Psychopaths and Moral Responsibility

Can psychopaths be morally responsible for their behaviors?  The question is a difficult one -- with different theories of moral responsibility yielding different answers -- and I'm sure there will be some interesting discussion of it at the exciting conference this weekend organized by Manuel Vargas.

Shaun Nichols, Daniel Batson and I were curious about how ordinary people would approach this issue, and we recently ran an experiment to see if we could figure it out.  All subjects were given a description of a disorder loosely based on psychopathy.  Subjects were then asked whether people who had such a disorder could be morally responsible for the immoral actions they performed.

But different subjects were asked this question in different ways.  Half of the subjects were simply asked in the abstract whether anyone who had this disorder could ever be morally responsible for the immoral actions they performed.  The other half were given a concrete story about a person who had the disorder and who therefore decided to kill his own wife and children.  They were then asked whether this particular person was morally responsible for what he had done.

By now, you have probably guessed the results.  Subjects who were asked the abstract question tended to say that people with the disorder could not be morally responsible for their immoral actions, but subjects who were given the concrete story tended to say that this particular person actually was morally responsible for killing his wife and children.

These results certainly don't leave us with a clear picture of people's ordinary understanding of moral responsibility and psychopathy, but they do provide some puzzling data that cry out for explanation...

Comments

[Chris]

Joshua, that is very interesting data. I wonder if you could account for it with something like Haidt's social intuitionist model. I suspect that people's emotional reaction is much stronger to the concrete situation, and you get more intuitive processing.

[Mark Smeltzer]

We have conflicting intuitions because the person's actions provide prima facia evidence that they are evil and deserving of blame, but our knowledge of the their inhibited rational faculties lead us question the warrant of that evidence.

Hence, we don't really know whether they are evil, and often won't have enough evidence to compell us to believe that they are evil. In those situations, we are inclined to punish those inviduals in a consequential sense (sending them to therapy, putting them in isolation, etc.) until such time that we are compelled to believe they are innocent or blameworthy.

I don't see any great mystery here...

[Heath White]

The results leave me with "a clear picture of people's ordinary understanding of moral responsibility and psychopathy." They're confused.

[T. Scrivener]

I don't think philosophers have to place too much weight on pre-reflective folk intuitions, it's folk intuitions once they have been brought to reflective equilibrium that matter and I think this is a perfect demonstration of why.

[Paul Torek]

The term "psychopath" is not a recognized diagnostic category in the fields of psychology and psychiatry. Here's a snippet from http://www.mental-health-matters.com/articles/article.php?artID=292:

"n response to our recent rash of sniper attacks in the Washington D.C area, and the term 'Psychopath' being used indiscriminately, we feel that it is appropriate to truly define this term for our valued members.

Psychopathy is not a clinical term in either the DSM-IV or the ICD-10. The nearest equivalent to it is, in the DSM-IV is Antisocial Personality Disorder,while the ICD-10 uses the term 'sociopathy' or 'Dissocial Personality Disorder'."

This leads me to wonder: how was a "psychopath" described in these questionnaires?

My wife is a psychologist, and from what I gather second-hand, and combining that with my understanding of our legal system, sociopaths generally would not even begin to qualify for the insanity defense. Given that our legal system reflects our moral norms, why would sociopaths be any more exempt from responsibility than anyone else?

[Joe Campbell]

One important issue is whether 'psychopaths' have a certain kind of control over their actions, the kind of control that is necessary for moral responsibility. (I put the term in quotes since you seem to doubt that this is a reputable psychological category. I don't know if it is or if it isn't.)

I don't think that the law reflects our social norms. According to our law, 14-year-olds are more culpable for their actions if the consequences of those actions are severe than if they are not. Thus, if a 14-year-old shoplifts he is unlikely to be tried as an adult yet if he commits murder he is likely to be tried as an adult. Clearly certain cognative capacities -- the ability to partake in means-to-end reasoning -- are relevant to attributions of moral responsibility but this is not reflected in our current law, as far as I can see.

I think that psychopaths -- and all functioning human beings -- have a basic set of relevant cogantive capacities and thus are morally responsible for their actions. However, there are mitigating circumstances that make 'psychopaths' less culpable for their actions than 'normal' agents.

