More Recent Comments

Monday, September 28, 2009

Naked Adaptationism

 
Most other mammals think that humans are excessively ugly. They probably see us in the same way we see naked mole rats. We (mostly) have no hair.

What happened to our hair? There are many explanations for human hairlessness but they all share one common characteristic—they are adaptationist just-so stories.1

If you're looking for the best example of an adaptationist then you need look no further than Elaine Morgan, author of the Aquatic Ape speculation [see Elaine Morgan and Aquatic Apes]. She has written an article for last week's issue of New Scientist: Why are we the naked ape?. It won't come as a big surprise to learn that she dismisses all of the speculations about the evolution of hairlessness, except one: we lost our hair because our ancestors lived in the water.

That's not the point I want to make. Here's what Elaine Morgan says in the first few sentences.
RIGHT from the start of modern evolutionary science, why humans are hairless has been controversial. "No one supposes," wrote Charles Darwin in The Descent of Man, "that the nakedness of the skin is any direct advantage to man: his body, therefore, cannot have been divested of hair through natural selection."

If not natural selection, then what?
The idea that our lack of hair might just be an accident is completely foreign to someone like Elaine Morgan. She's probably being deadly serious when she asks the question, "If not natural selection, then what?" For adaptationists, natural selection is the only game in town and no other sorts of explanation are possible.

If it's genetic and visible, then it must be an adaptation. If one just-so story is refuted then make up another one to take it's place. That's what the article is all about. One by one, she dismisses sexual selection, overheating on the savannah, neoteny, avoiding parasites, evaporating sweat, leaving only aquatic ape speculation that hasn't been refuted, or so she claims.

One of the problems with the adaptationist program was described by Gould and Lewontin (1979), "If one adaptationist argument fails, assume that another must exist ...." Why not start thinking about other, non-adaptationist explanations?

Why is that so hard?


1. The one exception is the idea that our nakedness is an epiphenomenon resulting from neotony.

Gould, S.J. and Lewontin, R.C. (1979) The Spandrels of San Marco and the Panglossian paradigm: a critique of the adaptationist programme. Proc. R. Soc. Lond. B 205:581-598.

31 comments :

Anonymous said...

Yes, perhaps we lost all those hair by accident. That explanation would be a non-adaptationist just-not-any-story.

-DG said...

Anonymous said...

Yes, perhaps we lost all those hair by accident. That explanation would be a non-adaptationist just-not-any-story.


Orrrr...

It forms the basis of the appropriate null hypothesis. Neutrality is, and should be, the null hypothesis. If we have evidence for an alternative explanation that allows us to reject the null hypothesis so be it.

Divalent said...

Larry, I love it when your casual reading triggers blog posts! :)

Anonymous said...

Come on Larry,she looks much better without hair.

Psi Wavefunction said...

Oh but there must be some adaptive value to not seeming obviously adaptive in the first place! =P

And then we wonder why evol psych thinks up adaptive explanations for...depression!

Thomas said...

What's Elaine Morgan doing on the cover of Vogue?

Corneel said...

Come on Larry,she looks much better without hair.

I second that, so hairlessness is not neutral. Isn't that so, Larry?

Anonymous said...

"Neutrality is, and should be, the null hypothesis."

A null hypothesis only have sense when there are alternative hypothesis to compare with. But if we disdain alternative hypothesis as invalid "just so stories", then the null hypothesis becomes equally invalid.

Larry Moran said...

anonymous says,

A null hypothesis only have sense when there are alternative hypothesis to compare with. But if we disdain alternative hypothesis as invalid "just so stories", then the null hypothesis becomes equally invalid.

I think you're missing the point.

Whenever we are faced with a trait that has a clear genetic component, like hairlessness, the first question we ask is, "How did this allele become fixed in the population."

We know there are two main possibilities: random genetic drift, and natural selection. If we begin our investigation with an open mind then it's best to keep both possibilities on the table. The best way to do this is to assume drift but look for evidence of selection. If you find such evidence then you can eliminate random genetic drift as a possible mechanism of fixation.

That's not what the adaptationists do. They assume adaptation right from the beginning thereby eliminating random genetic drift as a possibility. All of their efforts are focused on developing a plausible adaptationist explanation for the trait. They do this even though they usually have no evidence that the trait is adaptive.

That's not a good way to do science.

-DG said...

Thanks Larry,

Thats basically what I was also going to come in and respond with. No one is saying that we are unwilling to consider adaptation as an alternative hypothesis. In fact the appropriate methodology, at least when we are looking at molecular data, is to do a test for selection as you described.

I'm not sure if similar tests exist for morphological characteristics or not but in molecular evolution at least we have an appropriate framework set up where Selection (and thus adaptation) is an alternative hypothesis and we reject or fail to reject the null hypothesis (drift) based on the data.

