Gargi interrupts: the women of the Upanishads and the dharmashastra that later shut the door

Janaka of Videha had a thousand cows standing in his courtyard, gold fastened to the horns of each, offered to whichever of the assembled brahmins knew Brahman best. Yajnavalkya told his student to drive them home before the questioning had started. What followed was a public examination, challenger after challenger, and the person who came closest to stopping him was a woman. Gargi Vachaknavi questioned him twice, and the second time she told the assembly that if he answered her two questions, not one of them would ever beat him on the subject of Brahman.

That exchange sits in Brihadaranyaka Upanishad 3.6 and 3.8. Several centuries later the Manava Dharmashastra states that a woman is not fit for independence, and that her life-cycle rites are to be performed without Vedic formulas. Both records are genuine, and the second is not a distortion invented by hostile readers. So the honest position here is narrow. Not that Hinduism always honoured women, a claim its own scriptures dismantle. Only this: the later layer forbade what the earlier layer had already allowed. That makes the exclusion something the tradition did to itself, on a date, in texts with authors, and what a tradition did it can undo.

The contest at Videha, and the question she was told to stop at

The third chapter is built as nine rounds, one brahmana each, a fresh challenger every time. Gargi Vachaknavi takes the sixth round and then, alone among them, comes back for the eighth. Nobody else gets a second turn.

Her first intervention (3.6) works by a weaver's image. She asks what a thing is woven on, warp and woof, then asks the same of the answer. This world is woven on water, water on air, air on the intermediate regions, and those on the worlds of the gandharvas, the sun, the moon, the stars, the gods, Indra and Prajapati, until she reaches the world of Brahman itself. On what is that woven?

Yajnavalkya refuses. "Do not, O Gargi, push your inquiry too far, lest your head should fall off." He tells her she is asking about a deity concerning which one must not overquestion. She stops.

What that refusal means is genuinely disputed, and smoothing it over helps nobody. Shankara's commentary reads it as a limit on method: the subject has one proper means of approach, which is scripture, and Gargi is trying to reach it by reasoning, which cannot get there. Read flatly on the page, it is a threat from a man losing patience. Both readings survive contact with the text.

What settles nothing but is worth holding on to is that the same threat is made to a male opponent three brahmanas later, and carried out. Sakalya presses Yajnavalkya, is told his head will fall off if he cannot answer the question put back to him, fails, and the text records that it did (3.9.26). Gargi is not warned off because she is a woman. She is warned off the way the men are, and she comes out of it better than the man who ignored the warning.

The redactors could have cut this. A woman cornering the tradition's most formidable debater until he resorts to a threat is not flattering material. It was kept, along with her verdict on the contest, in a text memorised and transmitted syllable by syllable for centuries.

The second round, and the imperishable

Then she comes back. In 3.8 she asks the assembled brahmins for leave to put two more questions, and frames them with a warrior's image: as a prince of Kashi or Videha might string a slackened bow and rise holding two arrows made for the enemy's pain, so she rises against him (3.8.2). She asks what that is woven on which is above the sky and below the earth and between them, which men call what was and is and will be. Space, he answers. And on what is space woven?

The answer is the akshara, the imperishable, and Yajnavalkya can only describe it by subtraction (3.8.8). Not coarse and not fine, not short and not long, without shadow, without darkness, without air or space, without taste, smell, eyes, ears, speech or mind, without measure, with nothing inside it and nothing outside. It eats nothing and nothing eats it. Under its rule the sun and moon are held apart, the rivers run in their courses, and the giver of gifts is praised.

Gargi turns to the assembly and concedes on their behalf. You may count yourselves lucky, she tells them, if you get away from this man having lost nothing more than a bow of salutation, and none of you will ever beat him on Brahman (3.8.12). Then she sits down. She has ended the contest.

Maitreyi, and the use of a great deal of money

The second woman in the same text argues from a different position. Yajnavalkya is leaving householder life and settling his property between his two wives. Maitreyi asks whether, if the whole earth full of wealth were hers, she would become immortal by it.

