A Metaphysics for Freedom

Steward's book is an attempt to reorient the discussion of free will. Traditionally, the question has been what kind of freedom is required for moral responsibility, and the focus has been on humans. There has been little discussion of the freedom of such things as sharks and bats. One might defend this restriction on the ground that nonhuman animals do not have the kind of freedom required for moral responsibility. If that is the kind of freedom that ultimately interests us, the thought goes, then we can restrict our attention to humans.But Steward detects a bad argument here. Even if nonhuman animals are not candidates for moral responsibility, they may have a kind of freedom that is required for moral responsibility. A morally responsible agent must after all be an agent. And agency, she argues, extends well down the ladder of nature, vanishing only when we get to plants, sponges, and paramecia.Moreover, Steward argues that it is agency per se, and not just the kind of freedom required for responsibility, that conflicts with determinism. The very existence of animals, therefore, entails that determinism is false. The result is a form of libertarianism, the view that combines incompatibilism with the rejection of determinism. But the incompatibilism that drives the argument concerns agency as such, not just morally responsible agency. That such a view is available is one of the chief revelations of the book. Steward has shown that libertarianism need not be “the lunatic thesis” that humans alone are exceptions to “an otherwise more or less universal mechanism.”But Steward is not merely out to show that a more inclusive libertarianism is possible. She goes on to argue that since there is such a thing as action, determinism must be false. Determinism, for her purposes, is the idea that at any given moment there is just one physically possible future. The thought behind incompatibilism is that if determinism is true then our actions are not “up to us.” But how could something even be an action, Steward wonders, if it was not up to us what we would do? The very idea of an action, she argues, is the idea of a closing off of a previously open future. For an agent to act is for it to “settle” the answer to a question about what it will do, when or where it will do it, or how it will do it. The sense of settling in play, she says, is that of rendering the alternative answers physically impossible. Since settling the answer to a question entails that the answer was not already settled, action entails that there was more than one physically possible future. If there are actions, therefore, determinism must be false.For an event to count as an action, however, more than mere settling is required. In an indeterministic world an atom might, by emitting a particle, settle the answer to the question whether it will emit a particle. But its closing off of a previously open future would not count as an action. This is because an atom is not the sort of thing that can “move its body.” Steward does not deny that an atom might be a body, in the sense in which a lake is a body of water, or a tumor is a body of cells, or a planet is a heavenly body. But it does not have a body, in the sense in which a fox or a spider does. That is why it is not the kind of thing that can move its body. And for Steward, only the kind of thing that can move its body can act. This rules out not just atoms but also sessile sponges and paramecia. They do not support the “owner-body distinction,” and are therefore not “self-moving” in the sense required for agency.Steward is now within striking distance of the conclusion that determinism is false. She outlines her argument as follows:If universal determinism is true, the future is not open.If there are self-moving animals, the future is open.There are self-moving animals.Therefore, universal determinism is not true.Later I will raise questions about premise (3). But first I want to give some sense of how much more happens in this wonderfully rich book.While Steward has shifted the focus of the free will discussion to the topic of animal agency, she has not simply changed the subject. Animal agency remains a necessary condition of the kind of freedom required for moral responsibility. Steward therefore remains very much in contact with the familiar literature on free will. She takes up van Inwagen's well-known argument for incompatibilism, for instance, arguing that it is inconclusive, and noting that van Inwagen fails to question the assumption that determinism is compatible with action. She considers whether so-called Frankfurt-style cases undermine her analysis of action as a form of settling and argues that they do not. She discusses the possibility of a compatibilist analysis of settling in terms of causation by mental states, arguing that any such analysis will either omit or presuppose the very phenomenon that needs to be explained, and that animal agency does not in general require causation by such states. She responds to the worry that her libertarianism confuses agency with mere chance, or with a capacity to act in ways that make no sense against the background of one's existing motivations. And she addresses the