Water and Violence
on Arthasâstra III.9.31
fixing fines for violence
in conflicts related to irrigation

Francis Zimmermann, Water and Violence. On Arthaśāstra III.9.31 fixing fines for violence in conflicts related to irrigation, in Rajendran, Ed., Current Readings in Arthaśāstra (Dr. N.V.P. Unithiri Felicitation Volume & Proceedings of the International seminar held at Department of Sanskrit, Calicut University on October 18-19, 2005), Kozhikode, University of Calicut, 2008 (Calicut University Sanskrit series No. 33), pp. 69–83.

Downloadable PDF version (with accurate transliteration of Sanskrit citations) at the bottom of this page

As a European student of Things Indian, I shall be projecting my own cultural grid onto the texts I have been reading, and I shall emphasize particular aspects of the Arthasâstra learned tradition that are consonant with current debates in the contemporary world. Postmodern topics such as the idea of a Land Ethic or the emergent field of Environmental Ethics or the problematic of Governmentality have shed new light on this ancient text. Sanskrit words like janapada or danda (to mention two words among others that will be cited below) have taken new significance. Kautilya's Arthasâstra does not belong with any particular discipline; therefore it lends itself well to transdisciplinary configurations like feminine studies or environmental studies. These approaches allow us to explore ambiguities of Norms better than of Law, and to reconstruct the ancient jural system of claims and obligations better than the political system of rights and duties, as I would like to suggest in focussing on one example, water management.

Kautilya prescribes compensations to be paid when the ploughing and seeds sown in another's field are damaged by the overflow of a reservoir, channels or a field under water, and furthermore, when the damage affects the fields themselves, or gardens and embankments. The technical and geographical context is that of anicut-fed or bund-lined canals or catchment tanks. By anicut or bund I mean a weir or a dam built to retain or divert running water for irrigation purposes. In other words, we shall be dealing with weirs and dams, along with sluices or sluice-gates. We shall be reading a few words of Arthasâstra in the light of modern and contemporary issues. From this angle, a number of concepts and realities mentioned in the ancient text give a most graphic account of water management and the relationship of farming to Land in traditional settings.

We may like to historicize this description by setting back the political system of Arthasâstra in the context of Mauryan economics and society. The royal sîtâ lands mentioned by Kautilya were the economic foundation of Mauryan state power, and D. D. Kosambi, in discussing the king's investments for the benefit of his subjects, gave in one of his most brilliant papers a telling example of public works undertaken by the State to secure water supply to sîtâ lands in Mauryan times (Kosambi: 227-228 and note 12). Rudraraman finished rebuilding the shattered dam of lake Sudarsana below Girnar to thrice the original size, at his own expense, "without having plagued the regular town and country settlers (paura-janapadam janam) with corvee labour taxes, or voluntary contributions," as the relevant inscription says. This dam, now lost without trace, was begun under Candragupta Maurya, finished by Asoka's Persian satrap Tusaspha, rebuilt in greater size by Rudradaman, and again extensively repaired by Cakrapalita, son of Skandagupta's administrator Parnadatta in AD 456. "Thus a shrunken kingdom, Kosambi concludes, meant that regulation of water supply and conservation of the water—one of the few essential functions of a central power in India—deteriorated." This was the historical context in which regulations studied in the following pages took sense.

Arthasâstra, Book III is devoted to Civil Law: marriage, inheritance, sales and purchases, damages, debts, deposits, private property, slavery, etc. Water is dealt with as a resource for agriculture, from III.9.24 through III.10.34, where Kautilya prescribes action to be taken in case of encroachments and damages to fields and roads. In the colophon giving the subject of this section: iti ksetra-patha-himsâ, "Thus ends (the topic of) damage to fields and roads," the Sanskrit word for "damage" is himsâ, usually translated as "violence" since ahimsâ is the well-known Sanskrit word for "nonviolence."

