Sovereign Over Her Mother-in-Law: The Rigveda's In-Laws and the Map Inside Their Names

Four Relatives in One Line

In a Hindi-speaking household nobody says “my brother-in-law” and leaves it at that. The one English word covers four distinct Hindi ones, and which you reach for depends on whose relative he is: devar for a husband’s younger brother, jeṭh for the elder one, sālā for a wife’s brother, bahnoī for a sister’s husband. A woman’s husband’s father is her sasur, his mother her sās, his sister her nanad. English, which has to borrow a legal phrase to say any of this, sounds vague by comparison, and the vagueness has a history: the “law” in “in-law” is canon law, the medieval church’s table of relatives you were forbidden to marry. The earliest of the English compounds, brother-in-law, is recorded from about 1300; the free-standing noun in-law, meaning any relative acquired by marriage, waits until 1894. [1]

Four of the Hindi words are older than the English ones by roughly two and a half millennia, and they appear together, in the same four relationships, in one verse of the Rigveda.

samrājñī śvaśure bhava samrājñī śvaśrvāṃ bhava

nanāndari samrājñī bhava samrājñī adhi devṛṣu

“Over thy husband’s father and thy husband’s mother bear full sway. Over the sister of thy lord, over his brothers rule supreme.”

RV 10.85, verse 46, trans. Griffith (1896). [2]

The words are śvaśura (श्वशुर), śvaśrū (श्वश्रू), nanāndṛ (ननान्दृ) and devṛ (देवृ): sasur, sās, nanad, devar, worn down by three thousand years of use and still doing the same work in the same four slots. Nobody revived them out of a dictionary. They were simply never put down.

What makes the Rigveda’s in-law vocabulary worth an afternoon is not its antiquity, which is impressive but inert. It is that the vocabulary is badly lopsided. The poets had names for the people on one side of a marriage and almost nothing for the people on the other, and the whole collection of a thousand hymns never once names an uncle or a cousin. Count the words, note which positions have no name at all, and a book of liturgy starts behaving like a document about residence: who moved at marriage, who stayed put, and whose fire the couple woke up beside.

The Verse That Holds the Whole Vocabulary

RV 10.85 is the wedding hymn, the liturgy for the marriage of Sūryā, the sun’s daughter, which doubles as the script for a human wedding and still supplies verses to Hindu ceremonies today. Verse 46 comes near the end, among the blessings said over the bride once she has reached her new house. Its key word is samrājñī, the feminine of samrāj, “sovereign, universal ruler”, the title given to kings and to Varuṇa. Four times in two lines the bride is told to be samrājñī: over her husband’s father, over his mother, over his sister, and, with the preposition adhi, over his brothers.

Notice the grammar of that last phrase. Three of the four in-laws are singular. Only the brothers are plural, devṛṣu, as though the expected household contained one father, one mother, one unmarried sister, and several brothers who stayed. Hold on to that detail; a set of skeletons from the Bronze Age steppe will come back to it later.

Notice also what a blessing is. You do not bless what is already guaranteed. The verse is the clearest statement in the Rigveda of a woman’s authority in her marital home, and for exactly that reason it is evidence of where the pressure lay: an incoming bride, young, outnumbered, unrelated to anyone in the room, standing in a house whose other occupants had known each other all their lives. Stephanie Jamison’s study of Vedic women reads the whole institution of marriage in early India through the same lens as the institution of hospitality, as a system for managing relations between what she calls mutually obligated strangers. [9] An affine is precisely that: someone you have no blood tie to and every obligation toward. For the hymn’s broader argument about the bride’s authority and public voice, see the close reading of the wedding hymn; for the women who speak in their own voices elsewhere in the collection, see The Hidden Women of the Rigveda.

Fifteen Words in Eleven Verses

Here is the entire affinal vocabulary of the Rigveda, every occurrence of it.

Fifteen words across eleven verses, in a collection of roughly 10,550. For comparison, the ordinary words for “wife”, jāyā and patnī, turn up in some dozens of verses each, and agni occurs thousands of times.

Methods note. These counts come from searching the romanised Saṃhitā text of the complete Rigveda for each stem and then reading every hit. The filtering matters more than the searching. A string search for śvaśura also returns svasur, the genitive of svasṛ “sister”, which is what Pūṣan’s notorious epithet at RV 6.55.4 contains; a search for nanāndṛ returns ānanda “joy” at 9.113.6; a search for snuṣā pulls in dhṛṣṇuṣeṇa and aśnuṣe. Grassmann’s 1873 concordance remains the fastest way to check a Rigvedic stem by hand. [18] Every verse cited here was inspected in Sanskrit, not taken from a word list.

