Printer Friendly

The Dharmasastric debate on widow-burning.

Although scholars have long devoted substantial attention to analyzing and discussing the Indian custom of widow-burning or sati as it is popularly known, they have made surprisingly little use of a body of writings that deals in a very explicit fashion with the ethical and theological underpinnings of this practice. These woefully neglected writings to which I refer are those of the Dharmasastra tradition, the branch of Brahmanical scholarship most directly concerned with delineating the bounds of right behavior and, therefore, a decidedly unsurprising place to find discussions of sati. (1) Significantly, despite their general neglect, Dharmasastric writings on widow-burning should be of considerable interest to anyone interested in the history of this practice, for they provide clear testimony of a long, intricate, and pan-Indian debate on its very validity. Thus, they can give valuable insights into the intellectual justifications offered for and against the custom of sati in pre-modern India. Equally importantly, they allow us to track major changes in orthodox Brahmanical opinion on the subject. It is the purpose of this article to give a detailed account of this crucial Dharmasastric debate, which, although erudite and esoteric in appearance, shows the ways in which the broader Brahmanical community gradually shifted its outlook toward sati over time from widespread rejection to unqualified acceptance.

Before proceeding, I will briefly comment on my use of terminology. In the preceding paragraph, I have used the Sanskrit word sati and the English word widow-burning to denote a custom, historically practiced in parts of India, whereby a woman would commit self-immolation after the death of her husband, according to ideal rules, by voluntarily mounting his funeral pyre. These words, however, are not those used in the Sanskrit texts under discussion. Instead, these texts use a number of different words for essentially the same practice: anugamana ('going after'), sahagamana ('going with'), anumarana ('dying after'), sahamarana ('dying with'), and anvarohana ('mounting after'). In order to remain faithful to the Sanskrit and avoid introducing undue connotations or confusion, I will henceforth use the term sahagamana in preference to sati, widow-burning, and the other equivalent Sanskrit words.

As is fairly well known, there is no mention of sahagamana whatsoever in either Vedic literature or any of the early Dharmasutras or Dharmasastras. (2) In fact, of all the extant smrtis on dharma, only those two that are very likely the youngest, namely, the Vaisnava Dharmasastra (also known as Visnu Smrti/Dharmasutra [ViDh]) and the Parasara Smrti, make any mention of sahagamana. The relevant passages of these works read as follows: (3)
  mrte bhartari brahmacaryam tadanvarohanam va | (ViDh 25.14)
  When a woman's husband has died, she should either practice ascetic
  celibacy or ascend (the funeral pyre) after him.
mrte bhartari ya nari brahmacaryavrate sthita |
sa mrta labhate svargam yatha te brahmacarinah ||
tisrah kotyo 'rdhakoti ca yani romani manuse |
tavatkalam vaset svarge bhartaram yanugacchati ||
vyalagrahi yatha vyalam balad uddharate bilat ||
evam stri patim uddhrtya tenaiva saha modate || (Parasara Smrti
  If a woman adheres to a vow of ascetic celibacy (brahntacarya) after
  her husband has died, then when she dies, she obtains heaven, just
  like those who were celibate. Further, three and a half krores or
  however many hairs are on a human body--for that long a time (in
  years) a woman who follows her husband (in death) shall dwell in
  heaven. And just as a snake-catcher forcefully lifts up a snake out
  of its hole, so does this woman lift up her husband and then rejoices
  together with him.

From these passages it is clear that the authors of these two quite late Dharmasastras regarded sahagamana as a meritorious alternative to ascetic celibacy (brahmacarya) for at least some widows. Moreover, both passages provide some slight evidence that their authors regarded sahagamana as the superior of these two alternatives. In the case of Parasara, this evidence consists of the fact that the otherworldly rewards of sahagamana are elaborated in far greater detail than those of ascetic celibacy, which would seem to imply that sahagamana is the more meritorious of the two options. (4) In the case of Visnu, this evidence consists of the fact that the text (25.14) lists sahagamana second as an alternative marked by the particle va ('or'). At first glance, this by itself might appear to tell us nothing. However, Kiparsky (1979) has convincingly demonstrated that in the Astadhyayi Panini uses the word va to mark the more preferable of two alternatives. In other words, va does not simply mean 'or' for Panini; it means 'or preferably'. Hence, if Visnu is here following Panini's particular usage of va, then he must consider sahagamana to be preferable to ascetic celibacy as an option for widows. Unfortunately, however, it is unclear whether or not he is in fact following Panini's precise usage here and, as a result, the issue must remain unresolved.

In any case, aside from these two passages from extant smrtis, an examination of the commentarial literature reveals a number of passages advocating sahagamana ascribed to authors of Dharmasastras that are no longer extant. Such authors include Angiras, Usanas, Paithinasi, Vyasa, Harita, and Brhaspati. (5) Taken together with the previous citations from Visnu (25.14) and Parasara (4.29-31), these passages constitute the entirety of the Dharmasastric injunctions regarding sahagamana. However, a complete account of the scriptural injunctions related to sahagamana must also include several passages from the Puranas and Sanskrit epics. (6)

Considering the complete absence of any mention of sahagamana in both Vedic literature and the earliest works of the Dharmasastra tradition, it seems reasonable to conclude that this practice first gained enough popularity within Brahmanical culture to warrant mention at approximately the time when Visnu, Parasara, Angiras, Usanas, etc. composed their works on dharma. Consequently, in order to establish the period during which sahagamana first became a recognized custom within orthodox Brahmanical culture, it is necessary to establish the provenances of these works. Unfortunately, however, the available evidence does not allow us to do this within very narrow limits.

Recently, Olivelle (2007; 2009: 4-15) has put forth a convincing argument that the Vaisnava Dharmasastra was written in Kashmir sometime between the sixth and ninth centuries C.E. It is, therefore, fairly certain that sahagamana was a Brahmanical custom current in at least parts of Kashmir during the second half of the first millennium. This point, however, is far less illuminating than one might hope, for Enrica Garzilli (1997) has put forth another fairly compelling case that as early as the first century B.C.E., sahagamana was a recognized practice in Kashmir among members of the Kathaka school of the Black Yajurveda, to which, significantly, the author of the Vaisnava Dharmasastra almost certainly belonged (Olivelle 2009: 5-7). Hence, the advocacy of sahagamana in Visnu may not stem from the fact that this was a relatively new custom that had become popular in the author's day, but instead reflect the author's rather unique cultural and regional background.

Consequently, one is forced to rely upon the other Dharmasastras that prescribe sahagamana in order to establish when the practice first gained substantial and widespread popularity within Brahmanical culture; and regarding the dates and provenances of these texts, almost nothing is known. Perhaps the best that can be said is that they were composed somewhere in India probably during the second half of the first millennium C.E. The reason for this broad dating is that, based upon a comparative examination of their contents, scholars generally consider these works to postdate the Vaisnava Dharmasastra and certainly not to predate it by many centuries. (7) In any case, it is clear from citations found in later commentaries that by the twelfth century these works and the practice of sahagamana were widely known to orthodox Brahmins throughout India. (8)

Importantly, however, not all smrti passages that mention sahagamana endorse the practice, for, beginning with Vijnanesvara (c. 1076-1127), authors working within the Dharmasastra tradition cite a number of authoritative scriptures that explicitly prohibit Brahmin widows from performing sahagamana. The Two most frequently cited of these are ascribed to Paithlnasi and Angiras. (9) They read as follows:
mrtanugamanam nasti brahmanya brahmasasanat |
itaresam tu varnanam stridharmo 'yam parah smrtah ||
  Due to Vedic injunction a Brahmin woman should not follow her husband
  in death, but for the other social classes tradition holds this to he
  the supreme Law of Women.
ya stri brahmanajatiya mrtam patim anuvrajet |
sa svargam atmaghatena natmanam na patim nayet ||
  When a woman of Brahmin caste follows her husband in death, by
  killing herself she leads neither herself nor her husband to heaven.

Although all medieval exegetes who cite these passages and others like them manage to greatly reduce their proscriptive scope, to a neutral reader their intention is clear: they issue a general prohibition against sahagamana in the case of Brahmin widows. Hence, they inform us that while their authors, who were undoubtedly Brahmins, had no specific objection to non-Brahmin widows performing sahagamana, they strongly objected to this practice among widows of their own social class. This, in turn, may be taken to suggest that sahagamana was well established among certain other social groups at the time these scriptures were composed, but still relatively new and, therefore, controversial amongst orthodox Brahmins. Thus, these smrti passages add further support to the thesis compellingly put forth by Kane (vol. 2 [1962]: 626-27) that when Brahmins first developed the practice of sahagamana, they did so by adopting it from the royal/warrior class. At the very least these scriptures tell us that some authoritative Brahmin men felt deeply apprehensive about their own widows performing sahagamana.

