own forest troubles,133 is that just after Damayantī appeals to this mountain, she turns north and walks for three days and nights (56) only to come upon a miraculous “circle of hermitages” (93d) “looking like a heavenly park” (57) populated by Ṛṣi-Muni ascetics (tāpasaḥ) “the likes of Vasiṣṭha, Bhṛgu, and Atri,” who “lived on water or off the wind, or fed on leaves” (58–59). Just as the Ṛṣis are now protecting Pāṇḍu, Kuntī, and Madrī—but especially, as we shall see, the two women—these Ṛṣis (really, “the same” Ṛṣis) have apparently heeded Damayantī's call to the hundred-peaked mountain and have arranged for her to find them so that they can reassure her that she will find Nala, “the best of dharma's upholders,” restored to her and to his kingdom (88)—only to vanish, along with their hermitages, leaving her amazed and wondering whether she had seen them and their hermitages only in a dream (93).

But whereas Damayantī brings the Ṛṣis of the north to central India, Kuntī, Mādrī, and Pāṇḍu scale the Ṛṣis’ Hundred-Peak Mountain themselves. For four verses, the focus is entirely on Pāṇḍu as he “became dear to the sight of the hosts of Siddhas and Cāraṇas” (111.1):

To some he became a brother, to some he became a friend (sakhā), and other Ṛṣis watched over him like a son. (3)

(p.396) Let us note these “hosts” (saṅghas) of Ṛṣis, whose likes we shall meet again in chapter 12. In their midst, Pāṇḍu was “intent upon going to heaven by his own power (svargaṃ gantuṃ parākrāntaḥ svena vīryeṇa)” (2), and “after a long time he reached such stainless austerity that he became like a Brahma-Ṛṣi (brahmarṣisadṛśaḥ)” (4)—that is, he “looked like” or “resembled” one, yet did not become one, even if he seems to have momentarily convinced himself otherwise. For at this point, “wishing to cross to the other shore of heaven, heading north from Śataśṛṅga,134 he set forth with his two wives” (5). His wishing to make this crossing, even with Kuntī and Mādrī, is motivated before this last verse by a fairly widely found eight-line mainly Southern interpolation.135 Hearing from the Great Ṛṣis that they are starting out on a new moon-night's journey to visit Brahmā for a great gathering of the great-souled Gods, Ṛṣis, and Fathers, who are desirous of seeing the Self-Existent in Brahmaloka, Pāṇḍu suddenly gets up wishing to accompany them! The passage is cosmologically interesting for its new moon setting, its similarity to Buddhist instructions on the path to the company of Brahmā (see chapter 4 § A), and its suggestion that what motivates Pāṇḍu is to shortcircuit the system by joining the company of the heavenly Fathers before he is one. But it is clearly an interpolation, occurring throughout the Southern Recension, from which it seems to have entered all the Northern Devanāgari manuscripts but one, yet none of the others in the Northern Recension. It is another example of Southern ingenuity, and with or without it, the ascetics now tell Pāṇḍu enough is enough.

Without making it quite explicit, the ascetics describe the journey beyond Śataśṛṅga as reserved for immortals (as the interpolation spells out). And while saying nothing about Pāṇḍu's being childless, they state their concern for his wives:

Going higher and higher facing northward up the king of mountains, we have seen the peak's many inaccessible regions …. There are regions of perpetual snow where no tree grows, no deer or birds live; there are some great continents (kecin mahāvarṣā), some inaccessible passes. No bird could cross them, much less animals. Only the wind has gone beyond, and the Siddhas, the Supreme Ṛṣis. Not deserving the misery, how could these two princesses not sink on this king of mountains? Don't go, Bharata bull! (111.5e–6b, 8–10)

(p.397) The ascetics are of course politely telling Pāṇḍu, without mentioning his deficiencies, that his wives are not ready to ascend bodily to heaven. But Pāṇḍu also understands (again, as the interpolation has hinted) that they are saying he also is not ready because he is childless (11). In a brief Southern interpolation, he continues to defy the system and would have all three of them take heaven by storm:

By terrible tapas, together with my wives, my life abandoned, even without offspring, I will find heaven by terrible action (karma). (1.1177*)

But in the Critical Edition—in lines that the Southern Recension preponed when it transitioned from Gāndhārī's gestation to an earlier moment in Pāṇḍu's plight (see n. 122 above), leaving Pāṇḍu here with just these bold but hopeless words—it is a deflated Pāṇḍu who concedes his childlessness. He now admits that he has fulfilled three of his four debts: those to Gods, Ṛṣis, and Men, but not the one to his Fathers.136 Troubled that his Fathers will perish when he does, he asks the ascetics how he might have “offspring in my field,” meaning his wives (111.11–17). Surprisingly (at least, we would assume, to him), the ascetics are encouraging. By their divine eye (divyacakṣuṣā), they foresee him having “godlike, beautiful, flawless offspring” and urge him to apply his intellect and make the effort (18–20). Minimally, they must know that Kuntī has her boon. But note how they set Pāṇḍu up to think the solution could be at hand in some other way, perhaps with one of them—as if with a wink and a nod he should recall how he himself was conceived by the Brahmin Ṛṣi Vyāsa.137 Indeed, before Pāṇḍu speaks to Kuntī, the Southern Recension now has him bring both his wives near to explain how

lower persons in distress (āpadi) desire a son from the higher, and the straight (sādhavaḥ) desire offspring, the fruit of dharma, from the best—(1.1181*, lines 3–4)138

(p.398) whereupon, as “he thought about a qualified (guṇavantam) Brahmin, the dharma-knowing” Pāṇḍu “brought” Kuntī and Mādrī “right near a conclave of great Brahmins” (lines 5–6)! The Critical Edition (including the Southern Recension when it returns from this interpolation) allows Pāṇḍu to be a little (if not much) more subtle as he has his conversation with “his famous dharmapatnī Kuntī in private (vijane)” or off to the side (111.22ab).

So now we come to the point of having tracked the term niyoga and its verbal variations through so many vicissitudes. If other women have been appointed or involved in appointing themselves or other women to this or that contribution to the continuity of the Kuru line, none since Satyavatī got to voice an opinion on the practice of niyoga itself, not to mention its variations, overtones, and ramifications. That is saved for Kuntī, now a woman who knows to speak her mind on dharma, but one who also knew to do so when she was just a “free female” girl.139 Arti Dhand has done a good job bringing out the main import in Pāṇḍu and Kuntī's niyoga dialogue (2004, 40–43), but there are still things yet to notice. As Dhand says, usefully summarizing, “Pāṇḍu goes to some contrivance to convince Kuntī that what he is proposing is a meritorious act,” first “telling her about the six types of heirs that are possible,” then quoting Manu “that any child of hers would legally be considered a child of his,” before he finally, “to fully legitimize his proposal,” offers her precedent by a story (41).

It is in their dialogue through stories and counterstories that the dharma issues come alive, and Pāṇḍu's opener can justly be called a Lulu:

Listen, Kuntī, to this story (kathā) about Śāradaṇḍāyanī, that hero's wife who was appointed (niyuktā) by the elders to bear a child. Pious and bathed, at night, Kuntī, at a crossroads, having chosen an accomplished Brahmin with a flower, having offered into Fire in the rite to conceive a son, she lived with him when that rite was finished. She gave birth to three warlike sons, Durjaya and so on, and so you too, beautiful wife, must by my appointment (niyogāt) quickly rise to conceive a son from a Brahmin of superior austerities. (111.33–36)

In the name of niyoga, Pāṇḍu is pressing Kuntī to do something similar to what the young unmarried daughter would do who is driven to perform the “husband-finding spell” of the “Three Ambikā Homa”: she should go stand (p.399) at night at a crossroads seeking a way to get pregnant, and without parental or, it seems, spousal supervision! I do not press the matter of the crossroads being dangerous140 only because the Śataśṛṅga ascetics all seem to be of the holiest sort. Kuntī has her chance to say something important, and what she starts off with is, in my opinion, one of the highlights of the Mahābhārata:

On no account, dharma-knower, can you speak to me like this (na mām arhasi dharmajña vaktum evaṃ kathaṃcana), your devoted dharma-wife, O lotus-eyed one. (112.2)

As she makes up her mind whether and when to tell Pāṇḍu about her secret mantra, she is ready to stand her ground in a discussion that will turn on dharma, her affection for her husband, and her unreceptiveness to niyoga.

She is, to begin with, just as encouraging as the Hundred Peak Mountain ascetics:

You yourself, strong-armed Bhārata hero, will give birth to heroic offspring in accord with dharma. I shall go to heaven with you, tiger among men. And for offspring, you only come to me, joy of the Kurus! Surely I will not go even in my mind to any man but you. (112.3–5b)

He has come to her, but not on terms she can yet accept. We see that she holds back yet promises all. Kuntī, as Dhand says, is “quite as deft in debate as Pāṇḍu himself,” and she now cites a counterstory to match his (2004, 41). She calls her story (kathā) “purāṇic” (112.6ab, cf. 13b), and it is more ludicrous than his. As Dhand states briefly, “Bhadrā Kākṣavatī, … deeply aggrieved at the death of her husband, through fantastic resolve succeeds in having her husband return periodically to life to sire seven worthy sons on her” (41). Minimally, Kuntī tells a story of successful necrophilia. But in the last verse 34, she does not, as Dhand states, propose a “similar code of conduct for herself”; rather, if she sets anyone's course by this story, it is Pāṇḍu's, who, she explains, need not periodically rise from the dead:

And so you too, Bharata bull, are able with just your mind (eva manasā) to beget sons on me by the lasting yoga-power of your tapas. (112.34)

She is really saying that Pāṇḍu will be able to contribute mentally to the use of her mantra, but that still lies beyond his grasp. But he does grasp, even as he is (p.400) about to tell the next counterstory, that her lesson has been about the resuscitated husband and would somehow have to be for him:

Yes, so did (cakāra) Vyuṣitāśva of old, beautiful Kuntī, just as you have told. He was surely like an immortal (sa hyāsīd amaropamaḥ). (113.2)

I believe that necrophilia is a key to Kuntī's story, but that it is not only about a king who impregnates his queen from his corpse (112.29–33). Kuntī, I will now contend, answers Pāṇḍu with reminders of the circumstances of his own conception, and of dimensions of it that had less to do with niyoga, which she is refusing, than with its having been a cryptic and, at least to the “two mahiṣīs,” repulsive Aśvamedha. If so, all this would help to explain how Kuntī's answer to Pāṇḍu is a rejection of his explicit mention of niyoga without her ever mentioning it herself. Here are the clues to such a reading.