Then there is the distribution, which is the first genuine surprise. Seven of the eleven verses are in Maṇḍala 10. The other four are in Maṇḍalas 1 and 8. The family books, Maṇḍalas 2 to 7, the oldest core of the collection, contain not a single in-law.

The tempting conclusion, that the in-law system is a late development, is almost certainly wrong, and the comparative evidence below shows why. The real explanation is about genre. The family books are overwhelmingly concerned with gods, cattle, enemies and soma. The domestic material, the wedding, the funeral, the gambler’s confession, the dialogue hymns, the satires, is concentrated in Maṇḍala 10, the youngest layer, where the collection begins admitting subjects that are not liturgy. The in-laws were always there. The hymns simply had no occasion to mention them until the poets started writing about households. On how that layering was worked out in the first place, see the stratigraphy of the Rigveda.

The Shape of the Gap

Lay the seven terms out by whose eyes they look through and the asymmetry is impossible to miss.

Four terms name the husband’s immediate family, and they are the words a bride needs on her first morning: his father, his mother, his sister, his brothers. One term, snuṣā, names her from the other direction, and it does not even get a verse of its own; it survives only inside the compound su-snuṣā, “having a good daughter-in-law”, in the strange comic hymn of Vṛṣākapi at RV 10.86.13, where Indra’s wife is complimented on her son and her son’s wife in one breath. Two terms look out from the bride’s natal family toward the men who took her: jāmātṛ, the daughter’s husband, and syāla, the wife’s brother, which occurs exactly once in the entire Rigveda.

graph TD
    EGO["A bride enters a house"] --> A["śvaśura, his father"]
    EGO --> B["śvaśrū, his mother"]
    EGO --> C["devṛ, his brothers, plural"]
    EGO --> D["nanāndṛ, his sister"]
    HIS["His people look back at her"] --> E["snuṣā, only in a compound"]
    HER["Her people look out at him"] --> F["jāmātṛ, daughter's husband"]
    HER --> G["syāla, wife's brother, once"]
    GAP["Never named in the Rigveda"] --> H["father's brother"]
    GAP --> I["mother's brother"]
    GAP --> J["any cousin"]
    GAP --> K["wife's sister"]

A vocabulary is a record of which relationships had to be named over and over, every day, by someone. The Rigveda’s affinal vocabulary is built almost entirely from the point of view of a household that receives a woman, plus a thin set of terms for the household that gives one away. That is not a theological statement, and no hymn argues for it. It is the residue of a living arrangement.

What a Son-in-Law Was Worth

The two wife’s-side terms both occur in a single line, and the line is a compliment to two gods.

aśravaṃ hi bhūridāvattarā vāṃ vijāmātur uta vā ghā syālāt

“For I have heard that ye give wealth more freely than worthless son-in-law or spouse’s brother.”

RV 1.109, verse 2, trans. Griffith (1896). [2]

Read that as a piece of social information rather than as praise. A poet is telling Indra and Agni that they are more generous than a man’s son-in-law or his wife’s brother. The comparison only works if a congregation heard it and thought immediately of a particular kind of man: an affine with his hand out, or an affine from whom you could ask. Vi-jāmātṛ, with its pejorative prefix, has to be translated with something like “sorry excuse for a son-in-law”. The Rigveda’s only word for the brother of your wife turns up in a sentence about stinginess, and in the modern languages the same word, Hindi sālā, has gone on to serve as an insult in its own right.

The son-in-law surfaces twice more, and both passages are oddly domestic. At RV 8.2.20 the poet asks that a god not spend his evening somewhere else, away from the ritual, aśrīra iva jāmātā, “like an unlovely son-in-law”: a man sulking at a distance instead of eating at his father-in-law’s fire. [2] At RV 8.26.21 and 22 Vāyu, the wind god, is addressed twice as tvaṣṭur jāmātaram, the son-in-law of Tvaṣṭṛ the divine artisan, and the epithet is used as grounds for a request, as if the relationship itself were a form of credit. The gods, in other words, have in-laws too, which is how you can tell affinity mattered. On Tvaṣṭṛ and the artisan gods, see Tvaṣṭṛ, the Ṛbhus and the artisan gods.