With this in mind, it should perhaps come as no surprise that Medhatithi, who probably lived in ninth- or tenth-century Kashmir (10) and may have written the earliest extant commentary to discuss the issue, (11) staunchly opposes the practice. The relevant passage of his work reads as follows:
  pumvat strinam api pratisiddha atmatyagah | yad apy angirase patim
  anumriyeran ity uktam tad api nityavad avasyam kartavyam |
  phalastutis tatrasti | phalakamayas cadhikare syenatulyata | tathaiva
  syenena himsyad bhutanity adhikarasyatipravrddhataradvesandhataya
  satyam api pravrttau na dharmatvam | evam ihapy
  atipravrddhaphalabhilasayah saty api pratisedhe tadatikramena
  marane pravrttyupapatter na sastriyatvam | ato 'sty eva patim
  anumarane 'pi striyah pratisedhah | kim cu tasmad u ha na purayusah
  preyad iti pratyaksasrutivirodhe smrtir apy esa anyartha sakyate
  kalpayitum yatha vedam adhitya snayad ity adhyayananantaram
  akrtarthavabodhasya snanasmaranam | (Medhatithi 5.155 (12))
  (Author:) Suicide is prohibited for women just as it is for men.
  (Objection:) Yet one should also certainly carry out like a
  mandatory duty the following statement from the Dharmasastra of
  Angiras: "(Women) should follow their husbands in death." (Reply:)
  This statement praises the reward of performing this act. And since
  it qualifies a woman who desires that reward to perform it, the case
  is analogous to the syena sacrifice, indeed, even when a person who
  is qualified to "kill living beings by means of the syena sacrifice"
  actually engages in that rite when blinded by excessive hatred, it is
  not in accordance with dharma. It is just so here as well: When a
  woman who has an excessive desire for the result dies despite the
  fact that there is a prohibition against this and she is acting in
  violation of it, her reason is not sanctioned by the s'dstras. Hence,
  a woman is certainly prohibited also from following her husband in
  death. Moreover, since it is in contradiction with the perceived
  Vedic scripture "Therefore, one should not depart before one's
  natural lifespan"(Satapatha Brdhmana, one can construe this
  smrti text to have a different meaning. In this regard, it is just
  like the smrti "Having recited the Veda, one should bathe," which
  indicates that a person who has not learned the Veda's meaning should
  bathe after simply reciting it. (13)

Here Medhatithi puts forth two different arguments against sahagamana. First, he argues that the practice is adharmic, because it is analogous to the syena sacrifice, a Vedic ritual whose explicit result is the death of the sacrificer's enemies. According to the traditional interpretation given by Sahara in his commentary on Purvamimamsa Sutra 1.1.2, the performance of the syena sacrifice is not in conformity with dharma, since there is a general prohibition against violence. The Veda simply slates that if a person wants to kill his enemies, the syena sacrifice is one means of accomplishing his goal. It does not, however, enjoin the killing of one's enemies, so there is no specific injunction that would override the general prohibition against violence. Using the analogy of this sacrifice, Medhatithi argues that smrtis like that of Angiras do not actually enjoin sahagamana, because they explicitly mention its result, namely, heaven. They only state that if a widow wants to be reborn in heaven, sahagamana is one possible means. Thus, as in the case of the syena sacrifice, the general prohibition against violence still applies. Medhatithi's second argument against sahagamana is considerably simpler: those smrtis that evidently prescribe sahagamana as a means of attaining heaven are in direct contradiction with those statements in the extant, perceivable Veda that prohibit suicide. And since it is an accepted exegetical principle that sruti--the extant, perceivable Veda--is of greater authority than smrti, the various smrti statements that appear to advocate sahagamana can be construed to have a different meaning. On these two grounds, the earliest commentarial work within the Dharmasastra tradition to address the topic of sahagamana takes a position that is completely opposed to the practice.

However, beginning with the Mitaksara, Vijnanesvara's early twelfth-century commentary on the Yajnavalkya Smrti, this position starts to weaken in several distinct ways. To begin with, Vijnanesvara attempts to reconcile those smrtis that generally enjoin sahagamana with those that apparently prohibit this practice in the case of Brahmin widows. And importantly, he does this in a way that allows Brahmin women to perform sahagamana with little restriction. His statements on this matter read as follows:
yani cu brahmanyanugamananisedhaparani vakyani
mrtanugamanam nasti brahmanya brahmasasanat |
itaresu tu varnesu tapha paramam ucyate |
jivanti taddhitam kuryan maranad atmaghatini ||
ya stri brahmanajatiya mrtam patim anuvrajet |
sa svargam atmaghatena natmanam na patim nayet ||
ityevamadini tani prthakcityadhirohanavisayani
prthakcitim samaruhya na vipra gantum arhati |
iti visesasmaranat | (Mitaksara 1.86)

Those statements, such as the following, which prohibit Brahmin women from performing sahagamana, apply only to the ascending of separate funeral pyres:
  Due to Vedic injunction, a Brahmin woman should not follow her
  husband in death, but amongst the other social classes, this is said
  to be the supreme austerity. Living, she (= a Brahmin woman) should
  do what is beneficial to him (= her husband). By dying, she commits
  When a woman of Brahmin caste follows her husband in death, by
  killing herself she leads neither herself nor her husband 10 heaven.
  This is so due to the following specific rule given in this smrti:
  A Brahmin woman ought not to depart by ascending a separate pyre.

Here Vijnanesvara argues that those smrtis that apparently prohibit Brahmin women from performing sahagamana really only prohibit them from performing it on different funeral pyres from those of their husbands. In support of this interpretation, he cites another scriptural passage that appears to express precisely this idea. Via this argument, Vijnanesvara effectively does away with any objections aimed specifically at the right of Brahmin widows to perform sahagamana.

It is noteworthy that Vijnanesvara's line of argumentation on this point appears to have exerted considerable influence on a large number of later exegetes. For instance, consider the following statements from Apararka's commentary on the Yajnavalkya Smrti (c. 1110-90) and the Madanaparijata of Madanapala (c. 1350-1400): (14)
nanu ca brahmanya anvarohanapratisedham smaranti yatha tavat
mrtanugamanam nasti brahmanya brahmasascanat |
itaresam varnanam sridharmo 'yam parah smrtah ||


satyam evam smaranti kim tu smaranantarenaitesam visayo vyavasthapyate
prthakcitim samaruhya na vipra gantum arhati |
anyasam caiva nairnam stridharmo 'yam parah smrtah ||
tatas ca brahmanyanugamananisedhah prthakcitisamarohanavisayah |
(Apararka 1.87)

(Objection:) Isn't it the case that certain smrtis prohibit a Brahmin woman from performing sahagamana? For instance, Paithinasi states:
  Due to Vedic injunction, a Brahmin woman should not follow her
  husband in death, but for the other social classes, tradition holds
  this to be the supreme Law of Women.

[Here are cited four more verses with essentially identical meanings ascribed to Paithinasi, Viraj, Angiras, and Vyaghrapad.]

(Author:) It is true. Certain smrtis say this. However, their actual sphere of applicability is established by other smrtis. For instance, Usanas states:
  A Brahmin woman ought not to depart by ascending a separate pyre, yet
  for other women, tradition holds this to be the supreme Law of

Therefore, the prohibition against Brahmin women performing sahagamana refers only to ascending separate funeral pyres.
yani ca brahmanyanugamananisedhaparani paithinasyangirahprabhrtinam
mrtanugamanam nasti brahmanya brahmasasnat ||
ya stri brahmanajatiya mrtam patim anuvrajet |
sa svargam atmaghatena natmanam na patim nayet ||
ityevamadini tani prthakcitivisayani
prthakcitim samaruhya na vipra gantum arhati |
anyasam caiva narinam stridharmo 'yam parah smrtah ||
ity usanaso visesasmaranat | (Madanaparijata, pp. 197-98)

As for the following statements of Paithinasi, Angiras, etc., which prohibit Brahmin women from performing sahagamana, these apply only to separate funeral pyres:
  Due to Vedic injunction, a Brahmin woman should not follow her
  husband in death. When a woman of Brahmin caste follows her husband
  in death, by killing herself she leads neither herself nor her
  husband to heaven.

This is so due to the following specific rule laid out in the smrti of Usanas:
  A Brahmin woman ought not to depart by ascending a separate pyre, yet
  for other women tradition holds this to be the supreme Law of Women.

Here both authors clearly adopt Vijnanesvara's exact thinking. This is not only apparent in their conclusions: those smrtis which seemingly issue a general prohibition against Brahmin women performing sahagamana really prohibit them merely from ascending separate funeral pyres. It is also apparent in their supporting evidence, which comprises the very same smrti cited by Vijnanesvara, only here explicitly ascribed to Usanas. From the presence of such statements and the complete absence of contrary ones, it is evident that later commentators within the Dharmasastra tradition differ markedly from the authors of some earlier smrtis in that they feel no particular reservations about Brahmin widows performing sahagamana.

Moreover, some later commentators even go so far as to get rid of Vijnanesvara's restriction that Brahmin widows ascend the same funeral pyre as their husbands. For example, note the following passage from Kamalakara Bhatta's Nirnayasindhu, an influential digest composed in Benares in 1612: (15)
  nisedhavakyani prayascittartham mrtena patitena va saha
  marananisedhaparanity apy ahuh | asthidahe palasadahe va na
  prthakcitidosah angatvena sthanapattya va sariratulyatvat |
  (Nirnayasindhu, p. 439)
  Some even say that the statements prohibiting (a Brahmin woman from
  performing sahagamana) really mean to prohibit her from dying
  together with her husband if he has died for the sake of a penance
  or was an outcaste. When either (her husband's) bones or a piece of
  palasa wood is burned (with her), the sin of dying on a separate pyre
  does not result, for these things are equal to his body, as they
  either constitute a part of him (= the bones) or stand in his place
  (= the palasa wood).

