Pereboom Paraphrased

I evaluate Derk Pereboom’s argument in his 1995 paper "Determinism Al Dente". In Part 1 I paraphrase his argument and reasoning in favor of hard incompatibilism, the view that moral responsibility is incompatible with both determinism and indeterminism. In Part 2 I articulate a theoretical objection from a compatibilist standpoint. In Part 3 I consider how a proponent of Pereboom’s argument would respond to this objection. I conclude that Pereboom’s argument is successful.

Part 1: Pereboom’s argument

I formulate Pereboom’s argument as a simple syllogism:

1) If we are morally responsible, then our actions are self-determined. (Premise 1)

2) Our actions are not self-determined. (Premise 2)

3) Therefore, we are not morally responsible. (Modus Tollens)

The first premise says that moral responsibility requires self-determination. Why believe this first premise? For Pereboom, the fundamental incompatibilist intuition is the following: "moral responsibility requires that one’s action not result from a process that traces back to factors beyond one’s control." We might call this the “Traceability Principle” – if my actions can be traced to factors beyond my control, then those factors, not me, are to blame. For Pereboom, us not having moral responsibility means we “never deserve blame for having performed a wrongful act.” So for us to deserve blame for our actions, our actions must trace back to us, and the buck must stop with us. 

Premise 1 is easy to motivate. We feel it unfair when we are blamed for something beyond our control. Even a child will insist that something was an accident to avoid getting in trouble. We feel we do not deserve blame when we do something—or are something—we were forced to do or be. Consider someone who robs a bank. We naturally blame the bank robber for their actions, because their actions trace to their choices, and their choices trace back to their act of choosing. But say we discover that the bank robber was mind-controlled by a super villain. Now we trace the actions of the robber to the mind controller, and so we shift our blame accordingly. So it can seem like someone’s actions trace back to them when really they trace to something external (in this case, a super villain, but in a more realistic case, to the laws of nature). We thought the robber’s actions said something bad about the robber, but really they only speak to how unlucky the “robber” was to be a victim of mind control.

Why think Premise 2 is true, that our actions are not self-determined? While the bank robber’s actions were not self-determined, surely the actions of the mind controller were? Not so, argues Pereboom. Either determinism is true or indeterminism is true. But if determinism is true, our actions are not self-determined. And if indeterminism is true, our actions are not self-determined. So no matter what, our actions are not self-determined, not even the mind controller’s. Just as we shifted blame from the bank robber to the mind controller, we must shift blame from the mind controller to the laws of nature—things beyond anyone’s control. No person is at fault for anything; that’s the view to which free will skepticism amounts.

Why think that if determinism is true, then our actions are not self-determined? We might appeal again to the bank robber case. If the bank robber’s actions are determined by the mind controller, then the robber’s actions are not self-determined. But if the robber’s actions are determined by the laws of nature, then the robber’s actions are likewise not self-determined. What relevant difference is there between the two cases? In one case the robber is a victim to a malevolent agent. In the other case the robber is the victim of a rough childhood, poor opportunities, poor genetics, a lack of access to educational resources, or various traumas or social pressures, or some combination thereof. In both cases, the robber’s body is performing the actions of robbing a bank, but the robber as a thinking, feeling, conscious person is performing those actions not out of pure inner choice, but by force of circumstances. 

Why think that if indeterminism is true, then our actions are not self-determined? Here is where my language of “self-determined” proves useful. Some people argue that determinism is necessary for free will. Others argue it’s a detriment. Some argue that indeterminism is necessary for free will. Others argue that it’s a detriment. To keep things simple and clear, it’s best to talk about self-determination. Just by the term alone we immediately see the gist of the problem with appealing to indeterminism to rescue free will. If moral responsibility requires self-determination, then free will indeed requires a kind of determinism! So if indeterminism is true, this threatens moral responsibility.

But let’s give a specific reason for thinking that indeterminism threatens moral responsibility. If indeterminism is true, then it’s probably true at the level that current quantum mechanics describes. At least, we have no reason to ignore our current best science. But accepting this picture of things amounts to saying that when we deliberate between options, the option we end up with is, ironically, determined by random quantum events. Again, this is why language of self-determination is useful; on this language, indeterminism is actually a kind of (non-self) determinism. Put simply, if our choices are random, then our choices are not up to us. 

