www.water-alternatives.org Volume 5 | Issue 2 Shah, E. 2012.

Seeing like a
subaltern – Historical ethnography of pre-modern and modern tank irrigation
technology in Karnataka, India. Water Alternatives 5(2): 507-538 Shah: Tank
irrigation technology in Karnataka, India Page | 507 Seeing Like a Subaltern –
Historical Ethnography of Pre-Modern and Modern Tank Irrigation
Technology in Karnataka, India
Esha Shah Department of Technology and Society Studies, Faculty of Arts and
Social Sciences, Maastricht, the Netherlands; e.shah@maastrichtuniversity.nl
ABSTRACT: wE discuss here the historical ethnography of tank irrigation
technology in Karnataka that covers both medieval and modern periods AND how
the technical designs of this thousand years old technology significantly
transformed from the pre-modern to the modern times and how in each epoch
the reproduction of the technology implied the reproduction of radically different
social and cultural spaces and, most significantly, social and power relations.
KEYWORDS: Tank irrigation technology, pre-modern Knowledge, anthropology of
technology, Karnataka, India

These debates took an influential turn in India since the opposition against a
large dam on river Narmada in the 1980s. Among many other objections, the
large dam was opposed as a symbol of the domination of Western and modern
technology. However, the invocation of the pre-modern as a critique of the
modern has been widely criticized as modernity’s representation of its 'other'. It
has been widely shown how the agency ascribed to the idea of the pre-modern
(or traditional) is not only founded on a binary opposition of the modern and non-
modern but also embedded in the particular forms of imagination of the past and
present and hence far from being historically accurate.1 Despite the widespread
and longstanding criticism of the invented nature of the pre-modern, the idea has
not only continued to inform political and policy debates, but has also been
widely invoked in critical scholarship in the social sciences (Harding, 1994;
Lansing, 2006; Scott, 1998).
1 There is a significant body of literature that shows the 'constructed' nature of
the idea of the pre-modern. Although Agrawal does not talk about the
relationship of indigenous knowledge with the pre-modern form of knowledge,
Agrawal’s essay on dismantling the difference between the indigenous and
modern forms of knowledge has been widely cited (Agrawal, 1995). I will further
discuss David Mosse’s equally influential work on the way in which the colonial
government invented the idea of community for its own administrative purposes
(Mosse, 1999). In my own earlier work, based on the extensive folk literature, I
concluded that it would be difficult to find any fundamental ontological difference
between a pre-modern tank irrigation technology and a modern technology
(Shah, 2008). Based on archeological and historical evidence, Morrison comes to
the same conclusion that pre-modern irrigation systems differed in no
fundamental or essential way from modern ones (Morrison, 2010). Water
Alternatives - 2012 Volume 5 | Issue 2 Shah: Tank irrigation technology in
Karnataka, India Page | 508
Here, I aim to first review a few such representations of the idea of the pre-
modern invoked from politically diverse positions. The choice of these examples is
not exhaustive. I do not aim to review the whole range of such representations.
The examples discussed here are only to demonstrate a certain unifying
characteristic among these representations despite the politically disparate
vantage points from which they are invoked. I aim to show that while these
imaginings serve a wide range of different political purposes, they end up reifying
binary oppositions between the pre-modern and modern because they invariably
provide a 'view from the above', which Donna Haraway called "a view from
nowhere and everywhere" (Haraway, 1988). On the other hand, the critiques of
these forms of reified representation often end up erasing the difference between
forms of knowledge emerging from different social and historical contexts
(Agrawal, 1995). In throwing a reverse gaze – the modern gaze on the pre-modern
– these critiques fail to acknowledge the significant differences between the
knowledge emerging from different historical epochs. In the second part of the
essay, proposing a methodological alternative to go beyond the reductionist
opposition on the one hand and the 'dismantled difference' on the other, I join a
number of other critical voices that are redefining debates on water. Mollinga
reviews critical scholarship and policy debates on agricultural water use in India
which he thinks are caught between strategic essentialism and analytical
reductionism (Mollinga, 2010). The strategic essentialism is sustained by the
clamours on binary opposition between state-village and state-community,
(which referring to Kaviraj, Mollinga traces to the way pre-modern, colonial and
postcolonial states operated), and the resulting analytical reductionism
overemphasises the village and community institutions and pays scanty attention
to the way state functions at the local level. Mollinga’s proposes his own way
forward by proposing to focus on changing material conditions, especially the
impact of the processes of economic liberalisation and globalisation on water use.
What Mollinga calls silences, however, have been variously broken in the last
decade or so, the most noteworthy among these broken silences is the growing
literature on cultural politics of natural resources (Baviskar, 2003, 2007). This
growing body of literature goes beyond the "dull rigour of economic
determinism" and focuses on "lived experiences and sedimented histories of
situated cultural practices and places" (Baviskar, 2003). The focus in cultural
politics of natural resources is to understand the way in which multiple fields of
power manifest in contradictory and complex ways and the way in which
struggles over resources are simultaneously about identities and interests, and
about symbolic capital and material economy (Mosse, 1997, 2003). Joining this
line of critics and breaking what Mollinga calls a silence, I evaluate here the pre-
modern and modern tank irrigation technology to show how a situated position –
seeing like a subaltern – can provide a way forward from the mutually opposing
binary categorizations of the pre-modern and modern. I have used the category
of subaltern here in the broadest terms – to designate the subordinated interests
and positions of marginality. The term subaltern here is a short hand for denoting
the expression of power relations in design, construction and control of irrigation
works. The term broadly represents the position of marginality, especially the
social anxiety and environmental uncertainty caused by the technological designs
shaped by these power relations. In other words, seeing like a subaltern would
mean looking at tank irrigation systems from below, from the perspectives of
those whose interests and whose lives were compromised in the development,
spread and functioning of this technology both in pre-modern and modern
periods. Seeing like a subaltern would therefore mean explaining the ways in
which the social anxiety generated in the times of environmental uncertainty,
which in itself could be explained as a manifestation of the technological designs,
resulted into acts of violence against the marginal sections of society during the
medieval periods. In the modern period seeing like a subaltern explains the way in
which newer forms of power relations since the introduction of green revolution
have generated newer forms of marginality. Water Alternatives - 2012 Volume 5 |
Issue 2 Shah: Tank irrigation technology in Karnataka, India Page | 509
The discussion thus covers the historical ethnography of tank irrigation
technology in Karnataka that has sustained over both medieval and modern
periods.2 I have shown how the technical designs of this thousand years old
technology significantly transformed from the pre-modern to the modern times
and how in each epoch the reproduction of the technology implied the
reproduction of radically different social and cultural spaces and, most
significantly, social and power relations. By extensively referring to the folk
literature – songs and stories – I have done subaltern history of the tank irrigation
technology in the medieval period to show how the key technical designs of the
technology were reproduced by using institutionalised forms of forced labour. I
also argue that a number of popular folk songs on a woman being sacrificed in the
tank on the face of the looming threat of drought and water scarcity point out
that the tank technology was a source of social anxiety, uncertainty, and violence
during the medieval times. Moving on to the modern period, I further show how
several key designs of the tank irrigation structures in Karnataka radically
transformed as a result of the shifting power of the farming elites since the
introduction of the green revolution. In radically different ways, then, the tank
irrigation technology in both pre-modern and modern times has not been
amenable to democratic, communitarian or egalitarian forms of governance. In
both epochs, the technology has been the site upon which the complex power
relations are articulated.
2 I would like to clarify that the social and power contexts of tank irrigation
technology in the state of Karnataka is emphasised in this paper. This does not
mean that the literature on tank irrigation in other parts of South India – Tamil
Nadu, Andhra Pradesh and Kerala – is not relevant. In fact, given that the
Vijayanagara Empire covered much of South India during the medieval time, it
would be difficult to discuss the situation of Karnataka as a separate geographical
block. However, I have deliberately tried to focus only on Karnataka for the tank
irrigation in the modern period. This is also because since the British period an
overwhelming scholarly and policy attention is given to tank irrigation technology
in Madras Presidency and subsequently to Tamilnadu. There is often a tendency
to view the tank irrigation situation in other parts of South India from the lens of
the experiences in Tamilnadu. At some other time and place I would like to
compare evidences from Tamilnadu and Karnataka, but here I expressely wish to
discuss the nuances of the tank irrigation technology in Karnataka.
