More Recent Comments

Friday, February 16, 2007

Another Boring Just-so Story

 
From ScientificAmercian.com.

Child molestation and rape top the social taboo list, according to a survey of 186 people between the ages of 18 and 47, and smoking marijuana ranks lowest among the 19 choices of forbidden behavior. In the middle—worse than robbing a bank but better than spousal murder—lies incest between brothers and sisters. Given the deleterious genetic impacts of offspring from such mating, some researchers have suggested that there may be an evolved mechanism designed to prevent that from occurring. And now evolutionary psychologist Debra Lieberman of the University of Hawaii–Honolulu believes she may have elicited some of its functions from this simple questionnaire....

The evolutionary psychologists hypothesize that some form of mental mechanism assesses various cues to come up with an estimate of how related two people are. "The real question is: What are these cues?" Lieberman says. "A potent cue is seeing your mom caring for a newborn. That would have served as a great cue that the infant is a sibling, at least a half sibling." But for younger siblings, who would have no opportunity to make this observation, another cue might be the amount of time spent living with another child/potential sibling. Dubbed the "Westermarck hypothesis"—after the Finnish sociologist who first noted it in a book published in 1889—it posits that children reared together do not often end up being sexually attracted to each other.

11 comments :

Anonymous said...

Larry, I understand why you tend to be dismissive of explanations with a strong "adaptationist" angle, since you emphasize the importance of random genetic change in evolutionary history. But I've never heard you say that natural selection had *zero* impact on evolution. Do you believe that nothing in human behavior is adaptive, or just that it would be very difficult or impossible to prove it? Are there more rigorous procedures researchers should use to be able to include or exclude adaptation as a reasonable explanation for a behavior? What do you recommend?

Larry Moran said...

My position is that much of human behavior is an epiphenomeon. In other words it's a consequence of having a brain and living in a society.

The idea that there could be specific genes for each human behavior is contrary to everything I know about genes and how they work. The idea that specific alleles of these genes could become fixed in the human population because they confer selective advantage is also something that doesn't fit with my concept of how evolution works. I'm especially troubled by the lack of appreciation of how strong the selective advantage needs to be in order to achieve the required goal.

In this particular case, the selective advantage of the imaginary allele for avoiding mating with your sibling would have to be very powerful indeed in order for it to have become fixed in Homo sapiens. I see no evidence of this, do you?

We tend not to mate with children we've grown up with in the same household but mating with first cousins from an adjoining household has been very common throughout history. Do you really think that's because of your genes?

There's a very good reason why you've never heard me say that natural selection has zero impact on evolution. I'm a pluralist like Gould. The word "pluralist" means that we accept several different mechanisms of evolution. One of them is natural selection.

My beef is with those who: (a) don't even realize that there are other possibilities, and (b) can't tell the difference between memes and genes.

The "rigorous" procedure that researchers should use to exclude silly, adaptationist, just-so stories is called "common sense."

Tony Jackson said...

“In this particular case, the selective advantage of the imaginary allele for avoiding mating with your sibling would have to be very powerful indeed in order for it to have become fixed in Homo sapiens.”

Do you believe the same logic applies to other animal species as well? In which case, how do you explain this?

Frommen, J.G. and Bakker, T.C.M. Biol. Lett (2006) 2, 232-23: “inbreeding avoidance through non-random mating in sticklebacks”

“Here, we show that gravid female three-spined sticklebacks when given a choice between a courting familiar brother and a courting unfamiliar non-sib prefer to mate with the non-sib and thus avoid the disadvantages of incest. We controlled for differences in males’ body size and red intensity of nuptial coloration. Thus, females adjust their courting behaviour to the risk of inbreeding.”

Incest avoidance is a well documented instinct in animal behaviour. How it works of course is another question all together, but it seems unlikely to have no genetic basis.

Torbjörn Larsson said...

"The idea that there could be specific genes for each human behavior is contrary to everything I know about genes and how they work."