Water as Commons, and the Secularization of Violence

Traditional ideas should be understood in the light of modern ideas, and we should draw the list of all possible values of water in order to understand the significance of water management in an ancient Hindu kingdom. In writing or speaking about water, it is easy to get muddled because of its multiple aspects or dimensions. Water is perceived by different people (or by the same people in different contexts) in different ways. Only part of our modern perceptions of water are relevant to older times, when water was a common pool resource for a village community. In a remarkable inventory of the various Perceptions of Water, Ramaswamy R. Iyer delineates two basic concepts that for me will be keys to an understanding of traditional systems of water management: Water as Commons, and Water as Basic Right (Iyer: 78 ff.).

The conception of water as commons or as a common pool resource is better adapted to the context of a small lake or pond or tank on common land. The most relevant sociology of water as commons is that of local settings in closed polities, because, then, water management is simple enough and one can think of waterbodies as owned by the community. When we use expressions such as commons or community management, we usually have a local context in mind (one village or a cluster of villages constituting a watershed) and we shall at first confine ourselves to that context. This is not to say, however, that Arthasâstra ignored the more modern conception of Water as Basic Right, as we shall see below in the second and third sections. However, for the time being, let us remain at the local level which is the context to which prescriptions laid down in Arthasâstra III.9 and 10 are relevant. In the following quotations, I shall cite simultaneously the two translations by Kangle and Rangarajan, whenever I think they complement one another. In the two following precepts, I wish to emphasize the rhetoric of damage and compensation:


"In case of damage to the ploughing or seeds in another's field by the use of a reservoir, channels or a field under water, they shall pay compensation in accordance with the damage" (transl. Kangle).
"No one irrigating his field from a reservoir or tank shall cause danger to another's ploughed or sown field by letting water overflow from a reservoir, channel or field" (transl. Rangarajan: 232).


"In case of mutual damage to fields under water, parks and embankments, the fine (shall be) double the damage" (Kangle).
"If there is damage, adequate compensation shall be paid according to the extent of damage. In case the damage is extensive and affects fields, gardens and embankments, there shall also be a fine of double the damage" (Rangarajan: 435).

Let us reflect upon the rendering of himsâ (violence) as "damage", and danda (punishment) as "compensation," which is a shift of meaning that I would like to interpret as a secularization of violence:


(In case of himsâ…) yathopaghâtam mûlyam dadyuh /
"On shall give payment (mûlyam) in accordance with (yathâ) the damage (upaghâta)."


parasparahimsâyâm himsâd dviguno dandah /28/
"In case of mutual damage (parasparahimsayam), the fine (dandah) (shall be) double (dviguno) the damage (himsâd)."

The shift of meaning is still more obvious in the subsequent precept related to damages caused by an upstream tank to a downstream one. A tank constructed on a higher level shall not prevent the flooding with water of a lower tank, except when its use has ceased for three years…


tasyâtikrame pûrvah sâhasadandah, tatâkavâmanam ca /31/

"For transgression of that (regulation), (the punishment shall be) the lowest fine (dandah) of violence (sâhasa) and the emptying of the tank" (Kangle).
"A higher tank preventing the filling up of a lower one, in use for at least three years: Lowest SP (standard punishment) and emptying of higher tank" (Rangarajan: 232).

The concept of sâhasa (another word for violence) is defined in the section on Forcible seizure (sâhasam), AS III.17.1 ff. Sâhasa, which Kangle translates as "forcible seizure" and Rangarajan as "robbery with violence," is a deed of force (Kangle) in seizing forcibly a person or property in the presence of the owner (Rangarajan: 479); if the owner is absent, it is a theft. Rangarajan explains the figure of speech which is exemplified by Kautilya's use of sâhasa-danda, a figure of speech which Western students of classical Rhetoric would call catachresis, while in Sanskrit one would describe it as an example of svasamjñâ, that is, a terminological convention and short-hand expression according to which a newly coined "conventional proper sense" is given to these words. The literal or etymological meaning then is misleading, and sahasa-danda here does not have the pictorial meaning of "a punishment (danda) for forcible seizure (sâhasa)," but the more abstract meaning of "a standard penalty" or "fine" (Rangarajan: 112) imposed in courts as a compensation for damages.

