Set the Troughs, Fasten the Straps: The Rigveda's Well

A Work Order in the Tenth Book

Somewhere in the middle of the tenth book, the Rigveda stops singing and starts issuing instructions.

The hymn is RV 10.101, addressed to the All Gods, and it opens like any other invocation: wake up, friends, kindle the fire together. Then it drifts. Build a boat with oars. Yoke the ploughs, stretch the yoke-poles apart, drop the seed into the cut furrow. And then, at the fifth verse, it becomes something no other passage in the collection quite is: a parts list.

Make the troughs ready. Fasten on the leather straps. Let us pour out the well that runs with water, the good-pouring well that is never used up.

(RV 10.101.5, translated from the Sanskrit; compare Griffith 1896, “Arrange the buckets in their place, securely fasten on the straps”)

Two verses later the poet adds the fittings: droṇāhāvam avatam aśmacakram, a well with wooden troughs and a stone wheel, and aṃsatrakośam, a bucket described by a word that elsewhere means shoulder-armour, presumably a leather vessel reinforced the way body armour is. The verse ends by saying what the whole arrangement is for: nṛpāṇam, a drinking place for men.

This is a machine. Not a chariot, not a pressing-stone, not the fire-drill. A machine for defeating gravity, assembled out of wood, rope, leather and one dressed stone, and the Rigveda tells you how to put it together.

What follows is an attempt to take that machine seriously: to work out what its parts were, how much water it could actually raise, why the text sorts water into categories according to whether a human being dug for it, and why, once you have seen the rig, you start noticing it everywhere else, including in the sky.

Methods note. The counts above are my own, made by hand over the romanised Saṃhitā text, and they are the kind of number that rewards suspicion. Sanskrit word-joining hides avataṃ inside aśmāsyamavataṃ at RV 2.24.4, and a naive search also drags in avata as a verb, “help us”, which is a completely different word from avata the noun. Nine is the count after the verb forms are stripped out and the sandhi is opened up: RV 1.85.10, 1.85.11, 1.130.2, 2.24.4, 4.50.3, 8.72.11, and three times in RV 10.101. Anyone recounting should expect to land within one of that.

The Parts List

Vedic water vocabulary is small, and its smallness is informative. There are two words that matter, and they are not synonyms.

Utsa is a spring: water that arrives on its own. Avata is a well: water that arrives because somebody dug. Macdonell and Keith drew the line in exactly those terms in the Vedic Index, noting that avata “denotes a well, artificially made (khan, ‘to dig’) in contrast with a spring (utsa)”, while conceding that utsa sometimes drifts over and covers artificial wells too [1]. The distinction is not tidy, but the poets clearly felt it, and one verse (below) makes the categorisation explicit.

Around those two nouns sits a set of fittings, most of which appear only in this one hymn.

Read the table and a small, specific object assembles itself. There is a shaft. There are wooden troughs at the top, into which the raised water is tipped so that people and animals can drink without leaning over the hole. There is a leather bucket, and a leather strap, and a piece of stone shaped as a wheel or a ring at the mouth. There is a verb for the whole operation, sic, to pour, and the poets use it of the well as if the well itself were doing the pouring: siñcāmahā avatam, “let us pour out the well”, as though water came out of the ground the way it comes out of a jar.

What the vocabulary does not contain is just as pointed. There is no word for a dam. No word for a canal in the engineered sense; kulyā is a small runnel, the kind of thing you cut with a hoe, and it appears three times, twice in similes about milk and Soma. No word for a reservoir. And crucially, no word for a chain of pots on a wheel. The Rigvedic well is a hole with a rope over it.

Is the Wheel a Pulley or a Rim?

Everything hangs on aśmacakra, and the translators do not agree.

Griffith, in 1896, rendered RV 10.101.7 as “Pour forth the well with stone wheel, wooden buckets, the drink of heroes, with the trough for armour” [2]. Wilson, working from Sāyaṇa’s commentary, produced something noticeably flatter: “pour out (the water of) the well, having wooden cattle-troughs, having a stone rim, having a receptacle like armour, fit for the drinking of men” [3]. Griffith’s stone is a wheel that turns. Wilson’s stone is a kerb that sits still.

The difference is not trivial. A stone wheel with a strap running over it is a pulley, and a pulley is a genuine mechanical advantage: it lets you haul downwards with your body weight instead of upwards with your arms, and it stops the rope sawing itself apart on the well’s edge. A stone rim is a well-curb, a structural collar that keeps the shaft from collapsing and the topsoil from washing in. Both are real technologies. Only one of them is a machine.

