Emil Kirkegaard proposes a semi-objective definition of “mental illness”.

He’s partly responding to me, but I think he mangles my position; he seems to think I admit mental illnesses are “just preferences” but that which preferences are valid vs. diseased can be decided by “what benefits my friends”.

I mostly don’t think mental illnesses are just preferences! I’ve been really clear on this! But Emil is right that I don’t deny that there can be a few cases where it’s hard to distinguish a mental illness from a preference - the clearest example is pedophilia vs. homosexuality. Both are “preferences” for sex with unusual categories of people. But I would - making a value judgment - call pedophilia a mental illness: it’s bad for patients, bad for their potential victims, and bad for society. Also making a value judgment, I would call homosexuality an unusual but valid preference: it’s not my thing, but seems basically okay for everyone involved.

(I wouldn’t describe this as “benefiting my friends” - I’m against children getting raped whether they’re my friends or not. I think this dig was unworthy of Emil, and ask that he correct it.)

Emil proposes an alternate definition: a mental disorder is a mental trait which lowers reproductive fitness. This makes some sense: the brain, like other body parts, was optimized by evolution with reproductive fitness as the goal, so any behavior or preference which interferes with that goal is in some sense a “malfunction”. This definition would cover depression, where people might be too depressed to hunt or gather or woo mates. It would cover pedophilia, where people have sex with children (and not adults) and so can’t reproduce. But it would also cover homosexuality, which also lowers people’s chances of having children.

Emil admits that this definition isn’t very useful for making social decisions:

Just because we recognize that such behaviors are disordered, doesn’t mean they should be criminal or legal. That would depend on the usual harm considerations. Likewise, treatment options should be allowed and probably subsidized by the state (if it’s in the business of doing healthcare that is). I see no reason to force treatment in the same way we don’t force someone to get their broken bones fixed, or cancer cured.

…but thinks it has other advantages:

There is no need to inject politics into the scientific study of mental illnesses, in the same way we don’t inject politics into the study of many other natural phenomena.

Before we evaluate this argument: what exactly are we trying to do here?

We’re arguing about the definition of “mental disorder”, but in some sense this is a pointless fight. I’ve pointed out a useful category (mental conditions which are bad for people and society). Emil has pointed out a useful category (mental conditions which are bad for reproductive fitness). Why not just admit that these are two useful categories? Why not come up with separate names for them - disorder-(Scott) vs. disorder-(Emil), or maybe social-disorder vs. evofitness-disorder - and use the separate names, so everyone’s on the same page? If someone asked whether homosexuality was a mental disorder, we could confidently answer “It’s not a disorder-(Scott), but it is a disorder-(Emil)”.

Mental-disorder-(Emil) would be a more useful term in discussing evolutionary psychology. If you’re studying how mental disorders evolved, it’s useful to group homosexuality together with depression, anxiety, and the like, as examples of how evolution failed to optimize the reproductive fitness of certain individuals. Sometimes I use this definition colloquially when I say things like “Many different mental disorders are correlated with left-handedness”, including homosexuality in the list. Left-handedness is probably related to decreased canalization in brain development and so we should expect it to correlate with various deviations from the evolutionary norm. This is a useful thing to talk about, and I’m happy to keep mental-disorder-(Emil) around as a concept for these kinds of situations.

But mental-disorder-(Scott) would be more useful for a variety of ethical, medical, and philosophical problems. Statements like “if you have a mental disorder, you might want to consider seeing a psychiatrist” or “should companies give employees time off to recover from mental illness?“or “we should have a policy for replacing the President if he becomes severely mentally ill” rely on a definition where we’re talking about something socially dysfunctional, not something that affected reproductive fitness 100,000 years ago.

In case it’s not sufficiently clear that mental-disorder-(Scott) and mental-disorder-(Emil) are different, let me give seven examples:


People with ADHD have more children than people without ADHD, probably because the people with ADHD forget to use condoms (here is a source for teenagers, can’t find source for adults but I’ve definitely heard this). That means that by Emil’s definition, not having ADHD is a mental disorder!

Emil didn’t mention this in his post, but a common response to this complaint is that the definition should actually rely on conditions in the environment of evolutionary adaptedness (EEA), ie the African savanna 100,000 years ago, when there were no condoms.

But I’m not sure what the right amount of attention was in the African savanna. Plausibly, amounts of ADHD that would totally ruin your life today were just fine when the most demanding cognitive task around was hunting giraffes.

(I’ve heard claims that hunter-gatherers have more ADHD than farmers, for approximately this reason; evolution got 10,000 years to improve focus in peoples with a long history of sedentary living)

Either way, I think it’s pretty hard to call ADHD a mental disorder in Emil’s system, unless you come up with some jury-rigged comparison group for “normal functioning” (modern humans in a world without condoms? hunter-gatherers in a world where they all had to go to school until age 18?)

Again, Emil can bite this bullet, but it means his definition doesn’t match intuitive conceptions of who’s got mental problems and wants help with them.

2: Alcoholism

Who has more kids: alcoholics, moderate drinkers, or teetotalers?

I weakly predict that alcoholics have the most (they have lots of chances for drunken flings without contraception), moderate drinkers are in the middle, and teetotalers have the fewest (alcohol is a powerful social lubricant). So not only should alcoholism not be a mental disorder, but maybe non-alcoholism should be a mental disorder, and in any case teetotaling should definitely be a mental disorder!