SEEING FROM THE ABOVE: COLONIAL AND NATIONALIST REPRESENTATIONS OF
TRADITION
The beginning of the imaginings of the pre-modern could be traced to the advent
of modernity itself. Both the colonial and the nascent post-colonial state in-the-
making extensively referred to the agency of the traditional. The civilising mission
of the British in India, to dominate in order to liberate, was a profoundly
contradictory enterprise. Prakash argues that "the British positioned modernity in
colonial India as an uncanny double, not a copy, of the European original – it was
almost the same, but not quite" (Prakash, 1999). This positioning of modernity
involved translation between the modern and traditional and also the actors who
could do this translation. The process of translation, however, involved the
reinterpretation, and to some extent, the 'invention' of the traditional. In his
detailed historical ethnography of ideologies of tank irrigation development in
South India, Mosse shows how the colonial government in nineteenth century
'invented' the custom and tradition of 'village communities' for its minor irrigation
administration (Mosse, 1999). In the context of the recurring famines in the
Madras and Mysore presidencies in the late nineteenth century, the colonial
government was faced with a moral task of maintaining thousands of dispersed
minor irrigation structures. The financial and administrative burden this implied
was insurmountable. The British responded to this mammoth task by positioning
the tank systems in the state of decay and decline which was further related to,
on the one hand, the moral and administrative disorder of the preceding regimes,
and on the other, to the loss and decline of the traditions of communal labour
and village institutions for the tank maintenance and management (Mosse, 1999).
Mosse shows how the British administration went to a considerable length to
amass evidences for the existence of such institutions, rooted in pre-colonial
traditions, and how it framed laws and rules to reinstate such 'invented'
traditional collectivities (Mosse, 1999). The positioning of the rule of law of the
modern state was thus based on the incorporation of the language of custom and
tradition (Gilmartin, 1994). Water Alternatives - 2012 Volume 5 | Issue 2 Shah:
Tank irrigation technology in Karnataka, India Page | 510
Underlying this translation – the representation and invention of the tradition –
was the rearranging of power relations between the colonised and the coloniser
(Prakash, 1999). Another such rearrangement of power relations in the mid
twentieth century also happened around the reinterpretation of the traditional,
this time involving the nationalists – those Indians who wanted the British to
leave India (Zachariah, 2005). On the face of the impeding independence, the task
in front of the nationalists was to build an authentic nation that was modern but
not Western. The nationalists, who were largely western educated middle class
intellectuals, were engaged in debates on what path should independent India
take after the British left. These debates on 'possible nations' in the 1930s-1950s
happened around the futuristic themes of 'development' and 'improvement',
however, these constructions of the future were based on the reinterpretation of
the past in the same way as their colonial counterparts did, but with the
contrasting images of the past. The western-educated elites and indigenous
reformers reinterpreted the classical Hindu religious texts and traditions in the
light of the modern, scientific and western discourse (Prakash, 1999). These texts
were now reinterpreted as 'scientific', and rational and authentic, upon which the
universality of the modern Indian nation could stand. Not only that through the
interpretation of the archaic that the modern nation was cast as homogenous,
whole and Hindu (Prakash, 1999) but also that through the representation of the
tradition that the Indian modernity was cast as distinctly different from the
Western modernity and hence authentic.
SEEING FROM THE ABOVE: CRITICAL SOCIAL SCIENCE SCHOLARSHIP
In contrast to the governmentality of the colonial state and the nation-building of
the indigenous elites, the agency ascribed to the pre-modern in the critical social
science is rather different. Recent scholarship has posed important challenges to
the epistemological dominance of the European Enlightenment and Western
science, technology, and modernity. The pre-modern or 'outside of modern'
technologies and ways of knowing are not only posed as a counter-hegemonic
challenge, but, as some would insist, they form a viable or desirable alternative to
modernity. These forms of knowledge and practice are claimed to be
substantively, epistemologically, and contextually superior to modern science and
technology. At the same time, scholars also freely associate them with other
desirable traits, such as decentralization, democracy, bottom-up planning, and
even self-organizing spontaneity. Much of this literature supposes that the
superiority of pre-modern knowledge systems and artefacts originated in the
superior pre-modern cultural context from where modernity is perceived as
lacking, inadequate and insufficient. Below I review a few of such currently
influential literature.
James Scott, in his influential book Seeing Like a State, proposes the Greek
concept of metis to mean the local, implicit, embedded, practical experience of
knowledge and compares it with the abstract, general knowledge deployed by the
modern state agencies. Although Scott does not label this form of knowledge
'traditional' and often alludes to imply it as 'practical knowledge', his critique of
the modern state and his deployment of the concept of metis are nevertheless
based on an idea of the pre-modern state. Scott (1998) writes,
[t]he pre-modern state was, in many crucial respects, partially blind; it knew
precious little about its subjects, their wealth, their landholding and yields, their
locations and their very identity. It lacked anything like a detailed map of its
terrain and people. It lacked, for the most part, a measure, a metric, that would
allow it to 'translate' what it knew into a common standard necessary for a
synoptic view. As a result its interventions were often crude and self-defeating.
Scott further argues that in contrast of the pre-modern state the modern
European statecraft is devoted to creating a rational order through rationalization
and standardization. In creating the designed and planned social order, the
modern state operates through the manipulation and control of its subjects. What
is interesting is that while Scott relates two contrasting forms of knowledge with
the contrasting forms of statecrafts (as in the quotation above), he does not
further elaborate on how the Water Alternatives - 2012 Volume 5 | Issue 2 Shah:
Tank irrigation technology in Karnataka, India Page | 511
pre-modern form of knowledge is embedded in the form of statecraft. However,
the Greek concept of metis is later in the book elaborated as forms of informality
and improvisation based on practical and local experience, which, Scott argues,
emphasizes process, complexity, transiency, and open-endedness. Scott further
discusses this concept as representing the knowledge that involves rules of thumb
in contrast to the knowledge that is based on deductive principles transmitted
through book-keeping. In the rest of his book, through detailed discussions on
several case studies, Scott powerfully demonstrates the forms of knowledge
employed in the domineering and controlling practices of the modern state-craft.
The concept of metis is thus clearly 'invented' (my word) which is ascribed an
agency of critiquing the violence of 'seeing like a state'.
What is not convincing in Scott’s schema of knowledge is the absolute necessity of
the association of metis-form of knowledge with the ancient Greek-ness and even
with the pre-modern state. In social sciences many such concepts referring to the
modern society, for example, Pierre Bourdieu’s discussion on phenomenology of
tacit knowledge, and Levi-Strauss’s concept of bricolage, which Scott briefly
mentions, have similar derivatives. Scott’s own discussion in explaining the nature
of metis employs the examples of a boxer, a soccer player, and a war diplomat
which betrays the implied non-modern-ness of metis. In the absence of any
discussion on the kind of society, state and social order from which the metis
originates, I am implied to argue that the reference to the non-modern or ancient
roots of metis here is more driven by the search for the 'utopian outsider' of
modernity rather than the historically compelling evidence of a superior location.
Another influential scholar, Stephen Lansing, locates his arguments in the
historical and cultural context of Balinese society. He proposes that a holistic
cultural approach that is attentive to indigenous systems of order can solve many
contemporary political predicaments. Calling it a 'perfect order', Lansing argues
that the centuries-old traditional system of the water temple in Bali achieved a
balanced ecology without recourse to centralized state power. Balinese irrigation
is ecologically hierarchical but socially democratic, which, Lansing suggests might
have been developed through spontaneous self-organization (Lansing, 1991). For
Lansing, this culture, with its deep roots in the pre-modern times, is an alternative
rationality to the pervasive economization of modern times. In his subsequent
book, Lansing asks how Western social science might look like from a magical
Balinese perspective (Lansing, 2006). While the former is formed around the
emergence of modern, autonomous self-hood and the principle of equality, the
pre-modern, Balinese worldview, based on both the Hindu Brahminical and
Buddhist religious practices, is based on the principles of hierarchy on the one
hand, and the emotional self-mastery on the other. Different notions of self-hood
inaugurate different forms of governance. In other words, the subjective
awareness of individual connects to the objective conditions. The emergence of
the autonomous subject is associated with the rise of totalitarian rationality in
Western thought, whereas in contrast, Lansing explains, pre-modern Balinese
thought derives from magic and the ideas of self-mastery and awareness of
interdependencies which forms the egalitarian, self-regulating farmers’
organizations called subaks. The control of self through the control of emotions
and unruly passions go hand in hand with the ordering of the outer (natural)
world (Lansing, 2006).
Lansing thus contrasts the pre-modern and modern forms of subject formation
and methods of governance. Lansing bases his claim that the traditional irrigation
organisations such as subaks have been egalitarian and democratic on the
assertion that there was an absence of the Great Designer or a centralised,
despotic or bureaucratised pre-modern state invested in building irrigation
facilities. However, such absence of despotic centralised authority does not rule
out other forms of enactment, entrenchment, and diffusion of power. It would be
a great sweep to equate the absence of the Great Designer with decentralisation
or forms of self-organizing, democratic, community-collectives. The extent to
which the pre-modern Balinese state was involved in constructing and managing
irrigation facilities was a key question that drove the decade long controversy
among historians and anthropologists. The question whether Balinese irrigation
was entirely controlled by subaks and hence decentralised and democratised or
controlled by some form of kingly centralisation remains yet equivocally
unanswered. This debate was recently re-opened. Stephen Lansing and Clifford
Geertz’s Water Alternatives - 2012 Volume 5 | Issue 2 Shah: Tank irrigation
technology in Karnataka, India Page | 512
conclusion that the pre-modern Balinese state was weak or symbolic and
correspondingly the decentralised and democratic organisations of subaks
remained at the helm of the irrigation affairs is challenged by Hauser-Schaublin,
who argues that the pre-modern Balinese state was neither symbolic nor weak
(Hauser-Schaublin, 2003). She proposes a different model of the pre-modern
Balinese state, which, she argues, derived its power and authority from mobilising
a large number of neighbourhoods to repeated pilgrimage to temples, by means
of which people from unrelated areas were brought together to form localities.