Macdonell and Keith took the pulley reading, describing water “drawn up by a wheel of stone with a strap attached to a pail, and emptied into wooden buckets” [1]. Later handbooks have repeated it until it hardened into a fact, and it now appears in review papers written by hydrologists rather than Sanskritists, sometimes with additional equipment that is not in the text at all.

Three observations are worth putting on the table.

First, the word itself decides nothing. Cakra is the ordinary Rigvedic word for a chariot wheel, and a chariot wheel unquestionably turns; but cakra also names anything disc-shaped, and a circular stone curb is disc-shaped. The compound is a hapax, so there is no second occurrence to test the sense against.

Second, the context tilts slightly toward Griffith. The immediately preceding verses are about the strap, varatrā, and the same word is used at RV 4.57.4 for the traces that bind a plough team. Straps in this hymn are things that transmit force. A verse that names a strap, a bucket and a stone disc in sequence is at least describing a system in which the strap moves across the stone.

Third, and against Griffith, no true pulley has ever been recovered from a South Asian Bronze Age well. What the archaeology gives is the shaft. At Mohenjo-daro the wells are cylinders of wedge-shaped brick, a specialised brick made in a curved form so that the courses lock into a circle, and Michael Jansen’s survey of the excavated areas extrapolated to something on the order of 700 wells across the city, a density he described as unparalleled in the history of urban water supply [4]. Harappa, by contrast, may have had as few as thirty. Those are draw-wells, worked by rope, and a grooved or rounded stone lip does the same job for the rope that a pulley does, less efficiently and with no moving parts.

The honest conclusion is that aśmacakra is a load-bearing stone at the mouth of the well over which a strap ran, and that whether it rotated is unresolved and probably unresolvable from the text alone. That is less exciting than “the Rigveda knew the pulley”, and considerably more likely to be true.

Dug Water and Water That Made Itself

The most philosophically interesting verse about wells is not in the well hymn. It is in a short hymn to the Waters in the seventh book, and it sorts water into kinds.

Waters which come from heaven, or those that wander dug from the earth, or flowing free by nature, bright, purifying, speeding to the Ocean: here let those Waters, Goddesses, protect me.

(RV 7.49.2, Griffith 1896; see RV 7.49)

The Sanskrit is divyā, sravanti, khanitrimā, svayaṃjā: heavenly, flowing, dug, self-born. Sāyaṇa glosses khanitrima as water “formed by digging”, canals or reservoirs, and it is the only occurrence of the word in the collection [3].

Notice what the taxonomy is doing. It does not sort water by taste, by purity, by whether it is fit to drink, or by which god owns it. It sorts water by where it came from, and one of the four categories is defined by human labour. Dug water is its own kind of water. It is still divine, still addressed as a goddess in the same breath as rain, but its provenance is a man with a spade.

graph TD
    A["Waters, RV 7.49.2"] --> B["divya: from heaven"]
    A --> C["sravanti: flowing"]
    A --> D["khanitrima: dug out"]
    A --> E["svayamja: self-born"]
    B --> F["samudra: the gathering sea"]
    C --> F
    D --> F
    E --> F
    D --> G["avata: the well"]
    E --> H["utsa: the spring"]

That a sacred hymn should list human excavation alongside rainfall as a legitimate origin for holy water is a quietly radical thing, and it is not an isolated slip. RV 10.97.20, in the great hymn to healing plants, blesses the digger as well as the person dug for: “unharmed be he who digs you up, unharmed the man for whom I dig.” Digging is a category of religious action in this text, not merely an economic one.

Aside. The same instinct produces one of the Rigveda’s odder lines. At RV 1.179.6 the sage Agastya is described as khanamānaḥ khanitraiḥ, “digging with spades”, in a hymn otherwise concerned with a marital argument about celibacy. Whatever he was digging for, the poet reached for the technical vocabulary of excavation to describe an ascetic at work. The Rigveda’s holy men handle tools.

What a Bucket Can and Cannot Do

The recurring claim in popular summaries is that Rigvedic farmers irrigated their fields from wells. It is worth doing the arithmetic, because the arithmetic settles it.

Lifting water is work, and the work is exactly the weight times the height:

$$W = \rho\, g\, V\, h$$

where $\rho \approx 1000\ \mathrm{kg/m^3}$ is the density of water, $g \approx 9.81\ \mathrm{m/s^2}$, $V$ is the volume raised and $h$ is the lift. Take a shallow well, eight metres to the water.