What these passages have in common is that none of them is about kinship. They are similes and asides, thrown off in the course of praising somebody. That is exactly what makes them good evidence. Nobody in the Rigveda is arguing about who counts as a relative, so nobody is shading the terms.

What Delbrück Heard in 1889

In 1889 Berthold Delbrück, a Sanskritist at Jena best known for his work on syntax, published a two-hundred-page monograph on Indo-European kinship names. [4] His method was simple and, for its moment, startling. Line up the kinship vocabulary of Sanskrit, Greek, Latin, Germanic, Slavic and the rest; keep only the words whose correspondences are regular enough to be inherited rather than borrowed; and then ask what kind of family the surviving set describes.

The pattern in that table is the argument. The terms that reconstruct cleanly to the ancestral language are terms for the husband’s father, the husband’s mother, the husband’s brother, the husband’s sister and the son’s wife. The relationships that are poorly represented, or represented only by words each branch had to invent for itself, are the ones on the wife’s side. In several branches the inherited words narrow still further: Greek, Armenian and Balto-Slavic use the reflexes of the old father-in-law word only for the husband’s parents. [8] A Russian wife today calls her husband’s brother dever’; a Hindi wife calls him devar; the two women are thousands of kilometres apart and using one word.

Delbrück’s conclusion, still the standard one, is that the ancestral society was patrilocal: at marriage the woman moved and the man did not. It follows that she needed a name for each person she now lived among, while her own relatives, left behind, could be named loosely or not at all. The vocabulary is the shape of her new day.

Two details in the table are worth lingering on. The word for the husband’s sister is the odd one out: Greek gálōs, Latin glōs, Old Church Slavonic zŭlŭva and Armenian tal all continue one inherited term, and Sanskrit does not. Nanāndṛ stands outside that set, etymologically isolated, and has often been explained as a nursery formation. [17] Somewhere on the road to the Punjab, Indo-Aryan appears to have let the ancient name for the sister-in-law go and coined a new one; a rare lexical giri- is sometimes adduced as the last relic of the old word. The relationship survived the loss of its name. And jāmātṛ sits in a cluster, with Greek gambrós and Latin gener, that goes back to a verbal root meaning “to marry”: the son-in-law is named not for his position in a family but for the act that created him.

Aside. Reading social history out of word lists is a method with a long rap sheet, and it was already being argued over when Delbrück wrote. Paul Friedrich’s 1966 reassessment of Proto-Indo-European kinship and Oswald Szemerényi’s 1977 monograph both show how much of the apparent precision in these reconstructions depends on which later meaning you project backward. [7][6] A kinship term can outlive the arrangement that produced it by thousands of years, which is why English still says in-law in a world with no canon-law marriage tables. The safe claim is narrow: the inherited vocabulary is heavily skewed toward the husband’s kin, and skews like that are easier to produce in a patrilocal society than in any other kind.

The Skeletons Agree

For a century the patrilocal inference rested on philology alone. It no longer does, and the independent evidence has arrived from an unexpected direction.

In 2017 a team led by Corina Knipper published strontium and oxygen isotope data, with ancient DNA, for 84 radiocarbon-dated skeletons from sites in the Lech valley in Bavaria, spanning the Final Neolithic and Early Bronze Age. Isotope ratios in tooth enamel record the geology of the water someone drank as a child, so a non-local childhood is visible in an adult skeleton. Most of the women in these cemeteries had grown up somewhere else. Most of the men had not. The pattern held for roughly 800 years, from about 2500 to 1700 BCE. [13] A 2019 study of 104 individuals from the same region found households in which the sons stayed and the daughters are simply missing from the burial record, having presumably married out. [15]

Closer to the world the Rigveda came out of, a 2023 study sequenced 32 individuals from a burial mound at Nepluyevsky in the Southern Urals, a Bronze Age pastoral community of the Srubnaya-Alakul tradition dated to roughly 1800 BCE. The genomes let the authors reconstruct an actual extended family. Descent ran through men. Monogamy was the norm. Residence after marriage was patrilocal, with women transferring into the husband’s group. And the core of the household, the relationship the whole structure was built on, was consanguinity between brothers. [14]

Go back to samrājñī adhi devṛṣu, “sovereign over the husband’s brothers”, the one plural in the blessing. A Vedic poet composing somewhere in the Punjab and a steppe village far to the north west are not the same society, and I want to be careful here: Nepluyevsky is not a Vedic site, its people were not the composers of the Rigveda, and a convergence is not a proof. But the household that the Rigvedic blessing assumes, a father, a mother, an unmarried daughter and several resident brothers, into which a young woman arrives alone, is the household the genomes describe. Two bodies of evidence that cannot contaminate each other, a nineteenth-century word list and a twenty-first-century genome, are pointing at the same living arrangement.