According to her, state authority and power manifested in an entirely different
way than those modelled on modernist theories. Hauser-Schaublin alleges that
the earlier theories of the Balinese state being a theatre state as argued by Cliford
Geertz and a democratic irrigation model presented by Stephen Lansing were
based on the selective reading of empirical sources (Hauser-Schaublin, 2003). As I
already discussed above, in invoking the agency of the pre-modern Balinese
worldviews rooted in Hindu and Buddhist traditions, Lansing shows the third
alternative to centralisation-decentralisation dichotomy (Lansing, 2006), however,
the extent to which his own computer-simulated, conceptual model of self-
organising temple networks (Lansing, 1991) bore resemblance to the actual
history of Balinese irrigation remains ambiguous.
Another such powerful trend of post-colonial historiography of science and
technology, which is sometimes portrayed as a "second historiographic
revolution" (Harding, 1994) or described as a thousand flowers bloomed (Raina,
2003), relates to the idea of the pre-modern. A leading scholar in science and
technology studies, Sandra Harding, poses a probing question: whether natural
sciences are multicultural, and to what extent modern science has origins in the
non-modern cultures. She proposes that the indigenous and non-Western
scientific traditions should be strengthened, re-valued, and integrated as a model
for the global and democratic sciences of the future (Harding, 1994). In the
historiography of sciences, the post-colonial frame is the most recent and
internally diverse one (Raina, 2003). Among post-colonial science and technology
historians, the revivalist standpoint refers to the pre-colonial forms of knowledge.
Raina shows how this historiography often end up projecting the mirror-image of
Eurocentrism that post-colonial studies otherwise aim to critique (Raina, 2003).
Knowledge systems and artefacts of the pre-colonial era largely emerge in the
revivalist studies at best as fragments or facets that serve either as a benchmark
against which the level of pre-colonial civilization is measured or as a way of
exploring and charting the transformations brought about by colonizers (cf
Alvares, 1991; Kumar, 1997; Qaisar, 1998). Such studies largely approach pre-
colonial science and technology with the quintessentially colonial question: What
was the level of sophistication of 'Indian knowledge systems' before the British
conquest? Against the background of the European experience, pre-colonial
Indian science and technology emerge as a mirror image of the one generated by
the hegemonic project of the British, albeit turned upside down. The post-colonial
answer to the colonial question – namely, whether Indian science and technology
were sophisticated enough – invariably ends up being affirmative.
SEEING LIKE A SUBALTERN
In the above examples, despite the vast differences in the purpose and manner in
which the agency of the pre-modern is invoked, there is a tendency to make pre-
modern knowledge systems and artefacts into reified objects of virtue
irrespective of their social and historical location. In critical social science, the
detailed mapping of the social context of pre-modern forms of knowledge
remains elusive. While a number of critical social science studies raise significant
political and historical issues with respect to technological change and modernity,
they throw very little light on the social shaping of pre-modern knowledge
systems – a subject that barely exists in its own right. As a result, claims of
epistemic and cultural superiority – or even substantive equality – between pre-
modern and modern knowledge systems involve, in the words of Shigehisa
Kuriyama, "asserting comparability without even comparing". "The only way to
advance", Kuriyama further suggests, in response to Harding’s multiculturalism,
"towards a genuinely pluralistic appreciation of knowledge in different cultures is
actually to study Water Alternatives - 2012 Volume 5 | Issue 2 Shah: Tank
irrigation technology in Karnataka, India Page | 513
cultures, earnestly, humbly, in detail, over a long time. By itself, the most
thoroughgoing critique of western universalism contributes nothing" (Kuriyama,
1994). In her recent work, Harding makes a different argument. She wants to re-
evaluate the binary of tradition-modernity, which she thinks remains powerful in
shaping research and policy, because such binary renders the traditional cultures
as feminine and primitive (Harding, 2008). The success of the modernity and its
attendant values of objectivity, rationality, progress and science are measured in
terms of their distance from the feminine and primitive. Harding proposes that
the modern science should be considered only one among many ways of knowing
and also that this science must be examined from below – from the standpoint of
women and other least advantaged citizens (Harding, 2008). Below I aim to
examine tank irrigation technology from the below, from the standpoint of the
disadvantaged positions in both pre-modern and modern periods.
Seeing like a subaltern: Pre-modern tank technology, state and labour
organization
The kingdoms in early South India between roughly the eighth and sixteenth
centuries AD rose and fell on their ability to extract and sustain agricultural
production, which in turn rested on the effective control of water. The practice of
constructing and using tank irrigation systems started in the early centuries of the
first millennium AD.3 Technically speaking, a tank can be described as a miniature
version of a large dam. Water is impounded behind an earthen embankment and
is released through sluices into canals to irrigate downstream land. Excess water
from the reservoir is allowed to escape through waste weirs. Figure 1 provides a
schematic diagram of the technical principles of these tanks. Tanks usually receive
water from a seasonal rivulet, a drainage channel, a tributary, or a canal. Unlike
large dams, tanks are almost never constructed directly on perennial rivers.
Further, they are usually constructed in chains so that the overflow from an
upstream tank forms the inflow into a downstream tank. In other words, although
tanks are spatially dispersed, they are actually hydrologically linked. At present
there are thousands of tanks, irrigating anywhere from ten to a thousand
hectares, scattered across South India. Roughly 38,000 are in the state of
Karnataka alone, which, according to a survey done in 1986-87, supply 19 percent
of the total irrigated area in the region (Vaidyanathan, 1998). The period between
the seventh and tenth centuries saw wider utilization and effective management
of hydraulic technology in South India. This growth coincided with agricultural
expansion, especially with the spread of irrigated rice paddy cultivation (Ludden,
1985). Tank construction in Karnataka reached its zenith during the Vijayanagara
period (AD 1300-1700), when the tanks were expanded in upland regions across
South India, particularly in the ecological areas that some scholars identify as
'mixed zone' to distinguish them from other areas that were primarily wet or dry.
These areas are considered marginal because they were less preferred for human
settlement, and agriculture expanded to them only later (Breckenridge, 1985;
Heitzman, 1997; Ludden, 1985; Mosse, 1997).
3 The earliest written records available on the construction of tanks date from the
third century BC. The earliest record of a tank in south India refers to the tank
that existed at Inamgoan, near Pune, in 1500 BC (Dikshit et al., 1993; Gurukkal,
1986; Srinivasan, 1991).
Different dynasties ruled South India during the thousand-year period between
the seventh and seventeenth centuries. Some historians, however, argue that
there existed a fundamental continuity during this time with respect to several
important aspects of society and culture. Burton Stein called this a pyramidal,
segmentary state, one that was integrated culturally but decentralized politically.
It was religious and symbolic order, rather than military might, that bound the
state together. In this pyramidal state, tanks and temples linked localities with
central authority (Stein, 1980). The period of the Vijayanagara empire, which
claimed political hegemony over most of South India between AD 1300 and 1750,
was marked by the expansion and intensification of agriculture, a process in
which temples and tanks became linked together to create what Carol
Breckenridge calls social storage (Breckenridge, 1985). Social storage involved
two practices: creating tanks to store water for agricultural production, and using
temples to collect, share, and redistribute agricultural surpluses, keeping some in
reserve as Water Alternatives - 2012 Volume 5 | Issue 2 Shah: Tank irrigation
technology in Karnataka, India Page | 514
insurance against hard times. Temples were at the helm of moral and economic
transactions and provided an organizational framework for the creation and
redistribution of agricultural surplus. The local elite (made up of institutions,
corporate groups, and individuals) donated money and land to temples. These
donations, or endowments, were used to construct irrigation networks and
expand agriculture in the land granted to the temples. Part of the additional
income generated by this expansion was offered to temple deities in accordance
with donors’ stipulations, and part was claimed by local elites as returns on their
investments. Temples and tanks were therefore the backbone of social and
agricultural expansion into the mixed zone.
Figure 1. The technical principle of tanks (not to scale). Water Alternatives - 2012
Volume 5 | Issue 2 Shah: Tank irrigation technology in Karnataka, India Page | 515
Although the historical literature is thus rich when it comes to locating tanks
within a certain social order, it rarely discusses the techniques and labor
organization involved in tank construction. The first evidence of who might have
constructed the network of tanks comes from folk literature. Folk literature on
tanks – stories, legends, and songs – originates from three sources. The first set of
stories and myths I found were narrated to me by members of the artisan caste of
Voddas, who are rarely mentioned in the inscriptional sources or in South Indian
historiography but who figure prominently in folk accounts as tank builders.4 The
myths and stories remembered by Voddas are markers of their identity and thus
are usually not associated with any particular tank. A second set of stories
narrated by higher-caste farmers and temple priests, in which Voddas also figure
significantly, are usually associated with specific tanks. The third type of folk
literature consists of songs about the sacrifice of women in tanks, and these are
usually sung during times of drought and water scarcity by Dalit women –
formerly an untouchable, lower caste. Almost every tank in the semiarid region of
south Karnataka has a story or a sacrifice song attached to it. This literature
depicts quite a different picture from the collective, communal project of
construction some recent scholars assume. Instead, the labor for tank
construction seems to come primarily from the Vodda caste. The stories narrated
by higher-caste farmers and temple priests and the songs sung by Dalit women
commonly describe the large number of Voddas employed for tank construction.
