Skip to main content

Verified by Psychology Today

Cold Comfort In Compatibilism

The soft-serve version of free will seems to license punishment – unfortunately.


Father, forgive them; for they know [not] what they do.


Father, forgive them; for they know [not] what they do.

~Luke 23:34 [brackets added]

The only conception of free will that I find interesting is the libertarian one. According to this view, humans (but not other animals) can freely decide what to do regardless of the past or present state of the universe. The human will is seen as the only force that can set in motion a new causal chain, with itself being uncaused. This type of free will is radical and godlike; it creates something out of nothing; it does not have to answer to prior causes. Alas, this conception of free will is certainly false, impossible, and logically incoherent. No one (correct me if I’m wrong) has demonstrated that or how this type of will can triumph over the old soldiers of necessity and chance. Radically free will does not exist because it cannot exist.

Those who believe that responsibility and punishment (and praise) can be apportioned to acting humans only if the existence of a radically free will is assumed must hang on to this idea lest their moral judgments become groundless. Not being able to pass moral judgment seems abhorrent if not terrifying to them. Hence, the doctrine of the radically free will is moralistic and religious at its core. That’s why Nietzsche and the Hyperboreans objected to it.

Still others, most notably David Hume and some prominent contemporary social psychologists, believe they can have it both ways: accept determinism while also postulating a type of non-libertarian, straight-jacketed “free” will that still enables moral judgment [I put the “free” in quotation marks because the semantics are drained from the word].

This is the compatibilist view of free will (see here for an excellent article in the Stanford Encyclopedia of Philosophy). How is this supposed to work? First, we have to accept the view that prior events have caused the person’s current desire to do X. Wanting to do X is fully determined by these prior causes (and perhaps a dash of true chance). Now that the desire to do X is being felt, there are no other constraints that keep the person from doing what he wants, namely X. At this point, we should ascribe free will to all animals capable of experiencing desires (e.g., to eat, sleep, or mate). Yet, we don’t; and we tend not to judge non-human animals in moral terms. Exceptions occur, but are swiftly dismissed as errors of anthropomorphism.

Some aspiring compatibilists maintain that only humans are judged morally because only they could have acted differently. Those who try this argument must realize that they are not compatibilists at all; they are libertarians. The acceptance of determinism is a defining element of compatibilism. It forbids us to say that evil-doers could have done good if only they wanted to. Well yes, if they wanted to, but they were determined to not want to.

Hence, the compatibilist must find a defense for moral judgment that is applicable only to humans and that is safely nonlibertarian. He must look for a psychological feature that is presumably uniquely human and that is involved in the causal chain leading to action. The general version of this feature is self-consciousness and the specific version is intentionality. In other words, a person is judged to have acted freely and (ir)responsibly if he was aware of his desire to do X, foresaw the consequences (e.g., how moralists would judge him if he did X), and endorsed the desire (thereby forming an intention). Notice that a true compatibilist, who has gone on record saying that determinism is a fact of nature, must believe that the events of experiencing a desire, foreseeing the consequences of action, and forming an intention to act on the desire, are all determined. The causal chain leading a human to lift a finger is longer than the chain leading a squirrel to lift an acorn, but it is no less deterministic (he who says that it is less deterministic is not a compatibilist but a closet libertarian).

Hume and others have argued that humans can be held responsible for their actions because these actions reveal their character. In a particular situation, a human does exactly what he needs to do and nothing else. The action is not freely willed in the meta-physical, libertarian sense, but it reveals the person’s unique will, which is a manifestation of what and who he is. Moral approbation and recrimination become judgments of a person’s character with no pretensions that the person could be anyone but himself. Hume said that no other basis of morality is possible.

A rigorous compatibilist approach avoids the incoherences of the libertarian (incompatibilist) approach, but many social scientists and laypeople feel that punishing (or rewarding) people for who they are (what they do reflects who they are) is unfair. Formal law and folk law make allowances for actions that were not intended, i.e., based on desires not consciously endorsed. This notion of extenuating circumstances would make sense within the libertarian paradigm, that is, on the assumption that some actions could have been unchosen by the person. For a true compatibilist, however, it should not matter. If actions are fully determined, added features such as conscious awareness should only be “interesting,” but irrelevant for moral (or legal) judgment.