[Neil]

Joe,

The empirical evidence is overwhelming that psychopaths are unable to distinguish moral from conventional transgressions: they know that certain actions are prohibited, but they are unable to sort out those that are wrong only because they are against the rules from those that are wrong regardless of the rules. This seems to indicate that wrongdoing is not perceived by them in the same way as by us: they know they are breaking the rules, but are unable to see that there is any further wrong involved (btw, this empirical evidence is part of the reasons I am unimpressed by the failure of clinical psychology to recognize psychopaths). So their cognitive capacities differ from ours in a responsibility relevant way. Of course, the precise implications of the differences are far from obvious.

[Joe Campbell]

Thanks, Neil!

[Paul Torek]

Neil,
Your description fits sociopaths, whose lack of empathy would logically seem to lead to a failure to "get" morality. However, I don't think we should make getting morality a precondition of responsibility; as long as the defendant is capable of "getting" the illegality of his action - along with other normal broad cognitive capacities - that's enough. The law in my state implicitly agrees:

In response to the criticisms of the various tests for the insanity defense, the American Law Institute (ALI) designed a new test for its Model Penal Code in 1962. Under this test, "a person is not responsible for criminal conduct if at the time of such conduct as a result of mental disease or defect he lacks substantial capacity either to appreciate the criminality of his conduct or to conform his conduct to the requirements of the law."

Of course, maybe the law is a ass. Sometimes it is, as with Joe's 14-year-old criminal examples. But for the most part, not.

[Neil]

Paul,

I am not making an a priori argument based on how I think "sociopaths" would see morality. My claim is that the empirical evidence overwhelmingly demonstrates that psychopaths don't get the moral/conventional distinction. Unlike even very young children, their responses to the question "Xing is forbidden, but would X-ing still be wrong if it weren't forbidden" don't track the moral/conventioanl distinction (See, especially, the work of James Blair). Now I already said that the precise implications of this fact are not obvious. But as a matter of fact I think that this is a powerful mitigating factor. A psychopath doesn't see that there is any difference between double parking and assault (though he - overwhelmingly he - knows that both are against the rules); surely given the fact that *we* think merely infringing the rules - by, say, double parking - isn't sufficient for great moral wrongness shows that the psychopaths inability to distinguish the cases doesn't deserve the same response as a normal's assault? Yes, he may deserve punishment , but does he deserve the kind of punishment we reserve for specifically moral transgressions?

Moreover, I suspect that there will be some set of moral transgressions that the inability to get the moral/conventional distinciton prevents one from realizing are against the rules at all. If we typically generalize our moral knowledge to new cases by reference (inter alia) to the kinds of responses that are intact in us and missing in psychopaths, then there will be some new cases of this type.

[Paul Torek]

Neil,

You don't say much about the general approach to punishment that militates in favor of a different approach to these pathological individuals, or how it would be different. So let me go first.

I see punishment as importantly analogous to self-defense. In both cases we make exceptions to a general rule against force and violence. In both cases we reduce the harms that force and violence produce, even further than we would by attempting to promote an absolutely exceptionless rule against force and violence. In both cases we do so in a person-respecting way. (Notice, therefore, that we can regard deterrence as absolutely central to punishment, without being utilitarians.) In both cases we impose conditions of proportionality and (I would hope) necessity: for example, if a road-raging driver tries to shoot me, I should not shoot him and claim self-defense if I can simply drive away instead.

None of these considerations make much difference of the fact that the perpetrator does or doesn't "get" morality. The biggest exception is the point raised in your last paragraph: new and unlegislated cases of wrongdoing, which these pathological individuals cannot be expected to avoid.

[Neil]

Paul,

I don't have or want a theory of punishment. Responsibility is a question of the relationship beween an agent and her acts; what we should do to her afterwards is, quite simply, an entirely distinct question. It doesn't follow from "all things considered, we should reward (punish) agents who perform acts of type A" that "agents who perform acts of type A are responible for their A-ing". Psychopaths are not (fully) responsible for their actions. But incarcerating them may nevertheless be a good idea (indeed, b/c they don't get the moral/conventional distinction it might be an excellent idea).