For example, in the folksong 'Kanne Viramba', a village chieftain invites seven
hundred Voddas to build a tank. This song was sung to me by a group of Dalit
women from the village Dannayakankere, which is located close to Hampi, the
erstwhile royal capital of the Vijayanagara empire. In another such song, 'Kere
Hunnama', the chieftain enters into a lengthy and arduous negotiation with the
leader of three thousand Voddas (Rajappa, 1974). Voddas from seven villages
were invited to construct a tank in the story about folk hero Dannayakan Mudda
narrated as below. Once upon a time, Dannayakan Mudda, the king of the
Vijayanagara kingdom, went around his domain to find out about the welfare of
his people. When he passed by a village of Haragnur he felt that it was a beautiful
place, suitable for the construction of a tank. After twelve years of worship, the
Mudda convinced the deity of the local temple – Anjaneyyaswamy – to help him
build a tank. The god modified the Mudda’s plan, which would have displaced
twelve upstream villages and twelve downstream villages, to shift only three
villages upstream and three downstream. The people from these villages formed
a new village which was then called Haragnur. The Mudda employed Vodda
laborers from seven villages for the construction, but informed Anjaneyyaswamy
that he did not have the money to pay them. Anjaneyyaswamy instructed him not
to keep any accounts of the Voddas’ employment and not to pay them. It took
several years to complete the construction of the tank. On the day of payment,
according to the instruction of the Anjaneyyaswamy, the Mudda called the
Voddas together and made small sand mounds on the ground in front of them. To
each he gave a stick and instructed them to pass it through the mound in front of
them. As per Anjaneyyaswamy’s advice, the Mudda told all the Voddas who
complained about finding only a ghughari (a sweet made with wheat and coconut
powder) as their payment that this was because they had not worked hard
enough, with both physical and mental devotion. Had they worked hard, they
would have received instead a gold coin as per the god’s wish. The tank was
completed, but it did not receive enough water to fill it. The Mudda started to
worry and again contacted the god. This time Anjaneyyaswamy said that the tank
needed a human sacrifice and decreed that seven daughters of a family in the
nearby village should be sacrificed in the tank. The father was contacted and,
after some persuasion, agreed to sacrifice his daughters for the welfare of the
village. Accordingly all the daughters were taken to the embankment in a festive
procession. After worshipping the god, they were asked to enter small rooms
made in the embankment and the doors were shut behind them. An idol of the
youngest girl is installed on one of the sluices of the tank. The girl is worshiped
and called Kaneramma.
4 In the present days, Voddas no longer participate in tank construction; they are
largely employed as menial laborers for road building, setting telephone lines and
underground pipes, construction, stone work, and so on. They live in separate
villages and are not closely associated with agricultural activities. Water
Alternatives - 2012 Volume 5 | Issue 2 Shah: Tank irrigation technology in
Karnataka, India Page | 516
A temple priest told me this story in 1999 during my field work. The same folk
hero also figures in the songs sung by Dalit women. None of the stories and songs
suggests a collective, community-based construction of tanks. While villagers’ and
temple priests’ stories and songs clearly illustrate that Voddas built thousands of
tanks over the span of several centuries, the Voddas’ own narratives hint that
they may have been under some form of coercion when they did so. The recurring
and central theme of the legends narrated by Voddas pertains to their grievances
about not being paid, being paid inadequately, or receiving payments that have
been disguised. The theme of Voddas making unreasonable or insatiable demands
on their employers also recurs in other legends. One such story, recorded by a
British ethnographer, goes like this: One sultry day, Siva and Parvati (a god
couple) were walking upon the earth, when they got very hot and thirsty. The
drops of perspiration which fell from Siva were changed by him into a man with a
pickaxe and crowbar, while those falling from Parvati turned into a woman
carrying a basket. The new pair of man and woman quickly sunk a well, and with
the cooling waters god and goddess refreshed themselves. In gratitude, they
promised the labourers certain gifts. The nature of the gift did not satisfy the man
and the woman and both grumbled, which so incensed Siva that he cursed the
pair and their descendants to earn their bread only by digging wells and tanks.
Paraphrased from (Thurston, 1909). In another story, the god offers the Voddas a
disguised payment. Responding to the Voddas’ pressing demand for payment, the
god Siva hollows a measuring rod, fills it with gold coins, and gives it to the
maistry (leader of a group of Voddas), who unwittingly pawns it in exchange for a
toddy (a liquor). The god then plays another game with the unsuspecting Voddas.
He buries a pumpkin filled with gold coins in the ground where the Voddas were
digging. Noticing the raised mound caused by the buried pumpkin, they leave it
untouched to show the depth of earth they had dug. A buffalo, which was grazing
in the field close by, exposes the pumpkin, which the Voddas, unsuspecting of its
contents, sell to a local shopkeeper. In these stories, Voddas present themselves
as being eternally cursed by the gods: their situation is fated and not subject to
changes in dynastic rule. This folk literature describes characters and processes
that resemble certain real people and practices of the Vijayanagaran era, as
interpreted by traditional historiographic methods. Real-life historical entities
masquerade as folk characters. For instance, the folk hero Dannayakan Mudda
resembles the king of the Vijayanagara empire who constructed several important
tanks around the royal seat at Hampi. The gowda – the village chieftain who
either owns the village or has substantial rights over the produce from the village
land – whose daughters or daughters-in-law are sacrificed in the tanks behaves in
ways consistent with our understanding of the chieftains of the Vijayanagara
period. Gods appearing as local deities like Anjaneyyaswamy can be read as the
close involvement of local temples in the process of tank construction.
Various forms of these oral narratives point to the long-term continuity of the
social and ideological context within which the Voddas’ labor was extracted.
There are indications in the historical literature on South India that the labor for
construction activities was, in general, politically and ideologically controlled and
even coerced by the elites since the early medieval period. What are still being
debated are the nature and form of this forced labor. Should it be considered
compulsory – labor extracted in lieu of tax payments, for example – or simply
forced or coerced? For instance, Tejaswini Yarlagadda, in her study of inscriptions
from the western Deccan (the heartland of tanks now located in the neighbouring
state of Andhra Pradesh), suggests that the practice of visti – the imposition of
forced labor – began in the ninth century AD. The recipient of the land grant was
given the right to impose forced labor and was provided immunity for doing so.
Yarlagadda further observes that the frequency of the term visti in inscriptions
increased in the period between the ninth and tenth centuries and disappeared in
the period between the tenth and thirteenth. However, she argues, quoting other
sources, that the disappearance of the term may be due to the commutation of
forced labor into Water Alternatives - 2012 Volume 5 | Issue 2 Shah: Tank
irrigation technology in Karnataka, India Page | 517
money payments and not to the elimination of the practice altogether
(Yarlagadda, 1993).5 Therefore both the folk and the historical literature provide
strong evidence to suggest that tanks in the Vijayanagara period were not
constructed collectively by communities. Instead, elites extracted the Voddas’
labor through some form of coercion.
5 There is a large corpus of historical literature that point out the existence of
some form of coerced labour, although often the details of such forms of labour is
missing. "The largest of these irrigation works must have involved labor forces in
their construction", but "we have a limited understanding of how these laborers
were recruited" (Sinopoli and Morrison, 1995). "There are inscriptional references
to the demand for compulsory labour, but it is not clear whether this demand was
imposed on all members of subaltern classes" (Palat, 1987). David Ludden also
fleetingly mentions that "coerced labor was not unique to the nineteenth century
and perhaps was also convenient during temple and tank construction over the
centuries", but he does not provide any evidences (Ludden, 1985).