Coming back now to water, it is clearly a matter of natural resource for agriculture. As per public regulations enforced by the State authorities, the deprivation of this common thing entails compensation and fines. Kautilya sets up a kind of dialectics between the State and the individual. The State not only has dominion, but asserts a power of police over the water of all natural streams, recognizing at the same time a right of user by those who apply these waters to agricultural uses, and without regard to the location, riverine or not, of the lands thus irrigated. Water is conceived of as a commons and a natural resource for the State's agriculture. Nevertheless, a second value of water is also taken into account, at least implicitly and as a secondary aspect of things, when violence is considered, and natural environmental things such as the village's river are equated to the human body (the physical integrity of the individual): to spoil water or to deprive someone of access to water is to harm the body. We shall come to this meaning in the next section.

At the level of the State at large, encompassing a full-fledged system of streams and rivers on its territory, or in the context of the supply of irrigation water through canals from large reservoirs, whether state-owned or privately owned, we can still argue that the water source belongs to the community as a whole, or to civil society, and that the conflicts that arise can be resolved within that overall framework. But European jurists have appropriated to modern environmental legislation an ancient distinction, which was drawn in Roman law, between 'everybody's property' (res communis), that may not be sold nor purchased, and 'nobody's property' (res nullius), that may be reclaimed by a newcomer. This distinction will be useful in our reading of Arthasâstra. Kautilya's State territory (including the resources of the soil such as larger waterbodies), as the rest of our analysis will show, belongs with res nullius (nobody's) more than with res communis (everybody's).

Water as Basic Right, and the Full-fledged Concept of Violence

At the State level, water is no longer a village commons but a basic human right which announces the modern idea of individual rights of access to common things or natural resources like water. But the idea of an individual's right of access to water—Water as the support of Life—is very different from the idea according to which all farmers in the village are entitled to water their fields. When speaking of a commons, we stay at the level of a village. When speaking of a basic human right, we shift grounds and we refer to a few prescriptions in Arthasâstra, where the Land (the kingdom) and the Body (the human inhabitants' body) are compared. Water sustains life. In that respect, it is a basic need and therefore a basic right. Usually, in modern times, when we speak of the individual's right to water as a support of life, we are talking about drinking water. Kautilya did not envisage that aspect of water management, and consequently the conception that prevails in his work is that of water as commons. But, as soon as we link Arthasâstra to Ayurveda for example, the multi-faceted nature of water as a source of life becomes obvious. The ideal Hindu kingdom is a place and a community enjoying a sufficient amount of rains; in the middle of the land runs a river with its tributaries: the health and welfare of all inhabitants depend upon water as a basic right. We need both perceptions (water as commons and water as basic right), in order to give full justice to Kautilya's wisdom, and we must learn to harmonize them. As soon as we compare damages caused to the land with damages caused to the body, as Kautilya does in III.17 through 19, for example, we are faced with the dual status of water as commons and basic right.

Whenever water has been diverted, stolen or spoilt, and consequently someone is deprived of this water resource, the political and jural issue arises whether this water was the victim's property. Does the fine imposed on the culprit represent a compensation for a damage caused to private property? Probably not, since water is a commons. And again, probably not, since providing water to the inhabitants is the State's affair. Nevertheless, let me cite Kangle's Study (Volume III of his Kautilîya Arthasâstra), where he raises a permanent debate among historians and anthropologists:

"A question that has exercised the minds of scholars for long is whether all land belongs to the State or whether there is also private ownership in land. It seems that opinions were divided on this question in ancient times. So far as the present text is concerned, there can be little doubt that it recognizes private ownership in land, though it presupposes at the same time the existence of state-owned or crown lands. As we have seen, all unoccupied land is assumed to belong to the State. The existence of state farms is shown by the description of the duties of the sîtâdhyaksa in Chapter 2.24. Lands which are not under his supervision are clearly to be supposed to belong to private owners." (Kangle III: 169).