[Kip Werking]

Neil and Paul:

You raise a fascinating question. Let me say that I tend to agree with Paul, and I tend to disagree with Neil, that responsibility and punishment are entirely distinct questions. "To punish", on at least one meaning of the term, and perhaps the most prominent one, implies that the punisher considers the punishee responsible for some act, and hence deserving of punishment. There is another sense of punishment, according to which a maniac might "punish" an innocent person for hir own amusement, but I think it is important to keep these notions of punishment separate. In personal correspondence, Saul Smilansky has noted that, on his view, punishment entails desert (and, presumably, responsibility).

I think it is very important to note that punishment and responsibility are not entirely distinct questions, but rather that the former implies the latter, because the following reasoning about the behavior of free will deniers can confuse compatibilists and others (Dennett, amongst others, has criticized deniers for the following reason):

1. deniers deny that anyone is mr
2. punishment entails that people are mr
3. deniers sometimes punish people
4. THEREFORE deniers are inconsistent

One rhetorical tactic is to deny 2, as Neil does. But I don't think this tactic will have many converts. People instinctively associate punishment with desert and responsibility. A better tactic, I think, is to deny 3. Consider the following initial experience:

A. You see a police officer lashing another man.

Your first gut reaction is to think "surely, this man must have done something horribly wrong."

But then consider the follow-up experience:

B. The police officer tells you that a terrorist has bombs ready to destroy major cities all over the world unless the police publicly execute a random, innocent person.

Suddenly, the police officer is no longer "punishing" the innocent person. The notion that the subject is being punished is just part of a cognitive illusion. Free will deniers only "punish" people for the same consequentialist-type reasons that the police officer "punishes" the innocent person: creating a better future. There is a moral equivalence between the most atrocious serial killer and the innocent bystander. Both, ultimately, are victims of fate.

There is at least one important difference, of course, between the police officer analogy and "punishment" on the denier's view. "Punishment", unlike the public lashing, is intended to alter the subject: to render hir more or less likely to commit wrongdoings in the future. At this point, the denier can switch from talk-of-punishment and talk-of-wrongdoing to talk-of-illness and talk-of-medicine. "Wrongdoing", on this view, is like a subtle illness and "punishment" is just like its medicine. Sure, "punishment" is painful, but so is chemotherapy. One day there may be a medicine for wrongdoing that is not painful, just as one day there may be a cure for cancer that is not painful (one potential difference here: the cure for cancer, unlike the cure for crime, will not uncontroversially respect persons). This rhetorical shift helps observers to see wrongdoers for what, so says the denier, they are: victims, ultimately. Neil's suggestion, that we might punish those who are not responsible for their acts, does not have this advantage. Chemotherapy is not punishment.

So, the denier has a response to the allegation of inconsistency. Neil suggests one way: "Sure, I punish, but that has nothing do with responsibility." That, to me, does not seem quite right. Better, I think, is to say: "I do not even punish at all."

[Neil]

Kip, you're quite right that that's a natural way to use language. Rawls had the name "tellishment" for behavior that looked like punishment but did not reflect desert. My point was only that the question of whether an agent is responsible for an action is separable from how we ought to treat them. Suppose that I have stolen money from a poor old lady, but it so happens that punishing me will cause the end of the world. Clearly, we oughtn't to punish (equally, there are some circumstances in which tellishment is appropriate: eg, preventive detention of the very very bad). I just want to insist that we can't run the argument from how we ought to treat people to whether they are responsible, though we can, with more success, run it in the opposite direction.

[Tamler Sommers]

Hold on, as everyone knows, there's an entire tradition devoted to the justification of punishment (without scare quotes) that has nothing to do with moral desert: consequentialism Bentham never once mentions desert or moral responsibility in his treatment of punishment. All of his excusing conditions are based on forward-looking considerations--and not just deterrence. You're probably right that people now--since the retributivist revival--have desert-entailing MR in mind when they talk about punishment. But apparently this wasn't true, or at least was much less true, in the middle of the past centure in America--when rehabilitation and prevention were the primary goals of punishment.. So I think the two can be seperated and Kip's premise 2 can be denied.