Seeing like a subaltern: Social and environmental adaptability
Tank technology has been acclaimed not only for its cultural embeddedness in the
regions of South India, but also for its environmental sustainability. Tanks are
considered a more environmentally appropriate form of water management than
large dams, because they are small and store water in a dispersed and
decentralized fashion. Without disputing the argument regarding the comparative
environmental advantage of tanks vis-à-vis large dams, I argue that tanks were
also subject to serious environmental uncertainties (and still are), especially those
constructed in the mixed-zone region of Karnataka. As already discussed, tanks
were increasingly constructed in mixed habitats, largely in south Karnataka, to
expand rice paddy cultivation away from the fertile riverine areas during the
Vijayanagara period (AD 1300-1750). This period was marked by the expansion
and intensification of agriculture during a time of long-distance migration,
constant warfare, and the resulting social instability, especially during the later
centuries of the period. Tanks and temples during this highly volatile period
became a source of social storage (Breckenridge, 1985). However, the
requirements of social storage made tanks constructed in the mixed region
vulnerable to the hydrological vagaries of alternate drought and flood, due to the
highly erratic inflow of water. This uncertainty about water availability in mixed-
zone tanks was a major concern for British engineers. For instance, a British
officer named R.E. Playfair wrote that "the small tank systems on tributary
streams must fail: in the first place because the cost of storing the water is too
great… and secondly, that as the supply of such tributary streams is likely to fail in
seasons of drought, the tanks would be dry at the very time they would be most
wanted". Colonel Playfair wanted to demolish all existing tanks and build them
anew at more favourable locations (Playfair, 1866). In recent times, tanks in the
mixed-zone region of Karnataka routinely fill up only once in three or four years.
Those engaged in development policy routinely ascribe this hydrological volatility
to the poor performance, deterioration, and decline of tank irrigation since the
colonial period. Based on the assumption of deterioration and decline, a massive
program of tank reform and modernization is being implemented in Karnataka.
My argument is that the hydrological poor performance is not, as is often argued,
a result of the recent apathy of the state in maintaining the infrastructure, but
rather is more deeply historical in nature, an outcome of the social and
production requirements for which tanks were constructed in the first place. Oral
narratives suggest that since they were originally constructed for rice paddy
cultivation, tanks in the mixed-zone region have faced hydrological marginality
and the resulting uncertainty of water availability for a long time.
The tanks narrated in many stories and songs of folk literature are a source of
anxiety either because they have not received any water, or because they are
faced with the threat of flood. This hydrological irregularity had other
consequences as well. Earthen embankments and foundations subjected to
alternate drying and wetting become structurally weak and breach easily during
heavy downpours, creating yet another source for the social anxiety depicted in
stories and songs. The human drama in Water Alternatives - 2012 Volume 5 |
Issue 2 Shah: Tank irrigation technology in Karnataka, India Page | 518
these stories and songs stems from the fact that the artefact around which the
drama takes place is vulnerable to an unpredictable environment.
One genre of traditional ballads that tell the story of a woman about to be
sacrificed in a tank is testimony to this hydrological irregularity and the resulting
technological vulnerability.6 The woman about to be sacrificed usually belongs to
the family of the village chieftain, who has invested in constructing the tank and
whose fortune would be at stake if it did not receive any water. There are several
versions of these songs, each named after the woman sacrificed. One, called
'Kerege Haara', is famous in Karnataka and is considered an especially fine work of
poetry; it occupies a prominent space in the Kannada literature and is considered
a symbol of Kannada nationalism and identity.7 There are also other versions –
again, each named after the woman sacrificed and deified – such as Madagada
Kenchamma, Kere Hunnamma, Anajee Honnamma, and Kaneramma (Rajappa,
1974). Although these songs differ in terms of their content, style of narration,
and the name and context of the social actors depicted, they all broadly narrate
the pathos and plight, love, longings, and desires of the women about to be
sacrificed.
6 Alan Dundes describes the traditional ballad as a narrative song whose drama
depends upon acts of love and violence. He further argues that traditional ballads
and folk legends about human sacrifice have Indo-European origins, where the
foundation sacrifice motif tells of technological vulnerability in medieval times
across the Indo-European region. In western and northern Europe, there are
legends about the sacrifice of illegitimate or fatherless children in the foundations
of bridges, dikes, monasteries, palaces, and churches, whereas in southeastern
and eastern Europe it is always the sacrifice of women for similar purposes
(Brewster, 1996; Dundes, 1996). 7 In Kannada literature, the song 'Kerege Haara'
occupies a prominent place. For an English translation of it, see (Srikantaiah,
1996). For Kannada versions of two songs, 'Madagada Kenchamma' and 'Kere
Hunnamma', (see Paramshivaiya and Hanumanthappa, 1994; Rajappa, 1974).
A brief overview of one such song, 'Voddar Boyi', follows: The gowda (village
chieftain) invites the Boyi, a leader of a group of Voddas, and asks him to build a
tank in one week. The Boyi negotiates and agrees, but he returns in the third
week and complains that the bund (embankment) did not stand. The tank in fact
asks the Boyi for a human sacrifice in order for the bund to stand. When the
gowda asks who should be sacrificed, the Boyi first asks for the eldest son of the
chieftain. The gowda refuses to give up his eldest son, and what follows is a
lengthy negotiation. The Boyi then demands the youngest son, the youngest
daughter-in-law, the gowda’s wife, and finally the gowda himself. The two finally
settle on the eldest daughter-in-law. The rest of the song describes her life as she
awaits her impending death. Some of the songs also worship or deify the
sacrificed woman. For example, 'Kanne Viramba', a song sung by a group of Dalit
women from Dannayakankere village, begins with a eulogy to the great victory
and glory of a virgin goddess (a young girl sacrificed in a tank) and then tells the
story of the tank being built. "Once upon a time", it begins, "a snake with seven
hundred heads showed seven bushels of gold coins to the Muddanna [village
chieftain] who invited seven hundred Voddas to build a tank. Seven sluices were
built in seven years with seven human sacrifices and five more deaths". The song
then sings the glory of Muddanna, declares that the tank was built based on
moral truth and the principles of moral victory (dharma), and finally sings the
glory of the virgin goddess and considers her sacrifice the most glorious among all
the sacrifices in the tank. Although songs about sacrifice in tanks at the time of
floods are rare, I came across at least one popularly remembered story about the
sacrifice of a lower-caste man to prevent the breach of an embankment during a
time of heavy rain. A neerghanti (a lower-caste man employed to distribute water
and to watch the physical structures) realized that the tank embankment might
give way as a result of the heavy inflow of water due to the heavy rain. Worried,
he pleaded with the gangemma (water goddess) not to break open the
embankment. The gangemma, yielding to his prayers, promised that she would
not break the embankment until he went to the village, informed the village
leaders about the hazard, and returned. The neerghanti rushed to the village and
informed the leaders about the gangemma’s promise. The leaders immediately
held a meeting and quickly arrived at their decision: they killed the neerghanti,
cut his body in three pieces, and threw his head to the north, his torso to the
south, and his limbs to the west in such a way that his body could not Water
Alternatives - 2012 Volume 5 | Issue 2 Shah: Tank irrigation technology in
Karnataka, India Page | 519
be rejoined. To the gangemma’s rage, therefore, the man could never return to
negotiate. In order to keep her promise to the neerghanti, however, the
gangemma did not breach the embankment.
In certain parts of the mixed zone (the Kollar and Bellary districts), almost every
tank has attached to it a story of a sacrifice made to alleviate drought, to save the
tank from breaching, or to prevent the embankment from giving way. A small
shrine may also exist on an embankment, usually in the name of the higher-caste
woman sacrificed to the tank and then deified (almost never in the name of the
lower-caste man killed to save the tank). In rare cases, a temple may even exist on
the embankment or in the village. Upon entering the dry region of Bijapur in the
north and the wet region of Shimoga in the east, the shrines, temples, and
sacrifice stories disappear, as uncertainty about water availability for tanks also
significantly disappears.
The present-day uncertainty about water availability in the tanks of the mixed-
zone region may not be a recent problem, as is often claimed in the literature, but
rather may be one that dates back to the earliest construction of tanks in this
region. Traditional tank technology is therefore not necessarily entirely in
harmony with the environment of the region. Tanks were constructed in a certain
social and political context in order to support a certain form of agricultural
production, and certainly not with any self conscious concern over long-term
environmental and cultural suitability or sustainability. Tanks were instead
constructed on a terrain for which they were not entirely well suited in order to
expand rice paddy cultivation during a socially tumultuous period. Hence they
were highly susceptible to environmental uncertainties. Furthermore, while the
stories lyrically represented in the songs centre around a vulnerable artefact, the
social actors in these tales also seem to be acting under duress, to which they
respond with acts of social violence: the gowda asking the Vodda Boyi to
construct a tank in a week; the king leaving his kingdom and camping in a village
for twelve years to build a tank; a number of women murdered to alleviate water
scarcity or to ensure the safety of an embankment; the ghastly murder of a
neerghanti to save a tank from breaching. All of these suggest the presence of a
high degree of social anxiety regarding tank irrigation, anxiety which results in
acts of violence. The folksongs and stories popularly remembered and told by the
villagers do not even remotely resemble an idyllic picture of a blissful community
living on the banks of a bountiful tank in the pre-modern era.
TECHNOLOGY OF POWER: MODERN TANK TECHNOLOGY AND GREEN
REVOLUTION
In the modern times, especially since the introduction of the green revolution the
designs of the tank technology underwent a radical transformation. This
transformation happened differently in different socio-historical and ecological
zones in the state of Karnataka. Here I wish to discuss the case of Shimoga district
located in the wet zone of western Karnataka. I have extensively discussed several
such examples in the mixed and dry zones elsewhere (Shah, 2003).