I had the impression that for example the evo-devo people thinks the situation can be more complex than one gene-one trait.

"In this particular case, the selective advantage of the imaginary allele for avoiding mating with your sibling would have to be very powerful indeed in order for it to have become fixed in Homo sapiens."

This layman don't see that this follows either, since people discuss selective sweeps, bottlenecks and introgressions, and in the last cases I think with low differences in advantage.

I guess your description so far seems like a "just-so" story to me. ;-)

Larry Moran said...

Tony Jackson says,

Incest avoidance is a well documented instinct in animal behaviour. How it works of course is another question all together, but it seems unlikely to have no genetic basis.

Actually, evidence for incest avoidance is not all that well documented. The sentence right before the ones you quoted above is ...

However, experimental evidence for inbreeding avoidance through non-random mating in vertebrates is scarce.

In this case we're dealing with humans. If you're going to speculate about a genetic component to incest avoidance then it's best to suggest a mechanism before making up just-so stories. Also, some evidence of a genetic component would be helpful

Remember that your special gene has to discriminate between brothers and sisters (not allowed) and cousins (allowed).

Larry Moran said...

Tony Jackson says,

This layman don't see that this follows either, since people discuss selective sweeps, bottlenecks and introgressions, and in the last cases I think with low differences in advantage.

The probability of fixation of an allele by natural selection is approximately 2s where "s" is the selective advantage. If the selective advantage is about 1% the the probability of fixation is 2%—in other words the allele will be lost 98% of the time.

Most alleles do not confer a large selective advantage so they are more likely to be lost than fixed. It's a mistake to assume that all you have to do is imagine some selective advantage and the allele will automatically take over the population.

That's one of the problems with just-so stories in general and evolutionary psychology in particular. They make no attempt to justify their stories by giving us evidence that the selective advantage is significant.

Imagine a primitive human population that does not carry the anti-incest allele. (Such a state is a requirement for just-so stories.) How many members of this population will mate with their brothers and sisters on a regular basis? What is the strength of the disadvantage if they do? (It's actually very small, most consanguineous marriages produce normal healthy children.)

Now imagine a mutation(s) that causes incest-avoidance. How strong is the selective advantage to each individual carrying the allele given that they probably wouldn't mate with a brother or a sister anyway? (It's related to the idea that familiarity breeds contempt.)

Isn't it easier to imagine a cultural taboo that inhibits incest?

Tony Jackson said...

Torbjorn mentioned selective sweeps, not me. I’m still fishing for sticklebacks :-)

But if you don’t like sticklebacks, try mice, mole-rats, woodpeckers, toads etc. Cumulatively, there’s enough in the literature to indicate that some interesting biology is going on here, whatever the mechanistic explanation.

So with respect to humans, why can’t such a mechanism have been inherited from a common vertebrate ancestor? Yes I know, humans have a rich culture and that will likely interact with, modify, perhaps even subsume the biology altogether. Still, even if this is the case, it’s worthy of study from the broader zoological perspective and if in fact it continues to be important for us too, then that’s all the more reason to pay attention.

In fact of course, the intertwining of innate biology and culture is hardly a controversial idea. Pat Bateson (one of the more sensible researchers in this contentious field) wrote a short but good book about it (with Paul Martin): “Design for life”. This book pulled off the impressive trick of getting a positive endorsement from both Richard Dawkins and Steven Rose - and is probably worth a read for that fact alone.

Anonymous said...

How strong is the selective advantage to each individual carrying the allele given that they probably wouldn't mate with a brother or a sister anyway? (It's related to the idea that familiarity breeds contempt.)

Now you are putting the cart before the horse. "Familiarity breeds contempt" is the whole point of the Westermarck hypothesis.

