Freedom Within Necessity
Suppose the compatibilist is faced with the following objection: human actions can’t be free because movements of the body are not free. More fully, if we consider the nature of bodily movements, we see that they have sufficient physiological causes and that from the point of view of physiology no freedom can be discerned therein. If we examine a bodily movement as such, we find nothing that could suggest the idea of free will; but human actions are such movements; therefore, there is no free action. My arm going up is not a free action, but then the act of raising my arm cannot be free either, since it is that movement. The body is a machine with not a hint of freedom in it, but that is not consistent with supposing that we act freely, because actions are movements (not counting omissions). If we are trying to describe a body, we will not have any use for the concept of freedom; but that rules out the idea that we act freely, given the role of the body in typical human action. Bodily actions can’t be free if bodily movements are not.
We can imagine different responses to this argument. One response would be to give up the compatibilist position: human actions are not free after all. Another response would be to insist that bodily movements are free, odd though it may sound to say so (“My arm going up was a free action”). Here I want to recognize the possibility of a third response: actions are not movements. This allows us to hold that actions are free but movements are not free—which looks like the intuitive thing to say. My raising my arm was free but my arm going up was not, because these are not the same thing. That is, identity is not the relation between actions and movements. What the relation is precisely is not easy to say, but let us stipulate that the movement is a constituent of the action, not the whole of it. In addition to the movement we might include an act of will or a trying or a decision—something psychological. Then Leibniz’s law will not compel us to transfer freedom to the body or deny freedom to the will. The will is free, and so in consequence is action, though action is partly constituted by a bodily movement that is not free. The arm rising was a constituent of the arm raising and is indeed not free, but the arm raising brings in other psychological factors that render it free (it was prompted by a certain desire, say). Thus we reconcile the freedom of the agent with the un-freedom of his body, by distinguishing actions from movements (even when the action involve the movements).
We must beware of an ambiguity in the phrase “movements of the body”. It is clearly true that I can perform the action of moving my body in a certain way, and this act can be free (for a compatibilist)—in this sense actions just are bodily movements. I can obviously move my body freely, as when I raise my arm. But this notion is not the same as that of a bodily movement in the non-agential sense, as when my body moves because of a muscle spasm or the wind. That is my body’s movement not my movement—my will played no role in its production. It is movements in this sense that are not free, even when they are part of a free action. Muscle contractions and efferent nerve impulses are not free, even though they make up an essential part of a free action. Actions are not bodily movements in this sense.
We thus arrive at the following position: actions can be free even though they are composed (partly) of movements that are not free. No movement of the body (in the second sense) is free, despite the fact that the actions of agents are free. For actions can be in accordance with the agent’s desires, rendering them free, despite the fact that movements are the body cannot be in accordance with desires. Of course, we can transfer or extend the concept of action to the movements that compose actions, saying that they are derivatively free; but that is consistent with supposing that strictly speaking no bodily movement is free. What is interesting is how much this position concedes to the traditional incompatibilist without accepting his conclusion that human action is not free. For it allows that no movement of the human body (in the second sense) is ever free while insisting that human actions are free (or can be). The body can be as “mechanical” as you like, all its movements devoid of anything deserving the name of free, and yet human bodily action is free. This is because actions are not identical with, or reducible to, bodily movements. When I think of myself as a body I can find no room for freedom, but when I shift to thinking of myself as an agent with psychological properties, freedom enters the picture. It is almost as if the very same thing—a bodily movement—is both free and not free: but that contradiction is avoided by scrupulously distinguishing between what I do and what my body “does”. It moves; I act. The incompatibilist was quite right that movements of the human body cannot be free, but he overplayed his hand when he concluded that free action is impossible. Free actions do indeed incorporate un-free movements, so they have an un-free dimension; nevertheless, they are free. The embodied agent has a body whose movements (in one sense) are not free, but he and his actions are still free. The incompatibilist is pointing to an important truth, namely that the human body is not free; but he is wrong to infer from this that actions cannot be free, despite the involvement of the body.
Some libertarians seek to make the body free in order to make human action free—they postulate some sort of indeterminism in the movements of the body (or brain). This produces absurd results, but their instincts were not completely wrong, because it is easy to move from the un-freedom of the body to the un-freedom of the agent, given the close connection between actions and movement. We can prevent this freedom-destroying step by carefully separating action from movement—raising from rising. Raising my arm is something I do to satisfy certain desires, but my arm rising is something that happens in virtue of physiological events in my muscles and nervous system. The former can be free while the latter cannot—even though the latter is an essential part of the former. It is because I am more than a body that I can be free; the incompabilist is right that the body itself cannot be free. If actions were simply movements, then freedom of action would be impossible; but that is not an identity we are compelled to accept.
The resulting picture looks like this: the body is thoroughly deterministic and un-free, including all of its movements; actions are also thoroughly determined, both physically and psychologically—and yet actions are free. Freedom exists within necessity not in opposition to it. Freedom for the compatibilist is fundamentally a matter of acting on one’s desires, but that can happen even when encased in a thick shell of determinism that includes even the un-free bodily movements in which freedom is expressed. Nearly everything about us is not free, but not quite everything. To put it as paradoxically as possible: my moving my body is free but the movements of my body are not free. We thus have a new form of compatibilism: freedom of action is compatible with un-freedom of the acting body.