Where the Cattle Got Across

A Word With Mud On It

In the winter of 2025, officials in Uttar Pradesh tried to count the people bathing where the Ganges meets the Yamuna at Prayagraj. By the thirty-sixth day of the Maha Kumbh the state’s running tally had passed 540 million. The figure is an administrative estimate rather than an audited count, and the real number is unknowable, but nothing else people choose to do together happens on anything like that scale. The word for the place where it happens is tīrtha (Sanskrit: तीर्थ).

That word occurs ten times in the Rigveda. Not once does it mean a holy place.

It means a ford. Specifically it means the combination of features that lets a river be crossed: a bank that slopes gently enough to take a cart down, water shallow enough to wade, a bottom firm enough that a loaded ox will not sink to her belly in silt, and an exit on the far side that is not a cliff. In a landscape of seven large braided rivers, that combination is rare, local, seasonal, and worth knowing about. The Rigvedic poets ask the gods for it the way a modern traveller checks whether a bridge is open.

What makes this worth an afternoon is not the philology. It is that the Rigveda catches a metaphor in the act of forming. The same hymns that ask for a literal ford also start using the ford as a figure for anything that lets you get from a bad place to a good one, and once you can see both senses sitting side by side in the same collection, the later history of Indian religion reads differently. The Jain saviours are called tīrthaṅkara, “ford-makers”. The Buddhist goal is the far shore. The Hindu holy place is a crossing. Three traditions built their vocabulary of rescue out of one practical problem: how to get a herd across moving water without losing anyone.

This piece follows the word from the riverbank forward. The Rigvedic evidence first, because it is small enough to put on the table in full, then the vocabulary around it, then the comparative picture, then the point at which a ford turns into a shrine and someone starts telling you to travel there.

Ten Verses, No Shrines

Here is the whole Rigvedic file on tīrtha. Ten verses is few enough that nobody has to take my word for the pattern.

Six of the ten are literal or nearly so. Four are already similes, which is the interesting half, and we will come back to them.

Start with the most concrete. In a hymn to the Ādityas, the poet asks the gods to do something oddly specific:

“Look down on us, Ādityas, as a guide exploring from the bank. Lead us to pleasant ways as men lead horses to an easy ford.”

(RV 8.47.11, Griffith 1896) [2]

The Sanskrit is ādityā ava hi khyatādhi kūlād iva spaśaḥ, “look down from above, Ādityas, like scouts from the bank”. The gods are asked to do what a scout does: stand on high ground above the water and spot the crossing. Then sutīrtham arvato yathā, “to a good ford, as one leads horses”. This is not devotional language with a landscape attached. It is a job description.

Even better, because it is so unglamorous, is a verse to the Aśvins:

“Make a ford, Lords of splendour, where men well may drink: remove the spiteful tree-stump standing in the path.”

(RV 10.40.13, Griffith 1896) [2]

A ford with a good drinking place, tīrthaṃ suprapāṇam, and in the same breath, clear the stump out of the track. The gods are being asked for road maintenance. Nothing in the verse is interested in the riverbank as a site of presence or power. The riverbank is where the obstacles are.

Methods note. The counts here come from searching the Saṃhitā text itself for तीर्थ and गाध and reading every hit in context, not from a secondary list. Counting words in the Rigveda is less tidy than it sounds: compounds, variant sandhi and the odd disputed reading move a total by one or two, and the standard apparatus for the hard cases is still Grassmann’s lemma dictionary, Oldenberg’s textual notes and Renou’s commentary. [3] [14] [21] Treat “ten” as exact for the forms visible in the text and “roughly ten” for gādha.

Where scholars differ is on how much even this concreteness is about water. P.V. Kane, whose chapter on pilgrimage in the fourth volume of the History of Dharmaśāstra remains the fullest survey of the later material, read RV 1.169.6 and RV 4.29.3 as meaning simply “a way or a road”, and reserved “ford in a river” for RV 8.47.11. [5] Both readings land in the same semantic neighbourhood, which is the point: the word belongs to travel, not to cult. Kane’s own conclusion was blunt. The extension of tīrtha to a holy spot, any place made special by its water or by the presence of sages, belongs to a later stage of the language. [5]

The Problem That Would Not Go Away

To see why a word for “crossing” should be so load-bearing, look at what the Rigvedic people were trying to move and where.

