The paper examines the conceptual, ethical, and structural ambiguities which emerged in the Niger Range Livestock project, implemented between 1978 and 1983. Thinking around this time centered around pastoralists as superfluous to development schemes as they were nomadic, and in need of rescue from their own destruction. In this context, the introduction of participatory, beneficiary and socially sensitive development planning was surprising. The directive to ensure "local participation" is about channels of information, not content and is part of a democratic concern with government "by the people.' "Participation" assumes that everyone has an opinion to voice and that development should progress according to the aggregate will of the people, conditions which are not obtained in many pastoral societies. There is irony in the fact that anthropologists have favored the politically-loaded technique of participation. Involving local people is a complex issue: needs and policies which match the project capabilities have to be defined. There is no culture-free source of information on universal validity, and anthropologists need to exercise their ethical choices as much as their scientific judgment.