Abstract

Theories about ``target'' risk level are thought to explain why road casualty rates remain steady despite interventions to reduce them. Unfortunately, such theories appear inconsistent with the higher casualty rates associated with poor visibility because they imply that input to the road user is accurately and veridically processed. Motion perception for a driver, however, is highly dependent on relative motion in the retinal image: In poor visibility, the associated processing may be inaccurate and nonveridical. In this article, practical implications regarding conspicuity and lighting are outlined, and the difficulty of developing global theories to explain ``real-world'' activities is shown. Theories to explain the restricted aspects of such activities may be necessary, and the definition of risk may require considerable elaboration or change.