Where the Picture Frays

A tidy argument should be tested against the passages that do not fit it, and this one has several.

The gambler’s mother-in-law

The confession of the dice player at RV 10.34 is the collection’s great document of social ruin, and its third verse runs through the gambler’s relationships in order of collapse.

dveṣṭi śvaśrūr apa jāyā ruṇaddhi na nāthito vindate marḍitāram

“My wife holds me aloof, her mother hates me: the wretched man finds none to give him comfort.”

RV 10.34, verse 3, trans. Griffith (1896). [2]

Griffith’s “her mother” is an interpretation; the Sanskrit word is śvaśrū, mother-in-law. In a strictly patrilocal arrangement the gambler’s mother-in-law lives in another village, which makes her active hatred a slightly awkward fact. One can read it as evidence of marriage within short distances, or of in-laws who visited, or simply of news travelling. The better reading is the one the poet intends: the man has lost not only the people who are stuck with him by blood but the people bound to him by contract, which is worse, because affines can withdraw. Read the hymn next to the gambler’s hymn close reading and watch the sequence: dice, wife, mother-in-law, the ruined fields.

The other awkward passage opens RV 10.28, a dialogue hymn so compressed that scholars still disagree about who is speaking in which verse. Whoever it is begins: viśvo hy anyo arir ājagāma, mamed aha śvaśuro na jagāma, “every other stranger has come; my father-in-law alone has not come”, and then imagines him eating the grain, drinking the soma and going home satisfied (my rendering, keeping the Sanskrit order). An in-law who fails to turn up when he is owed a meal is a small domestic grievance of a kind the Rigveda rarely records, and it tells us that affinal visiting carried obligations in both directions.

The widow and her husband’s brother

The second occurrence of devara is a simile, and it is one of the strangest in the collection. The poets are asking the Aśvins where they have been all night:

ko vāṃ śayutrā vidhaveva devaraṃ maryaṃ na yoṣā kṛṇute sadhastha ā

“Who brings you homeward, as the widow bedward draws her husband’s brother, as the bride attracts the groom?”

RV 10.40, verse 2, trans. Griffith (1896). [2]

A widow drawing her husband’s brother to bed is used here as a figure for something ordinary and expected, which is the point. The verse is the oldest trace in Indian literature of what later law calls niyoga, the raising of a son for a dead man by his brother, and it sits alongside the funeral verse at RV 10.18.8 in which a widow is raised from beside the corpse and brought into the keeping of a man who “takes her hand”. [5] Neither passage legislates anything. They assume. For the funeral sequence itself, see the funeral hymns of the Rigveda.

The woman with no brother

If the brother is structurally important in such a system, the absence of one should be visible, and it is, twice, in a pair of similes that are hard to read comfortably.

At RV 1.124.7, Dawn approaching the world of men is compared to abhrāteva, “like a woman with no brother”, who “seeks men” and mounts her chariot as if to gather wealth. At RV 4.5.5 the poet reaches for the same image to abuse his enemies: they are abhrātaro na yoṣaṇo vyantaḥ, “like brotherless young women going astray”, like wives who betray their husbands. [2]

Two verses, two similes, and in both the woman without a brother is sexually unmoored. The structural logic is bleak and legible: a woman’s brother is the person who arranges her marriage and backs her afterwards, so a woman with no brother has no one to place her and no one behind her. That logic hardens over the following thousand years into law. Manu’s code warns a man away from marrying a woman who has no brother, and then provides the workaround, the putrikā or appointed daughter, whose son inherits from her father in place of the son he never had. [16] The Rigveda has none of this machinery. It has two throwaway comparisons, from which the whole later institution can be read backward.

No Word for Uncle

Now the absences, which are the best part.