This sounds like another one where we have to ignore the modern day to focus on what worked on the savanna. But there was very little alcohol on the savanna, so a tendency to become addicted to alcohol wouldn’t have decreased reproductive fitness. Do we have to throw the whole question out?

Chinese people seem to have some anti-alcoholism genetic adaptations, because alcohol was invented first in China, and only the Chinese had enough time to evolve advanced genetic defenses against it. I think it’s possible that a few other ethnic groups have weaker anti-alcohol adaptations that we just haven’t noticed yet, but still others - like the Inuit - clearly have none.

I think Emil’s theory classifies alcoholism as a severe mental disorder in Chinese people, a mild mental disorder in white/black/etc people, and not a mental disorder at all in the Inuit. This would be hilarious to try to get in the DSM, but it still doesn’t match our practical need to determine who has mental problems and needs help with them.

3: Ephebophilia

Which of these people, if any, have mental disorders?

  1. A 65 year old man who’s attracted only to 14 year old girls.

  2. A 65 year old man who’s attracted to girls of any age from 5 to 30.

  3. A 65 year old man who’s only attracted to adult women 40+.

Most people in our society would classify 1 (an ephebophile) and 2 (a non-obligate pedophile) as mentally ill or at least worrying edge cases.

But I think Emil’s theory rules that only Person 3 (the man attracted to people close to his own age) is mentally ill, since he’s ruled out mating with the vast majority of fertile women.

4: Plato

…never had children. “Platonic relationship” jokes aside, I guess he was too busy philosophizing. Great men (and women) who can’t slow down to raise a family seem to be a type.

Is an interest in philosophy (or science, or art, or any other worthy endeavor) that reaches the point where it consumes your life a mental illness? Kierkegaard bites the bullet and admits that the priests and monks who took vows of celibacy were mentally ill by his definition. But I think he has many more bullets of this type to bite.

Even if we agree that we should classify Plato as mentally ill, this again seems very different from the practical concept of “this person has mental problems and needs help with them”.

5: Chronic Pain, Panic Attacks, Or, If You Insist, Nightmares

Is chronic pain a mental illness? It seems pretty bad. But as long as it doesn’t impede your ability to hunt, gather, or have sex, I think Emil would have to say no. Same with panic attacks, anxiety, etc.

If it’s hard to imagine a form of chronic pain that doesn’t impede those things, consider nightmares. These surely don’t impede any daytime activity, but chronic nightmare disorders seem very unpleasant!

I think Emil has to bite the bullet that conditions which make people miserable and ruin their lives aren’t mental disorders as long as they don’t affect functioning.

6: Severity

In his post, Emil includes a few turns of phrase indicating we can talk about severity - ie some mental illnesses are more severe than others.

But by his framing, “severe mental illness” would indicate not schizophrenia and bipolar disorder, but homosexuality and asexuality. After all, schizophrenics are more likely to have children than gays.

Again, this is pretty different from the way you want to use words when talking about real-world problems around how to help people with mental problems get better.

7: Is Emil’s Definition Of Mental Illness Itself A Mental Illness?

Emil’s crusade to reclassify homosexuality as a mental illness doesn’t sound like it would be very popular in his home country of Denmark. Maybe there are even some nice Danish women who would be willing to date Emil otherwise, but are turned off by his un-PC opinions.

Willingness to violate taboos couldn’t have been very helpful in the environment of evolutionary adaptedness. I imagine some distant ancestor of Emil’s standing up in front of the tribe and saying “Me think Bear God stupid and ugly! Me piss on Bear Idol!” Might mean fewer Kirkegaards around today.

So is contrarianism a mental illness? I would say no, because it’s a matter of personal choice and serves a valuable social function. I’m not sure what Emil’s answer would be.

I don’t want to assert any of these too strongly. Maybe Emil knows something I don’t about the EEA, and can prove that actually ADHD would be maladaptive there, or ephebophilia would get you in trouble. If so, I think that would restore some concordance between our intuitive notion of mental disorders and Emil’s version, but that concordance would be coincidental, not necessary. The next day we might learn some different fact about the EEA that would make the two notions discordant again.

So to repeat my claim: mental-disorder-(Emil) and mental-disorder-(Scott) both describe useful concepts, but they’re not the same concept. Mental-disorder-(Emil) is useful for talking about evolutionary genetics; mental-disorder-(Scott) is useful for talking about present day mental health problems and what to do about them.

We won’t convince people to literally use the terms “mental-disorder-(Emil)” and “mental-disorder-(Scott)”. So who should keep custody of the current term “mental disorder” and who should have to make up a new word for their thing?

I think Emil should have to make up the new word, because:

  • There are a few thousand evolutionary psychologists, and a few hundred million normal people who want to talk about mental disorders for normal reasons (like because they have them).

  • Evolutionary psychologists are smart people and can probably coordinate on new terminology and move on, whereas normal people have brought the US to the brink of civil war over pronouns.

  • The pressure of normal people using words to discuss their own mental health problems is so strong that whatever term they use will dominate the discourse no matter what. Even if we tried to let evolutionary psychologists keep the word “mental disorder”, and forced the rest of the world to use something like “social disorder”, the concept of social disorders would be so much more salient that we’d have to get into fights over whether homosexuality was a social disorder, and then re-litigate this whole process.

So I propose Emil coin a new term for the thing he very reasonably wants to talk about - maybe “genetic maladaptation”. I will happily use it when discussing evolutionary psychology, and otherwise we can get back to the important work of talking about just how wrong Bryan Caplan is on this topic.