Here is a short description of the work of Arthur Wolf who has tested the idea:

Wolf has made a painstaking and most valuable study of marriage customs among the Chinese. One custom concerned so called "sim-pua" brides; female infants adopted into families and raised side by side with the biological offspring in order to eventually marry a son in the household. Wolf compares the outcome of such sim-pua marriages with those in which girls did not meet their prospective husbands in advance of the wedding day. Sim-pua marriages had a high incidence of adultery, low birth rates and high divorce rates. Thus, marriage to a childhood associate was 2.65 times more likely to end in divorce than an arranged marriage to an unfamiliar partner and 1.24 times more likely to end in divorce than marriages based upon personal choice. Wolf carried out very thorough analyses of over 14000 Chinese women, reaching the conclusion that association during a sensitive period of infancy, spanning approximately the first 30 months, effectively inhibits the development of later sexual attraction.

http://www.psgb.org/BookReviews/SexualAttraction1995.html

Now this hypothesis is far from confirmed, but since it is testable and currently being tested, it cannot be derided as a just-so story.

It must be noted that in the case of sim-pua the adaptation actually misfires, since the brides are not in fact related to the husbands. However, we do not say that imprinting in geese can't be an adaptation because goslings may become attached to German ethologists "by mistake".

Anonymous said...

I would like to attack the very notion of "just-so story". I claim that the term is operationally indistinguishable from the term "hypothesis".

Can anybody provide a set of criteria that would differentiate an evolutionary psychology "just-so story" from, say, the Big Bang theory (other than the use of math)?

Chris Crawford said...

First off, I'm the "anonymous" source of the previous post. I dug up my Google password and now I can decloak.

Larry, I'd like to jump all over the logic behind this statement of yours:

"If the selective advantage is about 1% the the probability of fixation is 2%—in other words the allele will be lost 98% of the time."

I find this statement logically indistinguishable from the assertion of creationists that evolution requires something similar to a tornado in a junkyard assembling a B747. You're using a simplistic probability calculation to deny the possibility of an outcome that is low-probability in a single trial -- but high-probability given enough trials.

Surely you have seen the population calculations the demonstrate that a tiny selective advantage, spread over thousands of individuals and thousands of generations, can indeed establish itself in the gene pool.

Indeed, almost EVERY genetic variation necessarily confers only a tiny selective advantage. Do you really think that the human brain gained mass in increments of 10% or 25%? While it's true that there have been some startling jumps in evolutionary history, you must agree that the fossil record is dominated by tiny incremental change, not mighty leaps.

Accordingly, your criticism of the hypothesis of the genetic basis of incest avoidance seems unsupportable to me.

Torbjörn Larsson said...

Larry:

Sorry for the late response, and thank you for yours!

"It's a mistake to assume that all you have to do is imagine some selective advantage and the allele will automatically take over the population."

But it seems it isn't impossible for low selective advantages to become fixated, which is what I argued. Even in the basic mechanism, it is allowed, though it is very unlikely for a certain allele. Over time several similar alleles may occur and be tested, of course.

"That's one of the problems with just-so stories in general and evolutionary psychology in particular."

I am no fan of evolutionary psychology either. But some such characteristics may be developed by evolution by chance. The problem is to test such hypotheses. Here it seems possible.

"Imagine a primitive human population that does not carry the anti-incest allele. (Such a state is a requirement for just-so stories.)"

If it is a complex trait, as in evo-devo explanations, it could be gradual and coincidental, as I understand it.

"given that they probably wouldn't mate with a brother or a sister anyway? (It's related to the idea that familiarity breeds contempt.)"

That seems like a competing loose assumption. Or as windy says, exactly what we want to explain.

ChepeNoyon:
"I would like to attack the very notion of "just-so story". I claim that the term is operationally indistinguishable from the term "hypothesis"."

I would say that a "just-so story" is (like) an ad hoc, namely descriptive. The assumed mechanism predicts exactly the known type of observations.

Contrats that with a testable predictive theory. The mechanisms incidentally suggest the known type of observations, and something more.