Saturday, May 30, 2009

John Hawks on "Adaptationists vs Pluralists"

 
John Hawks recently posted a comment about adaptationists [see Richard Lewontin: "[T]oo rapid for genetic adaptation"].

Hawks said ...
I don't really find the "pluralist versus adaptationist" debate very interesting. Despite the vocal complaints of some, I can't ever seem to locate the mythical "adaptationists" who deny that non-adaptive evolution ever happens. So the "debate" always comes down to whether particular adaptive hypotheses are true. Since no scientific hypothesis is true a priori, and since "those adaptationists are always saying stupid things" is not a scientific argument, I don't see the point.
I'm astonished that, after all these years, the adaptationists still don't get it.

First, the mythical adaptationist is a straw man that only exists in the minds of the adaptionists. This particular straw man was easily disposed of in the original Spandrel's paper. It is only resurrected by those who haven't been paying attention

Second, the debate does not come down to "whether adaptive hypotheses are true." It comes down to whether any adaptive hypothesis is true. When speculating on mechanisms, adaptationists tend to ignore any mechanism of evolution other than natural selection That's the problem. As a general rule, they don't seriously consider the possibility that the correct explanation may not be adaptation.

It's a difference in worldviews. Pluralists tend to look at an evolutionary outcome and ask, "What mechanism of evolution caused this?" Adaptationists tend to look at the same outcome and ask, "How can this be explained by natural selection?" Adaptationists know about random genetic drift—they just don't think it's an important player when it comes to the parts of evolution that they're interested in. I think that's a bad assumption.



24 comments :

  1. Hmm....

    How is:

    When speculating on mechanisms, adaptationists tend to ignore any mechanism of evolution other than natural selection. That's the problem. As a general rule, they don't seriously consider the possibility that the correct explanation may not be adaptation. different from

    "those adaptationists are always saying stupid things"?

    At any rate, whether someone "seriously considers" a hypothesis sounds like mind-reading to me.

    And whether genetic drift, selection, pleiotropy, or historical path-dependence is a priori likely to be the cause of any particular event seems a lot more like religious faith than science.

    Hence, I don't find the supposed "debate" to be very interesting. If any mythical adaptationists are to be found, I will happily volunteer to provide free primers to the work of Sewall Wright and Motoo Kimura. And if any mythical non-adaptationists need it, I can point them to R. A. Fisher and William Hamilton.

    ReplyDelete
  2. Adaptationists teach the 1st year evolutionary biology course at U of T (at least when i took it). Here's a concrete example: in the 'evolutionary and behavioural ecology' section of the course, Locke Rowe's formulation for evolutionary explanations of behaviour has an "axiom":

    "Natural selection has acted to maximize fitness"

    and we proceed from there to explain foraging, anti-predator, mating, social, parent-offspring behaviour, etc

    ReplyDelete
  3. I think Larry is screwed as long as he thinks drift is irrelevant to the origin of adaptations, according to the empirically false ( but alas, much circulated) dictum that "only natural selection explains adaptation"
    Spandrels are non-adapative traits and a source for exaptations. These are crucial in the evolution of complex adaptations. Read Gould, sometime.

    I'm afraid both Larry and Hawks, for basically cultural, non-scientific reasons, may be clingin on to Darwins old argument that"natural selection explains the appearance of design".

    ReplyDelete
  4. The evolution of opsins in fishes at different depths and light wavelentgths (PNAS, last year) is a beautiful example of the role of drift in the origin of adaptations. Substitutions without immediate effect nevertheless made future shifts in wave length absorvance possible.

    ReplyDelete
    Replies
    1. could you please send me the name of the paper.

      Delete
  5. This passage shows what an out of date relic Lewontin is:

    "Greg Gibson posits a revolutionary new hypothesis: Our genome is out of equilibrium, both with itself and its environment. Simply put, our genes aren’t coping well with modern culture."

    This is laughable. Lewontin has unknowingly described a tenet of evolutionary psychology. The fact that he thinks it is revolutionary just proves how clueless the once-great Lewontin has become.

    I have problems with all of you guys - Lewontin, Hawks (enabling Cochran's racialist hereditarianism - honestly, you ought to know better), Vargas (Goldschmidt cam be useful - in his proper place), Moran (citing seventies Gould as if it were state of the art instead of obsolete) - for different reasons.

    ReplyDelete
  6. "And whether genetic drift, selection, pleiotropy, or historical path-dependence is a priori likely to be the cause of any particular event seems a lot more like religious faith than science.

    Hence, I don't find the supposed "debate" to be very interesting..."

    I have also lost interest in this debate for a related reason; I can't see the fruitfulness of it.