Their world is the Punjab and the upper Indo-Gangetic plain: five to seven large rivers running roughly parallel, fed by snowmelt and monsoon, braided across wide beds that shift from year to year. [11] Their wealth is livestock, their mobility is carts and chariots, and their hymns are full of the logistics of going somewhere. A river in spate is not an obstacle you solve once. It is an obstacle that moves, deepens, and reappears in a different place next season. Wilhelm Rau’s reconstruction of early Indian material life from the Brāhmaṇa texts describes a society still organised around herds and movement rather than around towns. [12] The archaeology of the region gives no bridges worth the name for this period, and the text mentions none: setu, which in classical Sanskrit is the ordinary word for a bridge, in the Rigveda means a dam or a binding, something that holds water back or ties a person down.

So the Rigveda has a technical vocabulary for crossability, and it is larger than the tīrtha file.

Read down that column of glosses and notice what the poets are asking for. Not a miracle. A place to stand.

“May we obtain firm ground and room for resting.”

(RV 5.47.7, Griffith 1896) [2]

Aśīmahi gādham uta pratiṣṭhām. A shallow and a footing. In another hymn the request is almost comically practical: “find us this day a ford wide and extensive” (RV 10.113.10), vido ṣu ṇa urviyā gādham adya. Today. Wide. The gods of the Rigveda are being asked for a survey result.

The single most vivid crossing scene in the collection is RV 3.33, where the poet Viśvāmitra, travelling with the Bharatas and their carts, arrives at the confluence of the Vipāś and the Śutudrī (the Beas and the Sutlej) and talks to the rivers. They are described first as two mares let loose, then as two mother cows licking their calves, and then the negotiation starts. He asks them to subside. They answer. He insists. They agree, and the host fords the water.

“List quickly, Sisters, to the bard who cometh to you from far away with car and wagon. Bow lowly down; be easy to be traversed: stay, Rivers, with your floods below our axles.”

(RV 3.33.9, Griffith 1896) [2]

Adho akṣāḥ, “below the axles”. The unit of religious petition here is the clearance under a cart axle. I know of nothing else quite like it in ancient literature: a hymn whose request is a water level, stated in terms of vehicle geometry. Two verses later the crossing succeeds and the poet says so plainly: atāriṣur bharatā gavyavaḥ, “the Bharatas, seeking cattle, have crossed over” (RV 3.33.12).

A Miracle That Looks Like Engineering

The Rigvedic gods do intervene in water, and what they do is instructive. In the hymn on the Battle of the Ten Kings, where Sudās and the Bharatas face a confederation on the Paruṣṇī, Indra’s decisive act is not a parted sea:

“What though the floods spread widely, Indra made them shallow and easy for Sudās to traverse.”

(RV 7.18.5, Griffith 1896) [2]

Gādhāni akṛṇot supārā. He made them wadeable and easy to cross. The god’s miracle is delivered in the technical vocabulary of fords. Elsewhere Indra “made a ford for Turvīti” (RV 1.61.11), carried “those who could not swim” over safely (RV 2.15.5), and brought Turvaśa and Yadu across when they feared the water (RV 4.30.17). The Aśvins’ most famous rescue is lifting Bhujyu out of open water and getting him home in a ship (RV 1.116.5, RV 7.69.7), which I have discussed alongside the Rigveda’s thin but real sea vocabulary in the ships and the shipwreck.

Put those together and a pattern shows. A striking proportion of Rigvedic divine competence is ferry work. The gods are the ones who know where the crossing is, who lower the water, who carry non-swimmers. Long before anyone in India writes about liberation, the standard demonstration that a god is worth praising is that people on the wrong side of a river got to the right side.

Aside. This is a good place to resist a tempting slide. That the gods are ford-finders does not mean the Rigvedic poets were speaking allegorically about salvation and merely using river imagery. The Bharatas’ carts are real carts. Read RV 3.33 as allegory and you lose the one verse in the collection that tells you how high a Vedic axle sat above the ground.

The Figure Forms

Now the four similes, which are where the word starts to travel.

“The hymn is formed, poured are the allotted portions: as to a ford friends come unto the Wondrous.”

(RV 10.31.3, Griffith 1896) [2]

Tīrthe na dasmam upayanti. Companions converge on the generous god the way people converge on a crossing. The simile works because a ford is a place everyone must use: tracks from a wide region narrow into one line of approach. Anyone who has watched animals come down to water knows the geometry.

At RV 1.173.11 the ford is what pulls the thirsty: sacrifice draws Indra as the watering place draws an animal that needs to drink. At RV 9.97.53 the pressed soma flows “to the famed ford” of the god, a landing place where something arrives. And at RV 10.114.7, in one of the late riddling hymns, the ford becomes knowledge itself:

“Who will declare to us the ford Apnana, the path whereby they drink first draughts of Soma?”

