Abstract

This thesis examines compatibilist theories in relation to developments in the philosophy of action, beginning with the Humean "desire model" which dominated early in this century. It is argued that these "dissolutionist" forms of compatibilism failed to capture the meaning of "responsibility" and "free will" as ordinarily understood, partly because they depended on an inadequate theory of action and practical reasoning. Chapter III explores the role and meaning of "can" in compatibilist theories, arguing that the dissolutionist hypothetical analysis of "can" is also a failure. In chapter IV and V, I argue that the nature of practical reasoning presupposes distinctions between desiring and valuing as sources of motivation, and between the standpoint of the agent and the standpoint of the observer. The standpoint of the agent emerges as central to our notion of free action and to the fundamental meaning we attach to "can". While this does not mean that reasons cannot be causes, it shows that the language of action and the language of occurrences is incommensurable. The question becomes: whether what is affirmed within the agent standpoint can be logically reconciled with what is believed from the "observer" point of view without becoming "epiphenomenal". ;Chapters VI and VII argue that the "agent standpoint" is also central to our ordinary distinctions between actions and occurrences, and that Humean event causal theories of action fail to capture this distinction. Chapter VIII re-examines the question of compatibilism as an attempt to reconcile agent and spectator standpoints. Here I argue that because the languages of the two standpoints cannot be kept clearly apart, what looks like informal contradictions between the two levels do arise, and that while the incommensurability of the two languages does not permit a demonstration of formal contradictions between them, it also does not allow us to show how and why such informal contradictions can be resolved. Finally, I argue that unless we are prepared to distinguish ontological levels, we cannot explain how the two languages could be intelligibly "about" the same thing without exposing the agent standpoint to the threat of epiphenomenalism