Also, there is a difference between punishing a guilty (but not morally responsible) person for his crime and just lashing an innocent person for utilitarian reasons. The guilty person committed a crime and the punishment can help to reshape his character so that he doesn't do it again. The innocent person is just taking one for the team. The only time the two are comparable is when the punishment is the death penalty. Then the following quote from Oliver Wendell Holmes rings true:

"If I were having a philosophical talk with a man I was going to have hanged (or electrocuted) I should say, I don't doubt that your act was inevitable for you but to make it more avoidable by others we propose to sacrifice you to the common good. You may regard yourself as a soldier dying for your country if you like. But the law must keep its promises."

[Neil]

Well, it might be true that punishing the "guilty" but not responsible person reforms his character. Then again, it might not: depends on why he's not responsible. On the other hand, it might be true that punishing someone who's just taking one for the team helps reform his character too...

[Paul Torek]

Excellent discussion, Kip. Your comments give me a new way to express my conviction. It goes something like this: moral responsibility just is whatever (if anything) explains why our broad practices of praising and blaming, rewarding and punishing, are right, to the extent that they are. Or at very least, those are the absolutely central family members in the family resemblance that gives MR meaning. To make a long story short, I think Dennett is on to something.

I think that your separation between punishment and "punishment" only works as long as either you are discussing unusual cases, or you are proposing a radical change to a vast swath of morality. (Bentham wouldn't be radical enough. The views his harshest critics would impute to Bentham, on the other hand, would be. The key is to make the new practices clearly and abysmally fail to be "our broad practices" of punishment.)

If you take a paradigmatic, ordinary case of punishment, which is justified in roughly the way common sense says it should be, and say that's not punishment but only "punishment", you are in danger of creating the inverse of a weasel word. The opposite of sucking the meaning out of a word and leaving only the shell, is providing the white and the yolk but refusing to wrap it in the customary verbal exterior.

[Mark Smeltzer]

Paul,

While I do agree that the question of starting points is incredibly important to settling the free will debate, starting from punishment practices seems inappropriate for a number of reasons.

In creating my own account, I have chosen to start from reactive attitudes like loving and hating, and have drawn inferences that would support the application of those attitudes in response to questions like, "What would warrant loving/hating someone else?"

Afterward, I could not find a way to get from reactive attitudes to a theory of punishment. As a result, I recognize a deep running distinction between moral desert and legal desert. Legal desert takes into account many practical (consequential) considerations that moral desert does not.

As a result, there is a sense in which one can be a realist about legal desert without being a realist about moral desert. It seems Tamler and perhaps Kip would fall into that category. But from your comments here, it is unclear which category you fall into.

If one starts from purely legal considerations, one could end up with a robust theory of legal desert without having ever touched the question of moral desert (as Rawls does). Since the free will debate is ultimately about moral desert (a bold assertion I realize), to the extent that someone doesn't care about addressing the question of moral desert is the extent to which they do not care about the free will debate.

[Kip Werking]

Paul writes:

"If you take a paradigmatic, ordinary case of punishment, which is justified in roughly the way common sense says it should be, and say that's not punishment but only "punishment", you are in danger of creating the inverse of a weasel word. The opposite of sucking the meaning out of a word and leaving only the shell, is providing the white and the yolk but refusing to wrap it in the customary verbal exterior."

That is one alleged danger. But it is important to remember that, on the denier's view, there is a fundamental widespread illusion about *something*. As Pereboom says in LWFW, there is a limit to how much the compatibilist can appeal to certain widespread beliefs about moral responsibility, because on the denier's view, such beliefs are based upon a mistake. So the denier must reach *some* counter-intuitive conclusions, or else s/he just reinforces the folk and orthodox view of freedom and responsibility.

Drawing the precise line is a delicate matter. For example, in at least one version of the Consequence Argument, Van Inwagen suggests that nobody ever has any choice over anything. This seems to too strong. I make choices all of the time, but they are not free in the sense that entails robust moral responsibility. Similarly, in the new thread "A Question About Ability", Neal suggests a worry about whether we even have the ability to commit actions. Although I understand how the manipulation or design argument can motivate this view, I think such arguments show that the subject is not morally responsible for hir acts, but not necessarily that they are not able to perform the acts (indeed, they seem to be clearly able to perform them). So I draw the circle around counter-intuitive claims a little smaller than Van InWagen or Neal may.