In the opening sentence of this passage, Kamalakara introduces a novel interpretation of those scriptures that issue prohibitions against Brahmin women performing sahagamana: instead of referring to separate funeral pyres, they could refer to cases where a woman's husband either was an outcaste or had died in the process of repenting for some grievous sin. (16) If accepted, such an interpretation would virtually eliminate any special restrictions placed on Brahmin widows. Furthermore, Kamalakara proceeds to explain that in any case a woman can avoid the sin of ascending a separate funeral pyre by immolating herself together with either the bones of her dead husband or a palasa wood replacement for him. Again, such a position--even more so than Vijnanesvara's--effectively eliminates any special restrictions placed on Brahmin widows. Hence, taken in their historical and literary contexts, statements of this sort clearly reflect the further weakening of earlier restrictions on sahagamana. And this, as will be seen, is a salient trend throughout the commentarial literature.

To return to the Mitaksara, after severely limiting the prohibitive scope of those smrtis, that seemingly proscribe sahagamana for Brahmin widows, Vijnanesvara then attempts to demonstrate that the practice is dissimilar to the syena sacrifice and, therefore, in full conformity with dharma. In other words, at this point Vijnanesvara responds to Medhatithi's earlier discussion of sahagamana, with which he was clearly familiar, and specifically tries to refute the first of Medhatithi's arguments against the practice. Thus, here we have an excellent example of some of the ways in which Dharmasastric texts of the medieval period directly engaged with one another in a complex pan-Indian discourse. The relevant portion of Vijnanesvara's discussion reads as follows:
  yat tu kaiscid uktam purusanam iva strinam apy atmahananasya
  pratisiddhatvad atipravrddhasvargbhilasayan pratisedhasastram
  atikramantya ayam anugamanopadesah syenavat yatha syenenabhicaran
  yajeteti tivrakrodhakrantasvantasya pratisedhasastram atikramatah
  syenopadesa iti tad ayukiant | ye tavat syenakaranikayam bhavanayam
  bhavyabhutahimsayam Vidhisamsparsabhavena pratisedhasamsparsat
  phaladvarena syenasyanarthatam varnayanti tesam mate himsaya eva
  svargarthataya anugamanasastrena vidhiyamanatvat
  pratisedhasamsparsabhavad agnisomiyavat spastam evanugamanasya
  syenavaisamyam | yat tu matam himsa nama marananukulo vyaparah
  syenas ca paramarananukulavyapararupatvad dhimsaiva | kamadhikare
  ca karanamse ragatah pravrtisambhavena vidher apravartakatvat
  ragaprayuktahhimsarupatvat syehan pratisiddhah svarupenaivanarthakara
  iti tatrapy anugamanasastrena maranasyaiva svargasadhanataya vidhanan
  marane yady api rafatah pravrttis tathapi marananukule vyapare'
  gnipravesadav itikartavyatarupe vidhita eva pravrttir iti na
  nisedhasyavakasah vayavyam svetam alabheta bhutikama itivat | tasmat
  spastam evanugamanasya syenavaisamyam | (Mitaksara 1.86)
  Some have argued the following: "Since for women as for men suicide
  is prohibited, this instruction that prescribes self-immolation is
  for women who have an excessive desire for heaven and so violate the
  sastra prohibiting suicide. Hence, it is like the syena sacrifice, as
  the instruction to perform the syena sacrifice, namely, 'One who uses
  black magic should perform the syena' (Sadvimsa Brahmana 3.8.1), is
  for a person whose mind is overcome with excessive anger and so
  violates the sastra prohibiting violence."
  This argument is not proper. Now, some explain that the syena
  sacrifice is detrimental to perform because of its result (i.e., the
  death of one's enemies), it being the case that the force brought
  about through the instrument of the syena sacrifice results in
  injury to living beings. Therefore, the general prohibition against
  violence still applies, because no specific injunction applies.
  According to this opinion, sahagamana is quite clearly dissimilar to
  the syena sacrifice, but is instead like the agnisomiva rite (which
  involves animal sacrifice), for the sastra regarding sahagamana
  enjoins violence itself for the purpose of rebirth in heaven.
  Consequently, no prohibition applies (i.e., there is no prohibition
  against rebirth in heaven).
  Others, however, hold the following opinion: "What is called
  'violence' is any activity conducive to death; and the syena
  sacrifice itself qualifies as violence, because it consists of an
  activity conducive to death. Furthermore, since desire constitutes a
  qualification for the syena sacrifice, any injunction to perform it
  fails to result in a mandatory undertaking, as it is possible that
  one would undertake the part of the rite that is the instrument
  (i.e., the syena ritual) out of passion. Therefore, the syena
  sacrifice is prohibited and detrimental to perform by its very
  nature, since it has the form of violence employed out of a natural
  desire." According to this opinion as well, there is no occasion for
  the prohibition of sahagamana. Instead, the practice is analogous to
  the statement, "One desiring prosperity should sacrifice a white
  animal for Vayu" (Taittiriya Samhita This is because the
  sastra regarding sahagamana enjoins death itself as a means of
  attaining heaven. Thus, although one might undertake to kill oneself
  out of a natural desire to reach heaven, nevertheless one undertakes
  those activities conducive to death (i.e., violence), such as
  entering the fire, on account of the injunctions themselves, since
  such activities have the nature of necessary steps in the process.
  Therefore, sahagamana is quite clearly dissimilar to the syena

In order to refute Medhatithi's thesis that sahagamana is analogous to the syena sacrifice, Vijnanesvara here postulates two different arguments both of which aim to explain why the syena sacrifice is adharmic. He then attempts to show that these arguments do not work in the case of sahagamana.

In outline, the first argument he presents goes as follows: A) There is a general prohibition against violence. B) The syena sacrifice involves violence, since its outcome is the death of one's enemies. C) Only a specific injunction stating that one should kill one's enemies could override the general prohibition, but no such injunction exists. Therefore, D) the syena sacrifice is prohibited by the sastra?. Vijnanesvara holds that part C) of this argument does not apply in the case of sahagamana, since this practice is actually enjoined in the sastras. He points out that unlike the syena sacrifice, which results in violence, a prohibited outcome, sahagamana results in heavenly rebirth, a permissible outcome. The syena sacrifice is prohibited, because its violence is its result and its result is not enjoined. In other words, it is prohibited because the sastras never state that a person should kill his enemies and this is the violent part of the sacrifice. By contrast, the violence of the sahagamana rite (i.e.. the widow's suicide) is enjoined, as a means to rebirth in heaven, and although the sastras may not specifically enjoin rebirth in heaven, they certainly do not prohibit it. Vijnanesvara adds that sahagamana should instead be treated like the agnisomiya rite, which both involves violence to living beings (animals) and leads to a permissible outcome. Since the agnisomiya rite is permitted, sahagamana should be as well.

The second argument presented by Vijnanesvara is somewhat more complex. It begins by defining violence: violence is any activity that is conducive to death. Hence, according to this definition, the syena sacrifice itself, and not its outcome, is violence. Consequently, the general prohibition against violence does not apply to the result of the syena sacrifice, only to the rite itself. A difficulty then arises: the Veda enjoins the syena sacrifice (though not its result) and such an injunction would normally override a general prohibition, according to the standard exegetical principles of Dharmasastric discourse. If this is the case, then how can the syena sacrifice be prohibited? In order to solve this dilemma, the argument relies upon another accepted principle of interpretation: in order to qualify as dharma, an injunction (vidhi) must be adrstartha 'without seen purpose'. That is to say, a scriptural statement that recommends a course of action to which people are naturally inclined does not have injunctive force. This principle is used to divest scriptural statements concerning the syena sacrifice of injunctive force. It is argued that people naturally seek to injure their enemies and, therefore, those Vedic passages that mention the syena sacrifice are not injunctions. This being the case, the prohibition against violence still applies. Again, Vijnanesvara holds that this argument does not apply to the case of sahagamana. He concedes that a widow might seek to die out of a natural desire for heaven, but points out that, according to the proposed definition of violence, which is "activity conducive to death" (marananukulo vyaparah), dying itself does not constitute violence. On the other hand, the parts of the sahagamana rite that do constitute violence, such as entering the fire, are merely necessary steps in the process (itikartavyata). As such, one does not perform them out of a natural desire for heaven, but rather out a desire to complete the rite as enjoined. In this way, Vijnanesvara refutes Medhatithi's objection to the practice of sahagamana on the grounds that it is analogous to the syena sacrifice. Significantly, this refutation appears to have been quite effective, as not one of the later commentators within the dharma tradition, so far as I am aware, takes up this line of argumentation against the custom.

