Abstract

This paper criticizes Coliingridge’s arguments against an epistemology of technological control. Collingridge claims that because prediction mechanisms are inadequate, his “dilemma of control” demonstrates that the sociopolitical impact of new technologies cannot be forecasted, and that, consequently, policy makers must concentrate their control measures on minimizing the costs required to alter entrenched technologies. I argue that Collingridge does not show on either horn that forecasting is impossible, and that his criticisms of forecasting methods are self-defeating for they undercut his positive case for the control of entrenched technologies. Finally, I indicate an empirical base for forecasting risk that may define epistemic principles of technology assessment.