Others may draw the circle more narrowly than me. For example, Tamler is quick to emphasize that punishment might exist just as a deterrent. Consider the following claims:

RetConP: The concept of punishment entails both notions of retributivism and consequentialism.

OrdPun1: Ordinarily, it is right to punish someone for their wrongdoing.

OrdPun2: Ordinarily, it is right to punish someone just because doing so will deter further wrongdoing, and without consideration of (robust) moral responsibility.

DenPun: Because of the retributivist element of RetConP, it is never appropriate to punish someone, although it may be appropriate to do things that have a superficial similarity to punishment (just as chemotherapy has a superficial similarity to punishment).

Tamler is willing to preserve at least the appearance of OrdPun1, but in doing so he must deny RetConP, something that Honderich explicitly endorses in "How Free Are You?" and for which, I think, the folk would give robust support. However, upon inspection, Tamler is really denying RetConP, not for OrdPun1 but for OrdPun2, and OrdPun2 is a much stranger claim to make. So Tamler and I and others are torn between two strange and counter-intuitive positions: OrdPun2 and DenPun. Both positions are somewhat counter-intuitive but the line must be drawn somewhere.

It is interesting to see how Tamler draws the line around other things. On his view, resentment is incompatible with free will denial while the following are not: love, gratitude, forgiveness. Furthermore, certain aspects of guilt and pride survive free will denial but not other aspects. But surely, love involves a tremendous amount self-deception and illusion about the love-object's merits. It is not clear to me that we do not make the same sort of overreaching demands of love that we also make of moral responsibility (Fischer calls the latter a sort of "metaphysical megalomania).

To answer your original challenge, although I might not share Van InWagen doubts about "ever being able to choose anything" or Neal's worry about ability, I tend to endorse DenPun and therefore say that we should never punish. Consider the following claims:

MedWrong: Wrongdoing entails all of the essential attributes of being a symptom of a disease; from this one see how punishment is like a primitive painful medicine and prisons are like primitive hospitals.

NoFaultMed: Ordinarily, it is not a diseased person's fault that they have a non-self-induced disease.

FaultWrong: Ordinarily, we regard a normal person's wrongdoing as their fault.

I endorse DenPun because I tend to agree with Honderich that RetConP is right. But DenPun may gain additional support to the extent that MedWrong and NoFaultMed are true. Although NoFaultMed seem uncontroversially true, MedWrong is much more controversial. When considering MedWrong, I note that it seems controversial in the same way that "aging is a disease" is controversial. But the differences between aging and other uncontroversial diseases do not seem relevant. For example, the fact that everyone naturally ages does not seem to prevent aging from being a disease. So I suspect that our refusal to call aging a disease just represents a widespread performance error.

However, I can see even a further distinction between "aging is a disease" and MedWrong: aging represents the loss of faculties because of a loss of selection pressure, whereas much or most wrongdoing is a robust expression of our faculties and may even be (according to evolutionary psychologists) the result of strong selection pressures in favor of those behaviors (murder of rivals by jealous husbands, theft and deception for one's own benefit, some suggest even rape is an adaptation). What is motivating my conclusions here about MedWrong and "aging is a disease" is, in part, the ethical dimension: it seems so ethically right to regard aging and wrongdoing as diseases because they should be abolished.

However, I recognize that MedWrong raises the danger of making the sort of semantic mistake you describe. The question is whether our complacency during the suffering of the aging, and the suffering of wrongdoers (in, for example, prisons), represents a widespread but convenient performance error. Or alternatively, whether I am stretching the meaning of the word "disease" beyong its natural limits. Furthermore, I suspect that my ethical concern about whether it would be beneficial to treat X as an example of concept Y (as, for example, treating aging as a disease might be beneficial) is not a relevant factor when doing this sort of linguistic analysis.

So, recognizing that danger, but still motivated by that ethical concern, I'm willing to make a revisionist move. I'm willing to admit that I might be revising, slightly, the concept of a disease, but for perhaps good reasons. In this respect, my view is like that of Manuel Vargas. The difference, however, is that if aging or wrongdoing really doesn't represent a disease, then I won't insist on calling it a disease. I'll call it disease*.