This paper advances two arguments about environmental problems. First, it interrogates the strength and limitations of empiricist accounts of problems and issues offered by actor-network theory. Drawing on the work of C.S. Peirce, it considers how emerging environmental problems often lead to abductive inferences about the existence of hidden causes that may or may not have caused the problem to emerge. The analysis of environmental problems should be empiricist in so far as it is sceptical of the claims of those (...) who know in advance what the problem is, but it should also be alert to processes and things that are not readily traceable or perceived. Secondly, the paper’s contention is that environmental problems almost invariably involve an encounter between unlike or disparate materials or processes. In such circumstances, the challenge is to develop a form of inquiry that is alert to both the specificity of such encounters and to the specificity of the political situations in which they come to matter. (shrink)