Add offer

Bali’s Subak: A History of Irrigation and Farmers’ Cooperation

Discover Bali’s subak history through inscriptions, canals, water temples, farmers’ agreements and the UNESCO inscription of 2012, with sources.

9th–11th centuries · early evidence; UNESCO inscription 2012 · Bali, Indonesia; example: Jatiluwih and Catur Angga Batukaru landscape · Land & water

The history of Bali’s subak concerns the collective management of an agricultural landscape. Fields, channels, temples and rules of cooperation form an interconnected arrangement that cannot be understood by looking at rice terraces alone. Early medieval inscriptions document aspects of irrigation, while a text dated 1071 contains the term kasuwakan. There is no single securely established date for the invention of the entire system. The UNESCO inscription of selected landscapes in 2012 recognised an existing heritage; it did not create the institution or mark the beginning of its history.

An inscription is not an institution’s birthday

Early records refer to different components of water-based agriculture. An account published by Bali’s provincial agricultural administration identifies a tunnel specialist in the Bebetin A1 inscription of 896. The Pandak Badung inscription of 1071 contains kasuwakan. The first concerns technical expertise; the second is important for the documentary history of irrigation organisation and areas. Neither establishes that every rule associated with later subaks was created in one event.

Dating therefore needs to distinguish channels, associations of users, temples and the vocabulary surviving in texts. ICOMOS places the system’s development within Bali’s changing political history and acknowledges uncertainty in parts of that explanation. The oldest surviving mention establishes a boundary of documentation, while practice may have begun earlier. Conversely, a later description cannot prove that every rule familiar today operated in the eleventh century. Maintaining those distinctions avoids assigning an apparently precise foundation date to an institution whose components developed over time and whose history must be reconstructed from different kinds of evidence.

Gravity, terrain and dependence between neighbours

Terraced agriculture uses differences in elevation, but gravity alone does not allocate water fairly. Intakes, channels, tunnels and division structures guide flows towards fields. They must fit the slope and be maintained in working order. Gany’s technical account connects physical infrastructure with the organisation of its users. A view of green terraces therefore shows the outcome of continuing work, rather than an agricultural landscape operating automatically because the climate is wet.

A field’s position has social consequences. An upstream holding can influence the water available below it, and damage to a shared channel affects more than one user. Historically, this moves the analysis beyond the supposedly self-contained plot: an individual farmer makes decisions inside a wider network. A topographical map can show where water might flow, but agreements explain who is entitled to receive it and when. The difference is fundamental when interpreting engineering remains without a complete record of the communities that operated them.

Water rights, obligations and everyday maintenance

A subak is an organisation of users, not simply the name of a technique. Gany describes the relationship between rights, responsibilities, collective work and awig-awig rules. UNESCO’s 2012 decision likewise draws attention to codes defining membership and obligations. Access to irrigation should consequently be considered alongside contributions to maintenance, meetings and other collective commitments. A channel considered without its arrangements for upkeep provides only a partial picture of the institution.

Recognising the importance of rules does not require imagining a society without disputes. Cooperation offers a way to deal with interdependence; it does not demonstrate that everyone has identical interests. When repairs are required, or restricted flows make an agreed sequence necessary, engineering possibilities meet questions about recognised obligations. Historical analysis should distinguish the written ideal from actual practice. A code documents a norm, while observation and participants’ accounts help reveal how labour was apportioned and what happened when obligations were not met. The distinction prevents a formal rule from being mistaken for a complete description of social relationships.

Water temples and the larger organisation of the landscape

UNESCO describes temples as points linking farming communities, water and religious life. Tri Hita Karana connects relationships involving the spiritual realm, people and nature. Religious meaning should not be treated merely as decoration added to hydraulic equipment. For participants, ritual, obligation and agricultural practice may belong to an integrated whole, even when an outside researcher separates them into analytical categories.

Temple networks did not develop outside political history. The ICOMOS evaluation discusses periods of greater centralisation and later dispersal of power, while research debates the extent of farmers’ autonomy. A simple image of timeless democracy operating entirely independently of the state would therefore be misleading. Interpretation needs several scales: the holding, the local association, temple connections and the wider political setting. Each reveals part of the story, and none alone explains a millennium. This is also why the presence of a temple cannot, without further evidence, establish the exact authority exercised by its custodians over every surrounding field.

