Agent-Causation and Libertarian Free Will

Leave a Comment
In this installment on the philosophy of free will, I will be discussing libertarianism, which is an incompatibilist position. In previous posts, I discussed mainly the compatibilist attempts to show that free will and determinism are both true. While this position has its share of problems, the reader will soon see that the libertarian position is not without its own issues. As with the other posts in this series, much of the information contained here is taken from A Contemporary Introduction to Free Will by Robert Kane.

Libertarian Free Will and Its Critics

I'll begin here with a concise definition of libertarian free will: in short, a libertarian holds the position that free will and determinism are incompatible, that agents have free will and therefore determinism is false, and that this free will entails that agents are able to take more than one possible course of action under a given set of circumstances. Anyone arguing for the libertarian position, then, assumes responsibility for two tasks: they must first show that free will and determinism are incompatible (which we have already seen attempts of in the previous posts), and then they must provide a coherent positive account for how free will is possible. There are a number of prima facie objections to libertarian free will that I will try to summarize.

The first objection to libertarianism is that when events are undetermined, they happen by random chance, and are not under the control of anything. And if they are not under the control of an agent, how can they be considered free or responsible? If a choice is undetermined, then exactly the same circumstances can lead to multiple decisions. Suppose Katie is trying to decide which job offer to accept; the exact same deliberation process, including all logical reasoning that goes into her decision, could lead to several different decisions. It is difficult to make sense of this. If Katie's decision to accept a job with Google over a job with Microsoft, for example, is the result of an undetermined event, rather than being determined by her motives or desires, then this seems to be a non-rational choice. Gottfried Leibniz responded to this by saying that prior reasons or motives need not determine choice or action, they may merely "incline without necessitating."1  While Katie may have reasons which incline her to choose the job with Google over the job with Microsoft, those reasons do not determine her to do so. She could still choose to accept the job with Microsoft despite the reasons and motives that would incline her to accept the job with Google. But does this get the libertarian off the hook? It seems not. Were Katie to accept the job with Microsoft, despite her reasons to go with Google, her decision would be positively irrational and inexplicable.

This leads to a related objection which been called the "Luck Objection." Recall that libertarian free will requires that two agents with exactly the same pasts and circumstances could make different decisions in the present, and therefore have different futures. Suppose Mike is in a situation where he is faced with the temptation to commit some selfish crime for personal gain. It seems, assuming indeterminism, that if Mike chooses to give in to temptation and goes to prison for it, he is simply unlucky. There is nothing about his power, capacity, state of mind, or moral character that explains the different possible outcomes. Mike is just lucky if he happens to resist temptation and unlucky if he does not.

The "Extra Factor" Strategies

Libertarians have responded to these objections in a number of ways which posit some extra factor to explain how a choice can be undetermined without being merely random or arbitrary. One such popular extra factor is body-mind dualism. Body-mind dualism holds that humans are a dual system of a material body and immaterial mind. The stance is often called Cartesian dualism after Rene Descartes who popularized one version of this view during the Enlightenment. Those who appeal to body-mind Dualism as their extra factor for libertarian free will, argue in the following manner. Free will only allows one to choose differently, all past physical circumstances remaining the same. "But the activity of an agent's mind or soul would not be among the physical circumstances and would not be governed by the laws of nature; and the activity of an immaterial mind or soul could account for why one choice was made rather than another."2 But does this extra factor succeed in preventing undetermined choices from being random or arbitrary? The answer is no. If a choice is not determined by prior physical circumstances, is it then determined by prior mental circumstances? And if this is so, how can they be considered free in the libertarian sense? If not, how do they escape the charge of being random or arbitrary? Kane summarizes the problem: "Placing the agent's thought and deliberations in a disembodied mind or soul does not solve the problems about an undetermined free will. Dualism simply transfers these problems to another level, from the physical sphere to the mental."3

A second extra factor was posited by Immanuel Kant in the eighteenth century. Kant's philosophy drew a distinction between the way things appear to us, and the way things actually are in themselves. The former he called the noumenal and the latter he called the phenomenal. Kant argued that we could never directly perceive the nomenal world, the way things are in-themselves, but rather were limited to the phenomenal world, or the way things appear to us in space and time. As such, while things appear to us to be physically deterministic (in Kant's time, Newtonian physics was universally accepted), the noumenal world may not be determined. Moreover, libertarian freedom is needed to explain moral responsibility, so there must be a moral law exhibited in the noumenal world that is undetermined. Of course, the reader can see problems with this view right away. In order to explain the mysterious way in which a decision may be free without being random or arbitrary, Kant is appealing to a concept even more mysterious! He never seeks to explain how it is that the phenomenal world can appear determined while the noumenal world is not, instead appealing to mystery. While some libertarians may be okay with this approach, it will certainly not convince a compatibilist or hard determinist.


A third attempt to explain libertarian free will in a coherent manner seeks to clarify what it really means for an agent to cause something. Roderick Chisholm, an American philosopher who taught at Brown in the twentieth century, explains what that means:

If we consider only inanimate natural objects, we may say that causation, if it occurs, is a relation between events or states of affairs. The dam's breaking was an event that was caused by a set of other events-- the dam being weak, the flood being strong, and so on. But if a man is responsible for a particular deed, then... there is some event... that is caused, not by other events or states of affairs, but by the agent, whatever he may be.4

What Chisholm is getting at is that agent-causation is a different thing altogether than event causation. While event causation requires a continual chain of events, each one causing the next, an agent can cause an action directly, without being caused to do so by some prior chain of events, or by being uncaused or random. Thus, there is a unique kind of causal relation between an agent and an action "that is not reducible to, and cannot be fully explained in terms of, the usual kinds of causation by events, occurrences, and states of affairs, either physical or mental." If this is true, then human beings have a status often only attributed to God-- namely, the status of being a Prime, or Unmoved, Mover.

Is Agent-causation a coherent account of libertarian free will? There are certainly many philosophers who do not think so. I will discuss their objections to the Agent-causal view along with a reworking of the view developed by Kane himself in the next post of this series (I know, I know, this was supposed to be the last post. But Free Will is so darned interesting, why wouldn't I want to post on it again??).


1. Leibniz, Leibniz-Clarke Correspondence, 1715
2. Kane, 2005
3. Ibid.
4. Chisholm
Next Post Newer Post Previous Post Older Post Home


Post a Comment