And thats what people like Larry are arguing for in the context of the Adaptationist debate.

Joe Pickrell said...

Whenever we are faced with a trait that has a clear genetic component, like hairlessness, the first question we ask is, "How did this allele become fixed in the population."

it seems unlikely (though possible) that hairlessness in humans is due to a single fixation event at a single locus.

it seems more plausible to me that many alleles needed to fix. the fixation of many alleles (all in the same phenotypic direction) is very unlikely if the phenotype is evolving neutrally.

either it's relaxation of constraint (ie. for some reason humans no longer needed hair) or an adaptation. strict neutrality (along the entire primate lineage) is not an appropriate null hypothesis.

AJ Milne said...

Dispensing with Aquawoman's flogging of drowned horses, do we have a fix (pardon pun) on roughly when hairlessness was fixed?

Just looking for data, just curiosity... got no real horse in this race, speaking of those (does seem to me very naively as one of those cold-weather-loving-types who freakin' *hates* feeling even a bit too warm--and barely feels functional, in fact, in hot summers--that if it were during a warm period, there might be some mild pressure in that direction, and if it had occurred more recently, in ice-ageish conditions, you'd *think* there'd be a bit of pressure against the trait--making clothes does take effort, after all--But again, that's just wide-open naive speculation, seems to me, insofar as both notions do strike me as a bit simplistic... And yes, Dr. M., I do get your point that assuming either assumption is even necessary as an explanation here is probably assuming too much... It just set off the 'I wonder' circuits, is all).

Anyway, again: just curious. Does anyone here know if her statement about it not being known is really on? Is it at all known when? And how well pinned down?

Larry Moran said...

p-ter says,

it seems unlikely (though possible) that hairlessness in humans is due to a single fixation event at a single locus.

it seems more plausible to me that many alleles needed to fix. the fixation of many alleles (all in the same phenotypic direction) is very unlikely if the phenotype is evolving neutrally.


I never would have guessed that you are a fan of Michael Behe! That's exactly the argument he makes in The Edge of Evolution. Whenever there's more than one mutation required for a trait then each one has to be fixed separately and each one has to contribute separately to the fitness of the species.

At least that's what he claims. He leaves no room for the possibility that one, or both, mutations could have been effectively neutral—or even deleterious.

So, if it takes two mutations to produce a naked ape then either both have to occur simultaneously, which he argues is impossible, or else each one was selected.

Which option are you thinking of?

Anonymous said...

"They assume adaptation right from the beginning thereby eliminating random genetic drift as a possibility"

That's bad, I agree. I was trying to say that dismissing, or even laughing at adaptative explanations is also bad, and it's bad for the neutral explanation.

Anonymous said...

We lost our hair because our ancestors lived in the water?

Will seals and polar bears eventually evolve to be naked?

Buzz said...

The best way to do this is to assume drift but look for evidence of selection.

Why not the opposite? After all natural selection is more dominant than random drift. I agree that one should seriously consider drift but why should drift be assumed?

Joe Pickrell said...

So, if it takes two mutations to produce a naked ape then either both have to occur simultaneously, which he argues is impossible, or else each one was selected.

Which option are you thinking of?


i was thinking each was selected. This is what happened, for example, in the evolution of light skin in European populations. The difference there is that we can identify the exact genetic loci involved and show molecular evidence for natural selection.

See, eg.,
http://www.ncbi.nlm.nih.gov/pubmed/18046745

Larry Moran said...

Buzz says,

Why not the opposite? After all natural selection is more dominant than random drift.

No, it isn't. Not by a long shot.

I agree that one should seriously consider drift but why should drift be assumed?

Because random genetic drift is the only possibility in the absence of evidence for selection. When you postulate natural selection as a mechanism you need to have positive evidence of adaptation. You can't just assume the evidence exists and proceed from there.

The onus is on the adaptationist to provide evidence of adaptation because it's the adaptationist who is making a specific claim. The default assumption has to be no selection until there's evidence.

That's how science works.

Michael said...

Larry-

If I may make a suggestion:

The more important point is that genetic drift is always present in a finite population. It is just that change due to genetic drift is easier to distinguish from change due to natural selection is easier to distinguish natural selection, because the mean change in frequency becomes statistically significant.

Disclaimer: I am not a evolutionary biologist.

DiscoveredJoys said...

Larry,

I can think of several other possible causes for relative hairlessness but I don't know if you have included them in natural selection?

The first one is sexual selection, which I understand can often lead to unexpected phenotypes. Perhaps hairlessness is sexy?

The second one is one I've been thinking about recently - social selection. This is not group selection where genes are selected 'for the group benefit', but one where particular variations are selected because they confer a benefit to the individual - because of the nature of the social group he or she lives in.