It may not have been a sheer coincidence that water distribution infrastructure
such as sluices and distribution canals in many tanks of Shimoga district began to
either entirely disappear or were severely modified since the early 1970s.
Shimoga district produces the highest amount of rice and has the second largest
area under rice in the state of Karnataka. High yielding varieties were first
introduced in Shimoga district in 1966/67. According to one survey of two rice-
growing areas of the district, all farmers had adopted new varieties by 1972
(Krishnamurthy, 1975). A majority of the landowners of the tanks discussed below
adopted the new green revolution rice varieties since the early 1970s, but it took
a decade more for a complete shift to the intensive rice cultivation in tank
commands. This happened only after the national policy on procurement of rice
changed in the early 1980s.
The most apparent shift in the tank designs in this region was the disappearance
of water distribution canals and sluice operating mechanisms. A paddy growing
command area (locally known as atchakat) is usually so designed that it facilitates
the movement of water from field to field. Upstream and downstream fields are
levelled in such a way that they are irrigated in succession from head to tail. Not
only that the design of fields and the position and slope of canals facilitate
movement of water in the atchakat from upstream to downstream which gives
spatial advantage to the fields in the upstream Water Alternatives - 2012 Volume
5 | Issue 2 Shah: Tank irrigation technology in Karnataka, India Page | 520
but the direction of movement of water produces grades of different types of
lands in the atchakat. This spatiality of water movement assures repeated and
assured supply more to certain patches than others. The favoured patches
acquire certain features, owing to their submergence under water for longer
duration, which make them progressively more suitable for paddy cultivation.
These grades of land reflect the social relations of land holding in the atchakat.
During my field work I found that the best land in the atchakat invariably
belonged to the historically privileged groups of landowners. As I demonstrate
below the power of the landed privileged elites was reproduced in the tank
irrigated locality through the shift in the tank designs.
In addition to the disappearance or modification of the distribution network,
another major shift in tank designs since the intensification of rice cultivation was
the disappearance of sluice operating mechanism. Conventionally, tanks irrigating
paddy in Karnataka have the plug and pole type of sluice. See figure 2. Tanks in
the wet region like Shimoga with average annual rainfall in the range of 1500 to
2000 mm are relatively small in size and have lower bunds compared to the tanks
in the drier or mixed regions with average annual rainfall in the range of 600 to
900 mm and less.8 The sluice apertures in the tanks of the wet region are located
in embankments unlike in the case of bigger tanks in the mixed region where they
are located in the water spread area away from the embankments. The sluice in
the wet region has been conventionally operated from the platform provided
over the opening in the embankment. In several tanks in the Shimoga district, a
couple of examples discussed below, the sluice operating mechanism – plug and
pole – was stolen or had disappeared all of a sudden in the 1980s. Even when
they were replaced, they again went missing. Figure 2. An engineering line
diagram of plug and pole type of sluice (not to scale).
8 The categories of wet and mixed regions are constructed to include the
interaction between ecological and historical characteristics. See Shah, 2003 for
further discussion.
Take, for example, the case of Saulanga old tank in Shimoga district. Three canals
– two distribution canals located on the extreme edges of the atchakat that are
connected to two sluices and one drainage or seepage canal located in the middle
conventionally distributed water in the atchakat (see figure 3). After the
introduction of transplanted paddy of the high yielding varieties in the atchakat,
the section of the canal became narrower and finally disappeared completely. The
seepage canal collected Water Alternatives - 2012 Volume 5 | Issue 2 Shah: Tank
irrigation technology in Karnataka, India Page | 521
drainage water from the land on both sides and also served as a supply canal to
the tail end. Since the intensification of rice cultivation, irrigation in the atchakat
was entirely done from field to field, from head to tail. When earlier, water could
reach the tail end in a few hours, it now took at least seven to eight days. As a
result not only had the amount of tail end land receiving irrigation drastically
reduced but also the choice of crops had been severely curtailed. Tail end farmers
either plant semi-dry crops or plant early maturing, broadcasted paddy varieties.
Figure 3. A sketch of Saulanga tank atchakat (not to scale).
In the case of another tank called Chinnikatte Taverekere (figure 4), the left bank
canal (LBC) on the extreme edge survived, but the seepage canal in the middle
that earlier used to take water to the tail end had been heavily encroached upon
and silted up as a result its carrying capacity was reduced to one fourth of the
original. One more distribution canal on the right bank (RBC in figure 4), had
completely disappeared as a result of the encroachment. Sluice 1 of this tank,
which used to provide water to the RBC and which was located at a higher level
than the rest of the atchakat, had gone out of use. At the time of the field work it
supplied water to only a narrow strip of land on the extreme left edge. The entire
atchakat eventually received water from the LBC. Water was first distributed to a
part of the atchakat called vari jameen (upper land) from field to field. The
drainage from vari jameen was collected in the seepage canal and then
distributed to the lower parts of the atchakat called sara jameen (lower or
seepage land). The tail end received water once in eight days when it needed it
every day. Water Alternatives - 2012 Volume 5 | Issue 2 Shah: Tank irrigation
technology in Karnataka, India Page | 522
Figure 4. A sketch of Chinnikatte Taverekere atchakat (not to scale).
Figure 5. A sketch of Kumsidoddakere atchakat (not to scale). Water Alternatives -
2012 Volume 5 | Issue 2 Shah: Tank irrigation technology in Karnataka, India Page
| 523
Similarly, the plug and pole arrangement of the sluice disappeared soon after the
introduction of the green revolution varieties, ironically soon after it was repaired
and replaced by the PWD (Public Works Department). Before this period the
sluice had a heavy plug, which used to be opened three to four times in the
irrigation season. Each time the sluice was opened, it was kept open for a few
days until the entire atchakat was irrigated, and again closed until the next round
of irrigation. The plug and pole of the sluice existed for a long time, until only after
the intensification of paddy cultivation. The sluice since was stuffed with gunny
bags and paddy stems before the rainy season, opened in July, kept open for the
entire paddy season, again stuffed in October, opened in January, and closed
before the rainy season.
Yet another tank, Kumsidoddakere, had neither distribution canals nor sluice
operating infrastructure at the time of the fieldwork. The wooden plug and pole
existed, until the early 1970s, until the tank was managed by the village level
revenue officers – Patel and Shanbhoga. In 1971, the tank was handed over to the
PWD when officially the Patel and Shanbhoga stopped being responsible for the
tank. This happened at almost the same time the new paddy varieties were
introduced in the atchakat. The distribution canals existed at places shown in
figure 3. Tail end farmers of this tank alleged that soon after the tank was taken
over by the PWD the powerful farmers of the head reach, including the Patel and
Shanbhoga, first encroached upon the canals and later even destroyed the
remaining part. Destroying canals ensured that water was first supplied to the
head reach and reached the tail end only if it was allowed to.
This shift in the state of the crucial water distribution infrastructure, in the tanks
discussed above, does not seem to be the case of the deterioration of physical
structures as a result of the lack of proper management or maintenance. Instead,
in all tanks, the disappearance of canals and sluice operating mechanism were
sudden and related to the introduction of green revolution rice varieties and also
to the shift in the state irrigation policy. It is hard to comprehend why it would be
difficult for farmers to replace this part of the tank infrastructure if needed. The
plugs and poles were routinely replaced by farmers in other parts of the state,
and once replaced, they could easily last for twenty years. When I asked farmers
how difficult it would be to replace the plug and pole, the answer was, "it would
be futile to do so as they would be stolen again". Cleaning and mending of
especially earthen canals do not require either sophisticated technical assistance
or capital investment. Repair or reconstruction of any type of earthen or masonry
work such as the embankment, the stone revetment, the sluice platform or the
waste weir superstructure were difficult operations, hard for farmers to handle on
their own. Desilting of tank submergence area was another difficult operation,
which needed an organised form of technical assistance. Farmers, however,
routinely handled repairing and cleaning canals and replacing plugs and poles of
sluices.
The disappearance of the irrigation infrastructure was related to the political
economy of a changing cropping pattern and the shifting power relations at the
local level. Farmers I interviewed narrated the change in terms of "before, when
only broadcasted paddy was cultivated, and after, when all farmers began to grow
transplanted paddy of new varieties". I maintain the same distinction to narrate
the process of change over this period.
Tanks in the wet region of Shimoga district have a small capacity compared to the
size of tanks in other regions because tanks here fill up several times during one
monsoon season, which is also the irrigation season. Also, tanks in this region
have sufficient capacity for irrigating their atchakats once. Earlier, paddy was
usually sown in the last week of May by throwing dry seeds on the land that
sprouted after the arrival of the first showers. The sluices were opened by a
Talwar (a village servant) one and half months after sowing with the permission
from Shanbhoga and that too was postponed if there was enough rain. Water was
supplied three times during the whole season. Water thus was distributed
intermittently and not continuously and was also rotated between the tail end
and head reach. The presence of canal/s was absolutely crucial to ensure rotation.
