The Sarasvati River

← Encyclopedia

A river runs through the second book of the oldest of the Vedas, and the poet who sings it can find no higher word. He calls her ambitame, naditame, devitame — best of mothers, best of rivers, best of goddesses — and then, in the same breath, confesses that his people are as if of no repute, and asks her for distinction (Rigveda 2.41.16). That is the strange thing about the Sarasvati. She is the greatest of waters in the hymns, sung above all the rivers of the world, and she is also, by the end of the same scripture, a river that the singers can no longer find. She broke the ridges of the hills with her waves; she sank into the sand; she became a goddess and a memory and a dry channel argued over by satellites. The hymn praises her at the height of her flood. History meets her already vanishing.

The river of the hymns

The Sarasvati belongs to the Rigveda before she belongs to anything else, and the Rigveda is a text known with rare exactness. For more than two thousand years the hymns were carried not in writing but in sound — recited by priestly families with fixed accent and pitch and a redundant machinery of cross-checks that caught a dropped syllable before it could spread. So when the hymns describe the Sarasvati, the description itself is secure, even where the river it describes is not. The words have not drifted. Only the water has gone.

In the early books the river is enormous. The Sarasvati of the Veda comes down in flood from the mountains and runs to the sea — naditame, the most riverlike of rivers, the one against which the others are measured. The sixth book, the book of the Bharadvajas, gives her the image that fixed her in memory: “She with her might, like one who digs for lotus-stems, hath burst with her strong waves the ridges of the hills” (RV 6.61). The same hymn calls her seven-sistered, sprung from a threefold source — set among the seven streams of the northwest, a sister to the other rivers and the chief of them. She is not gentle here. She is a force that cuts rock.

The seventh book, the Vasishtha book, places her on the map. “Alone of the rivers Sarasvatī shows clear,” the verse runs in a recent rendering, “as she goes gleaming from the mountains all the way to the sea” (RV 7.95.2) — a single river running the whole span from the high snows to the ocean, distinct and pure among all the rest. The picture is consistent: a mighty perennial stream, fed from the mountains, broad enough to swallow its banks, the dominant water of the land the hymns describe.

The Sarasvati is more than scenery in these hymns. Her banks are where the tradition locates its own beginnings — the place where the seer-poets heard the hymns and the first fires were laid, the heartland later texts would call Brahmavarta, the holy ground between the Sarasvati and a second, smaller river, the Drishadvati. To sing of the Sarasvati was to sing of the cradle of the rite. She carried the libations; she nourished the five tribes; she was invoked, the sixth book says, in every deed of might. The river and the religion grew on the same ground.

There is one hymn that lets us see the river in its place among the others. The tenth book closes with the Nadistuti, the praise of rivers, in which the poet calls the streams of the northwest by name in a sweep that runs from east to west: Ganga first, then Yamuna, then Sarasvati, then the Shutudri — the Sutlej — and onward through the Ravi, the Chenab, the Jhelum, to the Indus, the mightiest, addressed again and again (RV 10.75). Read that order against the land and it traces a real progression across the Punjab. The Sarasvati sits in it exactly where a great river ought to sit: between the Yamuna and the Sutlej, holding her own place in the line.

The river that failed

The Sarasvati of the Veda does not stay at her height. Within the span of the hymns themselves, and more plainly in the texts that came after, the great river begins to fail. The flood thins. The stream that ran from the mountains to the sea no longer reaches the sea. By the time of the prose works that expound the sacrifice — the Brahmanas, the manuals composed as the Vedic age gave way to what followed — the Sarasvati is a river that ends in the desert. She does not pour into an ocean. She disappears.

The later tradition gave the disappearance a name and a place. Vinashana — the vanishing, the place of loss — is where the Sarasvati sinks into the sand and is seen no more. The Mahabharata tells of pilgrims who walk her course to the spot where she is lost, and the loss is not treated as a defeat but as a wonder: the river that dives beneath the earth. The Brahmanas already speak of her diving under, going hidden. What had been the best of rivers became the river that goes unseen — and the place of her vanishing became, in turn, a place to go on pilgrimage, a tirtha marked precisely by the absence of the water that had made it holy.