Nevertheless, despite his strong refutation of the syena analogy, Vijnanes'vara appears not to have wholly approved of sahagamana, for he cites two arguments against the practice that he regards as 'unobjectionable' (anavadya):
  yat tu tasmad u ha na purayusah svahkami preyad iti srutivirodhad
  anugamanam ayuktam iti yac ca tad u ha na svahkamy ayusah pran na
  preyad iti svargaphaloddesenayusah prag ayurvyayo na kartavyo
  moksarthina | yasmad ayusah sese sati nityanaimittakarmanusthanaks
  apitantahkaranakalankasya sravanamanananididhyasanasampattau satyam
  atmajnanena nityanairatisayanandabrahmapralaksnamoksasambhavah tasmad
  anityalpasukharupasvargartham ayurvyayo na kartavya ity arthan | atas
  ca moksam anicchantya anityalpasukharupasvargarthinya anugamanam
  yuktam itarakamyanusthanavad iti sarvam anavadyam | (Mitaksara 1.86)
  However, there is the following argument: "sahagamana is wrong,
  because it is opposed to the Vedic statement, 'Therefore, one who
  desires heaven should not die before one's natural lifespan'
  (Satapaiha Brahmtuia" And there is another argument:
  "Since heaven is indicated as the result in the following scriptural
  passage, 'Thus, it is not the case that one who desires heaven should
  not die before one's natural lifespan," a person who desires
  liberation should not relinquish his life before his natural
  lifespan. In other words, so long as life remains, it is possible
  that, through knowledge of the Self, one who has destroyed his mind's
  blemishes through the performance of perpetual and occasional rites
  and succeeded at learning, reflecting, and meditating (= the three
  stages of Vedic study) will attain liberation, which is defined as
  the attainment of brahman, which is eternal and unsurpassed bliss.
  Therefore, one should not relinquish one's life for the sake of
  heaven, which consists of only impermanent and miniscule happiness.
  And hence, like any other undertaking aimed at fulfilling a specific
  desire, sahagamana is proper only for a woman who does not desire
  liberation and instead seeks heaven, which consists of impermanent
  and miniscule happiness." All of this is unobjectionable.

The first of these arguments against sahagamana is essentially the same as Medhatithi's second argument, namely, that Vedic statements prohibiting suicide effectively negate smrti statements enjoining widows to immolate themselves on their husbands' funeral pyres. The second unobjectionable argument against sahagamana mentioned by Vijnanesvara goes as follows: although certain smrtis do enjoin sahagamana as a means of attaining heaven, heaven itself is an inferior goal to moksa ('liberation'). Therefore, since the alternative practice of ascetic celibacy (brahmacarya) can help a woman to attain moksa, this alternative is undoubtedly superior. Vijnanesvara's judgment of these two arguments as 'unobjectionable' (anavadya) shows that while he does not appear to share Medhatithi's strong opposition to sahagamana, he does have considerable reservations about the practice.

It is noteworthy that none of the commentators or digest-writers in the centuries following the Mitaksara takes up the position that Vedic statements prohibiting suicide negate those smrtis that apparently enjoin sahagamana. And this--it should be noted--is the only argument accepted by Vijnanesvara that would actually prohibit sahagamana and not just demote it to the lesser of two alternatives. Moreover, Apararka, who may have written only a few decades after Vijnanesvara, (17) explicitly refutes this line of argumentation. His refutation reads as follows:
  na canvarohanasmrtinam tasmad u ha na purayusah svahkami preyad iti
  srutivirodho vacyah bhinnavisayatvat | samanyena srutih svecchaya
  maranam nisedhati | smrtis tu mrte bhartari vahnipravesamaranavisesam
  vidhatte | ato bhinnavisayatvad avirodhah | svargakami yajeta yavaj
  jivam agnihotram juhuyad ityevamadikaya srutya samanyavisayaya na
  virodhah | tatha ca brahmapuranam
mrte bhartari satstrinam na canya vidyate gatih |
nanyad bhartrviyogartidahaprasamanam bhavet ||
desantaramrte tasmin sadhvi tatpadukadvayam |
nidhayorasi samsuddha pravisej jatavedasam ||
rgvedavadat sadhvi stri na bhaved atmaghatini |
tryaha[s]auce tu nivrtte sraddham prapnoti sasvatam ||
  ima narir avidhava ityadayo vedasrutayah | tasmad etah smrtayah
  pramanam | na ca na himsyadity anenanvarohanam sakyam niseddhum
  | anena hi hantuh sakasad anyasya hantavyasya himsa pratisidhyate na
  tu svavadharupa | sa tu tasmad u ha na purayusa ity anena nisidhyate
  | idam ca samanyvavisayatvad visesavisayayanvarohansmrtya
  tadvyatiriktasvavadhanisedhakatvena samkocyata ity uktam | (Apararka

And it should not be objected that the smrti passages that enjoin sahagamana are in conflict with the following sruti passage: "Therefore, one who desires heaven should not depart before one's lifespan" (Satapatha Brahmana The reason for this is that they have different spheres of applicability: sruti prohibits dying by one's own desire in general, but smrti enjoins the particular method of dying that is entering the fire when one's husband has died. Hence, there is no conflict, for they have different spheres of applicability. Likewise, there is no conflict with (other) sruti passages that have general spheres of applicability, such as "Desiring heaven, one should sacrifice" and "One should perform the Agnihotra rite as long as one lives." And thus, the Brahma Purana states:
  There is no other recourse (than sahagamana) for a good woman when
  her husband dies, (for) there is no other way to extinguish the
  burning pain of being separated from her husband. And when he dies in
  a distant place, a virtuous woman should place a pair of his sandals
  on her chest and, purified, enter fire. Due to the statement of the
  Rgveda, such a virtuous woman does not commit suicide. And when the
  three days' impurity has ceased, she eternally obtains the sraddha
  The Vedic statements referred to here are the verses beginning with
  the phrase "May these women, who are not widows ..." (Rgveda
  10.18.7). Therefore, these smrtis (that enjoin sahagamana) are
  authoritative. Furthermore, it is not the case that sahagamana can
  be prohibited by this rule: "One should not kill." For this rule
  prohibits killing another person who is to be killed by the hand of
  the killer. It does not, however, prohibit killing oneself. That is
  instead prohibited by this: "Therefore, one (who desires heaven
  should) not (depart) before one's lifespan" (Satapatha Brahmana And as I have said, because this passage has a general
  sphere of applicability, the smrtis enjoining sahagamana. which have
  a specific sphere of applicability, restrict it so that it prohibits
  suicide only in cases other than that (i.e., sahagamana).

In this passage, Apararka proposes what is called a visayavyavastha, that is to say, a resolution of apparently contradictory scriptures by ascribing to them different spheres of applicability. According to his analysis, those Vedic statements that proscribe suicide or otherwise stand opposed to ending one's life prematurely are only of a general nature, whereas those smrti statements that prescribe sahagamana are of a specific nature. Therefore, since it is a standard principle of Brahmanical hermeneutics (Mimamsa) that a specific rule overrides a general one, the smrth that enjoin sahagamana are of sufficient authority in this case to overrule sruti. In other words, Apararka concludes that those sruth that prohibit suicide apply everywhere except for the specific case of sahagamana, where the less general rules of the various smrtis apply. Significantly, a number of later authors accept the very same position on this issue as Apararka and none attempts to refute it. (18)

However, Apararka appears to have been more thorough than other Dharmasastric exegetes in his advocacy of this position. Perhaps the reason is that in his day many people still accepted the objection to sahagamana on the grounds that it contradicts express Vedic statements and, thus, a more thorough refutation of this objection was necessary than in later times. Whatever the reason, unlike later authors Apararka explicitly justifies the reconciliation of the following generally accepted exegetical principles: A) sruti texts override smrti texts, and B) more specific rules override more general ones. In the case of sahagamana these two principles clearly come into conflict, since principle B) leads to the endorsement of the practice, while principle A) leads to its prohibition. As a result, some reconciliation of these would seem to be necessary. Obviously, Apararka holds that principle B) should be of greater authority here as he supports sahagamana. And in order to justify his position, he states:
  na ca samanyavisayasrutiviruddha smrtir apramanam iti vacyam
  samanyavisesa-rupavisayabhedena virodhabhavat | visayabheda eva hi
  virodho na tu visesavisayabhede  tatas ca visesavisayaya smrtya
  svamulabhuta srutir visesavisayaivanumiyate | sa ca
  samanyavisayasruter batiyast sati tasyah samkocahetur bhavati |
  (Apararka 1.87)
  Moreover, it should not be objected that when a smrti text is
  contradicted by a sruti text that has a general sphere of
  applicability, it becomes unauthoritative, because there is really no
  contradiction between them, as these texts have different spheres of
  applicability: one is general and one is specific. For contradiction
  exists only when there is no difference between spheres of
  applicability, but not when there is a difference between specific
  spheres of applicability. And therefore, from a smrti text that has
  a specific sphere of applicability, one can infer a sruti text that
  has its same specific sphere applicability and that it is the basis
  of it. And that sruti text carries greater weight than a sruti text
  with a general sphere of applicability and, hence, causes it to
 be restricted.