A household’s daily water, say 200 litres for drinking, cooking and washing, comes to about 16 kilojoules of lifting. A person working steadily produces on the order of 75 watts, so that is a few minutes of hauling, before friction and spillage. Fifty cattle drinking forty litres each need about 160 kilojoules, well under an hour of somebody’s day. This is what the hymn says the rig is for: nṛpāṇam, a drinking place for men, and at RV 5.83.8, “for the cows let there be drink abundant”.

Now try a field. A hectare of barley wants roughly 400 millimetres of water across a season, which is 4,000 cubic metres. Lifting that eight metres takes about $3.1 \times 10^8$ joules. A person sustaining 75 watts for eight hours delivers about 2.2 megajoules a day. Even at perfect efficiency, with no leakage, no spilled bucket and no rest, that is roughly 140 person-days per hectare per season, and real leather buckets on real ropes are nowhere near perfect. Add oxen and the number improves, but nothing in the Rigvedic well vocabulary suggests animal traction: the varatrā is fastened by the same hands that will haul it.

The conclusion is not that Rigvedic people never watered a plant. A kitchen plot is entirely within reach, and the kulyā, the small runnel, is exactly the tool for spreading a few buckets across a vegetable bed. The conclusion is that irrigating fields from wells, at any scale that matters, was not physically available to them, which is consistent with everything else the text shows us: a society that ploughs where the rain and the rivers already put water, and that prays extremely hard for rain. The furrow hymn asks Parjanya for moisture, not for a well.

Aside. These figures are order-of-magnitude, and they are meant to be. The seasonal water requirement for barley varies with soil and climate, the depth to water in the Punjab varies enormously, and the sustained power output of a labourer is a range rather than a number. What survives all that uncertainty is the ratio. Drinking water and field irrigation differ by a factor of several thousand, and no plausible adjustment to the inputs moves a hectare of barley into the range of a leather bucket.

The Well That Nobody Dug

Return to RV 10.101 and read to the end. After the boat, the plough, the well, the cattle-stall and the armour, the hymn closes on a line that reads like a riddle:

Set the Lord of the Forest in place among the trees. Set down, well and truly, the spring, without digging it.

(RV 10.101.11b, translated from the Sanskrit; Griffith: “sink the well although ye do not dig it”)

Ni ṣū dadhidhvam akhananta utsam. Make a spring without excavation. Every other instruction in the hymn is literal. This one cannot be, and the standard reading is that the hymn has been building a sustained double image all along: the sacrifice is a field, the poets are ploughmen, the recitation is the digging, and the “spring” that appears without a spade is the flow of soma, or of song, or of divine favour, that the ritual is supposed to open.

That reading is almost certainly right. But notice the direction of the metaphor. The poet does not explain the well by comparing it to the sacrifice. He explains the sacrifice by comparing it to the well. The technology is the familiar term; the liturgy is the thing being illuminated. Which tells you how ordinary the rig was.

Turning the Well Upside Down

Here is where the machine stops being a footnote in material culture and starts explaining the Rigveda’s central myth.

The poets had a problem that every rain-dependent society has: rain is invisible in its mechanism. Something is holding water up there, and something releases it. To describe that, you need a model, and the model the Rigveda reaches for again and again is not a river, not a cloud, not an ocean. It is a well being worked.

Watch the Maruts do it:

With their vigorous strength they pushed the well up on high, and clove the cloud in twain though it was passing strong… They drove the cloud transverse, directed hitherward, and poured the fountain forth for thirsting Gotama.

(RV 1.85.10-11, Griffith 1896; see RV 1.85)

The Sanskrit verbs are precise. Ūrdhvaṃ nunudre ‘vatam: they shoved the well upright. Jihmaṃ nunudre ‘vatam: they shoved the well over sideways. Then asiñcann utsam, they poured out the spring. This is not a description of weather. It is a description of two men wrestling a heavy vessel up out of a shaft and tipping it into a trough, projected onto the sky at cloud scale.

The Aśvins do the same trick with an extra turn of the screw:

Ye lifted up the well, O ye Nāsatyas, and set the base on high to open downward. Streams flowed for folk of Gotama who thirsted.

(RV 1.116.9, Griffith 1896; see RV 1.116)

Uccābudhnaṃ cakrathur jihmabāram: you made it bottom-upward, with its opening tilted. An inverted well. That is a beautifully economical image for a raincloud, and it only works if the audience knows exactly what an upright well looks like.

The pattern runs through the whole pantheon.

The last of those is the one to sit with. RV 5.83, the Parjanya hymn, is as close to a rain-charm as the collection gets, and at its climax the poet tells the god exactly what to do:

Lift up the mighty vessel, pour down water, and let the liberated streams rush forward. Saturate both the earth and heaven with fatness, and for the cows let there be drink abundant.