Almost all tank-irrigated areas also had a seepage or drainage canal that carried
excess water from the head reach to tail end, which rotated water between the
head and tail end which also carried irrigation return flow to the Water
Alternatives - 2012 Volume 5 | Issue 2 Shah: Tank irrigation technology in
Karnataka, India Page | 524
downstream tank. One round of irrigation took 15 to 20 days depending upon the
size of the atchakat. Sluices were operated during the irrigation round – opened
in the morning and closed in the evening. For intermittent rounds of irrigation,
the sluice operating mechanism was kept in order, as each round of irrigation
would practically empty a tank. Sluice/s also had to be closed to store the next
round of inflow till the next round of irrigation. After the intensification, irrigation
was provided as per the requirements of transplanted paddy. Sluice outlets were
stuffed with gunny bags, paddy stems, stones and mud before the beginning of
the rainy season. They were opened at the time of land preparation and
transplantation when broadcasted paddy in the tail end would be one and half to
two months old and would require water. After the completion of the first round
of irrigation and depletion of water in the tank, sluices were again closed with
gunny bags, stones and paddy straw until the second round of irrigation. This
method of sluice control was fairly labour intensive, at times it even took two
days to entirely remove all the material stuffed inside the sluice openings. Only
farmers growing transplanted paddy closed and opened the sluice outlets with
the help of hired labour whom they pay from their own pocket. Once opened,
sluice/s were kept open for the rest of the season because closing them under
water was very difficult. Transplanted paddy needed continuous supply and
drainage during the maturing season except in the first month. Hence sluice/s
were again closed only at the end of the season. The disappearance of the canals
and the sluice operating mechanism ensured that water was supplied to the tail
end, field to field, only via the head reach and according to the requirements of
transplanted paddy, completely negating the possibility of rotation between head
reach and tail end. Thus, the disappearance of crucial water distribution
structures ensured that water was supplied according to the requirements of
transplanted paddy grown in the privileged patches in the atchakat. The
disappearance of essential infrastructure also signifies erosion of social relations
of power that previously maintained them. Nadkarni (1987), while discussing the
class character of the farmers’ movement of the 1980s, noted that, "the landlords
had a sense of security in the feudal order, which today’s rich peasants do not
necessarily have" (Nadkarni, 1987). The disappearance of the crucial physical
structures marked the breaking down of the old managerial order in which the
elites seemed to no longer have a stake. Their earlier involvement with the
management and maintenance of the physical structures earned them a right to
be part of the power-holding and decision-making authority, which in turn
ensured that their interests were protected. After the intensification, as long as
they had land that was located in the privileged part of the atchakat that received
assured irrigation, they became not only apathetic to the management and
maintenance of the physical structures but, as several examples of tanks in
Shimoga district show, they actually contributed to destroying some of the
structures in order to remain in the privileged position.
The erosion of such feudal power relations had another side too. On the one
hand, the farmers’ movement of the 1980s intensely affected the corridors of
power at the state level and since had also influenced the ruling governments and
the national policy on price of agricultural output. On the other hand, it also
resulted in the loss of external government authority to even nominally interfere
and create a normative structure of order at the village level. This perhaps was
the most evident in Shimoga district, the heartland of the new farmers’
movement. One of the important strategies of the movement in the 1980s was to
expose corruption in public life and oppose the attachment of farmers’ property
by government officers in lieu of loan recovery. In several occasions the
government officials were prevented from entering the village (Assadi, 1997).9
The farmers’ agitations were in response to the high handedness of government
officers and their corrupt practices. The success of the agitations
9 The local bureaucrats, especially from the Irrigation and Revenue Departments,
were targets of the farmers’ fury during these agitations. Their officers and files
were burnt at times as part of the agitation, for instance in the Malaprabha
agitation (Nadkarni, 1987). The government officers who came to attach farmers’
property to recover loans were hackled routinely, and were at times locked up
following the strategies adopted by Punjab farmers. Many villages were declared
as no-entry zones for the government officers (Assadi, 1997). Pitched battled
were fought between the KRRS (Karnataka Rajya Rayta Sangh – the farmers’
organisation) squad and government officers. At times some officers refused to
enter certain villages (Assadi, 1997). Water Alternatives - 2012 Volume 5 | Issue 2
Shah: Tank irrigation technology in Karnataka, India Page | 525
increased the power of the landed and elite farmers, especially sugarcane and
paddy-growers, to new heights. This farmers’ alliance succeeded in suspending
the normative structures of state authority at the local level.
CONCLUSION
The idea of the pre-modern holding an agency of hegemony’s other became
influential in India since the debates in the 1980s on the large dam on river
Narmada. The debate had many manifestations. A large part of the literature, on
the one hand, poses the pre-modern or traditional forms of irrigation as a viable
alternative to the modern forms of water management (Agarwal and Narain,
1997). A number of scholars on the other hand pointed out how the framing of
independent India’s dominant hydraulic models amplified colonial legacy which
relentlessly extinguished other forms of cultures and traditions for managing and
conserving water (D'Souza, 2008). This legacy, Zwarteveen traces to the
construction of hegemonic masculinities and irrigation cultures since colonial
times (Zwarteveen, 2011). In both these incarnations, the dominant modern is
evaluated and critiqued in the light of the 'outsider' – pre-modern – other.
In a similar vein, a significant part of the influential critiques of modernity, the
modern state, or modern science and technology – for example, in the works of
James Scott, Stephen Lansing, and Sandra Harding – are premised on the
assertion that the pre-modern was environmentally non-intrusive and
harmoniously embedded within a culture. The way these critiques pose the
substantive distinction between the modern and the pre-modern through the
reading of technology and knowledge systems insufficiently engages, I argue, with
the history of pre-modern technology and knowledge creation. Tank irrigation
technology, with all its presumed-appropriate attributes – small, decentralized,
traditional, Eastern, and local – is often touted as an ideal example of an
ecologically embedded and socially suitable pre-modern technology.
Going beyond the dichotomy of pre-modern and modern forms of knowledge
with mutually opposing attributes, I have shown here how a situated position –
seeing like a subaltern – can help build a complex history of tank irrigation
technology. In doing so, I have shown how in both pre-modern and modern times
the reproduction of technology implies the reproduction of cultural politics and
power relations in distinct ways. I want to emphasize the following points as a
way of conclusion.
Firstly, tank irrigation technology has not necessarily produced a democratic
social order, either in the past or in the present. Even when the state and polity in
pre-modern South India had little in common with the modern territorial state
and its concepts of sovereignty, exploration of the state-society-technology
relationship reveals an increasingly extractive statecraft involving coerced labor;
the expropriation of surplus by elites; and the spread of technological choices that
could be environmentally unsound – and that often resulted in forced
displacement, uncertainty, technological vulnerability, and social anxiety and
violence. Morrison in her recent work comes to the same conclusion by referring
to archaeological and historical evidence. By reconstructing the historical
development of irrigation and agriculture in sixteenth century, Morrison shows
how incidents of both catastrophic dam failures and construction of new ones led
to significant forms of devastation and displacement (Morrison, 2009). She also
shows how river-fed canals and canal-fed reservoirs were dedicated to the
production of commercial cash crops as early as the fourteenth century
(Morrison, 2010). Thus, tanks as artefacts were socially embedded in societies and
economies that were organized for warfare, sustained sharp social hierarchies,
and were often violent to women and people from lower castes. Clearly, the tank
as a techno-sociological artefact was integral to the forms of inequality and
violence of the pre-modern social order in South India. While not denying the
substantive differences between the pre-modern and modern forms of irrigation
technology in terms of size, scale and spread, what this essay asserts is that
seeing like a subaltern and based on the empirical inquiry into the social shaping
of tank technology in the pre-modern and modern historical periods, it would be
difficult to find any fundamental ontological contrast between a pre-modern and
a modern technology. Water Alternatives - 2012 Volume 5 | Issue 2 Shah: Tank
irrigation technology in Karnataka, India Page | 526
The fundamental problem with the great divide between modern and pre-
modern knowledge systems is that such a concentration attributes values and
ethics to the artefacts and knowledge systems and thereby reifies them, denying
their social shaping. This article argues instead that values emerge from
historically situated actors and not solely from the artefacts or knowledge
systems themselves.
Secondly, in modern times as well, tanks reinforce and reproduce the power
relations at the heart of the agrarian economy, as does any other modern
irrigation technology. Tank technology has been perfectly amenable to the
significant transformations heralded by the Green Revolution – one of India’s
most important state-sponsored, modern science- and technology-driven
development projects. In my own extensive field work on tank designs in
Karnataka, I found that the irrigation from the functioning tanks has intensified,
expanded, and thrived since the introduction of the Green Revolution. In these
tanks in many parts of Karnataka the designs of tank technology have undergone
modernisation initiated and implemented by the farmers themselves. Several
chapters in my earlier work describe in detail these processes of modernisation
and expansion (Shah, 2003). This shift in tank designs since the Green Revolution
has created, perpetuated, and reproduced new power relations.10 Therefore, in
radically different ways, but nevertheless, tank technology has reproduced
cultural politics and power relations in both pre-modern and modern periods.