Here Apararka argues that principle B)--specific rules override general ones--is of greater force than principle A)--sruti overrides smrti--because of a particular form of inference that is characteristic of Brahmanical hermeneutics, namely, the inference of Vedic texts. Classical Brahmanical thinking holds that there are basically three sources of knowledge about dharma: Vedic scripture (sruti), non-Vedic scripture (smrti), and the customs of cultured people (sadacara). However, of these, only Vedic scripture is a direct and ultimate source of knowledge. The other two (smrti and sadacara) are authoritative only because one can infer from them the existence of other, no longer available Vedic scriptures that express their essential meanings. in the above passage, Apararka uses this widely held belief to argue that one can infer from smrtis enjoining sahagamana the existence of some no longer available sruti text that enjoins the practice. And being specific in nature, this Vedic scripture would be of greater force than those Vedic scriptures that generally prohibit suicide. in this way, Apararka puts forth a uniquely thorough refutation of the objection to sahagamana on the grounds that it contradicts express Vedic injunctions.

In addition, both Apararka (1.87) and Laksmidhara's Krtyakalpataru (Vyavaharakanda, p. 634) approvingly cite the following line from the Brahma Purana, (19) which alludes to a Rgvedic sanction for sahagamana: rgvedavadat sadhvi stri na bhaved atmaghatini "Due to the statement of the Rgveda, such a virtuous woman [who performs sahagamana] does not commit suicide." These commentators both identify the Rgveda passage referred to here as RV 10.18.7:
  ima nair avidhavah supatnir anjanena sarpisa sam visantu |
  anasravo 'namivah suratna a rohantu janayo ynim agre ||
  Let these women, who are not widows, but rather have good husbands,
  enter together with fresh butter as ointment! Without tears or
  afflictions and possessed of fine jewels, let the wives ascend the
  womb first!

Given that neither Vedic literature nor the early Dharmasastras make any mention of the custom of sahagamana, it is extremely unlikely that this rather cryptic passage has anything originally to do with it. Indeed, although the hymn in general (RV 10.18) seems to refer to a funerary rite of some type, this verse clearly issues a command to women who are not widows (avidhavah), but rather have good husbands (supatnih). Despite this, however, the Brahma Purana, Apararka, and the Krtyakalpataru all apparently regard it as in some way sanctioning sahagamana. The underlying reason for this highly dubious interpretation is undoubtedly their desire to find Vedic support for this practice. Other facilitating factors presumably include that the verse A) pertains to a funerary rite, B) is ambiguous in parts, and C) contains the imperative verbs visantu ("'enter!") and a rohantu ("ascend!"), both of which take feminine subjects and, therefore, fit the performance of sahagamana. Beyond this, there is also the distinct possibility that within the context of the Dharmasastric debate on sahagamana, RV 10.18.7 underwent certain changes to make it appear more germane to the practice. Of these, the most important is the possible change of the word agre (first') to agne ('O fire') or agneh ('of fire'). (20)

In this regard, it is also significant that both the Nirnayasindhu and the Dharmasindhu attest to the use of RV 10.18.7 as a mantra in the sahagamana rite. in particular, the earlier of the two works, the Nirnayasindhu (1612), describes its recitation as a practice peculiar to Bengal, (21) while the Dharmasindhu (1791-92) cites it as part of the standard ritual. (22) Hence, these texts suggest that the verse may have been a rather late addition to the sahagamana rite in much of India. Whether or not this was the case, their testimony further indicates that certain elements within the Brahmanical community were not content to refute through strictly exegetical arguments the objection to sahagamana on the grounds that it contradicts the Veda. Instead, by including RV 10.18.7 as a mantra, they appear also to have constructed the sahagamana rite itself in such a w ay that it conveys an aura of Vedic authority. And if this is in fact so, we have here a further Brahmanical argument in favor of sahagamana, albeit of an oblique and unconventional sort.

To return one last time to the Mitaksara, although the argument against sahagamana on the grounds that it contradicts the Veda quickly falls out of favor after Vijnanesvara, the other argument accepted by him--namely, that sahagamana is an inferior option to brahmacarya, because it yields the inferior result of heaven--has at least one later supporter: Devana Bhatta, who probably wrote his Smrticandrika in South India sometime between 1150 and 1225 (Kane, vol. 1 [ 1962]: 738-41). His statements on sahagamana read as follows:
  yal tu visnuna dharmantaram aktam mrte. bhartari brahmacaryam
  tadanvarohanam veti | tadanvarohanam bhartaram anv arohanam
  agnyarohanam I tatha cangirah
  mrte bhartari ya nari samarohed dhutasanam|
  sarundhatismacara avargaloke mahiyate||
  ted etad dharmantaram api brahmacaryadharmaj jaghanyam
  nikrstaphalatvat | (Smrticandrika, Vyavaharakanda, pp. 596-97)

However, there is another dharma that Visnu (25.14) proclaims in the statement, "When a woman's husband has died, she should either practice ascetic celibacy or ascend after him." Here the phrase "ascend after him" means that she should ascend, i.e., ascend the (funerary) fire, after her husband. And likewise, Angiras states:
  A woman who ascends the (funerary) fire when her husband has died
  behaves like Arundhati and is honored in heaven.

This other dharma is lowlier, however, than the dharma of ascetic celibacy, for it yields a lesser reward.

Following Devana, however, even support for this weakened position against sahagamana seems to cease. (23) Furthermore, two rather late digests, the Nirnayasindhu (1612) and Dharmasindhu (1791-92), cite the following, previously unquoted smrti passage that lists liberation (makti) as an explicit result of sahagamana:
athanvarohanam strinam atmano bhartur eve |
sarvapaksayakaram nirayottaranaya ca ||
anakasvargaphaladam muktidam ca tathaiva ca |
janmantare ca saubhagyadhanaputradivrddhidam ||
(Nirnayasindlu, p. 438, Dharmasindhu, p. 384)
  When a woman ascends (the funeral pyre) after her husband, it
  destroys all her and her husband' s sins, leads to deliverance from
  hell, bestows the reward of many heavens, grants liberation, and
  gives good fortune, wealth, sons, prosperity, etc, in another

In light of the earlier commentarial debate, this passage appears tailor-made as a refutation of the last remaining objection to sahagamana. That is, the content of this smrti and its absence as a cited scripture in the preceding commentaries combine to strongly suggest that it is a relatively recent creation of the dharma tradition, a creation whose purpose was to refute claims regarding the inferiority of sahagamana vis-a-vis ascetic celibacy. (24) Moreover, this position finds further confirmation in the Dharmasindhu's conclusion that the practice of sahagamana results in liberation for a woman who lacks desires and rewards, such as heaven, for a woman who possesses desires. (25)

Nevertheless, despite the eventual disappearance of all objections to sahagamana within the Dharmasastra tradition, it is important to note that the practice never becomes an obligatory act (nityakarman). Instead, authors consistently regard it as optional (kamya). although given the especially lavish praise it often receives, for many it was probably not just optional, but in fact supererogatory. (26) The first commentator to explicitly mention the optional character of sahagamana appears to be Apararka, who writes:
  anvarohanam ca kamyatvad anityam | ata eva visnuh mrte bhartari
  brahmacaryam tadanvarohanam veli | (Apararka 1.87)
  And since it is optional, sahagamana is not obligatory. For this
  very reason, Visnu (25.14) states: "When a woman's husband has died,
  she should either practice ascetic celibacy or ascend (the funeral
  pyre) after him."

This point is also either explicit or implicit in the works of all other Dharmasastric commentators who write on sahagamana. Moreover, as Apararka himself notes, the Vaisnava Dharmasastra, probably the earliest Dharmasastra text to mention sahagamana, clearly regards the practice as nothing more than an option for widows.

A few other smrtis, however, might conceivably be interpreted to portray sahagamana as a mandatory act. Nevertheless, the Madanaparijata (pp. 198-99) explains these texts away as follows:
  nanu ca pativrataya idam anugamanam nityam eva | tatha ca haritah
  artarte mudite hrsta prosite malina krsa |
  mrte mriyeta ya patyau sa stri jheya pativrata ||
  ato mrte ya mriyeta pativratety anena nityatvam dyotyate | ato
  'nityatvavarnanam ayuktam iti cen maivam |
  mrte bhartari sadhvi stri brahmacarye vyavasthita |
  svargam gacched aputrapi yatha ta brahmacarinah ||
  iti manuna pativrataya ananugamanasyabhihitatvat | atra sadhvity
  anena pativratabhidhiyate anyatha brahmacarye vyavasthitety anena
  paunaruktyaprasakteh | tatha mahabharate bhagavan vyasah pativratayah
  satya agnipravesam darsayati na cagnipravesena pativratatvam
  pativrata sampradiptam pravivesa hutasanam |
  tatra citrangadadharam bhartaram sanupadyate. || iti |

(Objection:) This act of sahagamana is surely mandatory for a pativrata ('woman who is devoted to her husband'). And thus, Harita states:
  A woman should be known as a pativrata if she is pained when her
  husband is pained, rejoices when he's happy, becomes wretched and
  emaciated when he's gone abroad, and dies when he dies.

Hence, the mandatory nature of sahagamana is made clear by this statement, which says that a woman is a pativrata if she dies when her husband dies. Therefore, it is inappropriate to describe the practice as non-mandatory.