(RV 5.83.8, Griffith 1896)

Mahāntaṃ kośam ud acā ni ṣiñca, syandantāṃ kulyā viṣitāḥ purastāt. Raise the great bucket and tip it. Let the unstopped channels run. The two technical words in the verse are the two technical words from the well site: kośa, the leather bucket, and kulyā, the runnel that carries water away from the trough. The monsoon, in the Rigveda’s imagination, is a man at a well, hauling up a bucket, tipping it into the channels, and doing it on a scale that waters a subcontinent.

Once you notice that, the Vṛtra myth starts to look different too. At RV 5.32.1 Indra “split open the spring and released the cuttings”, adardar utsam asṛjo vi khāni, using the noun built on khan, to dig. Indra’s supreme deed, the one the collection returns to more than any other, is described with the verbs of well-work: boring, splitting, unblocking, pouring. The dragon is a stoppage. The god is a man clearing an obstructed shaft.

What the Rigveda Does Not Have

A caution, because this material attracts overclaiming.

The Rigveda has no araghaṭṭa, the geared wheel with a chain of pots that most people picture when they hear “Persian wheel”. The word is classical Sanskrit, not Vedic, and the device belongs to a much later technological world; it turns up in the Pañcatantra and becomes common in early medieval inscriptions from Rajasthan and Gujarat, more than a thousand years after the latest Rigvedic hymns. Nor is there a ghaṭī-yantra, a pot-wheel. Nor a dam, a sluice, a weir or a stepwell.

This matters because the retrojection is live in the current literature. A 2020 review in Hydrology and Earth System Sciences on water management in ancient India credits the Rigveda with both the aśmacakra pulley and a “drum-shaped wheel with earthen pots on an endless rope”, and supports well-irrigation with citations to RV 1.23.18 and RV 8.3.10 [5]. Look those up. RV 1.23.18 says “I call the Waters, Goddesses, wherein our cattle quench their thirst”; RV 8.3.10 says Indra sent the great waters to the sea. Neither mentions a well, a field, or a device. The authors are hydrologists compiling from secondary handbooks rather than Sanskritists reading the text, and the error is an honest one, but it propagates: a claim enters a review paper, the review paper gets cited, and within a few cycles the Rigveda has acquired a chain-pump.

Methods note. The defence against this is dull and effective. Every time a source attributes a technology to the Rigveda, open the cited verse in a translation you trust, and then open the Sanskrit. The corpus is small enough that a genuine occurrence of a technical word can be found and counted in an afternoon, and words like aśmacakra and khanitrima, which occur exactly once, are precisely the ones that carry the most interpretive weight and deserve the most scrutiny.

What the archaeology adds is a discontinuity rather than a line. Brick-lined draw-wells are a mature Harappan technology, standing in the streets of Mohenjo-daro in the third millennium BCE [4]. Terracotta ring-wells, the next form documented in any quantity, appear in the Ganga plain much later, in the urbanising phases associated with Northern Black Polished Ware from around the sixth century BCE onward. The Rigvedic hymns sit in the gap between the two, and their well is described in materials that leave nothing behind: wood, leather, rope, and one stone that would be indistinguishable from any other worked stone once the shaft had silted up. We have the instructions and not the object.

The Machine Under the Metaphor

There is a habit, when reading the Rigveda, of treating the technical passages as ballast: the interesting parts are the creation hymns and the riddles, and the bits about buckets are what you skim. RV 10.101 rewards the opposite approach. It is a hymn to the All Gods that spends a quarter of its length on ironmongery, and the ironmongery turns out to be the conceptual engine of the collection.

The Rigvedic poets did not have a theory of the water cycle. What they had was a working knowledge of one machine that moved water against gravity, and they used it, relentlessly and with real imaginative discipline, as the model for everything else water does. Rain is a bucket tipped out. A cloud is a well stood on its head. Drought is a blocked shaft. The greatest act attributed to the greatest god is, structurally, unclogging a spring. And when the poet wants to say what a successful sacrifice feels like, he says it feels like water arriving in a place where nobody dug.

That is what a good model looks like from the inside. It is not a metaphor decorating a thought; it is the thought. The reason the Rigveda’s rain imagery is so physical, so full of straps and stone and pouring and weight, is that the poets were describing the sky in the only vocabulary they had for moving water, and that vocabulary came from the well head, where somebody was standing every morning of every year, hauling.

Read RV 10.101 straight through, then read RV 5.83 immediately after. The first is a work order. The second is a prayer. They use the same nouns.