Private property is at stake in conflicts between individual farmers whose lands are located at different levels of ground. Those of them who are exploiting plots located upstream will get better access to the flowing water, while the others, situated downstream, depend upon their neighbours' good will. Whenever rivers flow through broad and flat plains, farmers are able to tap directly into main irrigation canals without worrying about their neighbours. But when farmers depend on water that originates far upstream and flows through canals crossing the boundaries of many neighbouring fields before reaching their individual ones, then collective values of equity and social control are at stake. Even a brief interruption in the flow will destroy a farmer's crop. An inordinate flow due to an unexpected breach in a weir or a bund will swamp the field and spoil the crop unless it is quickly shunted away. Weirs are made of earth, logs and stones that may easily be washed away by floods. Moreover, the amount of water that reaches the weir in the dry season may be strongly affected by the cropping schedules of upstream neighbours (Lansing: 41). For all these reasons, water management is as much a political matter as a technical one.

Riparian Lands, and the Issue of Forcible Appropriation

The issue of riparian rights is paradigmatic for any reflection upon the essence of neighbourhood. Take an individual farmer who needs access to water for wet-crop cultivation. His relationships with his neighbours upstream and downstream are of the utmost importance to define his riparian rights. Riparian rights are rights of access to the shore, bed, and water for someone owning land adjacent to the bank of a natural watercourse (as a river), a lake or backwaters. The upstream—downstream division is the key factor in devising rules of equitable water uses. While the upper riparian neighbour can actually control the flow, the lower riparian neighbour may only claim the right to an equitable share. Before reading Arthasâstra on these matters, let us briefly describe the situation prevailing throughout European history. The right of all neighbours to an equitable share has traditionally been protected by Law in England for example. Riparian rights were part of title to land adjoining a stream. They did not exist independent of the land, nor were they absolute. A riparian user could divert water for any useful purpose, and the amount could vary from day to day, month to month, or year to year. Lack of use did not affect the right. Except for the water they used for domestic needs and stock, riparian claimants could not substantially reduce the flow of a stream if downstream neighbours objected. Riparian rights were context-dependent, they were fixed in so far as they related to other riparian claims. If we may now come back to the relevant passages in Arthasastra, we do find similar regulations. Kautilya draws a distinction between farmers situated "upstream" (upari) and "downstream" (adhas).


"A tank on a lower level (adhara), constructed afterwards, shall not flood with water a field watered by a tank on a higher level (uparita). A (tank) constructed on a higher level shall not prevent the flooding with water of a lower tank, except when its use has ceased for three years" (Kangle).
"The water from a lower tank shall not submerge a field fed from a higher tank built earlier. A higher tank shall not prevent the filling up of a lower tank, except when the latter has not been in use for three years" (Rangarajan: 232).

At his own point of diversion on a natural water course, each diverter must establish some reasonable means of effectuating his diversion. He is not entitled to command the whole of a substantial flow of the stream merely to facilitate his taking the fraction of the whole flow to which he is entitled; in brief, he must behave as a good neighbour. A proper balance is to be achieved between neighbourhood and riparian rights.

A number of precepts in Arthasâstra III.9.24ff. revolve around situations in which a neighbour has indulged in obstruction, encroachment and damage upon a neighbouring pasture, dry land or wet-crop field, a vegetable garden, a threshing floor, a shed or a stall for vehicles. First and foremost, "All disputes regarding obstruction and damage shall be decided on the testimony of neighbours" (III.9.24; Rangarajan: 435). Then Kautilya's attention focusses on water management. The ownership of tanks shall lapse, if they have not been in use for a period of five years, except in cases of distress (III.9.32; Rangarajan: 231). Waterworks such as reservoirs, embankments and tanks can be privately owned and the owner shall be free to sell or mortgage them (III.9.34). Owners may give water to others in return for a share of produce of various kinds from sowings in fields, parks and gardens watered by dredging canals or building structures on adjacent rivers or tanks (III.9.35). In other words, all farmers in Kautilya's kingdom enjoyed riparian rights.

