Philosophy in Review XXXI (2011), no. 4 Joseph Keim Campbell Free Will. Cambridge: Polity Press 2011. 200 pages US$59.95 (cloth ISBN 978-0-7456-4666-4); US$19.95 (paper ISBN 978-07456-4667-1) The free will debate can be approached in several different ways. For some, it is a debate within moral philosophy, typically about the justification of punishment; for others, it is a debate about the mind and its causal powers. For yet others, it is a debate about the nature of fundamental reality, and therefore falls within the province of metaphysics. It is this last approach that Campbell takes in his excellent book. Campbell covers every major position in the debate, but he is especially concerned with arguments for free will skepticism. Free will skepticism, as Campbell understands it, is motivated by incompatibilism—the view that free will is incompatible with determinism—combined with the view that indeterminism cannot help. Since Campbell understands the debate in this manner, he centers the book around discussion of arguments for incompatibilism, especially the ‘Consequence Argument’ and the ‘Direct Argument’, both associated with Peter Van Inwagen. The Consequence argument, informally and roughly stated, argues from the claim that if determinism is true then our acts are the consequences of facts that predate our existence plus the laws of nature. Since neither the laws of nature nor the facts that predate our existence are up to us, their consequences—our actions—cannot be up to us either. This argument, in its several variants, has been much discussed and refined. Campbell covers this debate and adds to it. He points out that the argument rests on an assumption that is contingent: that there are facts that predate our existence. Worlds in which time is circular or in which agents are eternal are conceivable. Therefore the argument is not fully general and cannot serve as a proof of incompatibilism. Some philosophers will respond to this argument by shrugging their shoulders. So what if the argument is not fully general, they will say, we care about our freedom. That is a reasonable response, but it is not available to someone who is motivated, as is Campbell, primarily by the metaphysical issues. It is on questions like this that the different motivations driving different perspectives on free will really start to tell. Campbell takes a different approach to the Direct Argument, which resembles the Consequence Argument but does not assume that free will requires the power to actualize more than one possible course of action. He argues that the Direct Argument, like the ‘Manipulation Argument’ (which advances cases meant to demonstrate that responsibility-sapping manipulation is relevantly like determinism) and the ‘Ultimacy Argument’ (which holds that an agent must be the ultimate source of their actions in order to be responsible for them) all suffer from a common flaw: accepting them is incompatible with the (near universal) acceptance of the Frankfurt-style case as an 251 Philosophy in Review XXXI (2011), no. 4 argument against the principle of alternative possibilities. In Frankfurt-style cases, an agent (A1) performs a particular action on her own, without any interference from another agent (A2) who, unbeknownst to A1, would intervene to force A1 to perform that very action were A1 to show a sign that A1 would not perform it. Apparently, such agents as A1 are morally responsible despite not having alternatives to the action they perform. If Campbell is right, incompatibilists cannot accept this result while also accepting the validity of a major argument for incompatibilism. This is an important finding, because it shows that these arguments make more demanding assumptions than their proponents have thought—assumptions, moreover, that are widely rejected. Campbell notes at the outset that he is motivated by what he sees as the formal parallels between epistemological skepticism and free will skepticism. Thus, he often criticizes proponents of free will skepticism for having a hyperbolic conception of free will, akin to the Cartesian notion of knowledge as indubitability. Just as many philosophers think that we ought to respond to skepticism motivated by the failure to demonstrate that there are many significant truths that cannot be doubted by bringing our conception of knowledge down to earth, so Campbell thinks that the right response to free will skepticism is to bring the notion of free will down to earth. Perhaps we cannot have (say) ultimate sourcehood, but perhaps something less demanding is enough. This move is tempting and plausible, but it deserves more discussion that Campbell can give it here. Epistemic success is something that is clearly instantiated in the world: some people are better at predicting the course of events than others. This is a powerful reason for thinking that whatever knowledge is, it is something that human beings are capable of. But it is not obvious that we are ever free or ever morally responsible. It is not obviously more mature to think that whatever kind of control we actually have over our actions just has to be control enough to count as freedom. It might have to be control enough to achieve our ends as organisms, but we routinely think that free will is a distinctively human property, and therefore we apparently ought not to identify it with this adaptive property. Of course we can reject the assumption that free will is distinctively human, a move that has much to speak in its favor. But if we take that path, then it is not clear that this notion of free will can serve the purposes to which we—and Campbell—want to put it, such as justifying punishment. Campbell has aimed for concision in his text and achieved it admirably. He covers an enormous amount of ground in just 105 pages of main text. This strategy has costs as well as benefits. The major benefit is that the reader cannot conceivably be bored by this breakneck tour, but the major cost is that the arguments do not receive the attention they deserve. The book contains a number of promising and original new arguments, often presented in only a paragraph or two. Many of these arguments deserve a full, articlelength, elaboration. The result is a book that specialists in the debate will find endlessly rewarding, but that newcomers will find daunting. It would make an excellent graduatelevel text, but only in a context in which students have an expert instructor to guide them through its pages. Neil Levy Florey Neuroscience Institutes, Australia 252