Like Vicitravīrya with his two wives, King Vyuṣitāśva died of the extremely rare (in the epic) disease of “consumption” (yakṣmāṇam)141 after lusting madly (kāmasammataḥ) for Bhadrā (112.16). Also like Vicitravīrya, Vyuṣitāśva dies sonless. But unlike Vicitravīrya, this “most Law-minded” and “dharma-spirited” scion of the Pūru line (this would make him one of Pāṇḍu's ancestors) made his fame as a yajamāna or sacrificer (8) in numerous Soma sacrifices (9, 14). And the big event of his reign, by which he subdued the kings in each direction (11), was a horse sacrifice, which Kuntī sonorously links with his name and with his reputation for extraordinary strength:

At his Aśvamedha great-sacrifice, Vyuṣitāśva the majestic (aśvamedhe mahāyajñe vyuṣitāśvaḥ pratāpavān) surely became Indra among kings, endowed with the strength of ten elephants. (12)

The rite even occasions a song telling that he protected all the social classes like a father his sons (13)—perhaps a reminder that he probably had none. But we only know for sure that he was sonless with Bhadrā, who may or may not have been—but probably was—his mahiṣī. When he dies Bhadrā is not only his grieving wife (bhāryā; 112.17d); lamenting his demise and wanting to join him in the other world, she says, “Faithful as a shadow, king, I shall ever do your will, always loving to please you (nityam priyahite ratā)” (23). A mahiṣī is formulaically “dear” (priya) to her royal husband.142 When his corpse (p.401) impregnates her she is also a pativratā (32b). In any case, while there are no details on the rite, his name with -aśva means “the Daybreak Horse,” and possibly also “the One Who is Possessed by the Horse.”143 The first meaning clearly resonates with his story when Kuntī identifies him with a rising sun before his Aśvamedha (10) and with a setting one after it (17b), dying of lust-driven consumption from his lovemaking with Bhadrā, during which, if she was his mahiṣī, they would have been making up for a year's lost time after her cohabitation with the horse (recall that the king may utter the most erotic mantras to the mahiṣī while she is lying with the horse). The meaning “Daybreak Horse” would suffice to carry along an Aśvamedha innuendo, which calls only for something beside his “inner voice” (vāk … antarhitā) to arise from his corpse. But “possession” may also lie latent in the story since something beside or above the strength of ten elephants enables him to speak from his corpse and impregnate Bhadrā, who would already have had some familiarity with his Aśvamedha and its horse if she was his mahiṣī. A possession scenario may also be at play from her side, for before she “embraced the corpse” (taṃ śavaṃ sampariṣvajya) and heard its inner voice's directions on her timing and a bath (29–31), she uttered a long lament in which she vowed to lie from that day forward on a bed of kuśa grass “possessed (āviṣṭā) by sorrow and intent on seeing you (tvaddarśanaparāyaṇā)” (27). It thus seems that just as Pāṇḍu's story outrageously enjoins Kuntī to perform niyoga by recalling something similar to the “husband-finding spell” of the “Three Ambikā Homa,” Kuntī's story replies in kind, and equally outrageously, by getting Pāṇḍu to consider some of the deeper elements of what niyoga—which Kuntī will have no part of—meant for his own mother Ambālikā and her older sister Ambikā. Kuntī thus answers not in the name of niyoga itself but in the name of its Aśvamedha reverberations, which, it now appears, Pāṇḍu's conception has only doubled within his own dynastic line. Yet the matter is curious. Vyuṣitāśva is omitted from the Pūru–Bhārata–Kuru line's genealogies (see Brodbeck 2009a, 24–27). Could Kuntī be a better chronicler than Vaiśaṃpāyana? Could she be making this up as genealogy by invention? Brodbeck mentions her story (2009a, 176), but is silent on this question.

Kuntī has in any case mastered the art of deflection. But Pāṇḍu, while admitting that Vyuṣitāśva “was surely like an immortal,”144 still wants to return to the subject of niyoga, and takes a deep route back to it through the sources of Law. What he says Kuntī should now listen to is not precisely an old story but (p.402) what he calls, in impressive Vedicizing tones, an “ancient Law (dharmam … purāṇam) seen (dṛṣṭam) by the dharma-knowing great-spirited Ṛṣis” (113.3; cf. 6ef) that was repealed by none other than that Upaniṣadic enfant terrible and “spoiled brat” Śvetaketu.145 In former times women went around in the open (anavṛtāḥ), were free (svatantrāḥ), and took pleasure as it pleased them, but all that ended after young Śvetaketu, outraged at seeing a Brahmin making off with his mother, heard his father, the great Ṛṣi Uddālaka, defend the law: “This is eternal dharma” (eṣa dharmaḥ sanātanaḥ; 13d)! Śvetaketu did not concur, and laid down a new rule or limit (maryādā) that, once past the main point and its obverse, comes to a rider that is suspiciously convenient to Pāṇḍu's current cause. Śvetaketu ruled that women's faithlessness to a husband will from now on be a grievous sin of aborticide (or brahmanicide), as will seducing someone else's chaste wife, and that “a wife who is appointed (niyuktā) by her husband to conceive a child and refuses shall incur the same evil!” (Dhand calls this last point “a somewhat sinister twist” [2004, 41].) Although Pāṇḍu is a Kṣatriya, he is quoting a Brahmin—Śvetaketu—for whom bhrūṇahatyā (1.113.17), “killing an embryo,” can mean Brahmanicide as well.146 Brahmanical dharma texts are written from a Brahmin's viewpoint, and any interference to a Brahmin's seed is a murder of Brahmins. Pāṇḍu thinks Kuntī should understand her rejection of niyoga as a Brahmanicide because she won't perform it with a Brahmin! The account is interesting for its way of imputing change to dharma while abrogating even the “eternal dharma.” But Pāṇḍu is clearly less interested in legislative history than in suggesting that Kuntī is acting like she thinks the old law still applies to her. And in this, even if Pāṇḍu may sound like he wants to raise the stakes by “speaking Law” rather than just a story, and however convenient, and I think desperate, his argument now sounds, he has deepened the dialogue with an ironic glimmer of insight into his wife, who, ever since she was a “free” girl, has kept the secret of a mantra given her by a Ṛṣi “whose design in dharma was [as she knows] hidden”—a secret that the Ṛṣis of Mount Śataśṛṅga are now protecting along with her.

Indeed, imagine what Kuntī heard when Pāṇḍu began this narrative about ancient and eternal Law: “From youth on they were faithless to their husbands, but yet not lawless, for such was the Law in the olden days” (113.5). Kuntī will still keep that secret: that she was faithless to Pāṇḍu, her future husband, when she became a child-mother.

(p.403) The point about this secret that is now emerging, and the point of giving Pāṇḍu and Kuntī such a mutually probing exchange, is that she is still free to do with her mantra what she will, and to that extent at least, the ancient Ṛṣis’ eternal dharma that women are svatantra lives on in her.147 Once the kanyā who could “desire all,” she is now the woman who will tell her sons—speaking about another kanyā, Draupadī—to “share it all equally” and have it all turn out dharma, as it will now if Pāṇḍu will only calm down.

Pāṇḍu is not finished, but with this legal tale he seems to have run out of good ideas. He cites two more brief precedents including his own (113.21–23); and he makes a few more strident statements about dharma to continue pressing “the entire patriarchal establishment of ancient India into the service of his argument, alternately cajoling and bullying Kuntī” (Dhand 2004, 41). Finally, when he says,

Dharma-knowers also know, princess, that whatever a husband tells his wife, whether dharma or even adharma, that is to be done—(27)

he might seem to have pushed matters beyond the limit. But that is also his segue to saying a wife should “especially” do her husband's bidding if he is longing for sons and without the power to engender them,

as I am, flawless Kuntī, longing to set eyes on a son. Thus with folded hands, cupped like lotus petals with red fingers, this añjali is raised, lovely one, to my head for the sake of your grace (prasādārtham). (113. 28cd–29)

With his hands still, we may assume, so beseeching, he asks one last time that she go “by my appointment” (man-niyogāt) to a highly qualified Brahmin so that “on your doing, broad-hipped lady, I may go the way of those who have sons” (30). As Dhand says, it is this “truly extraordinary gesture, the parallel of which is hardly to be found in any of the Indian epics,” and Pāṇḍu's “finally actually pleading” (2004, 42)—and, we might add, his finally acknowledging that it is totally up to her—that convinces Kuntī to reveal the secret of her mantra.

For the rest, we can limit ourselves, while keeping in mind the women's rivalries and the question of succession, to three points: the mantra; some different features in the conceptions and births of Yudhiṣṭhira and Arjuna; and the initiative of the Ṛṣis in settling the Pāṇḍavas and Kuntī at Hāstinapura.

(p.404) One first hears about the mantra in the Mahābhārata's prologue, where one verse describes the conceptions of all five brothers:

Their two mothers were impregnated, in accordance with a secret Law (dharma-upaniṣadam), from Dharma, the Wind God, Śakra, and the twin Gods the Aśvins. (1.1.69)

“Secret law”—van Buitenen's translation—seems to call upon the esoteric character of the Upaniṣads. Upaniṣad can mean a secret teaching or spell, but it is in any case something “Vedic” and would have the authority of the Vedic Ṛṣis—like Pāṇḍu's “ancient law” endorsed by the Upaniṣadic sage Uddālaka—even if Kuntī gets her mantra only from Durvāsas and not his alter ego Vedavyāsa. A dharma-upaniṣad, however, is something new,148 for which “secret law” or perhaps “legal spell” may have to do while keeping such Vedic overtones in mind.149 As we have seen, Durvāsas gave this mantra to Kuntī as “one whose design in dharma is hidden (nigūḍhaniścayaṃ dharme)” (1.104.5a), a phrase (or verbal formula) that Kuntī knows well enough to repeat it exactly now to Pāṇḍu (113.33a). She also repeats that the mantra is “accompanied by sorcery” (abhicārasaṃyuktam; 34a), and further calls it a mantragrāmam (34c) or “canon of spells.” When that term is used in the longer version of Karṇa's birth, it is said to be something “heard in the Atharvaśiras” (atharvaśirasi śrutam; 3.289.20).150 Its hidden dharma is thus not only Upaniṣadic but Atharvanic, and sorcerous in that connection.