We might content ourselves with this paraphrase, but I think we can push the interpretation one step further, provided we make a clear distinction between two kinds of rights of access to such a mobile, wandering and elusive thing as water. Either water in its flow remains firmly attached to the geography of the Land, or it is diverted from its natural course and transported to distant fields for irrigation. Whenever settlers created weirs, canals and sluices, they changed the status of water, which was then deterritorialized, so to say. This process went against traditional views of Nature, in Europe as well as in South Asia.

The English common law originally reflected the belief that rivers were part of God's plan as revealed in Nature. God not only determined the course of streams. He dictated that water was a moving, a wandering thing whose very nature defied conversion to property. Consequently, any attempt to dam or divert a stream was at least an unnatural act. In such a worldview, water in general was definitely nobody's property. The owner of land along the stream had a right to the water in its accustomed flow, which was wholly independent of the user's will. The riparian farmer could at any time make use of riverine water for all reasonable purposes as a usufruct attached to the Land. The water so withdrawn became the object of a usufruct, so to say, but what remained in the stream continued to be nobody's property. Equally, the Roman civil law recognized only a qualified appropriation of water as property, and always "subject to a common right by natural law, where it is capable of being fully enjoyed without exclusive possession" (Gast: 75).

There have been places and times in the West where and when water truly became an individual property. Let us recall the story of the Gold Rush in mid-19th century California. When it entered the Union as the 31st State (1850), California adopted the English common law familiar to the eastern seaboard. One of the first actions taken by its law-makers was to adopt the common law of riparian rights. But the thousands of fortune seekers who flocked to California, following the discovery of gold, soon changed the rules. They built extensive networks of flumes and waterways to work their claims. The water carried in these systems often had to be transported far from the original river or stream. The self-governing, dissident miners applied the same 'finders-keepers' rule to water that they did to their mining claims—it belonged to the first miner to assert ownership (Shen: 151). To stake their water claims, they imposed what was labelled the Doctrine of Prior Appropriation; the catch phrase was "first-in-time, first-in-right." Controversies around water that have plagued North American western states like California, Arizona or Colorado from the middle of the nineteenth century are most relevant, for South Asia, since they flourished in the context of either arid or semi-arid landscape.

The overall water supply in such ecological settings is always insufficient to grow crops on all the land available. Neighbourhood, then, entails sharing scarce resources. The key point for us, in the North American example and the paradigmatic competition between Riparian Rights and the Doctrine of Prior Appropriation, is that the geographical link between land and water was broken. The appropriation system permitted the use of water where needed without tying use to riparian ownership. New settlers took water where they found it and used it where they needed it, by constructing weirs and channels that transformed the geography of the land (Gopalakrishnan: 62). Availing ourselves of this comparison to the West, let us now come back to Arthasâstra, and to the dialectic of Riparian rights and Appropriative rights. Provided we keep the two conceptions in mind, the ancient text speaks to us in very graphic terms:


karmodakamârgam ucitam sandhatah kurvato'nucitam vâ … /

"If one obstructs a customary (ucitam) water-course (udaka-mârga) in use (karma) or makes a new one that is not customary (anucitam)" (Kangle)…

The root of the damage is to destroy the customary link between the Land and water, either in obstructing or in diverting the flow by artificial means which transform the natural geography of rivers and streams. To give its exact sense to this phrase which begins a new chapter (III.10.1): "a customary water-course," we must refer to the lines preceding it in the Arthasâstra text. We are at the junction of two chapters. According to the concluding verse of the preceding chapter (III.9.38), farmers are fined for letting the water of tanks out to their neighbours at the wrong place or time or, conversely, for obstructing the flow of water to be shared by neighbours at the wrong place or time. At the wrong place, ie., at another place than a regular sluice-gate. At the wrong time, ie., letting out water out of turn, or obstructing the water of others when it is their turn. The interpretation of vara, repeted twice in this verse, has been the object of an interesting hesitation among scholars. Earlier translators like J. J. Meyer and R. Shamasastry (cited in Randhawa I: 363) read pâra, and they translated this word as "a sluice-gate," instead of vâra which to Kangle and others means "a turn" for receiving water from a canal or a tank used in common. But whatever reading and meaning you adopt, clearly one and the same social system of water management is described that entails the construction of sluice-gates, that is, an anicut-fed canal or a catchment tank, with sluice-gates from which the water is released in turns to be shared by neighbours. The root of this system of water management is the substitution of the natural geography of rivers and streams with artificial means of conveyance, storage and distribution.


