Libertarian Accounts of Free Will

by Randolph Clarke

This comprehensive study offers a balanced assessment of libertarian accounts of free will. Bringing to bear recent work on action, causation, and causal explanation, Clarke defends a type of event-causal view from popular objections concerning rationality and diminished control. He subtly explores the extent to which event-causal accounts can secure the things for the sake of which we value free will, judging their success here to be limited. Clarke then sets out a highly original agent-causal account, one that integrates agent causation and nondeterministic event causation. He defends this view from a number of objections but argues that we should find the substance causation required by any agent-causal account to be impossible. Clarke concludes that if a broad thesis of incompatibilism is correct--one on which both free will and moral responsibility are incompatible with determinism--then no libertarian account is entirely adequate.

Genres: Philosophy, Metaphysics, Nonfiction

256 pages, Paperback

First published October 16, 2003

Reviews

[Dan DalMonte · Rating 3 out of 5]

This book is very dense and hard to read. Clarke develops an integrated agent-causal account, in which agent-causation is intertwined with event-causation. He is not developing, though, a Leibnizian view in which the agent stands above the chain of event-causes. The agent retains its rationality in virtue of being caused by prior events. Clarke says that the agent causes, but does not decide, her action. It seems like this account is more a reducible form of event-causation. Clarke doubts the possibility of substance-causation and is sympathetic to physicalism. I'm not sure his view is that different from event-causal libertarianism. This view only introduces a negative modification on the compatibilist agent, by creating non-deterministic causation but not providing any positive capacity for control. The non-deterministic agent is still subject to probabilistic causation, and does not have any independent capacity for controlling the course of events. Without a substance that is not an effect, I am not sure how Clarke's agent really separates itself from mechanistic causation.