Existing studies of irrigation systems show that technical elements influence social processes, and also, that certain social relationships may have technical implications. However, little has been said about the precise content of this interplay. A better insight seems important, as irrigation systems, when put into operation, often have unforeseen organisational and social consequences, and a disappointing performance. The present thesis aims to answer the question: How do the social practices of people relate to the physical infrastructure of an irrigation system and its organisation, and how do thew practices change over time. To this end I studied the organising processes of an irrigation system, that of El Operado in the valley of Autlán-El Grullo, Western Mexico, from February 1987 through April 1989. To explore this theme I use the concepts of practice, social interaction and intervention. Through a focus on intervention I envisaged being able to single out the crucial factors for diagnosing irrigation organisation, and to assess viable intervention strategies.To capture the social dynamics that emerge around irrigation situations, I chose an actor-oriented research method, based on detailed observations (qualitative and quantitative), informal interviews, and (group) discussions. I also modestly participated in the stream of events concerning the El Operado's Water Users' Association.The present monograph on El Operado brings the system's social and technical dimensions under one common denominator, that of practice. To accomplish this, the thesis has the following structure. After an introduction to El Operado's historical and technical context, case studies follow on three irrigation activities. The focus then shifts to interventions in El Operado's management, and I analyse how, as a result of these interventions, people's practice changed. On the basis of this material I finally give an outline for an intervention perspective.The thesis is written for irrigation scholars and practitioners in the field. The study forms part of a multidisciplinary research project of Wageningen Agricultural University and the Colegio de Jalisco, Guadalajara, entitled 'Contrasting patterns of irrigation organisation, peasant strategies and planned intervention: comparative studies in Western Mexico'.The contextThe context of an irrigation system can be sketched through an appreciation of historical developments and the spatial-technical features of the system. An historical investigation of projects related to water and land development in the Autlán-El Grullo valley, from 1850 to the present, shows that such projects faced three recurring themes: a strained relationship between insiders (from the valley) and outsiders; contradictions between small farmers and the local elite; and the awkward position of intermediary actors and of farmer organisations, such as the ejidos. As we will see, the social dynamic found in the El Operado irrigation system, which was constructed in the 1950s, can also be characterised by these three sets of problematic relationships.Spatially, El Operado with its 9,000 hectares covers nearly half of the Autlán-El Grullo plain, and serves some 2,500 plots, 69 % of which planted with sugar cane. The 1,300 water users include a small number of entrepreneurs who control 40% of the land, and a majority of small farmers (1,100, mainly ejidatarios organised in 18 ejidos), who have less than 10 hectares. The third category of actors are the personnel of the Rural Development District ( Distrito de Desarrollo Rural ), a local office of the Ministry of Agriculture and Hydraulic Resources (SARH). The District, among other things, manages El Operado, and is responsible for distributing water to the farm plots, for canal maintenance, and for decisions concerning water fees, crops and conflicts.An important technical feature of El Operado is that all division structures are adjustable, which allows water delivery to follow the 'on request' method. Farmers are free, within limits, to chose the crop they cultivate, and to chose when to plant and when to irrigate.Irrigation practiceIn getting to grips with irrigation practice in El Operado, 1 focused on water distribution, canal maintenance and the making of the irrigation plan.The actor-centred study of the water guard, or canalero , revealed how, through properly operating the structures, he succeeds in tailoring water delivery to a variety of local demands. He has internalised the physical properties of the canal system. Water guards do not strictly adhere to the technical guidelines from District engineers, which are too general, but interpret them while drawing on their technical competence. Water guards are 'interfacers', linking farmers to the District, the District to the sugar refinery, and creating linkages among farmers.In El Operado problematic situations occur every year at the end of the rainy season, when canals silt up, and the District's maintenance department is unable to duly clean them. When water shortages become serious because of the silted canals, groups of farmers organise their own solutions, this despite the fact that farmers often regard group efforts with scepticism. Joint undertakings vis-à-vis the physical infrastructure (e.g. investing labour, money, and social capital in cleaning a canal) have an impact on the participants' sense of ownership over it; it shifts from 'the govemment's' to 'ours'. Canal groups repeatedly engaged in joint activities that took on a lasting character. Paired with farmers' new sense of ownership this provided a perspective for future farmer involvement in El Operado's management.The irrigation plan which the District's operation department formulates every year, appeared not to be an important instrument of decision-making for El Operado's management. Crucial decisions, for example regarding the water fees or granting certain fanners permission to grow sugar cane, are negotiated by a few important actors behind closed doors. In a meeting with the water users commission, District engineers did not provide sufficient information, so that the farmers' leaders were unable to critically