(RV 10.114.7, Griffith 1896) [2]

Griffith took āpnāna as a proper name. The form is more naturally a participle, something like “reached” or “attainable”, which would make the phrase “the ford one can get to”, and the commentators have never fully agreed. Either way the verse asks for directions to a place that is not on any riverbank. Somebody in the Rigvedic tradition has already worked out that the most useful thing you can tell a person is where the crossing is, and has started using that for things other than rivers.

The same transfer happens in a formula that is not about tīrtha at all, and which runs through the collection like a refrain. Two hymns use it in nearly identical words:

“Varuna-Mitra, may your path of Order bear us o’er trouble as a boat o’er waters.”

(RV 7.65.3, Griffith 1896; the same phrase at RV 6.68.8) [2]

Apo na nāvā duritā tarema: “may we cross over the hard passages as one crosses waters with a boat”. Durita, the thing to be crossed, is literally a “bad going”, a hard stretch of road. Agni is asked to “carry us across all hard places as across a river by boat” (RV 1.99.1). Indra is asked, in a line worth memorising, to “be a good crossing, carry us over” (RV 6.47.7), bhavā supāro atipārayo naḥ.

And then, in the tenth book, a verse that reads like the hinge of the whole history:

“Here flows Aśmanvatī: hold fast each other, keep yourselves up, and pass, my friends, the river. There let us leave the Powers that brought no profit, and cross the flood to Powers that are auspicious.”

(RV 10.53.8, Griffith 1896) [2]

Ut tiṣṭhata pra tarata sakhāyaḥ: stand up and cross over, friends. The first half is instruction for a group at a real river with a stony bed. The second half leaves behind what is not auspicious and crosses to what is. Nobody had to invent the metaphor of spiritual crossing. It was lying on the bank where people already stood.

Two Roots for One Riverbank

The comparative picture sharpens the point, because Indo-European languages have two separate inherited roots for getting across water, and Sanskrit kept both.

Two details in that table are worth sitting with. The English word ford, the German Furt in Frankfurt, the Latin portus behind every port in Europe, and the Sanskrit pāra, “the other side”, are all the same inherited root. And the Greek strait between Europe and Asia is the Bosporus, Bous-poros, the “cattle-crossing”. Herding peoples who spoke related languages, spread across thousands of kilometres, named their water crossings after the animals they were trying to get across, and then built their words for passage, trade and arrival on top.

Aside. Etymology of this kind proves a shared inheritance of words, not a shared religion. The Bosporus is named for cattle because cattle crossed there, not because the Greeks and the Bharatas shared a doctrine. What the comparison does establish is that “crossing” was a problem important enough, in the lives of early Indo-European speakers, to generate the vocabulary that later carried their most abstract ideas. Compare the Rigvedic use of yoking for discipline, which I traced in the word before yoga.

How a Ford Becomes Holy

If the Rigveda has fords and no pilgrimage, when does the holy place arrive?

Not in the Rigveda’s own ritual logic, which runs in the opposite direction. The Vedic sacrifice is portable. A patron engages priests, a ritual ground is marked out and a fire is kindled where the patron is, the gods are invited to come and sit on the strewn grass, and when the rite is finished the ground is abandoned. [16] Nothing accumulates. There is no building and therefore no place to return to. The sacred is a scheduled event, not an address. The Rigveda praises rivers lavishly, most famously in the river hymn at RV 10.75, which I have read closely elsewhere, and never once suggests that anyone should go and stand in one for the good of their soul.

The shift happens in stages, and the stages are visible.

graph LR
    A["Rigveda: find the ford"] --> B["Brahmanas: rites that move along a river"]
    B --> C["Epic: lists of tirthas to visit"]
    C --> D["Dharmasastra: rules, merit, itineraries"]
    D --> E["Today: tirthayatra, Kumbh, the ghats"]

In the Brāhmaṇa literature, ritual starts to be tied to specific stretches of ground. The Pañcaviṃśa Brāhmaṇa describes the Sārasvata sattras, long sacrificial sessions performed along the Sarasvatī, with the consecration at Vinaśana, the place where the river disappears into the sand, and the rite proceeding along the river’s course. [15] A sacrifice that walks is a new kind of object. The Sarasvatī’s hydrology, and what satellite imagery and geology have to say about it, is a story of its own, told here.