To summarize, values of collective water control are crucial when we set the scene at the regional level in a large engineered landscape. The notion of commons itself becomes questionable when we shift ground from local settings to a more global scene. To keep to the political world depicted in the Arthasâstra, when dealing with the streams and rivers of the State, this view of water as commons looses its legitimacy. As we have seen in the third part of this contribution, difficulties then begin to arise in the form of 'upstream versus downstream' issues, and it becomes unavoidable to draw a distinction between commons at the local level, private ownership at the individual level, and government at the State level. One may well bypass the issue of property rights (ownership) and talk about management rules (government), one may well speak in terms of 'community management of common pool resources', but the question of ownership remains open.

In the foregoing pages we met with three different concepts of violence, which Kautilya seems to have kept in mind simultaneously, although each one them appears in some specific context on a particular level of approach. The more common situation is that of damages and encroachments for which a compensation has to be given. Although the underlying rationale is that of forcible deeds and the words used, in describing these deeds and compensations to be paid for them, are words commonly used to speak of violence, on this first level we are merely dealing with accidents and repairs. By using words like himsâ and sâhasa, Kautilya simply gave a rhetorical dimension to the description of damage. Then, on a different level, the more serious conflicts about water control may lead to forcible appropriation and theft, in which case we are truly at the level of violence and punishment. Finally and on a third level of approach, the inhabitants' life is at stake and a link is explicitly made between the Land and the human body; the use of a word like himsa, then, is fully consonent with its occurrences in the Dharmasâstras and the other Sanskrit classical texts.


Charles E. GAST, "The Colorado Doctrine of Riparian Rights, and Some Unsettled Questions," The Yale Law Journal, Vol. 8, No. 2, November 1898, pp. 71-81.

Chennat GOPALAKRISHNAN, The Doctrine of Prior Appropriation and Its Impact on Water Development: A Critical Survey," American Journal of Economics and Sociology, Vol. 32, No. 1, January 1973, pp. 61-72.

R. P. KANGLE, The Kautiliya Arthasastra. Part I. A Critical Edition with a glossary; Part II. An English Translation with critical and explanatory notes; Part II. A Study. University of Bombay, 1960-1965; 2nd ed., 1972; three volumes.

D. D. KOSAMBI, "The Basis of Ancient Indian History (II)," Journal of the American Oriental Society, Vol. 75, No. 4, October-December 1955, pp. 226-237.

J. Stephen LANSING, Priests and Programmers. Technologies of Power in the Engineered Landscape of Bali, Princeton NJ, Princeton University Press, 1991.

Ramaswamy R. IYER, Water. Perspectives, Issues, Concerns, New Delhi, Sage, 2003.

M. S. RANDHAWA, A History of Agriculture in India, Volume I: Beginning to 12th Century, New Delhi, Indian Council of Cultural Research, 1980.

L. N. RANGARAJAN, The Arthashastra, Edited, Rearranged, Translated and Introduced by L. N. Rangarajan, New Delhi, Penguin Books, 1992.

Dajun SHEN, "Water rights and their management: A comparative country study and its implication for China," in Figueres (Caroline M.), Tortajada (Cecilia) and Rockström (Johan), Eds., Rethinking Water Management. Innovative Approaches to Contemporary Issues, London & Sterling VA, Earthscan, 2003; first South Asian edition, 2005.

Pre-print version in PDF format