"No," said Yajnavalkya. "Your life would be like the life of people who have plenty. There is no hope of immortality through wealth." (Brihadaranyaka Upanishad 2.4.2, in a rendering close to the standard translations)

Her reply is the line the passage is remembered for: what should I do with a thing that will not make me immortal, tell me instead what you know. He answers with the instruction later Vedanta built an entire method on, that the Self is to be seen, to be heard about, to be thought over, to be dwelt on in meditation (2.4.5). Shravana, manana and nididhyasana, the three stages of Vedantic practice, are read out of that one sentence, and it is addressed to a woman.

The dialogue appears twice in the same Upanishad, at 2.4 and again at 4.5. Recensions differ and the compilers preserved both. The second version opens with framing that anyone arguing this case has to face rather than skip. It calls Maitreyi a brahmavadini, one who discourses on Brahman, and describes the other wife, Katyayani, as having only stri-prajna, a woman's understanding. The text is not describing a world in which women generally did philosophy. It is naming an exception and marking her as one, in the earliest layer, by contrast with the ordinary case.

The hymns the tradition's own index gives to women

The Rigveda does not label its hymns with the names of their seers. That comes from the anukramanis, the indexes, of which the most complete is the Sarvanukramani, ascribed by tradition to Katyayana. Shadgurushishya, who finished his commentary the Vedarthadipika in 1184 CE, names Katyayana as the author, and A. A. Macdonell edited the text at Oxford in 1886 with extracts from that commentary. For each of the 1,028 hymns it gives the opening word, number of verses, metre, deity and seer.

Among those seers it names women. Lopamudra is given the dialogue hymn 1.179, where she and Agastya argue the claims of ascetic practice against those of marriage, and the opening verses are hers. Ghosha is given 10.39 and 10.40, addressed to the Ashvins. Apala is given 8.91, a hymn to Indra that reads like a folk tale, repetitions of oral telling and all. Vishvavara of the Atri family is given 5.28, one of the Agni hymns making up the opening block of the fifth mandala.

How many women the index names is not settled, and anyone quoting a firm number is overreaching. Counts in the secondary literature run from ten or fifteen up to twenty-seven, depending on whether disputed and joint attributions are admitted and whether personified speakers in the seer column are counted.

This has to be stated with its limits attached. The Sarvanukramani is a later catalogue, not a contemporaneous record of authorship, and its method sometimes appears to take the female speaker inside a dialogue hymn and register her as its seer. Whether Lopamudra composed 1.179 or is a character within it cannot be settled from the index. What it does establish is the more useful fact anyway: the tradition's own authoritative catalogue of shruti wrote women's names into the seer column, and no later authority went back and scrubbed them out.

What the dharmashastra actually restricts

Manu is explicit and there is no point in softening it. At 9.3 her father guards her in childhood, her husband in youth, her sons in old age, and a woman is never fit for independence. At 2.66 the whole series of samskaras is to be performed for women at the proper time and in the proper order, to sanctify the body, but without the Vedic formulas. At 2.67 marriage is stated to be a woman's Vedic sacrament, service to her husband her residence with a teacher, household work her tending of the sacred fire.

Then 9.18, which tends to get quoted in half. Renderings differ in force, but the verse has two clauses and the second is the harder one. Ganganath Jha gives it as: for women there is no dealing with the sacred texts, such is the rule of law, and being destitute of organs and devoid of sacred texts, women are false. Patrick Olivelle keeps the same shape, rites performed without Vedic formulas, then women, lacking strength and lacking formulas, called falsehood itself. Citing only the first clause is the tidying that gets an argument destroyed by the first reader who opens the page.

Read those four verses together and the shape is clear. The samskaras remain and the Veda is taken out of them. Marriage is made to occupy the place initiation and Vedic study occupy for a man. Independence is denied in general terms, and the justification offered is not merely procedural.

It is often said, in defences of the tradition, that women once received upanayana and that it was later withdrawn. That claim is weaker than the people making it usually admit. Its main support is a passage attributed to Harita, surviving only in quotation by later digest writers, dividing women into brahmavadinis who take upanayana, tend the fire, study the Veda and beg alms within the household, and sadyovadhus who receive the rite shortly before marriage without the study. Alongside it sits Manu 2.66, read as the residue of an older practice with the mantras stripped out. That is an inference from thin and contested material, and no date can honestly be attached to a withdrawal no text describes. The argument does not need it.