(Reply:) This is incorrect, for Manu prescribes a course of action other than sahagamana for a pativrata:
  When her husband dies, a sadhvi ('virtuous woman') who adheres to
  ascetic celibacy goes to heaven even if she is sonless, just like
  those who are celibate (MDh 5.160).

The word sadhvi in this statement expresses that the woman is a pativrata, for otherwise it would be redundant with the phrase that she "adheres to ascetic celibacy." Also, thus, in the Mahabharata venerable Vyasa shows a woman who is a pativrata entering fire rather than becoming a pativrata by entering fire:
  The pativrata entered the blazing fire. There she followed her
  husband, who wore colorful arm-bracelets (MBh 12.144.9-10).

Here the objection is raised that sahagamana must be obligatory for a pativrata since Harita, the author of a well-known smrti, says that a pativrata is a woman who, among other things, dies when her husband dies. Madanapala refutes this objection on the grounds that Manu (5.160) also prescribes a life of celibacy for a pativrata. But in order for this refutation to work, he must establish that the sadhvi mentioned in Manu is identical with the pativrata mentioned in Harita. Madanapala does this by arguing that if Manu's sadhvi is not the same as Harita's pativrata, it would be redundant for Manu to speak of a sadhvi "who adheres to ascetic celibacy" (brahmacarye vyavasthita). The point here is that if a pativrata is required to perform sahagamana, then in order for a sadhvi to be different, she must by definition be a woman who remains celibate after her husband's death (remarriage was not an option). Therefore, it would be redundant to mention a sadhvi who remains celibate. Hence, the sadhvi in Manu's statement must be identical with a pativrata. One crucial implication of this is that since a pativrata widow can either perform sahagamana or remain celibate, these acts cannot make a woman a pativrata. That is, a woman must be a pativrata prior to making the decision to either remain celibate or perform sahagamana. This, according to Madanapala, is why the Mahabharata (12.144.9-10) speaks of a woman (actually a female dove) who is already a pativrata before entering her husband's cremation fire. In this way, the Madanaparijata refutes one possible objection to the position, consistently held by authors in the dharma tradition, that sahagamana was strictly optional.

Furthermore, even beyond this, Apararka, the Nirnayasindhu, and the Dharmasindhu all specify a procedure for stopping the sahagamana rite, if the widow loses her resolve:
  devaradina tutthapanam tasyah putrakamanayam va jivalokakamanayam
  va satyam ahuh | (Apararka 1.87)
  However, they say that if a woman (who is to perform sahagamana) has
  a desire for sons or for the world of the living, her husband's
  younger brother or the like should cause her to get up (from her
  husband's pyre).
  kataram tu pretottaratah suptam devarah sisyo va udirsveti dvabhyam
  utthapayet | (Nirnayasindhu, p. 438: Dharmasindhu, p. 385 (27))
  However, if Iain down to sleep at the left side of her dead
  (husband), the woman (who is to perform sahagamana) becomes afraid,
  either her husband's younger brother or one of his students should
  cause her to get up (from the pyre) with the two verses beginning,
  "Rise ..." (RV 10.18.8-9). (28)

These statements show that even after committing to perform sahagamana, if a woman became terrified or otherwise regretted her decision, one or another of her husband's relations was considered obligated to stop the rite. Moreover, rather unusually, none of these texts cites any scriptural evidence in support of this rule. Hence, it seems a distinct possibility that here these Dharmasastric authors in ay for once be implicitly acknowledging the harrowing nature of the sahagamana rite and allowing a role for general compassion in what it is otherwise treated as a matter of dry scriptural analysis. At the very least, it shows that these authors believed strongly in the voluntary nature of sahagamana.

Finally, many authors in the dharma tradition explicitly stipulale that sahagamana is prohibited for women who are either pregnant (garbhini) or have young children (balapatya). For example, Vijnanesvara summarizes the injunctions regarding sahagamana as follows:
  ayam ca sakala eva sarvasam strinam agarbhininam abalapatyanam
  acandalam sadharano dharmah | (Mitaksara 1.86)
  And all of this constitutes the universal Law for all women right
  down to Candalas, provided that they are not pregnant and do not
  have young children.

Perhaps somewhat surprisingly, the earliest commentators to make this stipulation, including Vijnanesvara, cite no scriptural passages in support of their position. (29) From this we might infer that the justification lies not in specific scriptural prohibitions, but rather in the basic ethical principle that sahagamana may harm only the widow herself and no one else, including her born and unborn children. In other words, the prohibition against pregnant women and women with small children performing sahagamana is essentially an application of the general prohibition against violence.

Nonetheless, two later Dharmasastric authors, Madhava (on Parasara Smrti 4.31) and Kamalakara (Nirnayasindhu, p. 439), cite the following pair of verses that provide an explicit scriptural basis for this prohibition:
balasamvardhanam tyaktva balapatya na gacchati |
rajasvala s[u]tika ca raksed garbham ca garbhini ||
  A woman who has young children should not depart, thereby forsaking
  raising her young children; nor should a woman who is menstruating or
  has just given birth; and a pregnant woman should protect her fetus.
balapatyas ca garbhinyah adrstartavas tatha |
rajasvala rajasute narohanti citam subhe || (30)
  O beautiful princess, women do not ascend the funeral pyre when they
  have young children, when they are pregnant, when their menses have
  failed to appear at the regular time, and while they are

Madhava's commentary on the first of these verses is informative. It reads:
  atra balasamvardhanam tyaktveti vadan samvardhayitrjanantaravisrambhe
  balapatyaya apy adhikaro 'stiti darsayan raksed garbham ca garbhiniti
  raksam darsayan sambhavitagarbhasamdehaya apy adhikaram varayati I
  Here, by saying, "thereby forsaking raising her young children," the
  author shows that even a woman with young children is qualified (to
  perform sahagamana) if she entrusts other people to raise them; and
  by showing that there must be protection with the words "should
  protect her fetus," he excludes the qualification (to perform this
  rite) for any woman who might even possibly be suspected to be

Madhava's interpretation of this verse makes it clear that he accepts the general principle that sahagamana is permissible except in cases where a woman would thereby bring harm to her children. Consequently, women who are pregnant or even may be pregnant, such as those whose menses have failed to appear at the regular time (adrstartu), (31) are strictly forbidden from performing sahagamana. A woman with small children, however, may carry out the rite provided specifically that she entrusts the caretaking of her children to other people. Crucially, this general principle that a woman cannot perform sahagamana if it would directly harm her children goes unchallenged throughout the Dharmasastric literature.

To summarize, one can loosely arrange Dharmasastic writings on sahagamana into three historical periods. In the first of these, which roughly corresponds to the second half of the first millennium C.E., smrti texts that prescribe sahagamana begin to appear. However, during approximately this same period, other Brahmanical authors also compose a number of smrtis that proscribe this practice specifically in the case of Brahmin widows. Moreover, Medhatithi--our earliest commentator to address the issue--strongly opposes the practice for all women. Taken together, this textual evidence suggests that sahagamana was still quite controversial at this time. In the following period, opposition to this custom starts to weaken, as none of the later commentators fully endorses Medhatithi's position on sahagamana. Indeed, after Vijnanesvara in the early twelfth century, the strongest position taken against sahagamana appears to be that it is an inferior option to brahmacarya (ascetic celibacy), since its result is only heaven rather than moksa (liberation). Finally, in the third period, several commentators refute even this attenuated objection to sahagamana, for they cite a previously unquoted smrti passage that specifically lists liberation as a result of the rite's performance. They thereby claim that sahagamana is at least as beneficial an option for widows as brahmacarya and perhaps even more so, given the special praise it sometimes receives. These authors, however, consistently stop short of making it an obligatory act. Hence, the commentarial literature of the dharma tradition attests to a gradual shift from strict prohibition to complete endorsement in its attitude toward sahagamana.

The question then arises: what, if anything, does this esoteric and erudite Dharmasastric debate on sahagamana tell us about broader societal attitudes toward this practice in premodern India? Another more general way of asking this question is: what is the relationship between Dharmasastra and social reality? Obviously, scholars have written a great deal on this subject and I have neither the space nor the desire to discuss their writings in detail here. Nevertheless, it is worth pointing out that while scriptural exegesis generally presents itself as an endeavor to uncover the true meanings of sacred texts, the set of normative assumptions and moral intuitions that form a part of any exegete's background tend to have a major impact on the process of exegesis. In other words, commentators do not simply study scripture to learn the truth; they also study scripture to confirm the truth they already know from their social contexts. As a result, it seems reasonable to assume that for the commentators under discussion, one of these already known truths, which would guide their interpretations, was the relative morality or immorality of sahagamana. And indeed, the Suddhimayukha of Nilakantha (c. 1610-50) provides some evidence that supports this assumption, when it states:
  kalpatarau prthakcitim samaruhya na vipra gantum arhatiti vaco
  'ntaran nisedho bhinnacitipara iti vijnanesvaradayo yuktam utpasyanti
  | sarvajanina acaro 'py amum eva paksam anugrhnati | (Suddhinayukha
  p. 69)
  From this other verse quoted in the Kalpataru, "A Brahmin woman ought
  not to depart by mounting a separate pyre," (32) Vijnanesvara and
  others correctly derive that the prohibition (against Brahmin widows
  performing sahagamana) applies only to ascending separate funeral
  pyres. The custom of all peoples supports this position as well.