Why planting dates become a collective concern

Management concerns timing as well as quantities of water. In Lansing and Kremer’s model, related cropping schedules can limit opportunities for pests to move among neighbouring fields. Complete synchronisation over too large an area, however, can concentrate irrigation demand. The pressures pull in different directions: protection of the harvest encourages coordination, while limited flows require attention to location and the timing of water use. The practical problem cannot be reduced to asking every farmer to make an isolated choice.

A study published in 2017 compares a model of farmers’ decisions with spatial patterns of crop stages detected in satellite imagery. It helps explain how local actions can produce organisation over a larger area. It is not a record of discussions held a thousand years ago. The distinction matters: a modern model investigates a possible mechanism, while historical evidence documents particular institutions and events. A mosaic of differently coloured fields cannot by itself reconstruct the entire history of decision-making. Its value lies in providing evidence against which a clearly specified explanation can be examined.

Agricultural change and the limits of a simple tradition narrative

Intensification programmes in the later twentieth century altered the conditions in which local organisations operated. Lansing and de Vet treated the displacement of temple networks from ecological management as an important test of their explanation of those networks’ role. In that interpretation, new schedules could disrupt existing coordination. This is a research argument about relationships between institutions and environments; it does not establish that every new variety or technique necessarily caused failure.

The Stockholm Environment Institute’s 2015 report records a more complicated set of farmers’ expectations. Participants wanted technical assistance and production inputs, while connecting the continued cultivation of local varieties with costs and income. Maintaining heritage did not necessarily mean freezing every farming method. The historical question is therefore how opportunities to negotiate change developed. Assessing an agricultural programme requires following its effects on collective calendars, labour arrangements and the position of people who continued to bear the risk of a failed harvest. An account centred only on whether a tool or variety was old or new would miss those relationships.

A living institution: modelling and migration

Marco Janssen examined the robustness of the Lansing–Kremer model under changed assumptions. His analysis supported the possibility of effective coordination emerging from local decisions, while emphasising assumptions about pest dynamics and decision-making. Such criticism sharpens the interpretation: a simulation’s result depends on what the model represents and leaves out. It is neither universal proof that every community is perfectly coordinated nor a reason to dismiss farmers’ knowledge. Models contribute by making explanations explicit enough to question.

Dik Roth’s research among Balinese settlers in South Sulawesi examines an institution carried into a different setting. Organisations developed differently in relation to infrastructure, administration and local arrangements. The comparison demonstrates why copying the name subak does not reproduce an entire system. In agricultural history, transferring knowledge also involves negotiating rights and responsibilities. An institution can retain recognisable elements while changing its range of activity; continuity does not require every detail to remain identical. Migration thus provides a useful counterpoint to accounts that present tradition as permanently attached to one landscape and incapable of adaptation.

The 2012 inscription and a landscape where people still work

In 2012, selected Balinese landscapes associated with subak entered the World Heritage List, rather than every organisation and field on the island. The decision recognised the importance of connections between infrastructure, agriculture and culture. It did not transfer everyday channel maintenance to UNESCO. The difference between recognising value and sustaining its conditions is fundamental: a terrace remains an agricultural landscape because someone continues to work it and can support a holding through that activity.

The ICOMOS/ICCROM mission of 2015 emphasised direct participation by subak communities in heritage governance. SEI’s report from the same period documented institutional coordination problems and residents’ expectations. Together, these records illuminate a history of conservation itself: an inhabited, working landscape needs space for farmers’ decisions as well as protection of its appearance. Tourism, changing land use and infrastructure costs entered the debate as concrete economic matters. Subak consequently offers two connected historical perspectives—the long development of cooperation around irrigation and the much more recent process through which agricultural practice came to be recognised and administered as heritage.

Questions and answers

When was the first subak founded?

No single secure foundation date exists for the whole system. Early inscriptions document different aspects of irrigation; a record from 1071 contains kasuwakan. That dates a surviving reference, not a one-off invention.

Does subak mean only irrigation channels?

No. It also involves users’ organisation, rights and duties, shared maintenance and religious practices associated with farming. The physical equipment alone does not explain the institution.

Did UNESCO inscribe every rice terrace in Bali?

No. The 2012 inscription covers particular components of a cultural landscape representing the system. Sustaining its value also depends on continued farming and community participation.

Sources

History of agriculture: all topics