Abstract

This thesis presents a sustained case in favour of the plausibility and coherence of a realist conception of causality, as opposed to the anti-realist conception associated with Hume. Four foundational issues for the theory of causality are addressed: the detectability of causality in experience, the implicit generality of singular causal relations, the validity of the notions of natures and causal powers, and the reality of causal necessity. ;The thought of Thomas Reid, H. W. B. Joseph, Rom Harre, E. H. Madden, C. J. Ducasse, John Searle, Nancy Cartwright and others, is synthesized and extended to render a unified realistic treatment of the metaphysics of causality that avoids a number of objections that have been raised against traditional Aristotelian causal realism. ;The major findings of this study are as follows: Hume's causal anti-realism is an outcome of his sensationalism and phenomenalism, doctrines of mind and knowledge that are open to refutation. Without these foundational tenets in place, causal anti-realism loses considerable force. Causality is, in some cases, detectable--under ideal conditions, in visual perception, and in the proprioceptive detection of volitional efficacy. These cases of causality-detection provide the primitive empirical content of our causal concepts, which can be extended from the volitional context to non-volitional contexts in a way that avoids charges of animism. The detectable instances of causality support singularism--the thesis of the primacy of singular causes, against the regularity view. The concepts of "natures" and "powers" are shown to play indispensable roles in action explanation and in experimental reasoning in the sciences and are therefore valid. Drawing a distinction between type and token determinism, the necessity attaching to the time-evolution of causal processes is characterized. Causality so described is shown to be consistent non-linear systems, a causal-realist interpretation of quantum mechanics and with human volition