The Rigveda has no uncle. Not the father’s brother, who is a perfectly good Indo-European relative (pitṛvya belongs with Latin patruus and German Vetter), and not the mother’s brother. Macdonell and Keith’s Vedic Index, going through the whole Vedic corpus term by term, found pitṛvya first in the Atharvaveda and mātula only in the Sūtra literature, with a single odd compound, mātṛbhrātra, “mother’s brother”, in the Maitrāyaṇī Saṃhitā. Their summary is that little is heard of the maternal uncle in the Vedic period at all, and that his prominence appears only with the epics. [5]

For a reader in Chennai or Madurai this is where the piece should become interesting, because in Dravidian kinship the mother’s brother is not a peripheral figure. He is the pivot. Where cross-cousin marriage is preferred, a girl’s mother’s brother is also, potentially, her father-in-law, and his son is her expected husband; the relationship carries names, claims and ceremonies. Thomas Trautmann’s Dravidian Kinship is a book-length formal analysis of systems built on exactly this, and Irawati Karve mapped the resulting contrast across the subcontinent, Sanskritic kinship in the north, Dravidian in the south, decades before it became a standard textbook distinction. [11][10]

The Rigveda sits firmly on the northern side of that line, and its silence about uncles is consistent with everything else in it. A system that marries its women out, to households that are not kin, has no structural use for a named cross-cousin, because the cross-cousin is not a spouse. It is the same instinct the collection shows when it uses the word jāmi, “kin”, as a term of refusal: too much likeness between two parties is a reason they cannot produce anything together. That logic, traced from the Rigveda into the ritual manuals, is the subject of the fault called sameness.

Aside. An argument from absence in a poetic corpus is weaker than an argument from presence, and should be labelled as such. The Rigveda has no word for uncle in the way it has no word for “west”: the gap is real, and the inference from it is a probability, not a proof. On what the Rigveda does not say about direction makes the same methodological point at length. What strengthens the case here is that the gap is not random. It falls exactly where a patrilocal, exogamous system would leave it, and the terms arrive in the later texts in the order that system would predict.

Worn Smooth by Three Thousand Years

The last thing worth noticing about these words is that they are still in use, and not in the way Sanskrit words usually are.

Most Sanskrit vocabulary in modern Hindi, Marathi or Bengali is tatsama, borrowed back into the language in its classical shape by educated speakers: rāṣṭra, vidyā, saṃskṛti. The in-law words are different. They came down the other way, mouth to mouth through Middle Indo-Aryan, battered by every regular sound change on the route, which is why they look nothing like their ancestors and why Ralph Turner’s comparative dictionary of Indo-Aryan can line up the whole descent. [12]

These are not loanwords. They are inheritances, and they survived because the relationships they name survived. A word stays in a language when people keep needing it in a hurry, in rooms, at weddings, in arguments. Nobody preserved devar out of reverence for the Veda. It is simply the word you use when your husband’s younger brother will not get off the phone.

Which makes this small patch of vocabulary the strangest kind of continuity in the Rigveda. The hymns were preserved deliberately, syllable by syllable, by a transmission system built to lose nothing, described in how the Rigveda was memorised. The in-law words needed none of that. They were never curated. They came through on their own, carried by households rather than schools, and arrived at the present with their meanings intact in a way most of the collection’s theology did not.

The Blessing and the Complaint

Read RV 10.85.46 and RV 10.34.3 one after the other. In the first, a bride is told she will be sovereign over her husband’s father, his mother, his sister and his brothers. In the second, a ruined man says his mother-in-law hates him. Same institution, same vocabulary, two directions. The hymns give you the aspiration and the wreckage and leave you to work out the distance between them, which is roughly what any honest account of family life would do.

What I find hard to shake is how little of this the poets meant to tell us. They were composing praise, and the praise required a comparison, and the comparison needed a figure everyone would recognise: a son-in-law who sulks at a distance, a widow and her brother-in-law, a woman with no brother, a father-in-law who did not show up. The social world comes through in the similes, under the argument, where nobody was guarding it. This is the same reason the Rigveda turns out to be a usable source on wells, on smells, on dress and on house construction: the incidental material is the honest material.

So a reading recommendation, and it takes ten minutes. Open RV 10.85 at verse 46 and read the blessing. Then open RV 1.109 and read a poet telling two gods that they give more freely than a worthless son-in-law. Then RV 1.124.7, and sit for a moment with the fact that the oldest surviving simile about a woman with no brother is a simile about the dawn. Three verses, three registers, one kinship system, and not a word of doctrine anywhere in it.