I expect there are other possibilities too; I just wonder how broad your definition of natural selection is.

anthrosciguy said...

There are adaptationist stories and adaptationist stories. Morgan's doesn't fit what we know about when aquaticism and hair loss coincide, while Wheeler's and Jablonsky's view that it's good for more efficient sweating fits not just the physics but also the apparent timing (which we're starting to learn -- I wouldn't say it's nailed down -- from the recent lice DNA research). It's also clear that it also is either partly due to sexual selection or somehow, for some reason, mimics sexual selection exactly.

And that brings us back to Morgan, because the thing that really gets me about her in that article is that she's once again using that same Darwin quote in the same way creationists use his quote about the evolution of the eye. She quotes the beginning devil's advocate part (he often did that) and pretends his extensive case that it's due to sexual selection isn't there. That tells us something about her.

Larry Moran said...

Anonymous says,

I was trying to say that dismissing, or even laughing at adaptative explanations is also bad ...

It's not bad. In fact it's a lot of fun.

Have you heard of evolutionary psychology? Now there's a barrel of laughs. It can keep you entertained for hours.

Surely that's a good thing?

El PaleoFreak said...

"The onus is on the adaptationist to provide evidence of adaptation because it's the adaptationist who is making a specific claim"

The "neutralist" is making a claim too: hair coat became neutral at some point in our evolution, for a reason we don't know. Or perhaps the mammalian hair coat is and have always been neutral ;-)

Larry Moran said...

El PaleoFreak says,

The "neutralist" is making a claim too: hair coat became neutral at some point in our evolution, for a reason we don't know.

There aren't any "neutralists" that I know of. We prefer to call ourselves "pluralists" because we consider many different ways in which something could evolve.

In the absence of any knowledge, the most appropriate default position is that a trait is not adaptive. That isn't the same thing as advocating neutrality and fixation by random genetic drift. You need evidence in order to make that case as well.

I don't know whether hairlessness is adaptive or not. What I do know is that it is not good science to rule out random genetic drift by fiat.

In making the case against adaptationism, I often point out possible non-adaptationist explanations. That doesn't meant that I'm promoting neutrality. I'm just trying to make everyone understand that the adaptationist program ignores some very important parts of evolution.

Bryan said...

Since when are humans hairless? I, for one, am only hairless on the top of my head and the palms of my hands/feet ;-)

El PaleoFreak said...

"In the absence of any knowledge, the most appropriate default position is that a trait is not adaptive."

Exactly! I don't think there is absence of any knowledge in this case. We know that "nakedness" in mammals have happened few times, and is associated with specific features and adaptations: very big body size in land animals, full-aquatic life, subterranean life. Naked mammals are "extreme" mammals. If losing the hair were neutral, I would expect naked species being more evenly distributed among mammal diversity. Because this knowledge, a nonadaptive hypothesis of hair loss in the human lineage produces mysteries and can't be the default or "null" position.

A. Vargas said...

We are not naked, we have high hair density on our heads and other places.

Traits at any level (phenoytype, molecular) go into use or desuse, because adaptation is context-dependent. Thus a shift to neutral or adaptive can be expected.

I think the origin of human pattern was not directly adaptive but a side-effect of paedomorphosis. In fact the advanced fetus of the chimpanzee has the same hair distribution than a human baby, with high density only on the head, and with separate eyebrows (non- continuous with the hair on the head).

In fact, most other "naked" mammals species are also paedomorphic.

DK said...

I think the origin of human pattern was not directly adaptive but a side-effect of paedomorphosis.

Pleiotropy does not meet definition of neutrality. Seems like you are an evil adaptationist now. :-)

A. Vargas said...

Pleiotropy? what about when Paedomorphosis is environmentally induced?

Michael said...

DK said:

Pleiotropy does not meet definition of neutrality. Seems like you are an evil adaptationist now.

How does pleiotropy "not meet the definition of neutrality"?

More to the point how does making any statement about pleiotropy necessarily make any statement about neutrality?

While the concepts are related in so far as the both describe the relationships among genes, pleiotropy (or, for that matter, epistasis) does not imply anything specific about selective benefit or lack thereof. It just means that one gene can control the expression of many traits. If all the alleles for the same pleiotropic gene do not yield phenotypes that have different probabilities of reproduction, the alleles are neutral.

DK said...

@Vargas:

Environment has nothing to do with any of it. Not in the context discussed, anyway.

@Michael:

You are right and my comment was wrong. In my defense, I would like to think that I was thinking of increasingly low probability of neutrality with the increasing degree of pleiotropy. But I have to admit that it is possible that my evil adaptationist bias made me make that comment :-)