A more charitable interpretation of indeterminism says something like the following: if determinism is true, then we can imagine a machine that can predict all future events using the information of all past events plus the laws of nature. But let’s say that the past plus the laws of nature cannot entail the future until an important bit of information is added to our prediction machine: the actual choice of the free agent. So the prediction machine is “waiting” for that last input, the actual choice of the free agent, so that it can predict what the future will be. But the actual choice of the free agent does not occur until the present moment – the moment of the choice in the free agent’s mind. So the future can never be predicted. But surely a few moments before the actual choice of the free agent, the machine has enough information to predict what will happen only fractions of a second later.

But if the machine has enough information to predict that, then why couldn’t it predict what will happen just a few moments before? And a few moments before that? Now we have a regress problem. What is the non-arbitrary stopping point on this regress? 

Plus, are humans really that unpredictable that this machine would be unable to predict human choices? Given the fact that we can psychoanalyze and trace our choices back to obvious motivations and psychological inputs, it seems like human choice would not be so unpredictable.

Part 2: Objection

Recall Premise 1: If we are morally responsible, then our actions are self-determined. A compatibilist could reject this premise, giving the following syllogism in response: 

1) If our free actions speak to our moral qualities, then moral responsibility does not require self-determination.

2) Our free actions speak to our moral qualities.

3) Therefore, moral responsibility does not require self-determination.

Recall the question asked earlier about the relevant difference between the robber being manipulated by a super villain and the robber being determined by the laws of nature. One response is to point out that in the case of manipulation, there is nothing to be said about the moral quality of the robber. The robber is purely a victim in that circumstance. But there is something to be said about the robber’s moral character in the case of determination. If we define moral blameworthiness in terms of moral quality, then we are blameworthy, and hence morally responsible, for those actions that speak to our moral qualities even when determined. Put simply, if someone is determined to be a bad person, then they are still in fact a bad person and can be judged accordingly. Or put another way, what we really care about is what our actions say about us. And if our actions are free in the right way, then our actions do in fact reveal morally significant facts about our character, even if that character is determined and therefore not our fault.

Part 3: Reply

The free will skeptic can respond by distinguishing between two kinds of blame, what I would refer to as moral blame and critical blame. Someone is morally blameworthy just when their actions say something meaningful about their moral character. Someone is worthy of criticism just when their surface-level qualities are lacking. Put another way, we can blame the person as a subject who thinks, feels, and experiences, and we can blame the person as an organism or animal.

The free will skeptic must deny the legitimacy of moral blame. Nothing anyone ever does says anything meaningful about them as a person qua subject. Subjects are merely along for the ride, bearing witness to their own choices but not in control of them. Our actions speak only to a subject’s circumstances; the subject per se is merely lucky or unlucky to be attached to the body and brain to which they happen to be attached.

But there are many things we do that reflect our various qualities, or lack thereof, as an organism. A brain and a body are something we have, but not something we are, as far as this response is concerned. Brains and bodies have different qualities, qualities not up to us. Most fundamentally, we do not choose our intelligence, our rationality, our sensitivity to reasons, or our access to reasons for performing or refraining from actions.

Separating these kinds of blame allows me to take a critical (or praising) stance against a person (organism) and acknowledge their shortcomings (or excellences), while simultaneously attributing no blame (or credit) to the person (subject) for those shortcomings (or excellences).

Part 4: Evaluation

We started with the following argument:

If we are morally responsible, then our actions are self-determined. (Premise 1)

Our actions are not self-determined. (Premise 2)

Therefore, we are not morally responsible. (Modus Tollens)

I’m convinced that our actions are not self-determined. I’m further convinced that for our actions to say anything significant about us as subjects, our actions must be self-determined. Therefore, it logically follows that our actions say nothing morally significant, good or bad, about us as subjects. Our actions only reveal how lucky or unlucky we are as subjects to be attached to the brains and bodies to which we happen to be attached.

Luckily, we can still preserve common sense ideas of praise and blame on this view. While our actions say nothing significant about us as subjects, they do speak volumes about the qualities we have. We are highly interested in knowing which humans around us are rational, for example, as rational humans can be reasoned with, allowing us to improve their behavior. But we shouldn’t blame (or credit) people for having the qualities they do, as our qualities are not up to us. We can explain our common attitudes of praise and blame as acknowledgement of qualities.

I’ve heard those who believe in free will say that while we have no choice over the hand we are dealt, we do have free will over how we play our hand. But this metaphor is mistaken. How you play your hand depends on qualities you have, qualities you did not choose. How you play your hand is part of the hand you are dealt.