By the Mahābhārata, the thing is fully formed. The third book contains a long section, the Tīrthayātrā Parva (chapters 80 to 157 in the vulgate numbering), in which the sage Nārada recommends an itinerary of tīrthas to the Pāṇḍavas and the text catalogues what each one is worth. [17] The word now means a destination, the journey to it is a religious act, and the benefits are specified. By the time the Dharmaśāstra writers get to work, there are rules: who should go, when, in what state, with what results, and Kane’s survey of that literature fills a chapter and a separate published list of sites. [5]

Diana Eck’s analysis of this material, first in a 1981 article and later in a book-length treatment of Indian sacred geography, turns on the word’s literal sense. A tīrtha is a crossing, and the crossing works both ways: it is where human beings pass over to the far shore and also where the divine comes down. [6] [7] The Sanskrit for a god’s descent, avatāra, is built from the same root as tīrtha, with ava, “down”: a crossing-down. The architecture of a bathing ghat makes the old meaning concrete in stone. Steps down the bank, a place to stand in the water, a way back up. It is a built ford, and the fact that pilgrims no longer need to cross the river has not changed the shape of the thing they walk into.

Aside. One caution about arguments from absence. That the Rigveda never prescribes pilgrimage is an argument about a text, not about a population. The Saṃhitā is a professional poets’ collection, concerned with hymns commissioned for particular rites; ordinary religious behaviour that no patron paid a poet to describe would leave no trace in it. What we can say is narrower and still substantial: the organised religion of the Rigvedic elite had no sacred destinations, and its vocabulary for tīrtha shows no sign of having lost any.

The Ford-Makers

The payoff of this history is a set of titles.

In Jainism, the twenty-four liberated teachers of each age are tīrthaṅkara, “ford-makers”. The ford in question is a crossing over the flood of rebirth, and the fourfold community of monks, nuns, laymen and laywomen is itself sometimes described as the ford that the teacher establishes. [19] [20] Buddhist soteriology prefers the other Indo-European root: the goal is pāra, the far shore, and the perfections that take you there are the pāramitās, a word often analysed as pāram-ita, “gone to the far shore”. The raft that gets you across and should then be left behind is one of the best known similes in the canon. In Hinduism, the holy place keeps the word tīrtha itself.

Three traditions, three different technical vocabularies of release, and all of them are river-crossing words.

graph TD
    A["root tr: cross over"] --> B["tirtha: ford, access"]
    A --> C["duritā tarema: may we cross troubles"]
    A --> D["avatara: a crossing-down"]
    B --> E["tirtha: bathing place"]
    E --> F["tirthayatra: pilgrimage"]
    E --> G["tirthankara: ford-maker"]
    C --> H["para: the far shore"]
    H --> I["paramita: gone beyond"]

Why this metaphor and not another? Because of what a ford actually is in the life of a herding people. On a braided Punjabi river the difference between a crossing and a drowning can be two hundred metres of bank, and the knowledge of which two hundred metres is local, perishable, and held by particular people. Somebody who knows where the ford is can save your family, your cart and your animals in an afternoon. That is an unusually exact model for what a teacher or a god is supposed to do: not remove the river, not carry you personally, but know where the crossing is and point at it. The Rigvedic poets asked the Ādityas to behave like scouts on a bluff. The request has an afterlife.

Notice also what the image refuses to promise. A ford does not drain the river. You still get wet, the current still pulls, and people who are careless at fords still drown. The Rigvedic version of divine help is unsentimental in exactly this way: bhavā supāro, be a good crossing. The gods are asked to improve the odds on a passage that remains the traveller’s to make.

Reading the Bank

The Rigveda is usually approached for its cosmology, its ritual, or its historical puzzles. It also happens to be the only place in the Indian record where some of the tradition’s largest words can still be seen with the mud on them. Tīrtha is the clearest case. In ten verses it is a shoal, a slope, a watering place, a thing you ask the Aśvins to build and to clear of stumps, and in four of those ten it is already beginning to mean something else.

What I find bracing about this is the direction of travel. We are used to being told that ancient religious language is metaphorical and that the trick is to decode it. Here the opposite holds. The concrete came first and stayed concrete for a long time, and the spiritual vocabulary of three major traditions was quarried out of it. The Rigvedic poets were not thinking about crossing the ocean of existence. They were thinking about getting cows across the Sutlej in June, and because they thought about it so hard and said it so well, their grandchildren’s grandchildren had a word for everything else.

Read it in the order the text suggests. Open RV 3.33 next to a satellite image of the Beas and Sutlej where they meet, and watch a poet bargain a water level down to axle height. Then read RV 7.18.5, where a god does the same thing for an army. Then read RV 10.53.8 and notice how little distance there is between “stand up and cross over, friends” and everything that came after. If you want the next step in the same direction, the Rigvedic vocabulary of being hemmed in and getting free is the companion piece to all of this, and it is here; the measuring of the journeys themselves is here.