Here the Vedic frame and the documentary record run side by side without contradiction, and it is worth keeping them apart in the telling. Within the frame, the river did not die; she withdrew, choosing the hidden way, and her sanctity went with her underground. As a matter of record, the texts themselves register a change in the land: the early hymns describe a great flowing river, the later texts describe its end in the sand, and the two pictures are too different to be the same river in the same state. Something in the water of the northwest changed across the centuries the Veda spans. The hymns caught the river at its fullest. The later texts caught the silence after.

From water to word

The most consequential thing the Sarasvati did was to stop being only a river. Her name carried over, intact, to a goddess — and not a minor one. Sarasvati became the lady of speech and of learning, of the word and the song and the right ordering of sound, the patron of every utterance and every art that depends on it. The same four syllables that name the lost water of the Veda name the goddess before whom students and musicians and poets still set down their instruments to be blessed.

The bridge between the two is an older goddess, Vac — Speech herself, personified in the Rigveda as the divine Word, the power by which the hymn is uttered and the rite made effective. The Brahmanas repeatedly identify Sarasvati with Vac, folding the river into the goddess of the spoken sacred. It is a fitting marriage. The river had always been bound to the word: she ran past the place where the hymns were heard, she carried the offerings, she was herself invoked in measured speech. To make the river of the hymns into the goddess of the hymn was less a leap than a recognition. The water that nourished the rite became the goddess who governs its sound.

In the developed tradition Sarasvati is the consort and the mind-born daughter of Brahma the creator, called Vagdevi, the goddess of speech, and Vedamata, mother of the Vedas — the source from which the scriptures and all the branches of knowledge are held to flow. Her image is fixed: robed in white, seated with a veena across her arms, a book and a string of beads in her hands, a swan beside her. The swan carries its own teaching. It is said to separate milk from water when the two are mixed, and so it stands for the discernment the goddess grants — the power to part the true from the false. The river that once parted the ridges of the hills became the goddess who parts truth from error.

Goddess Saraswati, oil by Raja Ravi Varma, 1896 — the river of the Veda transformed into the lady of speech and learning, with her veena and swan. Wikimedia Commons, public domain.

The hidden third river

There is a second afterlife, and it keeps the river as a river even after it has vanished from the map. The doctrine of the hidden Sarasvati holds that she did not simply end at Vinashana but went underground and runs still, unseen, beneath the surface of the land — surfacing at last far to the east, at the great confluence where the two visible holy rivers meet.

That confluence is Prayag, where the Ganges and the Yamuna join. The tradition names it the Triveni Sangam — the braiding of three streams — and the third strand is the Sarasvati, present and invisible, joining the two waters one can see. She is the unseen thread that makes the holiest of confluences holy. The Ganges runs pale and the Yamuna runs dark and where they meet a third current, neither pale nor dark, is held to rise from below: the lost river of the Veda, come back to the surface for the length of a confluence and gone again. To bathe at the Sangam is to touch all three. The river that could no longer be found became the river that is everywhere and nowhere — the secret water at the heart of the sacred map.

This is the great work the name performs. It binds the geography of the oldest hymns to the sacred geography of the later faith. The Sarasvati of the Punjab, lost before history, reappears as the hidden current at a confluence a thousand miles to the southeast — not as a claim about hydrology but as a claim about continuity, a thread drawn from the cradle of the rite to the pilgrimage-ground of the living tradition. The river vanished, and the vanishing became the means of her survival. A river one can see can only be where it is. A hidden river can be wherever it is needed.

The search for the bed

And yet the question of the actual river — where it ran, and whether its bed can still be found — never went away. For more than a century scholars have looked for the Sarasvati on the ground, and the search has fixed on one candidate: the Ghaggar-Hakra, a great seasonal channel that runs through Haryana and Rajasthan in India and on into Cholistan in Pakistan as the Hakra, lying exactly where the Veda places the Sarasvati — between the Yamuna and the Sutlej — before it loses itself in the Thar desert. The Ghaggar today is a monsoon stream that flows only in the rains and dries before it reaches the sands. But satellite imaging and ground survey have traced, beneath and beside its present course, the broad bed of a much larger paleochannel running the same line toward the desert. That dry channel is what the identification rests on.

