Free will is a comforting fantasy

By Aristophanes

Robert Kane defends a theory of free will called libertarianism, in which freedom is incompatible with determinism, but still exists in humans due to the falsity of determinism paired with further argument. Although highly intuitive, Kane’s libertarian position, and others like it, fail to hold up under scrutiny.

Kane takes moral responsibility to require a concept of free will. To counter compatibilists, who believe that free will (or moral responsibility, for some semicompatibilists) is consistent with determinism, he appeals to the intuition of ultimate responsibility, which requires the existence of alternative possibilities for at least some actions. These alternative possibilities coexist with free will but do not wholly explain it. To counter hard incompatibilists, who believe free will cannot exist alongside determinism or indeterminism, whichever is true, he argues against the luck objection — the idea that the outcome of an undetermined event is up to mere chance and thus antithetical to freedom of decision.

Some philosophers have attacked Kane’s theory. John Martin Fischer, a semicompatibilist, refers to a strain of cases known as the Frankfurt examples in order to refute the supposed necessity of the availability of alternative possibilities in order for an agent to be held morally responsible for an action. Derk Pereboom, a hard incompatibilist, appeals to the luck objection to confront Kane’s understanding of an agent’s ultimate responsibility and self-forming actions. He claims Kane is wrong in asserting that indeterminism is somehow compatible with free will even while he takes determinism and freedom to be incompatible.

In this article, I will defend both of Kane’s opponents on their above objections to his theory of libertarianism. To do this, I will first present an account of a crucial component of Kane’s primary position, then proceed with an explanation of each objection, including, when necessary, further responses from Kane and my responses to those responses. Finally, I will conclude with a counterargument of my own, explaining what I view is a faulty assumption Kane makes in his explanation of self-forming actions.

Kanesian libertarianism

Kane believes determinism is incompatible with free will, making him an incompatibilist, but that free will can still exist because determinism is false. In his theory of libertarian free will, he appeals to the concepts of alternative possibilities, ultimate responsibility and self-forming actions. In Kane’s view, free will entails ultimate responsibility, which entails the existence of self-forming actions, which in turn entails an agent’s having access to, in at least some free decisions, a slate of alternative possibilities. Situations with alternative possibilities, where an agent has the ability to do otherwise, are not sufficient to fully explain how human individuals come to have free will, Kane says. Ultimate responsibility, or the idea that one is responsible for any sufficient cause of an action’s occurring, is required for this type of freedom, and this is due to the existence of self-forming actions. Particular cases of self-forming actions exist only when an agent has two opposing wills, when it is undetermined upon which the agent will act until the moment he or she actually does. These self-forming actions produce an inclination in the agent that can make him responsible for future actions committed as a result of the attitude he has formed for himself — even if the future actions in question are themselves determined. For Kane, this is the second way, after self-forming actions themselves, that an action can be done of one’s own free will.

To illustrate the idea of self-forming actions, a required component for this type of libertarian freedom, Kane presents the businesswoman example. Imagine an ambitious businesswoman on her way to an important meeting passes by a mugging in an alleyway. It is within her power to stop the assault, but if she takes the time to do so, she will be late for her meeting. Kane says that, in this moment, the businesswoman has two wills. Will A is to simply continue on to her meeting. Will B is to take the time to help the man. She cannot do both things, but she fully wills both at once, making it so that she is torn between the two options. When she has decided, we would say she is fully responsible for whatever path, whether A or B, she has chosen.

The woman’s freedom in taking whichever action she does is dependent upon us seeing the failed will as a sort of noise against which the successful will prevails. This concept of indeterminacy as an obstacle that does not undermine responsibility is explained further in Kane’s glass tabletop example. Imagine a man in an argument with his wife. In a sudden outburst, he swings his hand down on her favorite glass tabletop, aiming to break it. However, due to the nature of the compound, it is not determined whether the force will be enough to break the glass. Should the husband succeed in his effort, however, we would count him guilty; Kane says it would be ridiculous to exonerate the man by saying it was luck that broke the tabletop. Kane translates this intuition to the businesswoman example, only with the noise of the glass’ indeterminacy being replaced by the noise of an opposing will.