Yet we could not find much of sorcery in Kuntī's first tryout of the mantra with Suryā. Abhicāra comes explicitly into question only in the conception of Yudhiṣṭhira, and at Pāṇḍu's insistence. Making his mind up immediately, Pāṇḍu tells Kuntī, “Right now call Dharma” (39), and after explaining his choice, concludes, “With service and sorcery (upacāra-abhicārābhyām) propitiate Dharma” (42cd). Why the link between urgency and sorcery? For all his talk about Fathers, Pāṇḍu also wants a son quickly who will be in line for the throne:

Call Dharma, lovely woman, for he among the gods partakes of merit (puṇyabhāk). For Dharma could not join yoke with us if it were adharma and people will now think that this is the Law (dharmo ayam). And of a certainty this son shall become the standard of Law (dhārmikaḥ) for the Kurus. (39c–41b)

(p.405) If succession will hinge on this son's legitimacy, Dharma is the best he can do. As Kuntī now invokes Dharma, Vaiśaṃpāyana interjects that Gāndhārī has already been pregnant for a year (114.1); and soon enough (that is, before the birth of Duryodhana), Dharma's son is born to a brief announcement from a disembodied voice that this “firstborn son of Pāṇḍu called Yudhiṣṭhira” will be “the best of dharma's upholders” (5–6). Arjuna's conception and birth are quite different. Following another update that Bhīma was born on the same day as Duryodhana (114.14cd)—which would suggest that Gāndhārī's pregnancy lasted a long two years—the race is over and the couple can take their time for a son from Indra who will be Pāṇḍu's “choicest” (18). “Having consulted with the Great Ṛṣis,” Pāṇḍu “directed Kuntī to do an auspicious yearlong vow” (19cd)—exactly what Ambikā and Ambālikā did not have time for (cf. Brodbeck 2009a, 183)—and undertakes his own arduous tapas, until, after a long time, Indra announces his compliance (20–22) and foretells the grandeurs of his son-to-be, as does the disembodied voice, now that of the wind in space (vāyur ākāśe), once Arjuna is born (23–36). This divine wind is heard not only by Pāṇḍu and Kuntī but by the Hundred Peak Mountain ascetics, filling them with joy, and setting off such a vast clamorous celebration of the Gods, Divine Ṛṣis—a triṣṭubh verse is set off for the Seven Maharṣis of the Big Dipper (41)—and all manner of other celestials and even infernals (the snake sons of Kadrū appear along with the bird sons of Vinatā; 40a, 60–62) that those “best of Munis” (63b), who by now seem to include both the Hundred Peak ascetics and the celestial Ṛṣis, are astounded and all the more exultant over the Pāṇḍavas (37–63).

Pāṇḍu, however, is greedy for even more sons, and when Kuntī demurs, Vaiśaṃpāyana gives one last update that Gāndhārī has also now had her numerous sons (1.115.1b). Mādrī now coaxes Pāṇḍu to see if Kuntī will give her a chance with the mantra (1.115.1–18), thus activating the latent rivalry between these two that plays out from this point on.151 Kuntī's demurral comes with an interesting dharma adage, and she is surprised that Pāṇḍu is not respecting it:

They do not speak of a fourth son even in a time of distress. After three she would be a loose woman,152 after four a whore. Knowing this Law, which stands to reason (buddhigaṃyam), how do you transgress it and, as though besottedly, speak about offspring! (114.65c–66)

(p.406) Kuntī cites three sons as a dharma limit that she has already stretched. Manu makes one son the limit for niyoga, but allows that some say two (M 9.60–61). But we know that Kuntī now has four sons, counting her secret son Karṇa by Sūrya, and moreover that she has now had five “men,” counting Pāṇḍu. This will be the number of Draupadī's husbands when Karṇa uses the same term for “whore” (bandhakī) to revile her at the dice match for “submitting to many men” (Mbh 2.61.35). Again, Kuntī knows the Law, threads her way carefully, and is as free with her mantra to stop as she was to start.

H. Settling Mother Kuntī and Her Sons Back at Hāstinapura

Now that all the children are in place, the Pāṇḍava boys and their parents have a little time with the Śataśṛṅga ascetics. If Pāṇḍu and his wives hiked to Śataśṛṅga, they might have considered making the return trip, even if they would have been slowed down by the children. They did not. Maybe Pāṇḍu was embarrassed about his curse, or knew his story would be hard to explain. The journey to the plains is shrouded in mystery, and some have seized upon this to infer that the Pāṇḍavas are “inventions,” while others have made them out to be impostors, possibly Himalayan tribals from a polyandrous culture, and in any case not really the sons of Pāṇḍu, which is of course perfectly true. Compounding the mystery is the conundrum that Mādrī and Pāṇḍu seem to be cremated twice: first in the mountains and then outside Hāstinapura. I would just say that I do not think there is an ethnographic key to the Pāṇḍavas’ origins or their polyandry, but it would be nice to know more about royal funerals of the time or times of the epic's composition. The point we must satisfy ourselves with is that the Mahābhārata refers all these problems to the Śataśṛṅga ascetics.

These characters now reenter the narrative immediately after the births of Mādrī's twins:

Then those who dwelt on Śataśṛṅga gave [the five] their names, with devotion (bhakti) and ritual and benedictions, O king …. And as they grew up there on the holy Himalayan mountain, they brought wonder to the Great Ṛṣis who had foregathered there. (1.115.19, 27)

From here on, although both are repeatedly mentioned, the resident mountain-dwelling ascetics and the celestial Great Ṛṣis have merged and cannot be kept distinct. Love now springs its last time in Pāṇḍu's heart (116.4) as he succumbs to the “law of copulation” (maithunadharma; 9c) according to the “law of time” (p.407) (kāladharma; 12c), and Mādrī is the last to see desire on his face. As we saw in chapter 4, Pāli texts associate the powerful “law of sex” with village law and building houses to conceal it. The Mahābhārata now associates it with “law of time!”153 Kuntī and Mādrī, rivals nearly to the end, grieve and scream before working out which of them will lie with Pāṇḍu on the pyre and which one will take care of the children (13–31). Their last eloquent, noble, and sad exchange gives Mādrī the last word:

As he was lying with me the best of Bhāratas was cheated of his love. So how should I deprive him of his love in Yama's seat? Nor will I go on living treating your children the same as mine, noble lady, for evil would touch me that way! Therefore you, Kuntī, must treat my twins as your own sons. The king went to his death making love to me—let this carcass of mine be burnt with the body of the king that covers mine so well. Do this as a favor, noble lady! Watch over our children and think kindly of me—there is nothing else I could charge you with. (116.26–30b)

They do not mention (and must not know) that Mādrī's lot of joining Pāṇḍu in death was foreseen in Pāṇḍu's curse. It can be noticed that whereas the Kaurava–Pāṇḍava men's rivalries can be resolved only in heaven, their wives’ differences can be resolved on earth.

And now the text is where it wants to be, with the Great Ṛṣis and the ascetics of Śataśṛṅga helpmates together in securing the “work of the gods” and the “welfare of the world”:

After they performed Pāṇḍu's final bath, the god-like Great Ṛṣis then took counsel, those ascetics having assembled. (117.1)

Although it is hard to tell with the honorifics, it seems like the Great Ṛṣis are addressing the ascetics:

King Pāṇḍu has left his barely born children and wives154 in trust with your worships (bhavatām) here as he went to heaven. (117.3)

In any case,

Having thus taken counsel with one-another, those delighters in the welfare of all beings (sarvabhūtahiteratāḥ) put Pāṇḍu's children (p.408) before them and set off to the Elephant City. And the lofty minded Siddhas set their minds on journeying there to give the Pāṇḍavas to Bhīṣma and Dhṛtaraṣṭra. At that very instant, having taken them, all the ascetics set out with Pāṇḍu's wives, sons, and body. (117.4–6)

The arrival of thousands of Cāraṇas and Munis astonishes the people of the capital, and as the sun rises throngs of men and women from all four social classes come to see the ascetics, showing no trace of jealousy and becoming minded of dharma (dharmabuddhayaḥ; 12d), while members of the royal household, including Bhīṣma, Dhṛtarāṣṭra, Vidura, Satyavatī, Kausalyā (presumably Ambālikā), Gāndhārī, and Duryodhana and all his brothers come out to greet “all those hosts of Great Ṛṣis (maharṣigaṇān) with bowed heads.” Bhīṣma then, once “the mass of people all around had fallen silent, offered kingship and the kingdom to the Great Ṛṣis” (8–18)—a pro forma gesture that Mahābhārata kings often make to such visitors that requires no reply but acknowledges their higher authority (i.e., “real” rule).

Then the oldest of them, a Great Ṛṣi wearing the braid and deerskin, arose and knowing the Great Ṛṣis’ mind (maharṣimatam) spoke. (117.19)

Accountable for the collective “mind of the Maharṣis,” this unnamed senior Maharṣi155 now makes a fairly lengthy (117.20–31) speech, telling how Pāṇḍu arrived at Śataśrṅga “having renounced love and pleasure,” how the Pāṇḍavas were born despite that, how Pāṇḍu enjoyed his sons’ childhood and never strayed from the path of the good until he went to the world of the Fathers seventeen days ago, and how Mādrī joined him on the pyre faithful to the world of her husband. This senior Maharṣi speaks on three practical matters. First, saying “these are their two bodies,” the Kauravas should see to the “remaining rites” for Pāṇḍu and Mādrī (29–30). Second, while not quite saying it, he strongly hints that the preferred heir should come through Pāṇḍu:

This ancestral lineage was again uplifted by Pāṇḍu while the famous one dwelt in the forest always abiding by dharma …. And when the rite for the departed has been done, let famous Pāṇḍu, who knew every dharma, the bearer of the dynasty of the Kurus, receive the ancestral offering. (25, 31)

(p.409) Finally and accordingly, when it comes to describing the births of the Pāṇḍavas, it is Pāṇḍu's firstborn who really counts:

While he [Pāṇḍu] himself lived keeping a vow of celibacy, by divine instrument (divyena hetunā) this son Yudhiṣṭhira was born to him, begotten by Dharma himself. (21)

Enough said, and really, who would believe anyone else!156 I believe this is the first inkling of the divine plan made public in the course of the story, and it is the Ṛṣis’ role to make it so. And with that, as they did after imparting their message to Damayantī,

even as the Kurus were looking the Cāraṇas and Secret Ones (Guhyakas) all disappeared that instant. And upon seeing the host of Ṛṣis and Siddhas (ṛṣisiddhagaṇa) vanish there like a castle in the sky,157 they attained the highest wonder. (33–34)

It helps, however, to have Vidura—who incarnates the same god Dharma—there to be the first to speak, ordering that the rite for the departed be properly begun and confirming, while somewhat understating the matter, that Pāṇḍu's “five heroic sons were born like sons of Gods (surasutopamāḥ)” (118.4c).