    Unless the "pluralists" can generate more accurate testable predictions about the direction of evolutionary change than the "adaptationists", they can't do better than say that evolution happens. That doesn't have any novelty or impact.

    ReplyDelete
  7. I met a few "pluralists". They weren't asking "what mechanism of evolution caused this". They were asking "how can we build an explanation, no matter how unlikely or muddled, that totally exclude natural selection".
    Avoiding natural selection and adaptation seemed to be their main goal.
    I don't know if they were false pluralists, or mythical pluralists, or what.

    ReplyDelete
  8. John Hawks says,

    And whether genetic drift, selection, pleiotropy, or historical path-dependence is a priori likely to be the cause of any particular event seems a lot more like religious faith than science.

    I agree completely. Anyone who starts out believing that one particular type of explanation (e.g. adaptation) is the only possible explanation, isn't doing science.

    Many of you think that adaptationists never think like that but you are wrong. There are many adaptationists who believe that (almost) all visible phenotypic change must be due to natural selection.

    Here's an example from The Extended Phenotype by Richard Dawkins.

    If a whole-organism biologist sees a genetically determined difference among phenotypes, he already knows he cannot be dealing with neutrality in the sense of the modern controversy among biochemical geneticists.

    There have been lots of other examples on this bog.

    ReplyDelete
  9. If a whole-organism biologist sees a genetically determined difference among phenotypes, he already knows he cannot be dealing with neutrality in the sense of the modern controversy among biochemical geneticists.



    I'm sure you can find some Dawkins more unreasonable than this, which seems self-evidently true. Genetic drift has different effects on phenotypes than it does on allele frequencies. Among other things, it affects "genetically determined" traits more strongly than less heritable traits.

    ReplyDelete
  10. I may congratulate myself on my "pluralism" by accepting both creationism and evolution, for instance. As I've always said "pluralist" is a word that in itself means close to nothing, a hollow claim to virtue.
    In this regard, I'm not "pluralist": I reject an ideological panselectionism whose immediate most false notion is that only natural selection produces adaptation.

    ReplyDelete
  11. Tiresome indeed.
    Dawkins (and I guess I'm on his team in this "debate") knows that phenotypic traits impose costs. Whatever their proximate cost-currency (energy or monomer allocation, pleiotropy, survival, reproductive [frequency, fecundity, timing], etc.]), such costs can and do affect fitness. Traits that do not produce a significant fitness benefit are therefore selected against and weeded out. Population genetics models show that selection that is too subtle to measure in human lifetimes has nevertheless strong effects on populations over longer time periods. Insert the Dawkins quote here.

    Biologists that are not accustomed to thinking about trade-offs, energy budgets, populations, long time periods, and organismal-level phenotypes can keep waving aroung copies of Spandrels and championing drift and contingency under the banner of pluralism, but to what end? The examples they offer are expressions of personal incredulity. Show me some data.
    But "the pluralists" never assume the burden of proof. Neutral drift (or whatever) is privileged as the default assumption and adaptation has to explicitly demonstrated.
    And I think it should be, when possible; I have nothing against rigor and I'm always interested in data.
    But everything I know about biology (at the organizational levels that most interest me) suggests strongly that selection is powerful and adaptation ubiquitous. I see good theoretical reason to doubt that "neutral" traits should be common. That's why I remain one of those "adaptationists."

    ReplyDelete
  12. "Dawkins (and I guess I'm on his team in this "debate") knows that phenotypic traits impose costs'

    Which is precisely why adaptationists have problems explainig n the persistence of vestiges (the short answer: organisms are not "optimal")

    "Traits that do not produce a significant fitness benefit are therefore selected against and weeded out"

    Neutral traits are weeded out? This is just your adaptationist prejudice talking, what you're saying is pretty easy proven empirically false (think of the tongue-rolling example)

    "Biologists that are not accustomed to thinking about trade-offs, energy budgets, populations, long time periods, and organismal-level phenotypes can keep waving aroung copies of Spandrels and championing drift and contingency"

    This is just your chauvinistic stupidity showing now. Other biologsits who do not agree with you can be competent within all those areas of research.

    "The examples they offer are expressions of personal incredulity. Show me some data"

    I just mentioned a nice study showing a role for drift in the adaptive evolution of fish opsins (in PNAS last year).

    Certainly we can agree that most molecular change is neutral. Well, I've got news buddy. That's structural change and evolution, too. And it changes the possibilities for the evolution of the phenotype, including adaptations.

    ReplyDelete
  13. The tubercules on the fins of whales are a good example of why it is foolish to dismiss adaptation a priori. I would dollars to donuts that just about every pluralist assumed that these tubercules (if they gave the matter a moment's thought) assumed that these were phenotypic drift, developmental constraint, phylogenetic inertia, blah blah blah etc. But the adaptationists were right, just as they were on many other seemingly trivial or inconsequential features.