The Ghaggar-Hakra channel and its paleochannels, between the Sutlej and the Yamuna, ending in the Thar desert — the bed proposed as the lost Sarasvati. Wikimedia Commons.

The identification is genuinely contested, and the contest is sharp because it is knotted to a larger and more charged question. Two readings of the dry channel stand against each other.

On one side, the bed is the Sarasvati. A paleochannel that broad was once carved by a large perennial river; the river was fed from the mountains by the Sutlej and the Yamuna, which in an earlier age flowed into it before each turned away to its present course; when they diverted, the river was left to the monsoon, dwindled, and died in the sand — exactly the arc the texts describe, from mighty stream to disappearance at Vinashana. The dense band of settlement supports it: hundreds of sites of the mature Indus Valley civilization, roughly 2600 to 1900 BCE, cluster along the Ghaggar-Hakra’s course, the mark of a river that once watered a populous land. On this reading the lost Sarasvati is found, and the hymns remember a real river that ran in the northwest within human memory.

On the other side, the bed is not the Sarasvati, or not in the way the first reading needs. Recent geophysical work argues that the great paleochannel is an old course of the Sutlej, which diverted away between roughly fifteen and eight thousand years ago — long before the urban Indus period, and far too early to be the perennial river the hymns sing. On this view the Ghaggar-Hakra in the relevant centuries was already a monsoon-fed seasonal stream, not a glacier-fed torrent; the large rivers had gone before the cities rose. One influential study turns the settlement evidence around entirely: it was the river’s departure, not its arrival, that allowed the Indus towns to spread along a stable, no-longer-flooding bed. If the channel dried before the hymns were composed, then the Veda cannot be remembering it in flood — and the identification, however natural it looks on a map, fails on the dates.

The reason the argument runs so hot is that it touches the deepest and most disputed question about the people of the hymns: who the Indo-Aryans were, and where and when they composed the Veda. The two leading models pull in opposite directions, and the Sarasvati’s dating bears on both. One holds that the speakers of the Vedic language entered the northwest from the steppe in the second millennium BCE; the other, the indigenist or “Out-of-India” view, holds that they were native to the land of the hymns from far earlier. A perennial Sarasvati surviving into the time of the hymns would press the Veda earlier, toward the indigenist reading; a channel that dried millennia before would not. The hydrology has become a proxy for the history, and the dates of a dry riverbed are argued with a heat the dates alone would never generate.

A note of caution belongs here, because the older term for these people has a poisoned history. “Aryan” in the Veda is a self-designation of language and culture, not a race; the racial theories built on the word in the nineteenth and twentieth centuries, and the use the Nazi state made of them, were a later imposition that the texts do not support. The migration question is a real and open problem of archaeology and linguistics. It is not settled by the word, and the word should not be made to carry what it does not mean.

What flows now

The Sarasvati flows now in three places, and a river is in none of them. She flows in the name of the goddess, before whom the instruments are still set down and the books still blessed — the lady of speech who outlived her own water. She flows in the myth of the hidden confluence, the invisible third strand at Prayag, present wherever the tradition needs her and visible nowhere. And she flows, perhaps, in a dry bed across the northwest, traced by satellite and argued over by scholars who cannot agree whether it is her grave or someone else’s.

Of the great rivers of the tradition, she is the only one whose sanctity does not depend on her water. The Ganges is holy because it flows; pilgrims go to its banks to touch the living stream. The Sarasvati is holy because she is gone — because she withdrew into the earth and the name went on without her, the best of mothers become the mother of the Vedas, the river that broke the hills become the goddess who parts the true from the false. The hymns caught her once at her full flood, bursting the ridges with her waves. Everything after is the long life of a river that was already, when the last singers looked for her, sinking out of sight into the sand.

→ In the library: Woodroffe — Hymns to the Goddess (1913)

→ Related: Rigveda · Prayag · Kurukshetra · The Indo Aryans · The Rishis · Ganges · Varanasi · Indus Valley Civilization · Historical Vedic Religion · Hinduism · Indra · Agni · Varuna · Ushas · Yajna · Soma · Brahman · Mahabharata · Vishnu · Shiva