Causation and free will are clearly related, but the precise nature of their relation has been underexplored. My main goal in the book was to bring the two concepts together in a compatibilist view of freedom—a view according to which acting freely is possible in a deterministic world—and, in particular, a view that understands our freedom as exclusively grounded in the actual causes of what we do.

An important motivation for this kind of view is a thought that drove Frankfurt to reject the traditional model of freedom based on alternative possibilities in a highly influential paper (Frankfurt 1969). Frankfurt’s insights in that paper extend beyond the reasons for rejecting the principle of alternative possibilities (commonly known as ‘PAP’), in that they also include the basis for a new model of freedom, one that rests just on actual causes. Frankfurt noted that it seems inappropriate to excuse your behavior by pointing to factors that in no way explain why you acted; if those factors didn’t at all drive you to act, they are arguably irrelevant to how responsible you are when you acted in that way. Although Frankfurt mainly had in mind explanatorily irrelevant factors that close off your alternative possibilities, such as powerful neuroscientists whose presence you never become aware of because they don’t intervene, the thought generalizes to other explanatorily irrelevant factors. Thus, this motivates the view that all that matters to responsibility—or, more precisely, to the freedom required for responsibility—is the actual sources of our behavior.

In the book I cash out this thought as the thesis that freedom is exclusively grounded in facts about actual causes (and whichever other facts may ground those facts). A supervenience claim also follows from this: facts about freedom supervene on facts about actual causes; in other words, any difference in freedom is reflected in a difference in the actual causes. What’s distinctive about this view is not the claim that facts about actual causes are relevant to freedom (since I take it that most people would agree with this), but the claim that this is all that matters. Although this is the most natural interpretation of the reasoning by Frankfurt, and although it results in a view that is particularly attractive given the simplicity and elegance of its core claim, I note that the main conceptions of freedom in this tradition [such as the “actual-sequence” view developed in Fischer and Ravizza (1998)] threaten to depart from this idea in important ways. This strikes me as a mistake. The value of the project, I believe, lies in preserving the basic, intuitive motivation. So I stick with it, and I explain how it can be developed into a plausible and attractive view of freedom: an “actual-causal-sequence” view or ACS.

The view works, I argue, because the facts concerning actual causes are rich enough to do the required work. We know that simple actual-sequence requirements, such as acting on the basis of our actual desires or reasons, aren’t enough for freedom. More is needed. In particular, the idea that a more robust notion of reasons-responsiveness is required for freedom is very compelling. However, it’s hard to make this idea compatible with the claim that freedom is just a function of actual causes, for reasons-responsiveness seems to be a counterfactual or dispositional concept: one that appeals, not just to actual causal histories, but also to other possible circumstances. I argue that this impression is misguided, and I offer in its place an “actualist” account of reasons-responsiveness, one that is only based on actual causes and their grounds.

The main idea is this. When we act freely, our acts have complex causal histories, consisting in patterns of reasons (the actual reasons for which we act) plus the absences of (many other, non-actual) reasons on the basis of which we also act. As a result, causal histories are substantially richer than they appear to be at first sight: the reasons we typically see, but the many absences of reasons we tend to miss, because they are absences and because we’re often not conscious of the fact that we’re responding to them when we act. I illustrate this point with the example of addicts who lack the relevant kind of control when they decide to take a drug under certain circumstances, and I compare it with the control enjoyed by non-addicts. I argue that, whereas both addicts and non-addicts may be similarly motivated by a strong desire to take the drug, there are still important differences in their actual sensitivity to reasons. These differences concern the role played by the absence of reasons of a certain kind. For example, when the non-addicts decide to take the drug, they are responding to the absence of certain emergencies, but the addicts are not (at least not to the same extent). Considerations of this kind account for the difference in control between addicts and non-addicts, and they do so purely on the basis of the actual sources of their behavior.

I also argue that the view works thanks to the fact that, in claiming that freedom is based on actual causes, one is thereby drawing on the metaphysical concept of causation itself. Without relying on any particular theory of causation, I argue that causation has some important features that help ground the moral responsibility of agents. (I explain that, if there is more than one concept of cause, as some people nowadays believe, this claim should be restricted to the relevant concept.) Causation is, I argue, an extrinsic and intransitive concept, and one that is tied to a particular form of difference-making. I argue that these and other features help address some important challenges that arise for the view, and I illustrate with some key examples. Throughout the book I assume that causation involving absences is possible (as when I claim, as explained above, that reasons-responsiveness is partly a matter of causally responding to absences of reasons). However, I note that this is mainly a simplifying assumption, for the main tenets of the view could in principle be reformulated in terms of other metaphysical notions that do similar work in grounding our moral responsibility—the example I use in the book is the concept of “quasi-causation” developed in Dowe (2001), but there could be others.

The book is divided into five chapters. In the first chapter I explain the basic commitments of a view like ACS, in particular, the grounding and supervenience claims on which it rests, as well as other preliminaries. In the second and third chapters I explain the work done by the metaphysics of causation in supporting ACS. In addition, chapter 2 contains an account of derivative responsibility or responsibility for outcomes, one that draws heavily on the concept of causation. In the fourth chapter I develop my account of basic (non-derivative) responsibility in terms of the actual causal sensitivity of agents to reasons and absences of reasons, and I compare the view with other reasons-responsiveness views that have been offered in the free will literature. In chapter 5, the final chapter, I respond to the main “source incompatibilist” objections to compatibilist views of freedom like ACS (these are arguments that attempt to show that determinism undermines freedom given the particular kind of causal sources that our acts would have if determinism were true). I discuss ultimacy arguments, direct arguments, and manipulation arguments—this last class of arguments, in particular, has become quite popular in recent years.

As can be seen from the way in which the book is structured, my ultimate goal was not to give a direct argument for compatibilism, or for a particular form of compatibilism. Instead, the aim was to articulate the best possible version of a compatibilist view within a certain family, and to then respond to some of the main incompatibilist objections that arise for a view of that kind. Still, it seems to me that, if one can accomplish all of this, one will thereby have made substantial progress towards a defense of compatibilism. Also, some would argue that this is the type of view that makes free will most clearly compatible with the truth of determinism, given that it doesn’t require alternative possibilities of action (which they see as threatened by determinism) but just actual causes of a certain sort. So, perhaps in this sense too, the articulation of a plausible version of this view is an important step in defending compatibilism.