Here Nilakantha unambiguously cites popular custom (sarvajanina acarah) as additional support for his position on sahagamana. In broad agreement with Richard Lariviere's provocative thesis that "dharmasastra literature represents a peculiarly Indian record of local social norms" (1997: 98), I believe that Nilakantha is not alone amongst Dharmasastric writers in his reliance on custom when analyzing sahagamana, only in his explicit statement to this effect. And if this is in fact the case, then by following the intricate details of this Dharmasastric debate one can trace a shifting moral attitude toward this practice within the greater Brahmanical community.


Primary Sources

Apararka. Commentary on the Yajnavalkyasmrti. 2 vols. Anandasramasamskrtagranthavali, vol. 46. Pune: Anandasrama, 1903.

Devana Bhatta. Smrticandrika, ed. L. Srinivasacharya. 5 vols. Bibliotheca Sanskrita, vols. 43, 44, 45, 48, 52, 56. Mysore: Government Branch Press. 1914-21.

Kamalakara Bhatta. Nirnayasindhu, ed. Narayana Rama Acarya. Vidyabhavana Pracyavidya Granthamala, vol. 31. Varanasi: Caukhamha Vidyabhavana, 1991 [rpt.].

Kasinatha Upadhyaya. Dharmasindhu. Sri Garibdas Oriental Series, vol. 14. Delhi: Sri Satguru Publications, 1986 [rpt.].

Laksmidhara. Krtyakalpataru, ed. K. V. Rangaswami Aiyangar. 14 vols. Gaekwad's Oriental Series. Baroda: Oriental Institute, 1941-79.

Nilakantha. Suddhimayukha, ed. Anant Yagneshwar Dhupakar. Bombay: Gujarati Press, 1949.

Madanapala. Madanaparijata, ed. Madhusudana Smritiratana. Bibliotheca Indica. vol. 114. Calcutta: Asiatic Society of Bengal, 1893.

Madhava. Parasaramadhava (commentary on the Parasarasmrti), ed. Candnikanta Tarkalahkara. 3 vols. Calcutta: The Asiatic Society. 1973-74 [rpt.].

Medhatithi. Manubhasya (commentary on the Manava Dharmasdstra), ed. and tr. Ganganath Jha. 10 vols. Delhi: Motilal Banarsidass, 1999 [rpt.].

Raghunandana. Suddhitattva, ed. Candicarana Smrtibhusana. Calcutta: Victoria Press, 1913.

Sabara. Subarabhdsya (commentary on the Purvamimamsa Sutra), ed. Gajanan Shastri Musalgaonkar. Chowkhamba Sanskrit Series, vol. 115. Varanasi: Chowkhamba Press, 2004.

Vijnanesvara. Mitaksara (commentary on the Yajnavalkyasmrti), ed. Narayana Rama Acarya. Delhi: Nag Publishers, 1985 [rpt.].

Vaisnava Dharmasastra. ed. and tr. Patrick Olivelle. Harvard Oriental Series, vol. 73. Cambridge: Harvard Univ. Press, 2009.

Secondary Sources

Colebrooke. Henry Thomas. 1795. On the Duties of a Hindu Widow. Asiatick Researches 4: 205-15.

Garzilli. Enrica. 1997. First Greek and Latin Documents on sahagamana and Some Connected Problems. 2 parts. Indo-Iranian Journal 40: 205-43, 339-65.

Hall, Fitzgerald. 1868. The Source of Colebrooke's Essay "On the Duties of a Faithful Hindu Widow." Journal of the Royal Asiatic Society n.s.-3: 183-98.

Kane, P. V. 1962-75. History of Dharmasastra. 5 vols. Puona: Bhandarkar Oriental Research Institute.

Kiparsky. Paul. 1979. Panini us a Variationist. Cambridge, Mass.: MIT Press.

Lariviere, Richard. 1997. Dharmasastra, Custom, "Real Law" and "Apocryphal" Smrtis. In Recht, Staat und Verwaltung im klassischen Indien, ed. Bernhard Kolver and Elisabeth Muller-Luckner. Pp. 97-110. Munich; R. Oldenbourg Verlag.

Lingat. Robert. 1973. The Classical Law of India, tr. J. Duncan M. Derrett. Law in India Series. Delhi: Oxford Univ. Press [rpt. 1998].

Olivelle. Patrick, ed. and tr. 2005. Harm's Code of Law: A Critical Edition and Translation of the Manava-Dharmasastra. New York: Oxford Univ. Press.

__. 2007. The Date and Provenance of the Visnu-Smrti: On the Intersection between Text and Iconography. Indologica Taurinensia 33: 149-64.

__, ed. and tr. 2009. The Law Code of Visnu: A Critical Edition and Annotated Translation of the Vaisnava-Dharmasastra. Harvard Oriental Series, vol. 73. Cambridge: Harvard Univ. Press.

Rocher, Ludo, and Rosane Rocher. 2007. The Making of Western Indology: Henry Thomas Colebrooke and the East India Company. London: Rout ledge.

Author's Note: I wish to thank Patrick Olivelle and Ludo Rocher for their many useful and insightful comments on an earlier draft of this article. I also wish to thank Joel Brereton for his help in understanding the Vedic passages discussed herein.

(1.) Very early on, in a 1795 article, H. T. Colebrooke translated and briefly discussed an array of Dharmasastric texts on sati (Sanskrit: sahagamana). Since then, a number of later Indologists (e.g., Hall 1868, Kane, vol. 2 [1962]: 624-36; Garzilli 1997) have also cited and analyzed these and a handful of other Sanskrit tests on the topic. To the best of my knowledge, however, these scholars' writings have uniformly failed to present a systematic, thorough, and historical account of Dharmasastric statements on widow-burning.

(2.) By "early Dharmasutras or Dharmasastras" I refer specifically to both the early Dharmasutras of Apastamba, Hiranyakesin, Gautama, Baudhayana, and Vasistha and the later Dharmasastras of Manu, Narada, and Yajnavalkya. There is also no mention of sahagamana in Kautilya's Arthasastra.

(3.) In addition to ViDh 25.14, the Vaisnava Dharmasastra contains another passage that appears to refer to sahagamana:
  mrte 'pi bandhavaih sakyam nanugantum naram mrtam I
  jayavarjam hi sarvasya yamyah pantha vibhidyate II (ViDh 20.39)
  Even when they have died, relatives cannot follow a dead man, for
  the path of Yama is cut off for all save his wife.

However, the later commentarial literature of the Dharmasastra tradition never cites this passage in connection with sahagamana.

(4.) The commentator Madhava readies precisely this conclusion, for he introduces Parasara Smrti 4.30 with the statement: uktabrahmacaryad apy adhikaphalam anugamane darsayati "Now he [= Parasara] shows that the reward for sahagamana is even greater than that just stated for ascetic celibacy." This strongly suggests that Madhava considers sahagamana to he the superior alternative, although certainly not mandatory.

(5.) See Medhatithi 1.157, Vijnanesvara 1.86, Apararka 1.87, Madhava 4.30-31, Madanapala, pp. 196-200, Devana Bhatta, p. 596, Nilakantha, pp. 68-69, and Kamalakara Bhatta, pp. 438-39.

(6.) For instance, Apararka (p. 111 = 1.87), Laksmidhara's Krtyakalpataru (Vyavaharakanda, p. 634), and Kamalakara Bhatta's Niranayasindhu (p. 438) all cite an identical passage that they ascribe to the Brahma Purana; the Nirnayasindhu (pp. 438-40) cites passages ascribed to the Skanda, Vayu Brhannarada, and Brahmavaivarta Puranas; Apararka (p. 112 = 1.87) mentions, but does not cite, a story from the Ramayana, which purportedly approves of Brahmin widows performing sahagamana; Vijnanesvara's Mitaksara 1.86 cites Mahabharata 12.144.9-10 and 12, which refers to the story of a female dove performing sahagamana; and Madanapala's Madanaparijata (pp. 199, 197) and the Nirnayasindhu (p. 439) also cite Mahabharata 12.144.9-10, as well as another passage ascribed to the Mahabharata, but not found in the critical edition of that text.

(7.) For some generally accepted assessments of the relative chronology of the Dharmasastras, see Kane, vol. 1 (1962), and Lingat 1973. However, thanks in large part to the excellent and dedicated work of Patrick Olivelle (2005, 2009, etc.), the absolute chronology of Dharmasastra literature presented in these works has in recent years undergone serious revision. It is generally now accepted that scholars of previous generations dated the major Dharmasutras and Dharmasastras too early by at least several centuries.