To Kane, the cases are analogous because both involve an obstacle that must be overcome for the will to be manifest as reality — this is the supposed noise in the system. The key difference, however, is that in the businesswoman’s case, the obstacle is an opposing will, whereas in the glass tabletop example, the obstacle is the undetermined outcome of the table’s either shattering or not shattering.

In the next two sections, I will explain and support counterarguments that other philosophers have raised to Kane’s theory. The first, from John Martin Fischer, attacks Kane’s requirement that at least some actions in an agent’s life possess alternative possibilities in order for the agent to have moral responsibility.

Alternatives are irrelevant

Frankfurt examples, as explained by John Martin Fischer, are meant to show that the existence of alternative possibilities is irrelevant in determining whether an agent has free will. The cases do this by way of presenting examples in which an agent does not have the ability to do otherwise, but who our intuitions say is responsible, regardless.

A popular Frankfurt example is the case of the undecided voter. Imagine a man named Jones has not yet decided which candidate to cast a ballot for in a certain general election. Jones goes into the polling place, deliberates for a moment, then votes for the Democrat. However, he is not aware that, were he to have attempted to do otherwise, maybe by voting for the Republican, a liberal neuroscientist would have forced him to vote for the Democrat anyway, employing electric signals emitted by a well-placed computer chip in Jones’ brain. As the case stands, none of this was necessary, for Jones voted for the Democratic candidate voluntarily. Thus, as Fischer explains, Jones still acted with moral responsibility despite his not possessing any true alternative possibilities.

But the case as it stands is simplistic. Jones could have done otherwise, as the neuroscientist would need some sort of signal in order to trigger the mind control. This would present an alternative possibility, however inconsequential, in that Jones could begin to choose to vote for the Republican, in which case he would be forcefully diverted. But, as Pereboom proposes, the case can be enhanced in a manner that avoids this worry, yet triggers a separate response — do sufficiently extensive illustrations lack the capacity to elicit trustworthy intuitions?

I think this last worry is unnecessary, but even granting that it holds, there is still a way to avoid the issue entirely. Imagine the neuroscientist is just an incredibly good-guesser — he is able to trigger the device every time it is needed and never unnecessarily. If his record is perfect, as we are supposing, then Jones really could not have done otherwise. In this case, no prior signal need occur to alert the controller, taking Jones’ previously described alternative action out of the equation. This scenario retains simplicity, thus mitigating the above question of the reliability of intuitions derived from convoluted hypotheticals. And because the intuitions here are similar to those in the original case, we can see that, unlike Kane’s assertion, alternative possibilities do not seem to be a necessary condition for the existence of moral responsibility.

Getting lucky

Derk Pereboom brings another objection to the table, attacking the claim that indeterminism is somehow a more acceptable companion to free action than determinism. The luck objection, as Pereboom calls it, although the general idea did not originate with him, says that agents cannot control indeterministic events. Lacking this control, they cannot have moral responsibility and thereby free will in connection to such occurrences. Specifically, agents can have no prior causal determination over undetermined future events, which, by their very nature, are not dependent upon resolving in a way that cleanly extrapolates from prior events and the laws of nature. This uncertainty prevents an agent from maintaining a relevant manner of control over the scenario — a type of control which moral responsibility seems to require. Because the event’s occurring is basically a matter of luck, it cannot foster the properties required for an agent’s being morally responsible.

Kane replies that such a component of undetermined self-forming actions is not sufficient to mitigate moral responsibility, and this is due to the nature of the scenario, which requires two opposing wills to manifest in an agent simultaneously. Just as with the noise present in the glass tabletop example, one will serves as an obstacle to be overcome by the other. Surmounting such probabilistic noise does not destroy our intuition of moral responsibility in the single-will example, so neither should it rule out such responsibility in the case of the businesswoman’s self-forming action.

