Over the past half century, I have been developing and defending a libertarian view of free will that is incompatible with determinism. In the past decade, I have made changes to this view in response to the large critical literature that has developed around it since the publication of my book The Significance of Free Will (Oxford University Press, 1996). This paper describes and defends some of the more significant of these new aspects of my view. Section 1 describes the problem of free will as conceived traditionally and in modern times, as well as the various competing positions on it (compatibilist, libertarian, skeptical, and so forth). In section 2 I concede to compatibilists that there are many meanings of freedom, and that many of them could exist in a thoroughly determined world. But there is one kind of traditional freedom, I argue, that could not exist in a thoroughly determined world—freedom of will. Sections 3 and 4 explain what this freedom of will requires and why it is important. Sections 5–11 then discuss and answer many arguments that have been made against the possibility of such a free will requiring indeterminism.

A satisfactory construal of the subject matter of free will debates must allow for disagreements along two axes. First, it must allow for the possibility of higher order disagreements, or disagreements about what concepts, phenomena, or practices an account of free will is supposed to capture or explain. Second, it must allow for the fact of variation in the extent to which theories are bound by antecedent pre-philosophical thought, talk, and practices. A promising way of accommodating these two thoughts is to treat free will in broadly functionalist terms. On the account proposed here, free will is a power that makes sense of everyday responsibility practices. This construal is not widely shared, but it allows for the possibility that we might have false beliefs about the nature of free will, while still making sense of central philosophical debates about free will.

Moral responsibility skepticism has traditionally been dismissed as a nonstarter, but because of the important work of Derk Pereboom, Gregg Caruso, and others, it has become increasingly influential. I lay out this doctrine, and I subject it to critical scrutiny. I argue that the metaphysical arguments about free will do not yield the result that we do not deserve (in a “basic” sense) the attitudes and actions definitive of moral responsibility. Further, I argue that skepticism leaves out crucial components of our considered views about moral responsibility, making it seriously problematic.

John Martin Fischer has recently critiqued the skeptical view that no one is ever morally responsible for their actions in the basic desert sense and has defended a view he calls semiretributivism. This paper responds to Fischer’s concerns about the skeptical perspective, especially those regarding victims’ rights, and further explains why we should reject his semiretributivism. After briefly summarizing the Pereboom/Caruso view and Fischer’s objections to it, the paper argues that Fischer’s defense of basic desert moral responsibility is too weak to justify the kind of retributive blame and punishment he wishes to preserve. It then turns to the issue of victims’ rights and argues that Fischer is mistaken that victims want retribution above all else, and that the public health-quarantine model is better able to deal with the concerns of victims. It concludes by offering two additional objections to Fischer’s semiretributivism.

Benefiting from recent work in neuroscience, this paper rebuts a pair of neuroscience-based arguments for the non-existence of free will. Well-known neuroscientific experiments that have often been cited in support of skepticism about free will are critically examined. Various problems are identified with attempts to use their findings to support the claim that free will is an illusion. It is argued on scientific grounds that certain assumptions made in these skeptical arguments are unjustified—namely, assumptions about the times at which decisions are made and the times at which the point of no re­turn for the making of a decision is reached. It is argued as well that alleged findings about decisions of the kind featured in these experiments—arbitrary decisions made in the absence of conscious reasoning about what to do—fail to support certain crucial claims about non-arbitrary decisions made after careful deliberation. The paper also examines a fallback position for skeptics about free will. Some scientific skeptics about free will contend that free will, by definition, depends on the existence of immaterial souls. Survey-style studies that bring free will down to earth are brought to bear on this contention.

Many authors treat freedom and responsibility as interchangeable and simply apply conclusions about responsibility to freedom. This paper argues that the two are distinct, thus allowing for a “semi-compatibilist” view, on which responsibility but not freedom (in the sense of freedom to do otherwise) is compatible with determinism. It thereby avoids the implausible features of recent compatibilist accounts of freedom without alternative possibilities—as if one could make oneself free just by accepting the limitations on one’s choice. In particular, the paper puts forth two main arguments that responsibility does not imply freedom: an argument from “the stakes,” meaning what is at stake in a given action, or how seriously wrong it is, as affecting responsibility more than freedom, and an argument from temporal standpoint, that responsibility is often assessed from a standpoint farther back in time, when the agent could have taken steps to prevent being unfree later, whereas freedom is typically assessed at the time of action. These arguments are applied to the well-known case of Robert Alton Harris as in Watson (1988), in contrast to a Watsonian account of psychopaths as lacking moral responsibility because of “moral blindness.” Instead, what makes responsibility moral is the reasons for our blame, not necessarily what motivates the agent we take to be blameworthy.

It is often held that P. F. Strawson endorsed a radical and groundbreaking priority thesis according to which holding someone morally responsible is prior to (or more fundamental than) being morally responsible. I do three things in this paper. First, I argue for a novel interpretation of Strawson according to which he did not endorse a priority thesis that is radical or groundbreaking or original; instead, Strawson’s “priority thesis” is just a consequence of his view that the meanings of our words are determined by our usage and intentions and practices concerning those words. Second, I argue against the radical priority thesis that is often (erroneously) attributed to Strawson. Third, I argue that while Strawson’s view does not involve a radical priority thesis, it does imply that debates about the nature of moral responsibility (and many other debates about normative ethics, metaethics, and conceptual analysis) are trivial in a certain sense.