When this ceremony has settled Pāṇḍu and Mādrī among the line's honored forebears, Vyāsa then brings our skein to its end:

When the śrāddha had been completed, Vyāsa looked upon the grieving people and said to his mother Satyavatī, who was blinded by pains of sorrow, “The times of happiness are past and times of trouble lie ahead. The days grow worse every new tomorrow. Earth's youth is gone. A dreadful time is at hand, confounded by much trickery (māyā), beset by many vices, when all conduct and acts of dharma shall be soiled. Go now, leave it all. Yoke yourself and live in the wilderness of austerities, lest you must witness the ruination of this your own family.” (118.5–8)

Satyavatī consents, tells Ambikā she has heard that her son's “bad policy” will bring about the destruction of the Bhāratas, and Ambikā consents to leave as well. Bidding adieu to Bhīṣma, Satyavatī and Ambikā then set off with Ambālikā, (p.410) who is still distraught over the death of Pāṇḍu, for the forest. There, after the fiercest austerities, their lives end (9–12). Kuntī and Gāndhārī are now the line's only living mothers.

I. Conclusions

Looking back over this chapter, we see that dharma is spun out from all variety of Brahmanical sources: regional and family customs, anecdotes and proverbs, Veda—especially in relaying it through the great Vedic Ṛṣis, and in some sense what is pleasing to the self as regards both men and women. As we have suggested, it may be that the Mahābhārata picks up on an implication of the dharmasūtras, which, like our skein, use the enigmatic term dharmatantra, “loom of dharma.” But unlike the dharmasūtras, this skein presents women who take initiative in the threading and are indispensable to the texture. They are back-clothed by female prototypes like the weavers of night and day and the snake and bird mothers Kadrū and Vinatā. Great things could also be said individually about other women's interventions in briefer crises of the Paurava–Bhārata–Kuru line (Devayānī and Śarmiṣṭhā, the two rivaling wives of Yayatī; Śakuntalā; Tapatī158) or about princesses married into other lines (Damayantā, Sītā, Sāvitrī) whom the Pāṇḍavas and Draupadī hear about in the forest. But nowhere else in either epic is there a skein that gives women such repeated prominence not only in the succession of their lives but in the ramifications of their lives for each other's lives, or does so with such recurring focus on a primordial Law of the Mother. Each of these women plays her part in this textualization of dharma, bringing home its nuances—whether in questioning it, interpreting it, raising questions by her silences, or even by a slip of the tongue. Though one sees it most artfully in Kuntī, each one, pace Manu, raises the question of her svatantra, “whether she is a child, a young woman, or an old lady.”


(1.) Although I take a different route through this material in focusing on the women's agency with regard to dharma, it is worth quoting Biardeau's take on the “generations that precede those of the protagonists”: “It is always by women that the foreign element of disorder in a dynasty that must remain interrupted is introduced (Devayanī, Śakuntalā, the two Satyavatīs, Ambā, Kuntī …), but the women are also the obligatory mediatrices of the restoration of good order (once again, Devayānī, Śakuntalā, etc.): the Puruṣa remains immobile facing its Śakti who does not stop manifesting the phases of her yoga” (1979, 125; my translation). See also Chakravarti 2009, examining much of this skein through the lens of transition in “gendered social and economic processes” (5).

(2.) See Olivelle 2005 a, 287 on this “cause celebre” verse and on svatantra as a legal term; cf. 9.2–3.

(3.) Cf. Jamison 2006, 200; on the contrary Ā 2.15.9, 2.29.11 and 15, giving some latitude to women.

(4.) M 8.70: “When there is no one else, even a woman, a child, an old man, a pupil, a relative, a slave, or a servant may give testimony.” But negatively, see 8.77: “Even one man free from greed may be appointed as a witness, but never women, even if they are many and honest, because the female mind is unsteady ….”

(5.) Draupadā, Sītā, and Śakuntalā, insofar as all three take their cases to court. Draupadī and Śakuntalā do so literally; Sītā presents her case to accompany Rāma to the forest during the “court intrigue” of Rām 2, and calls on divine witness in her two ordeals.

(6.) On the semiotics of women's desire, this chapter is affected by the work of Julia Kristeva.

(7.) Were this chapter short enough to include discussion of Buddhist sources, as originally conceived, it would take note especially of the theme of debt to one's mother that provides one of the main motivations for the Buddha to permit the founding of an order of nuns (see Ohnuma 2006). Cf. also the practice by both monks and nuns of making gifts to the sangha in the name of their mother and father, as recorded on early donative inscriptions (see Schopen 1997, 35–42, 56–71).

(8.) I thank Mark Pizzato for his stimulating comments. I was at the time supervising a dissertation on a Lacanian reading of the Mbh (Custodi 2005). On Lacan, law, and legal theory, see Caudill 1997.

(9.) In most of these cases the real father is human, but the Pāṇḍavas and Karṇa's fathers are gods. This complicates any use of Lacan's distinction between Real, Imaginary, and Symbolic fathers.

(10.) On Bhāṣma from Freudian perspectives, see Goldman 1978; Fitzgerald 2007. Sax's critique (2002, 78–92) of Goldman's psychoanalytic interpretation in favor of a local ethnography of Rajput filial piety, which Fitzgerald rightly rejects, has in any case little bearing on Bhīṣma's epic complexities—not least because it has so little to do with women.

(11.) “The Symbolic father is to be distinguished from the Imaginary father (often surprisingly distant from the Real father) to whom is related the whole dialectic of aggressivity and identification. In all strictness the Symbolic father is to be conceived as ‘transcendent,’ as an irreducible given of the signifier …. The real father takes over from the Symbolic father” (Wilder 1968, 271). Cf. S. J. Sutherland 1991, 47: Gāndhārī, Kuntī, and Mādrī “function almost as if they were the wives of one husband rather than two. After all, both husbands can be said to be complimentarily dysfunctional.”

(12.) Cf. Lacan 1977, 311: “The fact that the Father may be regarded as the original representative of this authority of the Law requires us to specify by what privileged mode of presence he is sustained beyond the subject who is actually led to occupy the place of the Other, namely, the Mother. The question, therefore, is pushed still further back.”

(13.) As with the semiotics of women's desire (see n. 6 above), this chapter reflects on Kristeva's departures from Lacan, with whom she studied, in her treatment of “the law before the Law,” the presymbolic, and the abject mother (see Oliver 1993, 46–68). My own caution, however, has been to bear in mind that although Kristeva and the Mahābhārata can be said to be looking for ways to talk about something similar, Kristeva starts with the child–mother dyad whereas the Mahābhārata starts more with the child-wanting couple and the impressionable girl.

(14.) Along with the cloth (paṭa) (1.3.147, 167) woven by the female weavers Dhātṛ and Vidhātṛ in the epic's opening story of Uttaṅka, major Mahābhārata scenes are compared to pictures (citra) drawn on a woven cloth (paṭa): the outbreak of fighting just after the Bhagavad Gītā (6.42.25); the pause in the battle on the fourteenth night when the warriors, horses, and elephants sleep (7.159.40); Gāndhārī, Dhṛtarāṣṭra, and Vidura's collapse upon hearing that all hundred Kauravas have been killed (9.1.40); Bhīṣma's closing silence on his bed of arrows, having finished his immense dharma talk (13.152.1); and Dhṛtarāṣṭra, Gāndhārī, and others seeing their slain sons rise from the Gaṅgā thanks to Vyāsa's boon of “divine sight” (15.40.20). These usages for such vivid scenes may anticipate the paṛ (a cloth depicting epic events) used in Rajasthani folk epics (Smith 1991, 5–9, 54–70, 504).

(15.) This is admitted with reference to “the putrikā/pativratā dichotomy” (Brodbeck 2009 a, 261), by which sons belong lineally to a woman's father or to her husband (49–56). These categories do not pose overt oppositions in our skein, in which Gaṅgā “is no putrikā” (155) and “her father is absent” (225); the overall intent, even with the anomalies presented by Gaṅgā and Satyavatī, is to make “good lineal wives” (160) and find dutiful pativratās (168). Yet Brodbeck does imply that Gaṅgā should have a father; hints at further possibilities in the sufficiently enigmatic story of King Vasu Uparicara's paternity of Satyavatī (163); and offers speculation linking Kiṃdama (the Ṛṣi disguised as a mating deer whom Pāṇḍu killed) with Kuntī and Mādrī as potential putrikās, which Brodbeck himself seems to reject (175–76). Brodbeck is less inclined to invent lineage connections and merge characters’ identities in this skein than he is in discussing earlier and later generations—perhaps because this is more inescapably the “full story.” In any case, he grants that “characters in this plot may seem slightly independent of those surveyed above” (151), who lived in prior generations.

(16.) This section draws some phrasing and insights from Hiltebeitel 2001 b, which now reappears slightly revised as Hiltebeitel 2011 a, chapter 13. From another angle, see now also Hiltebeitel forthcoming-g.

(17.) See Hiltebeitel 2001 b, 270–71, 274–75. I pass over most of the details, but Vasiṣṭha's hermitage is on “a side of Mount Meru” (meroḥ pārśve; 93.6)—the “cosmic mountain” that defines the axis of the Pole Star.

(18.) Ā 1.1.1–2 uses it to describe “accepted customary laws” (sāmayācārikān dharmān) and “the authority for their acceptance by those who know the Law” (dharmajñasamayaḥ pramāṇam). See chapter 5 at n. 28; Olivelle 1999, xl, citing also G 8.11.

(19.) Cf. Ā 1.13.10; 3.18.13; M 8.218–21; 7.202. The Mahābhārata tells of many samayas, including the Pāṇḍavas’ agreement to share treasures and their sleeping protocol with Draupadī (1.2.90; 1.204–5). Two of the first mentioned are the primordial bet on a horse between Kadrū and Vināta, the mothers of snakes and birds (1.18.5), and the prenuptial agreement between the husband and wife both named Jaratkāru (43.28).

(20.) This is the point made by the serpent woman Ulūpī to Arjuna, who is married and vowed to some kind of spurious celibacy: that it is a “subtle” matter of dharma to satisfy a ready woman (1.206.26–33).