However, it is not clear that Kane’s analogy holds, and we can see why if we divert from Pereboom’s counter to pursue a separate line of thinking. The glass tabletop example seems to be fundamentally different from the case of the businesswoman. When it comes to the former, the event willed by the husband, the table breaking, either occurs or does not occur, whereas in the businesswoman’s case, one or other of her wills must occur — there is no outcome in which everything we know she currently wills fails. If Kane believes that the indeterminacy of the event has no bearing on the responsibility we must assign to the agent, he is cornered into a position in which he must accept that, if we were omniscient, we would hold the husband equally responsible for his will to break the glass tabletop even if it did not actually break. If the presence of indeterminacy cannot exonerate him when the action does occur, as Kane mentions, this same manner of indeterminacy cannot exonerate him when the action does not occur, either, as the will in question has not changed. (The only manner in which such an excuse would be acceptable is in an instance of partial knowledge, such as a legal argument in a court case, where, of course, the judge and jury are not all-knowing. But, in these examples, that is not the case.)

Were this case actually equivalent to the businesswoman’s, and the observers were once again omniscient, we would have to hold the businesswoman accountable for both her will to help the man and her will to rush by and get to her meeting concurrently, regardless of which resulting option played out in the real world. The outcome here would have been a mere result of the indeterminacy of the woman’s brain particles and, as we established with the tabletop example, this would not pertain to our ruling of responsibility, as only the will to accomplish something would. Thus, if we accept all of Kane’s premises, it seems self-forming actions are irrationally conceived in that they force us to hold someone accountable for both doing and not doing a certain thing.

Simplifying will

In addition to Fischer’s and Pereboom’s criticisms of libertarianism, which I support, I offer a further objection, which specifically pertains to the way in which Kane composes the diametrically opposed wills in his case of the businesswoman. According to Kane, the woman in his example has two separate wills: (a) to help the man and (b) to arrive at the meeting on time. However, I am not convinced that this is indeed the best account of such a scenario. I see no reason why each of the wills must be thought of as separate simple sentences. In fact, I think there is a much simpler formulation — one that allows us an easier explanation of the agent’s brain states and incorporates information in alignment with the problem I raised in the final paragraph of the previous section. Mine being a leaner description with no less evidence supporting it, and one that sidesteps a rather troubling and illogical consequence, I believe it should replace Kane’s own description of the businesswoman’s opposing wills.

Perhaps we can understand what is going on here more easily if we assume the businesswoman’s wills stand as conjunctions: (a) to help the man, thereby missing the meeting and (b) to be on time to the meeting, thereby not helping the man. Because the “thereby” condition is explicit in each (as it should be, for the woman is surely aware of the undesirable consequences of each action as she weighs them against one another), accepting Will A seems to, on the very face of it, entail rejecting Will B, and vice versa. This creates a system in which the two wills cannot truly exist within a single agent concurrently, as the truth of one entails the falsity of the other; it would create a logical contradiction if we demanded both be true at the same time. There is no torn decision to be made. The wills together are both irrational and not possible. While it might seem to the outside world — and, perhaps, even to the businesswoman herself — that she holds two opposing wills, she can only truly hold one at a time. It should be obvious she cannot in one moment will herself to both succeed and fail at a given task, but in Kane’s example she must. This makes a genuine self-forming action strictly impossible.

To counter my objection, Kane must give a satisfactory answer to why he describes the businesswoman’s wills in the particular manner he does. He must also seek more convincing rejoinders for the Frankfurt-style attacks as explained by Fischer, as well as the luck objection as explained by Pereboom. Otherwise, his theory asks for too much. With several unwarranted assumptions required to fully explain Kane’s ideas, and the seeming logical contradictions contained within the current formulation of undetermined self-forming actions, free will libertarianism does not yet seem an adequately robust theory. ■


Fischer, John Martin, Robert Kane, Derk Pereboom, and Manuel Vargas. Four Views on Free Will. Blackwell, 2007. Print.

Kane, Robert. “Libertarianism.” Philosophical Studies 144.1 (2009): 35–44. Web.

Home → Philosophy

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s