(8.) The following Dharmasastric works/authors all discuss sahagamana at some length: Medhatithi 5.155 (c. 825-1000, Kashmir); Vijnanesvara's Mitaksara 1.86 (c. 1076-1127. Maharashtra/Kalyana); Apararka 1.87 (c. 1125, North Konkan): the Vyavaharakanda of Laksimdhara's Krtyakalpataru, pp. 632-36 (c. 1 110-50, Kannauj); the Vyavaharakanda of Devana Bhatta's Smrticandrika, pp. 596-97 (c. 1150-1225, South India): Madhava 4.30-31 (c. 1330-60, Vijayanagara); Madanapala's Madanaparijata, pp. 196-201 (c. 1350-1400, U.P.); Kamalakara Bhatta's Nirnayasindhu, pp. 438-40 (1612, Benares); and Kasinatha Upadhyaya's Dharmasindhu, pp. 384-86 (1790-91, Maharashtra). For detailed discussions of the provenances of these works, see Kane, vol. 1 (1962).

(9.) See Apararka 1.87, Mitaksara 1.86. Madanaparijata, p. 197, Suddhimayukha, p. 69, and Nirnayasindhu, p. 438.

(10.) For Medhatithi, Kane (vol. 1 [1962]: 583) gives the date 825-900 C.E., but in my opinion there is no reason to rule out a slightly later date than this. However, it should be stressed that his commentary was clearly well known throughout much of India by the early twelfth century. For instance, the Danakanda of Laksmidhara's Krtyakalpataru (p. 10), a massive digest probably composed in Kannauj between 1110 and 1150, cites Medhatithi by name. Moreover. Vijnanesvara's Mitaksara (1.86), which was composed roughly in modern-day Maharashtra between 1076 and 1127, appears to contain a direct response to Medhatithi's position on sahagamana. Thus, it would seem unlikely dial Medhatithi wrote much after 1000. For arguments in favor of Medhatithi's being from Kashmir, see Kane, vol. 1 (1962): 574-75.

(11.) However, a single defective manuscript of an early and previously unknown commentary on the Yajnavalkya Smrti has recently come to light in Kathmandu and is now being edited by Prof. Yasuke Ikari. Based upon the orthography of this manuscript, this early commentary could quite possibly predate Medhatithi and certainly predates the Mitaksara, which appears to draw upon it heavily. Much like the Krtyakalpataru (Vyavaharakanda, pp. 632-36), the text (on YDh 1.87) is definitely in favor of sahagamana, but fails to address or even mention any possible objections to the practice. Hence, although it discusses sahagamana, it does not seem to be seriously involved in the theological debate on that topic.

(12.) The verse commented on here is the same as MDh 5.157 in Olivelle's critical edition of Mann (2005) and a number of other editions of that text.

(13.) I have been unable to identify the text to which this smrti passage belongs. However, here Medhatithi is likely referring to Sahara's commentary on Purvamimamsa Sutra 1.1.1: athato dharmajijnasa "Next, therefore, is the inquiry into dharma." Sabara interprets the word atha ('next') in this sutra as indicating that the inquiry into dharma (dharmajijnasa) should occur after one has learned to recite the Veda. The hypothetical objection to this interpretation is raised that the scripture cited by Medhatithi enjoins a person to take a bath--presumably the graduation bath marking the end of Vedic study--after reciting the Veda, not to then seek to understand its meaning. In response to this objection. Sahara construes this scripture in the following manner, which is compatible with his position: na va idam snanam adrstartham vidhiyate I kim tu laksanayasnanadiniyamasya paryavasanam vedahyayanasamakalam ahuh I "This scripture does not enjoin bathing that has an otherworldly purpose [i.e., the graduation bath]. Instead, through indirect indications it simply states that the restrictions [placed upon a Vedic student], such as that he should not bathe, cease to apply at the same time that he recites the Veda."

(14.) Also see Nilakantha's Suddhimayukha, p. 69, and Kamalakara Bhatta's Nirnayasindhu, p. 438.

(15.) Similar arguments are also presented in the Mudanaparijata (p. 198), Suddhimayukha (p. 69) and Dharmasindhu (p. 385). In addition, the Krtyakalpataru (Vyavaharakanda, p. 635) likewise permits a Brahmin widow to perform sahagamana by ascending a separate funeral pyre from her husband's. However, it presents no real argument in support of this conclusion and, indeed, never even mentions the possibility of any special restrictions placed on Brahmin women.

(16.) Note that the successful performance of certain penances prescribed in the Dharmasastras entails the death of the person who performs them. See, for instance, Manava Dharmasastra 11.74, 100-101, and 105-6.

(17.) Kane (vol. 1 [1962]: 721-23) puts forth a generally convincing argument that Apararka was familiar with the Mitaksara (c. 1076-1127) and wrote his commentary in the northern Konkan area around 1125.

(18.) For example, in his commentary on Parasara Smrti 4.30, Madhava writes:

nanv idam anugamanam pratyaksasrutivirudham tasmad u ha na purayusah svargakami preyad iti srutya atmahatyapratisedhat | asurya nama te loka andhena [t] amasavrtah | tams te pr[e]tyadhigacchanti ye ke catmahano jana iti srutyantarac ca | maivam anugamanasmrter niravakasatvena prabalyat | atmahatyanisedhasrutis tu svargakamiyosito nyatra savakasa |

(Objection:) Surely this act of sahagamana is contrary to the perceived Veda, because suicide is prohibited in the following Vedic text, "Therefore, one who desires heaven should not depart before one's natural lifespan" (Satapatha Brahmana, and also due to this other Vedic text: "Those worlds are called 'Sunless' that are enveloped in pitch darkness those people who kill themselves go to them after death" (Isa Upanisad 3). (Reply:) This is not so, for the smrti enjoining sahayamana is of greater force, as these Vedic texts do not apply here. Instead, the Vedic texts that prohibit suicide apply only to people other than women that desire heaven.

19. The entire passage in question does not occur in the version of the Brahma Purhana edited by Peter Schreiner and Renate Sohnen-Thieme (Wiesbaden: Harrassowitz, 1987).

(20.) See Hall 1868 Kane, vol. 2(1962): 634; and Rocher and Rocher 2007.

(21.) Nirnayasindhau (p. 438): gaudas tu ima narir avidhava iti om pativratah punyah striyo papah susobhanah | saha bhartuth samvisantu vibhavasum iti ca viprah pathed ity ahuh | "But the Bengalis say that a Brahmin should recite the verse beginning, 'Let these women ...' [= RV 10.18.7], and also the following verse: 'Om, let these auspicious, sinless, very beautiful women, who are faithful to their husband, enter the fire together with their husband's body!'"

(22.) The Dharmasindhu (p. 385) contains a passage virtually identical to that given in the preceding footnote, but without the phrase: gaudas tu ... ity ahuh ("But the Bengalis say that ..."). Thus, this text, which was composed in Maharasthra (not Bengal), clearly understands the recitation of RV 10.18.7 to be a standard part of the sahagamana rite rather than a regional peculiarity.

(23.) I have been unable to find any Dharmasastrie texts later than the Smrticandrika (c. 1150-1225) that argue that sahagamana is inferior to brahmacarya.

(24.) It is noteworthy that the creation of new smrtis, which is most clearly demonstrable in this particular case, may also have played a role in other aspects of the Dharmasastric debate on sahagamana. For instance, there is the verse ascribed to Usanas that explicitly prohibits a Brahmin widow from mounting a separate funeral pyre and is, therefore, used by commentators to limit the more general prohibits placed on such women in other smrtis. Authors within the dharma tradition may well have created this verse with the intention of achieving precisely this end.

(25.) Dharmasindhu, P.385: atra niskamatve muktih sakamatve svargadiphalaniti vyavastha | "The resolution here is as follows: When a woman is without desire, she attains liberation; when she possesses desire, she attains rewards such as heaven."

(26.) See, for instance, n. 4.

(27.) In the case of the Dharmasindhu, the word dvabhyam is replaced by mantrabhyam.

(28.) The recitation of RV 10.18.8-9 in this context would seem to stem from two factors: A) It comes immediately after RV 10.18.7, which according to certain authors is recited as part of a successful sahagamana rite (see above); and B) it is apparently addressed to a woman who has lain down beside a deceased man (gatasum etam upa sese) and is beckoned to rise up to the world of the living (udirsva nary abhi jivalokam).

(29.) These are the Mitaksara 1.86, Krtyakalpataru (Vyavaharakanda, p. 635), and Madanaparijata, p. 196.

(30.) The Suddhimayukha (p, 69) also cites this verse, but in a somewhat different version.

(31.) Tellingly, Madhava (4.31) explains this term as follows: adrstartavah rtvadarsanena sambhavitagarbhasamdehah "The phrase 'when their menses have failed to appear at the regular time' denotes women who might possibly be suspected to he pregnant, because their menses have failed to appear at the regular time."

(32.) I have been unable to locate this verse in the extant edition of the Krtyakalpataru.


COPYRIGHT 2010 American Oriental Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2010 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Brick, David
Publication:The Journal of the American Oriental Society
Article Type:Essay
Geographic Code:9INDI
Date:Apr 1, 2010
Previous Article:On the Qur'anic accusation of scriptural falsification (tahrif) and Christian anti-Jewish polemic.
Next Article:On some who are not allowed to become Buddhist Monks or nuns: an old list of types of slaves or unfree laborers.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters