Monday, January 14, 2008

Moral Instinct: A Case Study

Yesterday's posting concerned a New York Times article called, "The Moral Instinct". In it, I argued that an inference assumed in the article - that a sense that an instinct to make the judgment P is wrong implies P is wrong - is an invalid inference.

Today, I would like to look at one of the specific examples used in the article and demonstrate this fallacy.

The article reports on some experiments that were performed.

Julie is traveling in France on summer vacation from college with her brother Mark. One night they decide that it would be interesting and fun if they tried making love. Julie was already taking birth-control pills, but Mark uses a condom, too, just to be safe. They both enjoy the sex but decide not to do it again. They keep the night as a special secret, which makes them feel closer to each other. What do you think about that — was it O.K. for them to make love?

When presenting this case to people, the researchers point out the fact that they can block any argument that the test subjects give for saying that it is wrong. They can drive their subjects into a situation where the subject says, "I can't argue against anythng that you say. But, I know it's wrong, and nothing you can say will convince me otherwise."

The fact is, I have met people who would say exactly the same thing about interracial relationships - or homosexual relationships for that matter. This is a descriptive fact about how a lot of people think - about how they reach moral conclusions.

However, it leaves an important set of questions unanswered. Is this form of sibling incest, or homosexuality, or interracial relationship, wrong in fact? What does it take for homosexuality (for example) to be wrong in fact?

Is it the case that - if there are people who can be driven to the conclusion, "I cannot argue against anything you say, but I know that homosexuality is wrong and that it is permissible to take actions that harm the interests of homosexuals, and there is nothing you can say to change my mind," then in fact homosexuality is wrong and these people really are justified in harming the interests of homosexuals?

Just because you can find a lot of examples of people doing something, this does not mean that they are justified in doing it. No matter how many instances we find of people committing the fallacy of 'ad hominem' - this does not make 'ad hominem' a valid form of reasoning. No matter how many instances we find of people firmly anchoring their claims that they are justified in doing harm to others on a particular set of personal likes and dislikes, this does not mean that the particular set of likes and dislikes actually justifies those harms.

I think that I can offer an account of what is going on in the case of incest. This research starts off entirely on the wrong foot. It is making the false assumption that morality is primarily concerned with evaluating actions. So, it is trying to get people to evaluate an act of incest. When, in fact, morality is primarily concerned with evaluating desires, and what agents are evaluating in fact is a universal aversion to incest.

They then force their perspective on the subjects - forcing the subject to come up with a justification in terms of acts being the primary object of evaluation. Which the subjects cannot do - because the assumption is false.

Here is how the argument would go if we take desires to be the primary object of moral behavior.

What I am looking at in evaluating this act is whether there are desires or aversions inherent in the situation that people generally have reason to promote throughout a population.

People generally have reason to promote an aversion to incest. Even though there may be a few cases in which incest does not thwart desires, by and large incestuous relationships are ultimately desire-thwarting; they cause a great deal of harm. If everybody has an aversion to incest, then for the most part we would be preventing a great deal of harm, and if everybody has an aversion to incest we would not be thwarting any desires (there would not be a desire to engage in incest for us to thwart).

A universal aversion to incest implies a universal aversion to the type of behavior described in this case.

If I were to say that their behavior were justified, then I would be saying that nobody should have this aversion to incest. However, a culture in which nobody has this aversion to incest would be a culture in which abusive incest would be far more common. We certainly have no reason to create a culture in which abusive incest is far more common. Therefore, we certainly have no reason to say that nobody should have an aversion to this type of situation.

The case clearly states that the couple keep their act a secret, so that it has no affect on society.

However, I am not talking about the act itself. I am talking about the judgment of the act. When you ask me to judge the act as morally permissible or impermissible you are asking me to judge whether everybody should or should not have an aversion to that type of situation. Even if, in this rare case, no harm was done, people generally still have many good reasons to promote an a culture in which people are averse to these types of relationships. People generally still have reason to judge the absence of an aversion to incest negatively.

The researchers who write about these types of cases point out that their subjects generally cannot articulate a defense that the researchers cannot counter. Therefore, they say that all moral justifications are 'rationalizations'. However, that conclusion requires something stronger than saying, "We have been able to counter every response." It requires saying, "There can be no response that we cannot counter."

That, I argue, simply is not true. Try the above response on for size.

If this response fails, then there is still the possibility that there is another response not yet thought of.

If it is the case that all responses fail, then the conclusion is not to say that the prohibition on incest is justified in virtue of some 'moral instinct'. The correct conclusion to draw is that humans have an inherent impulse to do unjustified harm to others in these types of cases. If incest is not wrong in fact, then a moral instinct to treat incest as wrong is, itself, wrong - a sickness driving people to do harm to those that it has no real reason to harm, and to merely imagine that the harms they inflict are justified.

One of the things that I must note is that subjects often cannot give a conscious defense of their moral conclusions. However, this does not require any sort of ‘moral instinct’. A lot of our reasoning is subconscious. Nature does not have a particularly compelling need to make it the case that all reasoning is conscious. Nature is well served by coming to the correct conclusion. Making the method of reasoning conscious is, sometimes, a waste of energy and a waste of time. One of the things about morality is that it must often weigh in on our actions the instant we decide to act. We do not have time to consciously deliberate the morality of every action.

Morality, on this account, is still an object of reason, though some of it is subconscious. When moral reason is done poorly, it yields moral conclusions that are wrong. There is no ‘moral instinct’ or ‘innate morality’ – it is simply a fallacy to argue from the fact that we have certain innate desires to do harm and feel justified in doing so that we are, as a matter of fact, justified in doing so. In order to be justified in doing harm there has to be a fact of the matter – a way in which ‘we are morally justified in doing harm’ can be true as a matter of fact.