Thirdly, a plea to abolish the great divide between knowledge systems is not
meant to suggest that technologies may not differ in their social implications. In
fact, this article argues that technologies shaped in radically different social and
political systems have different impacts upon society. A rigorous understanding of
these differences is critical for comparing, assessing, and evaluating technologies
of different eras. Ultimately, technological artefacts are not inert objects. They
articulate with a complex mix of social tensions, relations, power, and ideology
and transform with historical processes. Those who advocate pre-modern
artefacts and knowledge systems as objects of value and virtue fail to capture
their social and political scripting and ultimately deny the historical
embeddedness of technology.
10 See several chapters in Shah (2003) for further discussion on the way in which
tank technology and tank irrigation policy have changed as a result of the
recursive – that is, mutually transforming – roles played by the state and society
in the political economy of agrarian change.
REFERENCES
Agarwal, A. and Narain, S. 1997. Dying wisdom: Rise, fall and potential of India’s
traditional water harvesting systems. New Delhi: Center for Science and
Environment. Agrawal, A. 1995. Dismantling the divide between indigenous and
scientific knowledge. Development and Change 26(3): 413-439. Alvares, C. 1991.
Decolonising history: Technology and culture in India, China and the West from
1492 to the present day. Goa, India: The Other India Press. Assadi, M. 1997.
Peasant movement in Karnataka: 1980-94. New Delhi: Shipra Publications.
Baviskar, A. 2003. For a cultural politics of natural resources. Economic and
Political Weekly 38(48): 5051-5055. Baviskar, A. (Ed). 2007. Waterscapes: The
cultural politics of a natural resource. Ranikhet: Permanent Black. Breckenridge, C.
1985. Social storage and the extension of agriculture in South India: 1350 to 1750.
In Dallapiccola, A.L. (Ed), Vijayanagara city and empire, pp. 44-72. Stuttgart:
Steiner Verlag Wiesbaden Gmbh. Brewster, P. 1996. The foundation sacrifice
motif in legend, folksong, game, and dance. In Dundes, A. (Ed), Walled-up wife,
pp. 35-62. Wisconsin: The University of Wisconsin Press. D’Souza, R. 2008.
Framing India’s hydraulic crisis: Politics of the modern large dam. Monthly Review
60(3): 112-124. Dhanagare, D.N. 1987. Green Revolution and social inequalities in
rural India. Economic and Political Weekly 22 (19/21): AN137-AN144. Dikshit, G.S.;
Kuppuswamy, G.R. and Mohan, S.K. 1993. Tank irrigation in Karnataka: A
historical survey. Bangalore: Gandhi Sahitya Sangha. Dundes, A. (Ed). 1996.
Walled-up wife. Wisconsin: The University of Wisconsin Press. Water Alternatives
- 2012 Volume 5 | Issue 2 Shah: Tank irrigation technology in Karnataka, India
Page | 527
Gilmartin, D. 1994. Scientific empire and imperial science: Colonialism and
irrigation technology in the Indus basin. The Journal of Asian Studies 53(4): 1127-
1149. Gurukkal, R. 1986. Aspects of reservoir system of irrigation in the early
Pandya state. Studies in History 2(2): 155-164.
Haraway, D. 1988. Situated knowledges: The science question in feminism and the
privileging of partial perspective. Feminist Studies 14(3): 575-99.
Harding, S. 1994. Is science multicultural? Challenges, resources, opportunities
and uncertainties. Configurations 2(2): 301-352.
Harding, S. 2008. Introduction: Why focus on modernity? In Harding, S. (Ed),
Sciences from below: Feminisms, postcolonialities, and modernities, pp. 1-19.
London: Duke University Press.
Hauser-Schaublin, B. 2003. The pre-colonial Balinese state reconsidered. Current
Anthropology 44(2): 153-181.
Heitzman, J. 1997. State formation in South India: 850-1280. In Kulke, H. (Ed), The
state in India, 1000-1700, pp. 233-262. Delhi: Oxford University Press.
Krishnamurthy, A.N. 1975. Shimoga, Mysore. In International Institute of Rice
Research (Ed), Changes in rice farming in selected area of Asia, pp. 117-132.
Manila: International Rice Research Institute.
Kumar, D. 1997. Science and the Raj. New Delhi: Oxford University Press.
Kuriyama, S. 1994. On knowledge and diversity of cultures: Comment on Harding.
Configurations 2(2): 337-342.
Lansing, S. 1991. Priests and programmers: Technologies of power in the
engineered landscape of Bali. Princeton: Princeton University Press.
Lansing, S. 2006. Perfect order: Recognising complexity in Bali. Princeton:
Princeton University Press.
Ludden, D. 1985. Peasant history in South India. Princeton: Princeton University
Press.
Mollinga, P. 2010. The material conditions of a polarized discourse: Clamours and
silences in critical analysis of agricultural water use in India. Journal of Agrarian
Change 10(3): 414-436.
Morrison, K. 2009. Daroji valley: Landscape, place, and the making of dryland
reservoir system. Delhi: Manohar Press.
Morrison, K. 2010. Dharmic projects, imperial reservoirs, and new temples of
India: An historical perspective on dams in India. Conservation and Society 8(3):
182-195.
Mosse, D. 1997. The symbolic making of a common property resource: History,
ecology and locality in a tank-irrigated landscape in South India. Development and
Change 28(3): 467-504.
Mosse, D. 1999. Colonial and contemporary ideologies of community
management: The case of tank irrigation development in South India. Modern
Asian Studies 33(2): 303-338.
Mosse, D. 2003. The rule of water: Statecraft, ecology, and collective action in
South India. New Delhi: Oxford University Press.
Nadkarni, M.V. 1987. Farmers’ movement in India. Ahmedabad: Allied Publishers.
Palat, R. 1987. The Vijayanagara empire. In Claessen H.J.M. and van de Velde P.
(Eds), Early state dynamics, pp. 70-186. Leiden: E.J. Brill.
Paramshivaiya, G.S. and Hanumanthappa, B. (Eds). 1994. Madaga Kenchamma.
Bangalore: Karnataka Janpada Mattu Yakshagana Academy.
Playfair, R.E. 1866. A paper on irrigation in the Deccan and southern Maratha
country. Bombay.
Prakash, G. 1999. Another reason: Science and the imagination of modern India.
New Delhi: Oxford University Press.
Qaisar, A.J. 1998. The Indian response to European technology and culture: AD
1498-1707. New Delhi: Oxford University Presss.
Raina, D. 2003. Images and contexts: The historiography of science and modernity
in India. New Delhi: Oxford University Press.
Rajappa, T.S. (Ed). 1974. Kere Hunnamma Mattu Ittare Lavanigalu. Bangalore:
Kannada Adhyayana Samstha (Mysore University).
Scott, J. 1998. Seeing like a state: How certain schemes to improve the human
condition have failed. New Haven: Yale University Press.
Shah, E. 2003. Social designs: Tank irrigation technology and agrarian
transformation in Karnataka, South India. Hyderabad: Orient Longman.
Shah, E. 2008. Telling otherwise: Historical anthropology of tank irrigation
technology of South India. Technology and Culture 49(3): 652-674.
Sinopoli, C. and Morrison, K. 1995. Dimensions of imperial control: The
Vijayanagara capital. American Anthropologist 97(1): 83-96. Water Alternatives -
2012 Volume 5 | Issue 2 Shah: Tank irrigation technology in Karnataka, India Page
| 528
Srikantaiah, T.N. 1996. 'Kerege Haara' – A tribute. In Dundes, A. (Ed), Walled-up
wife, pp. 126-132. Wisconsin: The University of Wisconsin Press.
Srinivasan, T.M. 1991. Irrigation and water supply: South India, 200 B.C.-1600 A.D.
Chennai: New Era Publications.
Stein, B. 1980. Peasant state and society in medieval South India. New Delhi:
Oxford University Press.
Thurston, E. 1909. Castes and tribes of South India. Bombay.
Vaidyanathan, A. 1998. Tanks of South India. Chennai: Madras Institute of
Development Studies.
Yarlagadda, T. 1993. Social groups and economic change: 7th-13th century A.D. In
Parasher-Sen, A. (Ed), Social and economic history of early Deccan, pp. 158-181.
New Delhi: Manohar.
Zachariah, B. 2005. Developing India: An intellectual and social history. New Delhi:
Oxford University Press.
Zwarteveen, M. 2011. Questioning masculinities in water. Economic and Political
Weekly 61(18): 40-48.
THIS ARTICLE IS DISTRIBUTED UNDER THE TERMS OF THE CREATIVE COMMONS
ATTRIBUTION-NONCOMMERCIAL-SHAREALIKE LICENSE WHICH PERMITS ANY NON
COMMERCIAL USE, DISTRIBUTION, AND REPRODUCTION IN ANY MEDIUM,
PROVIDED THE ORIGINAL AUTHOR(S) AND SOURCE ARE CREDITED. SEE
HTTP://CREATIVECOMMONS.ORG/LICENSES/BY-NC-SA/3.0/LEGALCODE