Abstract

Although Hume successfully showed that causation could not equal necessary connection, he did not himself relinquish the idea of necessity. His own psychological analysis of causation was therefore unsuccessful. The insistence on necessity has continued to haunt many of the responses to Hume, and remains the downfall of other regularity-based accounts. While such Humean solutions have failed, Hume's problem is genuine, and any satisfactory theory of causality must answer both his ontological and his epistemological worries. ;Physical theories of causality do not rely on necessity, and when supplied with a proper account of spatiotemporally extended events, can satisfy Hume's ontological concerns. They do not, however, address how observers come to know causes. I show how one such theory---the conserved-quantity theory advanced by Dowe and Salmon---can be made to answer that question, if philosophers attend to the empirical details of perception. ;Working with the Dowe-Salmon theory but relying mainly on the chief insight that causation is closely linked to CQs, I have tied this theory to recent work in experimental psychology. Studies of causal perception and event perception have demonstrated observer sensitivity to velocity and relative mass. Working from this tradition I show how it is possible for observers to perceive other dynamic quantities such as energy and momentum. If observers are indeed sensitive to normal exchanges of such CQs, then at least some instances of causation can be directly perceived. ;I present two experiments which demonstrate such sensitivity in dynamical simulations of simple physical events. Furthermore, observers were more sensitive to anomalous increases of energy or momentum than to anomalous decreases, indicating a sensitivity to causal asymmetry in the form of entropy