Most metaphysical debates over free will start with the problem of compatibility of causal determinism and free will. Philosophers who hold that causal determinism and free will are incompatible fall into two camps: hard determinism that denies Free W ...
Most metaphysical debates over free will start with the problem of compatibility of causal determinism and free will. Philosophers who hold that causal determinism and free will are incompatible fall into two camps: hard determinism that denies Free Will and libertarianism that attempts to account for Free Will by means of causal indeterminism. The most powerful critique of libertarianism is that appealing to indeterminism rather impede than help to resolve the problem. To be more precise, anyone who argues for the freedom of actions of an agent faces dilemma as follows.
P1. Every action of an agent is either deterministic or indeterministic.
P2. If actions of an agent are deterministic, there is no alternative possibilities in them.
P3. If an action of an agent is indeterministic, then it is random.
P4. If an agent has no alternative possibilities in his actions, then his or her actions are not free.
P5. If actions of an agent are random, then his or her actions are not free.
C. Actions of an agent are not free.
This argument shows that what a libertarian who attempts to account for free will by means of indeterminism should do. A libertarian should persuasively deny (4). It is obvious that, if an agent freely acts, he or she must control his or her actions. However, if an event randomly occurs, it is not controlled by anything. Accordingly, it seems that libertarianism relying upon indeterminism cannot be successful.
Some philosophers suggest the theory of agent causation to make the libertarian approach successful. According to them, the notion of agent causation is required to resolve the problem of an agent's control of his or her actions, which I mentioned above. The basic idea of the agent causation can be summarized as follows. First, an agent as a substance is the direct cause of his or her actions. Second, an agent as a substance cannot be the effect of other events. Third, such entity-event causal relations are not reducible to event-event relations. Such theory of agent causation has ontological burden, because generally causal relations are regarded as those between events. Despite such ontological burden, if it is conducive to explicate an agent's control, the theory has attraction. This is because it is very important to explicate the freedom of an agent. The theory of agent causation apparently has such attraction. If an agent as a substance is the cause of his or her actions, it is not reasonable that they are produced by randomness. Moreover, actions caused by an agent are not governed by causal determinism, because an agent is not an effect caused by an event. If so, it can be said that actions of an agent are indeterministic but non-random. Accordingly, it seems that, if one accepts the theory of agent causation, he or she can reasonably deny P3.
In this paper, however, I argue that the theory of agent causation fails to deny P3. To do so, I explore the theories of agent causation proposed by O’conner and Clark, concluding that it is no use of denying P3. As for O’conner, agent causes cannot provide any appropriate reason-explanation of events as actions, which is required to show that the actions of an agent are not random. As for Clark, if he invoke reason-explanation in terms of mental events to complement the theory of agent causation, agent causes have no role in eliminating randomness. Instead, I suggest a way to deny P3 without appealing to the notion of agent causation. Of course, it is not that there are no problems in my approach. However, I argue that such approach is at least better than appealing to the notion of agent causation, in that the ontological burden of the theory of agent causation is too great to carry.