charge that she is committed to the existence of a mysterious form of causation, treating us to a fascinating discussion of how living things generally might exhibit causation that is “top down.”One possible avenue of escape for the compatibilist would be to accept Steward's criticisms of mental-state-causal accounts of action but propose that substance causation is compatible with determinism. On such a view, animals, considered as substances or things, could exhibit agency in a deterministic world by causing their bodies to move. To my mind, wisely, Steward does not work up and discuss a version of this idea. There is a time to speculate about what others may say and a time to wait and see what they do say. It is by no means obvious just how a compatibilist analysis of the notion of an animal's moving its body might go.Now I want to return to the argument that determinism is false, and specifically to premise (3), the claim that there are self-moving animals.Steward opens her fourth chapter by asking, “Do any nonhuman animals deserve to be called agents?” To this point, she has proposed that the concept of animal agency is the concept of a certain kind of closing off of a previously open future. One might expect the fourth chapter to address the question whether anything in the world actually answers to this concept. That is the question one needs to answer if one wants to substantiate premise (3).Instead, Steward argues that our indeterministic agency concept is essential to our capacity to experience animals in the way that we do, and in particular to our experience of them as centers of some form of subjectivity or mentality. She makes it plausible that the development of this concept begins in infancy and that the concept is part and parcel of our folk psychology. None of this, however, addresses the question whether anything in the world corresponds to our agency concept. None of it, therefore, substantiates premise (3). The argument against determinism thus remains incomplete.How might one complete the argument? One might argue that since, in a deterministic world, there can be no animal agency, neither can there be animals. There might be plants, sponges and paramecia, but that would be all. This would provide us with a powerful argument that determinism is false. For we need only spot a squirrel in the garden to know that plants, sponges, and paramecia are not the only living things.This way of arguing, however, relies on an assumption about what sorts of living things could emerge in a deterministic world. The assumption is that in such a world, evolution could not serve up anything with the kind of “complexity” and “integration,” as Steward calls it, that we find in a squirrel, or indeed in a worm. It could not serve up anything, in other words, that we might mistake for an animal. There are two problems with this assumption. First, we don't currently know whether it is true. Second, whether it is true does not seem to be a question for philosophical inquiry.Suppose, then, that we do without the assumption of a kind of ceiling on what sorts of things could evolve in a deterministic world. In that case, it becomes hard to see how we could know whether the things we take to be animals are animals. We have learned from Steward that they count as animals only if their movements are in fact settlings of matters that were previously open. But how do we decide whether their movements are such settlings if we have not already ruled it out that determinism is true?It is quite unclear, I think, how we can complete Steward's argument against determinism. But it is a mistake to regard that as the argument of the book. Indeed the value of the book does not at all depend on the success of the refutation of determinism. Steward's primary aim is to rescue libertarianism from its image as an ad hoc solution motivated solely by a desire to make room for moral responsibility in humans. Her libertarianism is a principled philosophical response to a very general problem about freedom.I want to close with a question about the principled response. In the cetology chapter of Moby-Dick, Ishmael famously defines a whale as “a spouting fish with a horizontal tail.” One can imagine a time when it would have seemed that the concept whale was the concept of a kind of fish. And one can imagine it turning out that the concept was not so rigid as to force us to conclude, from the fact that there are no spouting fish with horizontal tails, that there are no whales. (In another much-discussed example, Dalton defined ‘atom’ in terms of indivisibility.) Suppose it seems to us now that the concept of agency is the concept of a kind of closing off of a previously open future. Might it turn out that the concept of agency was not so rigid as to force us to conclude, from the fact that there are no such closings off, that there are no actions? It is worth asking whether we should expect a philosophical analysis of the concept of action to tell us, one way or the other, whether action is compatible with determinism

Contemporary philosophy · Epistemology · General interest · Metaphysics · Free Will and Agency · Philosophical Ethics and Theory · Philosophy · War, Ethics, and Justification