Research Methods In the field work for this research we combined rapid rural appraisal methods with more formal individual farmer interviews. Where possible, we worked in teams of three people: an agricultural scientist, a social scientist and a person skilled in communicating with farmers. We worked in a total of 12 villages in four states of India's SAT; the villages varied by agroclimatic conditions such as rainfall, soil type and slope.4 We began by transecting the fields, covering different aspects of the landscape. We spoke casually with farmers whenever possible to learn about their perceptions of erosion and efforts to control it. We then conducted open-ended group interviews to help formulate hypotheses about the determinants of investment. Finally, we carried out detailed individual interviews to gather data to test those hypotheses. During the course of this process we continually repeated steps to cross-check our findings. The approach is summarized in Figure 1. Figure 1: Field Research Methods In all of the surveyed villages we observed three categories of fields in regard to SWC measures. First, those where indigenous practices were implemented and maintained on a regular basis, resulting in satisfactory conservation of soil and water resources. Second, those which were badly neglected, resulting in severe erosion of soil. Third, those somewhere between the first two categories. It is important to note that we easily could spot the neglected, eroded fields from a distance, even while driving on the road. Fields protected by indigenous technologies, on the other hand, could not be appreciated until we visited them individually. GATEKEEPER SERIES NO. SA34