If notions of responsibility – compatibilist or otherwise – fail to provide a justification for punishment, what does? The main alternative is the consequentialist concept of deterrence. Killing a murderer is ok, according to this view, if his execution lowers the number of future murders committed by others. Those who apply the logic of deterrence rigorously and purely would tell the murderer, “Look, your impending decapitation is nothing personal; it is not retributive; it is not even directly motivated by your evil deed [which occurred in the past, which is dead and gone – pardon the poor choice of words], but exclusively by our rational calculus, which says – on the basis of empirical evidence – that there will be fewer murders in this state if we decapitate you now as opposed to letting you rot in prison.” In its pure form, the logic of deterrence is exactly that: a matter of logic and hence rationality; it is not a matter of morality, and that’s why few people like it. Most people want judgments of punishment (and praise) to be moral. What you get from this is a lot of muddled arguments, e.g., the attempt to justify moralistic punishment with evidence for its deterrent properties [We really want her to suffer for what she did, and by the way, her suffering will teach others a lesson].

The hard version of deterrence is difficult to sell because the concept of deterrence does not care about the past crime and who committed it. It only cares, by definition, about lowering the future incidence of crime. If it were found that killing a murderer’s child would be the most effective deterrent, then that would be the way to go. Of course, this scheme would not play, at least in most democracies, and that is my point. Deterrence must be shored up with moral (deontological) principles that make it more acceptable to the public, but also less rational. Deterrence also faces serious practical problems. Punishment works if it is swift and sure. In the legal world, it is neither. That is so partly because of sloth, incompetence, or venality, but also because the accused has a (moral) right to due process. And that takes time.

Now for a final discontent with the discourse on morality. Compatibilists and other moralists love to point out that we must hold people responsible for their actions. If we don’t, society will break down. The examples given typically include the gravest misdeeds: rape, pillage, and burn – and of course murder. The more heinous the acts are, the greater is our sense that by executing punishment we are morally good.

If we failed to punish a true crime, we would be committing a Type II error. Moralism’s goal is to minimize this type of error. There is, however, another type of error, conveniently called Type I. This type of error consists for false positives, which in our context is the punishment of a deed that is not “really” wicked. The contemporary discourse in moral psychology typically conditions its analysis on the assumption that a bad deed has occurred, one in which one person harms another. Here, a Type I error can still occur, but only if there is a fine constellation of exculpating circumstances. Otherwise, the only question is how harsh the punishment should be.

The other, and more common, type of Type I error occurs when there is no harm at all, but only communal disapproval. Consider masturbation. In Genesis 38:9, Onan breaks the law by interrupting coitus with his brother’s widow. The law, presumably divinely inspired, commands him to produce an heir for his deceased brother, but this is against his interests. By not completing the act, Onan becomes a criminal and God (who is Love) kills him for his wickedness. Harsh, right? Still, it is noteworthy that the ancient Hebrews were far more rational than contemporary moralists. Their community was governed by laws designed to regulate property rights and keep the group from breaking up. Onan’s sexual pleasure was not the point (he was actually expected to have it). Modern moralists object to masturbation because it is pleasurable, and they have to invent harm in order to bamboozle the naïve.

Masturbators can also be inventive. Suppose Joey fantasizes about Rachel. Moralists will object and the intensity of their condemnation will be a concave function of fantasized Rachel’s age; it will be an increasing monotonic function of her genetic relatedness to Joey. Why condemn Joey at all? Why indeed? As a latter-day Hyperborean, I have advocated the abolishment of punishment, unless it is swift, sure, and just harsh enough to modify behavior. Moral condemnation is for suckers and by suckers.

Compatibilism, character & shame

Compatibilism breeds shame, not guilt. Recall that according to Humean compatibilism, you are responsible for your actions because they reveal your character, not your free will. If responsibility demands an emotion, it is shame, not guilt. You feel bad about who you are, not about what you have done. There is no way of changing who you are and you are stuck with your shame. A libertarian has an easier time. He can feel guilty about having done a bad deed, but he can promise (credibly to himself and other libertarians) that he will freely mend his ways.

Shame is a far more destructive emotion than guilt, in part because of its non-fixability. The author of Genesis seemed to be aware of this. He -- or she -- notes that Adam & Eve felt shame after the apple, but not before. There is little talk of guilt, other than Eve blaming the serpent. To a compatibilist, Adam & Even were evicted from the garden for shame, not because they were guilty.


From time to time (too often actually) I find myself scratching my head after reading a commentary to a post. The current post stimulated this from Anonymous:

"as a young, well-educated individual..i must say that i have absolutely no idea what the fuck you're talking about."

Right. I deleted the commentary using the "Unwanted, taunting, off-topic" category as justification (but then again, I just put it right back in -- love ironies).

So what advice can I give those who don't know what the profanity I am talking about?

[1] Log off without writing a comment.

[2] Ask specific questions.

[3] Do your homework (read materials linked in the post; here the article in the Stanford Encyclopedia; find past posts on the same topic). Oh there are so many ways.

Taking my own counsel regarding moralism, I will not blame Anonymous or even be mad. He (she?) could not help it, right? Let's hope that nature will deliver more pleasant news tomorrow.