Either there is a justification for the harms that we inflict on others under the heading ‘morality’, or the harms that we inflict on others under the heading ‘morality’ are unjustified.


Anonymous said...

Is it possible for someone to have both an aversion to incest and a desire to engage in a specific act of incest?

Uber Miguel said...

I might butcher this, but let me take a stab at it.

There are a couple types or levels of desires, and some of them are compatible with aversions while others are not. I have trouble thinking of a scenario where someone can be simultaneously averse to incest yet have a desire-as-ends to engage in incest.

Alternatively, I can very easily think of a scenario, say: someone puts a gun in your mouth and tells you to engage in an incestuous act, where one may be averse to incest yet may engage in incest as a desire-as-means to satisfy other desires, such as the desire-as-ends to survive. Also, false beliefs can play a heavy role - perhaps you are not aware that the gun is a water pistol or that the other person is a relative.

Of course, other less dramatic desires-as-means would also allow such a seeming violation of one's aversion, such as desires to satisfy a curiosity, to be a rebel and break rules, to win a bet, etc - provided that these other desires-as-ends are strong enough to engage in incest as a desire-as-means to reach a different, stronger desire-as-ends.

However, if one has a simple desire-as-ends to engage in an act of incest, then it doesn't make much sense to suggest that they are also averse to incest. In such a situation, it would make more sense that they are not averse to incest but are actually averse to others finding out about their taboo desire for incest.

Matt M said...

I read the incest example as challenging such beliefs rather than validating them: If you can't provide a coherent argument for why X is wrong then how do you really know that it is?

But aside from that, I'm in more or less complete agreement with what you've written.

Uber Miguel said...

"If you can't provide a coherent argument for why X is wrong then how do you really know that it is?"

Good thing this is horrible logic - as Fyfe quite throughly points out. A sort of combination God of the Gaps, reverse Argument from Ignorance (you can't explain to my satisfaction), and with the occasional bit of Moving the Goalpost tap dancing as needed.

Does anyone else feel that the framework is strikingly similar to that of anti-evolution trolls? It's almost as if they're trying to artificially create a gap in our concept of morality and magnify it to the point of a permanent foothold on ambiguity.

Any explanations of morality, no matter how sound and thorough, are dismissed almost without rebuttal if your "intuition" isn't fully convinced or they can conjure a particular moral perspective that makes the situation even more complex.

Then the gapists simply proceed to declare victory for themselves and insist that we should just act as we've evolved to act - or to embrace religion depending on whether the person feels that religion is generally a good thing for humanity.

Anyone else get that vibe - or is just me?

Alonzo Fyfe said...

Doug S.

The short answer to your question is yes.

A person can have an aversion to having sex with her brother, and a desire to have sex with a male having a particular set of physical characteristics. It may be the case that her brother has that specific case of characteristics.

This is 'ambivalence'. She would both, at the same time, desire and be adverse to having sex with her brother.

Martin Freedman said...

You have in the past made a critique of these type of experiments as being artificial and not really related to day to day issues that morality deals with. Hauser answers that challenge in the first chapter of this book, adequately I might say.

Now here this is a peculiar experiment and I do agree with your analysis. Some key points

1) Hauser and by implication Pinker would be arguing for a Rawlsian intuitionist where the inability to verbalize one's views is evidence for this. I agree with you this is insufficient, much of our reasoning is not conscious period. Still their goal is not to derive prescriptions but descriptively understand how moral reasoning works. We might agree that they are mistaken here but I would say that this does not alter the fact that this is a useful empirical approach to take. New and better experiments need to be designed and maybe you could propose some?

2) You often use Nozick's Experience Machine in your arguments against Happiness based morality. What Hauser et al have done is taken these type of philosophical mind experiments and turned it into psychology experiments. Given your use of the experience machine, you surely cannot be against doing that in principle?

3) This particular experiment - harmless incestuous sex study - I agree is problematical and I agree with your analysis. Nonetheless we would hold that a universal aversion to incest is good and one to homosexuality or other races is bad. Granted you do not claim to be a psychologist, still how can one design an experiment to make this distinction or should or can that be done and if not, why not?

Anonymous said...

After some consideration I would like to add my voice to the request to drastically reduce the Experience Machine metaphores. I think it is a great case example that demonstrates that most people value truth over happiness (ie: almost everyone would take the Red Pill if they were Neo, even if they knew what life is like outside the Matrix), but it has too many downsides as an argument to be over relied on.

The primary downside being that it is again one of those outlandish hypotheticals with no bearing on the real world. I occasionally field questions like "If you could end all human suffering forever by torturing one child do death, would you do it?" or "If a race of beings had a strong desire to be enslaved, and a different race had a strong desire to enslave others, wouldn't this make slavery moral?" The fact that they are outlandish hypotheticals with no bearing on reality often takes quite a bit of explaining. The Experience Machine is too close to these.

Also, with certain caveats (such as: no one else would be hurt, and I can bring the people I care about with me into it) a lot of people would seriously consider entering the experience machine (I know I might).

Finally, it's not the best explanation available for countering (for example) happiness theory, or demonstrating desire for truth over desire for happiness. When I am confronted with happiness theories, I simply point out that it does a poor job of explaining why I smoke, whereas DU can explain that quite easily. The "not knowing your child has been murdered keeps you happier than finding out s/he has been" is an ample and strong demonstration of how people value truth over happiness. Etc.

So in summary, I think the Experience Machine is unnecessary, and introduces more problems than it solves.