assess the irrigation plan and the proposed water fees, nor formulate alternatives.In conclusion, the analysis of irrigation practice revealed that different groups of actors, such as water users, farmers' leaders, District engineers and field personnel, develop different practices. These practices occur in different domains: 'the field', the domain of the farmers; 'the office', the domain of the District engineers; and 'the ejido', the domain of farmers' leaders. In El Operado, it appeared that these domains of interaction co-existed alongside each other and were relatively autonomous.Intervention and change of practiceThe following step in the analysis dealt with intervention and change in El Operado's management. Of three interventions reviewed, one intervention concerned the formation of a new Water Users' Association, which took eventual charge of canal maintenance.The intervention resulted from power struggles within the intervening institutions, but its outcome was shaped by the interface encounters that unfolded during implementation: encounters and negotiations between actors stemming from different social worlds or holding different positions (e.g. engineer - farm leaders). Importantly, the intervention left particular actors out, namely the ordinary farmers, which is crucial to our understanding of what happened. Although the intervention had some immediate effects (improved canal maintenance for example), it remained confined to the formal level and did not change the practices of the people concerned.Practices finally started to change when the Water Users' Association also got involved in water distribution. The position of 'gate leader' was instituted, who supervised water distribution along one minor canal. On a higher level, farmer delegates of the Association together with the water guard coordinated water distribution in each of El Operado's six zones. With this structure a linkage between board and delegates on the one hand, and water users and water guards on the other, was established. Put differently, domains got interconnected, and farmers, for example, could now better check their delegates.This new organisational structure means that the basis of the Association not any longer is formed by ejido representatives but by water users who represent a canal. The Association can now discontinue obsolete practices originating from the ejidos, can redirect the Association's lop-sided organisation structure, and can concentrate on practical issues rather than on political wheeling and dealing. Having touched down at canal level, the Association links up with farmer practices found there. As we saw earlier, this has already led to farmers appropriating their canals. The association now follows suit.PerspectiveThe analysis of El Operado, summarised above, describes the actions and interactions of people involved in the irrigation system, while taking seriously their material settings. Practice appeared a sensitising concept which helped to pin down the interplay between people and things in this micro-society.In answer to the research question, different groups relate to the canal system in specific ways: District engineers have the overall responsibility for it, field personnel have to work with it, and farmers depend on it. These groups develop particularpractices as people cope with the constraints put forward by the system. Practice articulates with physical infrastructure through the actions of people in specific domains. The irrigation system's arena is made up of different domains, but some domains are disconnected. This helps to explain some of the operational problems. Interventions aimed at solving these problems succeed if they change existing practice accordingly. Changes of practice occur when new settings of interactions are put in place, and when different domains formerly separated become interconnected. New linkages imply that practice at one level feeds practice at another.On the basis of these findings, 1 suggested a type of diagnosis of irrigation organisation which is based on a descriptive account of practice concerning key activities; of the different domains present, and of the existing linkages between domains. Practice can be described through examining the context in which an irrigation activity takes place, the social interactions it entails, the interpretations people give of this practice and the strategies the different groups involved develop. When the diagnosis detects problems, these can be discussed with the people concerned. With them possible solutions can be reviewed and intervention strategies designed.The type of interventions which I have in mind aims at changing irrigation practices, through strengthening some practices found in specific domains while breaking down others. The approach thus is both about continuity and discontinuity. This can be pursued through adapting formal organisational arrangements to informal initiatives and experiences. Adjustments on the formal level could include defining more precisely ownership relations with respect to the canal system, and to irrigation water; (re-) defining responsibilities that groups of people have towards the system and towards each other; re-organising the system's arena through establishing new linkages between domains while breaking down others; and adapting the settings of interactions. This approach to intervention starts from the idea that irrigation management implies 'interfaces', that is, interactions between actors with structurally different positions, and it acknowledges that these interactions are of an emergent character, which means that their outcomes can never be wholly planned. The approach thus aims to carefully manage interfaces, but without the illusion of controlling them.Finally I turn to my own experience with the research methodology which centres around actors, and tries to understand people's actions and practices. Interviews with the 'research population' often resulted in interface situations, in which the attitudes and actions of both myself as well as of my discussion partners were questioned. This is a first small step in an intervention process.