The ritual theorists did not treat the matter as closed either. The sixth book of Jaimini's Purva Mimamsa Sutra takes up adhikara, entitlement to perform the Vedic sacrifice, and argues the case of women rather than treating it as settled, working from criteria about property, health, marriage, maintenance of the fires and knowledge of the relevant Veda. Ritualists were still debating it in a literature that treats shruti as its authority and Manu as one voice among many.

The restriction lives in smriti, not shruti. That distinction is the tradition's own, not a modern convenience: remembered texts of human authorship, ranked below revelation, disagreeing with one another so freely that the commentarial literature exists to reconcile them. Manu's modern prominence owes a great deal to the colonial courts, which took Sir William Jones's translation, published at Calcutta in 1794 as the Institutes of Hindu Law, and used it as a statute book for Hindus. Jones said plainly that he meant to make Manu the Justinian of India. A text human beings composed, ranked and argued with can be argued with still.

Why goddess worship proves nothing here

The standard reply, that Hinduism worships the Divine Mother and reveres Saraswati as the source of learning and therefore honours women, does not work. A society can revere a goddess and restrict its own daughters at the same time, and many have. Athens kept Athena on the Acropolis and its women out of the assembly. Theology is not legal status, and nothing in the argument here depends on the move.

The streams that never accepted the closure

In the Tamil country, in the eighth century by the usual academic dating though Srivaishnava tradition places her far earlier, a young woman of Srivilliputhur composed the thirty verses of the Tiruppavai and the hundred and forty-three of the Nachiyar Tirumoli. Andal is the only woman among the twelve Alvars, and her verses did not stay in a private notebook. They were taken into the Nalayira Divya Prabandham and into Srivaishnava liturgy, and are chanted in temples through Margazhi every year. A woman's composition became recited sacred text inside a tradition whose law books had already denied women the recitation of Vedic mantras. The temples did not resolve the contradiction. They simply chanted her.

Akka Mahadevi, in twelfth century Karnataka, left vachanas signed with the name Chennamallikarjuna, counted at between about 350 and 430 depending on the collection, and by the traditional account argued philosophy at the Anubhava Mantapa in Kalyana alongside Basavanna and Allamaprabhu. The honorific "Akka", elder sister, was theirs to give and they gave it to her. Lal Ded in fourteenth century Kashmir composed vakhs that are among the earliest surviving compositions in Kashmiri, putting nondual Shaiva teaching into a vernacular that had carried almost none of it. Mirabai in sixteenth century Rajasthan sang her padas to Krishna as Giridhar in Braj and Rajasthani, songs still in circulation among people who have never read a dharmashastra.

Two things need saying. The first is that none of these women claimed Vedic authority or asked for it. They composed in Tamil, Kannada, Kashmiri and Braj, and their standing came from the devotional stream, not from shastra. They are not a refutation of Manu. They are evidence that his ruling never held everywhere in practice.

The second is the cost. The traditional biographies of Akka Mahadevi and Mirabai are stories of women leaving households, facing hostility from their families, walking out of the roles the law books assigned them. The opening existed. Its width can be measured by what walking through it required.

What can be claimed, and what has to be conceded

The concessions first, because they are real. Gargi and Maitreyi are two named individuals in a very large corpus, and two exceptional women tell you nothing reliable about how ordinary women lived in Vedic India. The Upanishads record the arguments of a small learned elite, and that elite was overwhelmingly male. The Brihadaranyaka itself marks Maitreyi out as unusual. The later restriction is not drift that crept in while nobody was looking. It is written down, in plain language, in texts the tradition preserved and taught. And the tradition produced both records. It did not have one imposed on it.

What survives all of that is still worth stating. In the Brihadaranyaka and the Chandogya there is no rule barring a woman from the pursuit of Brahman-knowledge, and in Manu there is one barring her from the mantras. Gargi is not told to be quiet because she is a woman. She is told to stop where the men are told to stop, and one of the men who kept going lost his head for it. The tradition's own index of the Veda names women among the seers of revelation. The closure came later, from named human authors, in a category of text the tradition itself ranks below revelation and treats as subject to time and place.

That is a smaller claim than the apologetic version. It is also the one that holds up when somebody opens Manu and reads chapter nine.