(21.) Hopkins [1915] 1969, 6 has two Mahābhārata references to Gaṅgā being Śiva's wife, but they seem to come from interpolations. Śiva does break Gaṅgā's celestial descent at Mbh 3.108.9–14 and 5.109.6, from which their later (?) marriage results.

(22.) Cf. Brodbeck 2009 a, 157 and n. 15; there is another etymology for Śaṃtanu's name at 1.90.47b–49. Within our skein, from this point to the death of Pāṇḍu, there are fourteen of the Mahābhārata's forty-eight usages of saṃtāna, seven with kula and one with vaṃśa, and there is no other such cluster. The Rāmāyaṇa uses saṃtāna only five times, two with kula.

(23.) Both “were not [yet] satisfied” (nātṛpyata) by having seen each other. On Lacan's view of the objectifying disembodied male gaze and the possibilities these two scenes open for seeing in Gaṅgā's musing and coyness an exemplum of an embodied female gaze charmed by male vulnerability and “the sense of seeing-oneself-being-seen,” see Custodi 2005, 165–78.

(24.) Cf. Biardeau 2002, 1: 221: All this would have a “fonction avatārique” since Gaṅgā is doing devakārya, “ce qui est à faire pour les dieux”—which I translate as “the work of the gods.”

(25.) See Doniger 2009 a, 225, on Daśaratha's propensity to kāma over artha and dharma, and Rāma's criticism of this.

(26.) This unnamed wife does not seem to have any trait that would identify her as the Pṛthivī—Earth, Dyaus's Vedic wife. Indeed, note that the Vasu who most helped Dyaus steal the cow was Pṛthu, Earth in the masculine (93.26).

(27.) The Critical Edition follows the Southern Recension with śāstras (scriptures) rather than śastras (weapons).

(28.) Earlier, at 91.12, all Vasiṣṭha said was “Be born in a womb,” and that is all that Gaṅgā has reiterated so far (92.50; 93.31).

(29.) With typical recursivity, the Mahābhārata has begun that “prehistory” earlier with the story of Vasu Uparicara (1.57), which, at Mbh 1.1.50–51, is introduced as one of the three starting points from which some Brahmins learn the epic—the other two being “from Manu onward” and “from Āstīka onward.”

(30.) Cf. Brodbeck 2009 a, 158 n. 18, resisting such an “astronomical cosmological interpretation” while attending to genealogical matters, which the Mahābhārata actually subordinates to its divine plan and its cosmologically worked out sense of “history.” Indeed, Gaṅgā's intervention in the Bhārata genealogy is similar to the “descent of the Gaṅgā” (gaṅgāvataraṇa) in the Rāmāyaṇa, where she solves a genealogical crisis of the Ikṣvāku lineage posed by the disappearance of the sons of Sāgara. On these points, see Hiltebeitel forthcoming-g.

(31.) Ratim aprāpnuvan strīṣu; 94.18c. The alpha privative has been overlooked by van Buitenen, 1975, 223 and Ganguli 1: 230.

(32.) See Brodbeck 2009 a, 159 for some interpretive options, among which I stick to mine and Randy Kloetzli's, that the eternal river of time is momentarily arrested. I appreciate also Brodbeck's view that it signifies Bhīṣma's “movement back to his father's house,” but see as counterproductive his notion that Gaṅgā's “father is absent” (224). She clearly interrupts the lineage as a goddess, and her father Jahnu has not been shown to be at all relevant genealogically.

(33.) Nārada attests to a similar list of Bhīṣma's sagely sources at 12.38.7–13, when Bhīṣma gets the divine eye toward the beginning of the Rājadharmaparvan. See Hiltebeitel 2001 b.

(34.) See van Buitenen 1973, 47: “The reason of this demand is not clear: for purposes of offering?” A deer hunt will figure again in Pāṇḍu's incapacitation to satisfy his ancestors.

(35.) Śuśrūṣa (obedience, literally “liking to hear”) is the characteristic of women and Śūdras (see chapter 5 § E). Satyavatī would presumably have been raised as a Śūdra.

(36.) See Oliver, 1993, 64–65: “Kristeva takes us back to the milky way of the primary dyad. She analyzes the pulses and jolts of this primary universe made up of only mother and child” (64–65). The heavenly (ākāśa) Gaṅgā is the Milky Way. Though I know of nothing to tell us that it is milky, it may have some relation to the Milky Ocean (kṣīroda) that is above Mount Meru and has a northern shore where Brahmmā,the Gods, and Ṛṣis go, seeking the welfare of the worlds (Mbh 12.322.8; 327.39). See chapter 6 § C on this Nārāyaṇīya cosmography.

(37.) See Oliver 1993, 61. I do not push Vyāsa too hard on gynophobia, but a case can be made; see Dhand 2004, 44–45.

(38.) See Biardeau 2002, 1: 201. Tārī as “ferryboat” means “something that carries across” to the other shore, “saves.” Cf. Biardeau 219: Satyavatī's “link with the world of fish places her near the origins and she puts into the world the essential components of good order.” Biardeau takes her name to mean “She who has within her all beings.”

(39.) Bhāṣma will make his vows “while the earth-protectors were listening (śṛṇvatām bhūmipālānām)” (77c; cf. 83c, 86c). Cf. Brodbeck 2009 a, 162: Satyavatī's “adoptive father seems to know her true paternity.”

(40.) Van Buitenen 1973, 447 imagines the father to be “the king of a probably aboriginal fisher tribe” to square the two stories and make Satyavatī “obviously a fishing and ferrying wench for the taking, [who] is obviously legitimized as a king's daughter.”

(41.) Cf. Biardeau 2002, 1: 214: the fisher king seems to know his daughter's royal origins.

(42.) See van Buitenen 1973, 446: “south of the rivers Yamunā and Carmaṇvatī.”

(43.) Since the sonless have woeful destinies, this could suggest that Bhāṣma has an intimation of his having been Dyaus, “Father Sky,” or at least of having been there. Dyaus belongs to the same noun stem (div/dyu) as div[i]. My thanks to Stephanie Jamison for reinforcing this point (e-mail, June 8, 2007).

(44.) On madhyadeśa, as last cross-referenced see chapter 5 n. 34, and further chapter 7 §§ A.2, B.2.

(45.) On these points, cf. Biardeau 2002, 1: 218.

(46.) The phrase kāladharmam upeyivān is formulaic: see 1.70.46d for Yayāti; 12.31.45 for “Excretor of Gold”; Rām 1.41.9. Kāladharman is also common in the instrumental singular, as with the death of Pāṇḍu (Mbh 1.116.12d).

(47.) Later, Kṛṣṇa mentions to Yudhiṣṭhira that Bhīṣma told Duryodhana that, when he installed Vicitravīrya as king, he “became his retainer below him” (bhṛtyo bhūtvā hy adhaś caraḥ; 5.145.21d).

(48.) Unlike English “consent,” “permission,” or “concurrence,” anu + mati implies an assenting mind. Anumati is one of the four Vedic lunar phase goddesses.

(49.) See Biardeau 2002, 1: 220–21; 225 and n. 2. Kāśi might not yet be a name for Vārāṇasī, and it or its people may belong to the kingdom of Kosala from the fact that the two younger daughters are both sometimes called “Kausalyā,” woman of Kosala.

(50.) Cf. Jamison 1996, 299 n. 38: a “pun” amounting to Bhāṣma's “self-identifying announcement” that he “formally wooed” the maidens as Vicitravīrya's “proxy wooer” (220).

(51.) As Jamison 1996, 218–26 shows, the chariot-mounting is among the features by which this violent marriage mode is regulated by sanctified protocols found in the higher marriage modes.

(52.) I follow van Buitenen's parsing (1973, 228, 456) and translation of the “extremely elliptical” 1.96.8 as covering three modes (the brāhma, daiva, and “doubtfully” the prājāpatya) and also his handling of 96.10b as referring to the svayaṃvara rather than, as per Dumézil, the gāndharva mode. Cf. Brockington 2006, 36; Dumézil 1979, 37, who sees only one mode (the brāhma) in verse 8, and reaches eight by counting a northern interpolation (1.999*) that mentions the ārṣa mode, even though it is accounted for in 9ab.

(53.) G 4.6–15, B 1.20.16, and M 20–35 each mention eight. See Dumézil 1979, 35–39 on variations in other enumerations. As Heesterman 2001, 254 points out, most gṛhyasūtras mention only two types: bride-price (śulka) or unconditional bride-gift.

(54.) Jamison (1996, 305 n. 99), noting that “there seems to be no particular evidence for the general autonomy of Kṣatriya women,” rightly questions Dumézil's view (1979, 43–44) that the “conflict between the woman's right to bestow herself in a Gāndharva marriage and the legal doctrine denying her ‘independence’” is resolved by an extension to Kṣatriya women of “the ‘autonomy’ characteristic of Kṣatriyas.” But the epics also give Kṣatriya women a narrative autonomy not found in the legal tradition as wider role models. Cf. Jamison 2001 on Ṛgvedic formulary evidence for the antiquity of the svayaṃvara.

(55.) Mbh 1.67.8–14; see Hopkins 1882, 256–57; Jamison 1996, 249–50.

(56.) I think van Buitenen is right that “The present series misses the Gāndharva and possibly the Prājāpatya modes” (1973, 456), whereas Dumézil thinks it only “suppresses the prājāpatya” (1979, 38). Ā 2.5.11; 17–20 and V 1.28–35 omit the prājāpatya and paiśāca. On the prājāpatya as “the most difficult to distinguish of the four ‘gift’-marriages,” defined only by a mantra that joins the couple in practicing dharma, see Jamison 1996, 217–18, saying “probably MBh 1.96” omits it—but without clarifying. Somehow Jamison says Bhīṣma explains “the eight forms of marriage in wearisome detail” (299 n. 38).

(57.) See Dumézil 1979, 33 and n. 2 (on the svayaṃvara, “Je suis porté à y voir un dérivé ‘policé’ du mode gāndharva, propre à une société chevalresque”), 37–38, 54, 67–69, 82.

(58.) See Jamison 1996, 225 and Heesterman 2001, 256–57 on the vīryaśulka svayaṃvara: in Jamison's terms “a pseudo self-choice in which the maiden ‘chooses’ the man who accomplishes a feat of strength or manly skill set by her father.” But it is unclear how this would apply to the Kāśi princesses, one of whom had preselected her suitor with her father's agreement.

(59.) See Satyavatā's joyful consent (anumati; Mbh 5.171.4) to the marriage once the three girls arrive, just before the eldest declines.

(60.) See Jamison (1996, 224) on the chariot ride to the groom's home in an ordinary marriage, “now ensconced on a tiny island of his property.”

(61.) The eldest of the three, Ambā, later wishes she had jumped off and gone to her fiancé (5.173.4).

(62.) Regarding the “spectacular violence of the Rākṣasa marriage, Manu speaks of ‘hitting,’ ‘cleaving,’ and ‘breaking’ and of a ‘screaming,’ ‘wailing’ bride” (Heesterman 2001, 256, citing M 3.33; cf. Jamison 1996, 211, 218, 225).

(63.) In its violence to women, this “superhuman feat” may be comparable to his stopping the Gaṅgā.

(64.) The father should be irrelevant to the gāndharva mode (M 3.32; see Dumézil 1979, 34, 43, 53). But see Mbh 13.44.5 where Bhāṣma says he rejects his own preference in that mode to give the bride to the man she loves (see Dumézil 1979, 39).

(65.) Biardeau 2002, 1: 221. Moreover, her father has agreed (Ibid.), which Biardeau suspects may reflect a connection of Kāśi and Vārāṇasī with Buddhism. Biardeau relies on a widely found northern variant here; but the CE at 1.61.17 has the demon Ajaka incarnate not in Śālva but in a king named Malla.

(66.) Ghosh 2000, 42, tending to read disinterest into Satyavatā's decisions, I think gratuitously; cf. 2000, 42 on her agreement to seek a Brahmin to undertake a proxy siring.

(67.) See chapter 7 § B.3 on Kṛṣṇa's version of this story, which is told from the different angle of regeneration in out-of-the-way places.

(68.) “‘The son is his who took the hand,’ so it is decided in the Vedas” (pāṇigrāhasya tanaya iti vedeṣu niścitam; 98.5ab). That is, the son belongs to the woman's husband, dead or alive.

(69.) Translating it as “the Dark Woman” (1973, 234), van Buitenen compares it to her supposed name Kṛṣṇā, “the black one,” as reflecting her “low origins” from an “aboriginal fishing tribe” (447). But he cites no instance of Satyavatī's being called Kṛṣṇā, and there is none in Sörensen's Index of Mbh names ([1904] 1963).

(70.) Biardeau links epic usages of “Kālī” with the killing of the Buffalo Demon as part of an argument that it appears in two hymns to Durgā in Books 4 and 6 (2002, 2: 467) that are rejected by the CE as interpolated.

(71.) Manu already knows of Kālī as Bhadrakālī in connection with the householder's Bali offerings (M 3.89).

(72.) Hiltebeitel [1976] 1990, 326. See now Johnson 1998, 115–16 for a better assessment.

(73.) On Kṛtyā as parallel to Draupadī and later associated with Kālī, see Hiltebeitel 2001 a, 190–92; Türstig 1985.

(74.) Biardeau 2002, 2: 464–65: “Kālī, au visage et aux yeux rouges, … C'est Kālarātri ….” Recall n 38: Biardeau 2002, 1: 219 explains Satyavatī's name as “she who has within her all beings.” This is really an etymology of her name as based on sat, “being,” which could give us the name Satī, Śiva's first wife, whom the purāṇic Kālī becomes after she sheds her darkness (Hiltebeitel 1999c).

(75.) Vyāsa is frequently called the son of Gandhavatī, “the Fragrant,” which could, of course, recall either's smell.

(76.) The theme occurs in the Devī Māhātmyā and widely in the Purāṇas.

(77.) Only once do I find the name not referencing Satyavatā as “mother,” when Bhīṣma tells Duryodhana that he informed “Kālī Satyavatī” and various courtiers that he had given Ambā leave to go to Śālva (5.172.1–2).

(78.) On “ambā” and a suggestive discussion of Ambā as a third “mother” of Bhīṣma along with Gaṅgā and Satyavatī, see Fitzgerald 2007, 191. Along with the other mother-triads that occur in successive generations, this set gives Bhīṣma different mothers across at least two generations.

(79.) Cf. 96.45: all three maids were guṇasampannā when abducted. Even though Ambikā is a dharmacāriṇī (99.49; cf. Gāndhārī at 103.11), she is vaiguṇya—suggesting a disjunction between her dharma and her guṇas.

(80.) See M 9.57–70, 167. Vyāsa would seem to be in violation of Manu's stipulation that the man should never approach the same widow twice (60), but then he sleeps not with the widow but her Śūdra maidservant. Indeed, at M 9.167, niyoga can be a woman's svadharma! Although the Mahābhārata may imply this when it takes recourse to dharma as its justification for niyoga, it seems to spare its women characters—and particularly Kuntī—the argument that it is their svadharma. On Vidura as a karaṇa, see MW 254: a son of mixed class variously defined; a writer, scribe! In later terms, a Karṇam. One might say that in contrast to Yudhiṣṭhira, about whom one keeps hearing that it is his very nature or svabhāva to be or embody dharma (see chapters 9 § D.2.b and 10 § C, and Mbh 6.115.63–65 cited in chapter 1 § C), Vidura is a kind of moral accountant—a contrast I owe to a conversation with Gurcharan Das (October 2010). Das writes of this difference as follows: whereas Yudhiṣṭhira offers “a view of the world, based on dharma, which he explains as a universal duty of righteousness,” Vidura's “moral thinking is based on the consequences of actions rather than duty” (2009, 15–16).

(81.) Most of adhyāya 1.102 describes a time of plenty in which these four are said to have rolled along (avartata): the Kṛtayuga or Golden Age (5d), the supreme Law (dharmam uttamam; 7d), the Wheel of Dharma (dharmacakra; 12f) held by Bhīṣma, and a saying that went around in all kingdoms that said, “Of mothers of heroes (vīrasūnām) the daughters of Kāśi, of countries Kurujāṅgala, of dharma-knowing princes Bhāṣma, and of cities the Town of the Elephant” (21–22).

(82.) See van Buitenen 1973, 243, with “bastard” for pāraśavī: in Gautama's possibly earliest account of mixed classes (Olivelle 1999, xxxi), a child born to a woman three classes below the man (G 4.17). This would describe Vidura, but be problematic for a king's daughter if there is “no fifth” class, unless the king were a Brahmin.

(83.) The usual opening line ambe ambāly ambike has, as Jamison observes (1996, 274 n. 107; 304 n. 87), a precise Mahābhārata nomenclature and sequence in VS 23.18: ámbe ámbike ‘mbālike, with the same but for the initial vocative in MS 3.12.20. According to Jamison (243), “the three vocatives are taken as variants on affectionate terms for “mother,” but together they also add up to the “three Ambikās” of Rudra Tryambaka and its vṛddhi derivative, the Traiyambakahoma.”

(84.) Biardeau says, “It seems that the queen complains at being looked at in the sexual act” (2002, 1: 220). See Jamison 1996, 67: before the Adhvaryu covers the mahiṣī and the horse with a linen blanket, she “lies down beside the horse and invites it to stretch out its forefeet along with hers.” How that would help achieve a position for sexual contact is not made clear.

(85.) VS and MS apparently do not have a “crude version” with the precise Mahābhārata names.

(86.) See chapters 7 § A.1 and 12 § A on the Yuga Purāṇa's allusion to the closing episode of the Harivaṃśa where Indra animates the horse to have intercourse with King Janamejaya's wife.

(87.) Jamison 1996, 242–44 (it uses a mantra to Tryambaka found in ṚV 7.59.12 that the girl modifies to request a husband). As she points out, both rites resemble the movements of fire-circling servant maids on the Mahāvrata day of the Gavāmayana.

(88.) Jamison 1996, 304 n.94 “assumes that some version of the epic story of Ambā and her sisters already existed in early Vedic and that these girls were associated with the three Ambikās belonging to Rudra. Despite the difficulties …, I think this is more likely than assuming that a later epic poet simply made up the story of the abduction and its aftermath and named the female protagonists by plucking some designations out of the onomastic repository of Vedic ritual.” I see the alternative more positively as part of the Mbh's project of knowing Vedic allusion (see Hiltebeitel 2001 a, 42 and passim).

(89.) This rite (Rām 1.8–17) is undertaken by the potent Ṛṣi Ṛyaśṛṅga for the “continuity” (saṃtāna; 10.5; 11.2) of the aged and sonless Daśaratha's line, and is doubled by a putrīyā iṣṭi (or putrakāmeṣṭi) or “son-producing sacrifice” (14.2–3; 15.24) that yields the porridge that the three queens then share to get pregnant with portions of Viṣṇu. Cf. Goldman 1984, 60, 74–77.

(90.) Rāma proposes a Rājasūya sacrifice, but Bharata tells him a horse sacrifice is less destructive and that the Aśvamedha removes all sins and purifies (Rām 7.75.2). Lakṣmaṇa tells of Indra's Aśvamedha to recover from the Brahmanicide of killing Vṛtra, which could allude to Rāma's Brahmanicide in killing Rāvaṇa, a descendent of the Brahmin sage Pulastya.

(91.) See Hiltebeitel 2007 a, 126–35 on the Mahābhārata's postwar Aśvamedha; 2011a, chapter 9, on this episode along side prewar and postwar Aśvamedhas in both epics.

(92.) Note that Sātā, the would-be mahiṣī of this piece, is brought to the Aśvamedha not to lie with the horse but to make the disagreeable attestation to her purity that results in her earthen engulfment, but also confirms the legitimacy of her two sons.

(93.) See Jamison 1996, 84: “During the year when the horse is journeying, the king lies nightly with his favorite wife (Vāvātā), but does not have sex.” In ŚB, he thinks, “May I, by this austerity (anena tapasā), reach the end of the year successfully.” Similarly, the horse on its travels is to be kept away from mares, and if it should mate, an expiation is required.

(94.) Elsewhere in the Mahābhārata, I can find the two gods occurring alone together only with reference to Agastya as their son (12.329.38; 13.151.3), and at 14.59.14.

(95.) Vyāsa's Horse's Head story (on which see chapter 6 § B) is called a “purāṇa equal to the Veda (purāṇaṃ vedasammitam)” (12.335.7b). Recall that the Nārāyaṇīya gets to it by dipping to the outer dialogue frame so that Śaunaka can ultimately hear what Vyāsa said to Janamejaya about the Horse's Head.

(96.) I say this having once been devoted to the idea ([1976] 1990, 57–59; 1982b), and knowing that interesting connections can still be posited on its basis (see Allen 1996, 2005, 2007a, 2007 b). But the question of genre is too important (see Hiltebeitel 2001 a, 6–8), and I would agree now with Wulff 2008, 24 that “culture contact” is usually and over all the better answer. See also chapter 1 n. 28 and Hiltebeitel forthcoming-a on the improbability that an Indo-European “tribal” epic could lie at the base of the Mahābhārata.

(97.) Both epics use the term mahiṣī almost exclusively for single chief queens, and where they do otherwise it is with similar discordant notes for the real mahiṣī. See, for example, Arjuna, already married to Draupadī, telling Kṛṣṇa his good fortune that Subhadrā will be his mahiṣī (Mbh 1.211.19); Mantharā telling Kaikeyī that she (rather than Kausalyā) is Daśaratha's mahiṣī (Rām 2.7.19); Sumantra, as her messenger, calling her the mahiṣī when he tells Rāma that she and Daśaratha want to see him (14.11); Rāvaṇa inviting Sītā to be his mahiṣī, with the odd qualifier that she will be his agramahiṣī, “primary chief queen” (3.4.24; 5.18.16)—a term used nowhere else in the epics’ Critical Editions, although the Southern Recension finally makes Śakuntalā the agramahiṣī of Duṣyanta (Hiltebeitel forthcoming-d), and Mandodarā would seem to have that position among the wives of Rāvaṇa. On the other hand, in two plural usages it means mainly “woman” (Mbh 1.187.26; Rām 2.36.7; see 36.1).

(98.) Jamison 241, 303 n.76. Although “Tryambaka” may—as usually translated—refer in the epic to Śiva's having three eyes (see Scheuer 1982, 237–36, 255–56), the matter is uncertain (Hopkins [1915] 1969, 220), and in the one case where the Mbh gives an etymology (Vyāsa is telling Arjuna how Śiva preceded his chariot in battle), it refers to his having three goddesses: “And since the Lord of the universe possesses three goddesses—Sky, Waters, and Earth—he is remembered as Tryambaka” (7.173.89).

(99.) See Jamison 1996, 243. The Mahābhārata never mentions the pati-vedana or the Traiyambaka Homa. Generally, the epics seem to overlook the Caturmāsyāni rites. It does mention a Traiyambaka Bali (7.56.1–4, esp. 3d): according to Scheuer 1982, 255–63, it is probably offered nightly throughout the war on Arjuna's behalf by his servants, after which Arjuna sleeps on the ground surrounded by weapons. Cf. also Scheuer 258 n. 23 on the Pāṇḍavas’ offering (upahāra, bali) to Rudra Tryambaka (14.8.23–24; 64.1–8) after a night's fasting to get hold of the wealth needed to perform their postwar Aśvamedha.

(100.) The Mahābhārata does know Śiva as “Ambikā's husband” (ambikabhartre), yet a brahmacārin (7.57.53).

(101.) Biardeau, however, proposes that Kāśi would index Buddhism “where the Buddha preached his first sermon”; the two sisters’ distaste for Vyāsa would reflect a Buddhist preference for monks, who would not be drawn into such a scene (2002, 1: 221–22). This seems strained, but, if so, reference to Śiva could cover, as elsewhere, for Buddhism.

(102.) Allowing for overlap and not counting descriptions of their sons at birth, from 103.9 to its end at 119.12 our skein gives 195 verses to Kuntī, eighty-five to Mādrī, and thirty-six to Gāndhārī.

(103.) See Mbh 1.13.35–14, 18–23, 30.11–35, and 49.3–16. Kadrū and Vinatā's parallels with Gāndhārī and Kuntī, widely noted, include rivalry, use of pots for gestation, and the servitude of the one's sons to the other's. Although the parallels crisscross (most notably, it is Vinatā who has a long and aborted gestation), Gāndhārī has more in common with Kadrū (mother of a thousand snakes) and Kuntī with Vinatā (mother of two birds). But the key theme of a mother (Kadrū) cursing her sons is unparalleled, unless one reads Gāndhārī's boon as a parallel, or perhaps Kuntī's abandonment of Karṇa.

(104.) Karṇa mocks the low conduct of Madras at Mbh 8.27–30, once mentioning Gāndhāras, Madras, and Bāhlīkas (Bactrians) together (27.55cd), particularly deriding Madra women. See Biardeau 2002, 2: 313–15 and n. 11 on the Vāhīka-Bāhlīkas, inclusive of Madras, as (by Karṇa's etymological tale) those from outside Āryavarta, descended not from Prajāpati but from two “ghastly” Piśācas who now inhabit the former Vedic heartland of the Punjab (cf. Hiltebeitel [1976] 1990, 272–73). This Āryavarta corresponds, in fact, to the Buddhist majhimadesa described in chapter 7 § B.2.

(105.) The Yādavas descend from Yadu, son of the lunar dynasty ancestor Yayāti, who cursed Yadu and his descendants to have “no share of royalty” (Mbh 1.79.7). A Northern (N) passage (1.1129*) compounds the story at this point, having Kuntibhoja arrange the svayaṃvara when no king had asked for Kuntā's hand!

(106.) Cf. Brockington 2006, 38. We see again in this formation from √vid that a svayaṃvara could be a kind of “husband-finding” (pati-vedana).

(107.) A lengthy N passage (Mbh 1, App. I, No. 61) has Śalya, already king of Madra, stipulate that his sister's purchase is a matter, whether good (sādhu) or not, of what his family (kule; lines 24–25) has always done.

(108.) See Hara 1974, 304–5; Jamison 1996, 235; and Heesterman 2001, 255 n. 31, on the Kṣatriya's “warrior ethos, which forbids him to accept, let alone ask for gifts,” as an underlying explanation for the Rākṣasa mode of marriage.

(109.) A hundred sons is not inherently a bad boon. Sāvitrī gets Yama's boon of a hundred sons to trick him into keeping her husband alive, and the same number for her husband's parents (Mbh 3.381.44–58). Here the number seems to signal the spread of a martial clan (as it does where the Kauravas are worshiped as clan deities in the Tamil cult mentioned in § B above).

(110.) Perundevi Srinivasan is gathering variants of this story. I thank her for her thoughts on the banana leaf.

(111.) Of course, whereas Gaṅgā “immersed” (amajjayat; 92.44d) her first seven sons as they wished so that they could return to heaven, Kuntī “abandoned,” “cast,” or “poured out” (utsasarja; 104.13c) Karṇa to a life of earthly resentments.

(112.) Their adoptions are curious. Both have their premarital affairs apart from their “real” parents (Satyavatī's mother being an Apsaras-turned-fish). They also attenuate their already loose ties to royal lines. And Kuntī's foreshadows that of Karṇa.

(113.) Van Buitenen captures this sense in translating what Kuntā says when she tells Karṇa shortly before the war how she conceived him: “Then, out of curiosity and childishness, … I, being just a girl (kanyā satī), made the Sun god come to me” (5.142.21, 23–25; van Buitenen 1978, 452). Van Buitenen 1973, 241 missed the opportunity to translate kanyā satī in this fashion at 1.104.8cd. Cf. Oliver 1993, 54 on Kristeva's view of “virginity.”

(114.) The causative kāmayate (√kam) reads either way; √kan means “to be satisfied or pleased; to agree to, to accept with satisfaction; to please, be liked or wished for.”

(115.) A name for Duryodhana: Biardeau (2002, 1: 886 n. 4, 900) suggests, “Easy to combat.”

(116.) Śvaśura (plural): usually “fathers-in-law” (see van Buitenen 1978, 370), but “also applied to a maternal uncle and any venerable person” (MW 1105). This is surely Kuntī's meaning. She is complaining about her father's deal with a maternal uncle, and had no fathers-in-law yet (she calls Vyāsa her śvaśura at 15.38.1a).

(117.) She repeats much of 1.104.4–6 at 1.113.31–33, condensing “gods and guests” to the essential “guests,” and speaks of her “sacrifices” (yajñair) rather then “efforts” (yatnair).

(118.) Cf. Biardeau 202, 1: 230: to wish for sons like their father is “the wish of every femme dharmique, but is it wise in this case?” As with Gāndhārī's utterance when she puts on her blindfold, her everyday wifely sentiments sound foreboding. Brodbeck 2009 a, 168–69, however, implies that Śiva's and Vyāsa's boons to Gāndhārī are the same.

(119.) In a widely found adhyāya (omitted only in four N mss., including the important Śāradā one [on which see Sukthankar 1933, lxv]), Gāndhārī interrupts Vyāsa to ask for a daughter (1 Appendix I, no. 63, after 1.107; Ganguli [1884–96] 1970, 1: 260–61). I stress this daughter's gratis appearance because it parallels Draupadī's unasked-for birth (see Hiltebeitel 2001 a, 187), which—in Draupadī's case—makes it part of the divine plan, for Draupadī is born specifically to do surakārya, “the work of the gods” (Mbh 1.155.45; see chapters 10, 12). I would not agree that Duḥśālā “is apparently no part of Vyāsa's plans for the fleshball” (Brodbeck 2009 a, 170).

(120.) As Brodbeck has noted (personal communication), and, though I take the possibility that these allusions are to Karṇa from him, he credits them (Brodbeck 2009 a, 170 n. 12) to Bowles 2008, 1: 44 n. 5.

(121.) Mbh 1.107.13–23. Biardeau 2002, 1: 233 finds it tempting to link Gāndhārī's name Saubaleyī here with the bali offerings (one of the five mahāyajñas) made to inferior divinities and thrown on the ground.

(122.) As Sukthankar 1933, 474–75 shows, the Southern Recension's “entirely different” handling of 1.106.11–114.15ff. makes the Pāṇḍava and Kaurava princes’ births follow one temporal line. Beginning with an excision of 109.1–4 where Janamajaya asks Vaiśaṃpāyana to return to the story of the Pāṇḍavas’ partial divine incarnations, it does not divide Pāṇḍu's story by Gāndhārī's, and continues with Pāṇḍu's from 109.5 through Yudhiṣṭhira's birth before it gets to Gāndhārī's pregnancy. Meanwhile, as S joins 107.8 to 111.12 to make Pāṇḍu's story continuous, it drops the first reference to Gāndhārī's learning of Kuntī's sunlike son (107.10), and once it has recounted her pregnancy, supplies six lines (1135*) to take one back to Pāṇḍu's plight. Gāndhārī's remaining mention of Kuntī's sunlike child (107.15) is thus deferred to follow Yudhiṣṭhira's birth (114.1–7), eliminating any ambiguity as to who he is and diminishing her rivalry with Kuntī. The extensive revision is another example of S's replacement of ambiguity and literary experimentation with a flatter and more easily communicated orderliness. See Hiltebeitel 2006 a, 252–53; forthcoming-a; forthcoming-d; Mahadevan 2008.

(123.) Cf. Brodbeck 2009 a, 175, making do with having Pāṇḍu “imitate aspects” of only two fathers, Vyāsa and Vicitravīrya. The additional two would be Bhīṣma and Citrāṅgada.

(124.) See 102.12; 105.19–23 and 26: he reclaims kingdoms (rāṣṭrāṇi) the Kurus had lost, and takes captives from them.

(125.) Mbh 1.106.6–11. Biardeau 2002, 1: 235 says Pāṇḍu is now living an irresponsible life, having given up his responsibilities in the capital.

(126.) See Biardeau 2002, 1: 226 n. 3: This Ṛṣi “is not a model of asceticism,” his name meaning “Who has some self-mastery,” “More or less master of himself.”

(127.) Vyāsa sires his first son Śuka by ejaculating after seeing a nymph transformed into a female parrot (12.311.1–10; see Hiltebeitel 2001 a, 286–87), and also interrupts his tapas to sire Dhṛtarāṣtra, Pāṇḍu, and Vidura.

(128.) Mbh 1.109.28–30. The description suggests that this double Liebestod under Yama's sway could evoke Yama's connections with possession; see chapter 8 § D.2.b.

(129.) Mbh 12.9.12–28; see Klaes 1975, 113, 129 n. 10; Fitzgerald 2002, 670 n. to Mbh 12.308.36. Yudhiṣṭhira first projects himself as a solitary vānaprastha (12.9.1–11; Fitzgerald 2004 a, 685); but then, as his begging shows, he sees himself as a renunciant. The parallels are as follows: 1.110.7–9 ≈ 12.9.12–14; 1.110.10 ≈ 12.9.17; 1.110.11 ≈ 12.9.16 (with a verse-order reversal); 1.110.12ab ≈ 12.9.23ab; 1.110.14 = 12.9.25 excepting a verb; 1.110.15 = 12.9.24 except for a word-reversal; and 1.110.16–18 ≈ 12.9.26–28, with 16ab = 26ab, 17b = 27b, and 18cd = 28cd about seeking “the Law (dharman) of the wind.” Most likely Pāṇḍu's is the trunk version while Yudhiṣṭhira is given some intermittent fancies (e.g., he will be the last after-hours guest on his begging rounds [12.9.22]) before each concludes on a distinctive note: doggish lust in Pāṇḍu's case, perhaps his way of recalling the mating deer he just shot; and massive guilt in Yudhiṣṭhira's.

(130.) Compare the situation with Yudhiṣṭhira mentioned in the previous note. He starts out envisioning the vānaprastha mode and then, perhaps aware that his wife and brothers have had their fill of the forest, raises the stakes to a renunciatory mode.

(131.) On Gandhamādana, prominent in many Mbh episodes and in the Nārāyaṇīya, see Hiltebeitel forthcoming-a; it will also be the retreat of the last Arhat in the Tibetan Candragarbha Sūtra (chapter 7 § B.3.b).

(132.) Its only other mention occurs as Arjuna and Kṛṣṇa pass over it—along with the Śaryāti Forest and the holy places of the Horse's Head and Ātharvaṇa (7.57.28)—on the way to asking Śiva for the Pāśupata weapon. The Rāmāyaṇa seems to have relocated Śataśṛṅga to the west, “where the Sindhu river meets the ocean” (4.41.12)!

(133.) See Biardeau 2002, 1: 412–13 (mentioning it in this category along with the Rāma- and Sāvitrī-upākhyānas). On upākhyānas in the Mahābhārata, see chapter 9 § B; Hiltebeitel 2005 a.

(134.) Damayantī's three-day journey north from a hundred-peaked mountain before the Ṛṣis appear to her may echo this route.

(135.) Mbh 1.1171*. See Biardeau 2002, 1: 227; Ganguli [1884–96] 1970, 1: 267. This is similar to the one at the beginning of our skein where King Mahābhiṣa offends Brahmā by gazing at Gaṅgā. Note the interpolation there as well.

(136.) Pāṇḍu improvises to combine the three debts and the five mahāyajñas (see chapter 5 § A), leaving out the offering to Bhūtas and introducing the novel note that one performs the offering or debt to men by noncruelty (ānṛśaṃsyena mānavān; 111.14d), which is what Kiṃdama said Pāṇḍu lacked when he shot a mating deer (109.18d), as Pāṇḍu soon remembers (111.26). On this virtue, see further chapter 9.

(137.) Dhand 2004, 41 misses that the ascetics foresee the outcome and must know Kuntā's boon, saying that these verses “propose the timely solution of niyoga.” But yes, Pāṇḍu “seizes on” this implication “with enthusiasm.”

(138.) The Southern Recension, both flat and arch here, cribs the gist of these lines from what Pāṇḍu soon tells Kuntī at 111.30c–31b.

(139.) In the fullest account of her impregnation by Sūrya, Kuntī asks him, “But how can I make a gift of myself that is surely not to be made?’ … But if you think that this is dharma, best of burners, I will do your wish without being given away by relatives. Having made you the gift of myself (ātmapradānam), I shall remain virtuous (satī)” (3.291.5cd, 10–11ab).

(140.) See 5.37.26, where, Vidura tells Dhṛtarāṣṭra: “A sensible man will never feel free to enter a stranger's house at the wrong hour, nor stand at night concealed at a crossroads, nor solicit a woman of baronial rank.”

(141.) The epics use it only for Vyuṣitāśva and Vicitravīrya (1.96.57d; 5.145.23f)—with the same verb samapadyata.

(142.) Draupadī is the “dear” mahiṣī of Yudhiṣṭhira (Mbh 4.15.31, 16.12; 10.11.17) and of all five Pāṇḍavas (4.20.19, 5.80.22); so is Sītā to Rāma (Rām 4.48.18, 5.12.43–44, 13.46; Mbh 4.20.10 according to Draupadā); Śakuntalā finally to Duṣyanta (1.69.43); Indrāṇī to Indra (5.11.13; 12.22, while coveted by Nahuṣa); and Tārā was dear to Vālin according to their son Aṅgada (Rām 4.54.3). Both Draupadā (Mbh 4.19.10; 20.10, 19; 5.80.9) and Sītā (Rām 6.38.3) speak of the title with pride, and Vālmākā uses it even when he welcomes the banished Sātā to his ashram (7.48.8).

(143.) See MW 1040–41. As a neuter noun, vyuṣita, “daybreak,” derives from vi-2.√vas, “to shine forth” (from vi-uṣ, “to dawn”). But as an adjective from vi-5.√vas, “to abide, dwell, live,” it can mean “inhabited by” in compounds.

(144.) 1.113.2 as cited. I see nothing to suggest that Pāṇḍu “dismisses Kuntī's narrative as naïve” (Dhand 2004, 42).

(145.) See chapter 4, n. 51; Olivelle 2005 b, 13–51. Olivelle does not discuss this Mbh story (which makes his “spoiled brat” additionally into a mommy's boy) in that piece, but does in connection with Manu's injunction (M 9.5–12) that women should be guarded (2005b, 257 and n. 32)—though mistaking the free partnering of Śvetaketu's mother for the rape of his wife. See now Black 2011, 137, 146–47.

(146.) See Olivelle 1999, 363, n. to Ā 1.19.15; Fitzgerald 2004 a, 695, n. 50 to Mbh 12.15.55.

(147.) To be sure, she is a special case: “Kuntā was fortunate: unlike Ambikā and Ambālikā, she had a magical formula …. Ordinary women did not possess special mantras to call on beautiful, well-scented, well-behaved gods” (Dhand 2004, 42–43).

(148.) It occurs only here. See Scheuer 1982, 74–75 with discussion and bibliography, notably Katre 1943, 122, who, drawing on the commentaries of Nīlakaṇṭha and Devabodha, says that “Dharma stands for Āpaddharma and upaniṣad for mantragrāma,” and takes mantra-upaniṣad as a parallel, which he translates as “secret mantra.”

(149.) To conceive Yudhiṣṭhira from Dharma, Kuntī “muttered what was to be muttered according to rule (jajāpa japyaṃ vidhivat)” (114.2cd), thus intoning it as a softly muttered Vedic prayer (japa).

(150.) For other usages, see 1.53.4; 3.290.1 and 2 (also in the account of Karṇa's birth); cf. Rām 1.21.10c; 1.26.21d.

(151.) See 1.115.23, where Kuntā refuses Pāṇḍu's request to let Mādrī have another use of her mantra: “I said to her,’ For this once,’ and she got two! I was deceived. I fear that she will best me. That is the way with women!”

(152.) Paraṃcāriṇī, one who would now have a reputation for “moving with others.”

(153.) See chapters 4 § C.1, n. 151; C.3.a and 6 § B on Mbh 12.200.35–37, where maithunadharma begins in the Dvapāra yuga.

(154.) Dārān, masculine plural, means both wives here and below, although only Kuntā is now alive.

(155.) One is reminded of Śaunaka, “the Mahābhārata's anchorman” (see chapter 6 n. 29). Also, if the compound maharṣimatam is read as “the great Ṛṣi's mind,” one could take it as echoing passages that call the Mahābhārata Vyāsa's “thought entire (kṛtsnaṃ matam)” (1.1.23; 55.2; Hiltebeitel 2001 a, 12; cf. 54), in which case this Ṛṣi who might remind us of Śaunaka could be cryptically tuned into the thought of the author.

(156.) The short account of this scene in the epic's prologue has the Ṛṣis speak together, before disappearing, telling only that the boys are the sons of Pāṇḍu. But after this, “Some said, ‘They are not his.’ Others, ‘They are his.’ Others again, ‘How can they be his when Pāṇḍu has been dead long since?’” But all agreed, “They must be bid welcome!” (Mbh 1.1.72–74). See Brodbeck 2009 a, 172, 175.

(157.) Van Buitenen 1973, 262 for gandharvanagara, “a city of the Gandharvas,” the heavenly musicians.

(158.) On Śakuntalā, see Biardeau 1979; Hiltebeitel forthcoming-d. On Tapatā, note that Jamison 1996, 305 n. 98 remarks on how she “dutifully cites man's views on women's lack of independence” at Mbh 1.161.14.