    Often there is no more evidence for constraintist/structuralist/driftist theories than adaptationist ones, but that doesn't stop anyone from getting smug and reciting their sacred texts (Gould, Fisher whatever).

    Not to say that there aren't dogmatic adaptationists too.

    The intense passion on the part of advocates and critics of pluralism, structuralism, selectionism-adaptationism and so on demonstrates that faith abhors a vacuum. (I'm an atheist myself, but I don't turn to scientific theories to provide myself a secular meaning of life like some 'rationalists' do.)

    ReplyDelete
  14. To Sven DiMilo,

    "Energy costs" are often thrown around as a way to refute genetic drift arguments, including the old "junk DNA" argument.

    However, it seems that organisms with genes that impair mitochondrial function (therefore, reducing cellular energy production) survive just fine. This continues until the mitochondrial DNA in the organism carries a large excess of mutant DNA versus wildtype, usually over 80% mutant allele for the more common human mitochondrial diseases. One can assume these mutations have significant energy costs, probably measurable decreases in OXPHOS capacity, but not clinically obvious until at very high levels. Yet carriers of these diseases are not that rare.

    How does this observation fit in with huge selective cost on energy inefficiency in an organism?


    To state my potential bias, I use drift and neutral theory as a null hypothesis, and prefer some gene or biochemical confirmation of how the story is adaptive.

    (For those not read up on mitochondrial genetics, each cell contains 100's to 1000's of copies of the mitochondrial DNA, and replication is no coordinated or cell cycle synchronized, so you can get varying levels of wildtype to mutant allele between tissues or even cells).

    ReplyDelete
  15. Painfully.

    The arguments seems pretty stupid.

    Biochemistry will do what biochemistry will do.

    And selection -- of any type -- will do what selection will do.

    That is all. The rest is detail. Important detail -- but just detail.

    Labels trying to tie things down which are not clearly understood and then arguing which label is the best label seems completely counter-productive.

    ReplyDelete
  16. I agree with John Hawks about how ultimately this debate is uninteresting at any meaningful level. At the end of the day, if the evidence shows that selection was likely, the pluralists will not object (all agree they are fine with adaptations), and if all reasonable adaptive hypotheses are ruled out, even the uber-adaptationists will concede that drift remains a viable possibility.

    The only danger of the "adaptationist program" (even the mythical one) is if one assumes that merely proposing an adaptive explanation is sufficient to settle the question. In the main, for those actively studying some question in a rigorous way, this is not a the case. (Speculations in the lay press don’t count.)

    As Sven DiMilo points out, drift is always the null hypothesis to account for a phenotypically visible trait or characteristic. Establishing drift requires (among other things) one to rule out *all* reasonable adaptive hypotheses. Consequently, as a tactical matter, taking an adaptive hypothesis as the test case is most likely to lead *directly* to the most fruitful lines of inquiry.

    For >99.99% of all allele fixations in the past that affect a phenotypically visible trait or characteristic, we will never conclusively know the definitive answer, as the necessary information is irretrievably lost. Based on my own qualitative survey of the literature, the number of such fixations known or strongly suspected of having been fixed by drift appears to be (far) less than the number similarly known or suspected of having been fixed by selection. If true, the best working hypothesis to start with is that selection was the culprit.

    Larry: “There are many adaptationists who believe that (almost) all visible phenotypic change must be due to natural selection.”Do you have evidence to contradict this? This seems to be a key point of difference. Your position seems to be that it is not true, but I've never seen the evidence and (as mentioned above) my own survey seems to agree with the proposition. So, of such traits where a consensus has been reached, what is the ratio of adaptive vs drift explanations? (And why can't one agree with this statement and yet still call oneself a "pluralist"?)

    ReplyDelete
  17. Jim:

    Junk DNA and mitochondrial inefficiency are perfectly explicable from a selectionist-adaptationist perspective - as long as the level focused on is the gene, not the whole organism.

    Parastic DNA, meiotic drive, selfish DNA, intragenomic conflict, ERVs, parental conflict in genomic imprinting - all of which can reduce the fitness of the organism - are all predictable in terms of genic selection. Only a dogmatic ORGANISMIC adaptationist thinks that selection will necessarily result in organisms being optimally adapted. (Even worse than an organismic adaptationist is a group adaptationist. There is group selection, but group adaptation can be undermined by individual selection, just like individual adaptation can be undermined by genic selection.)

    Similarly, cancers are explicable from the vantage point of cellular selection.

    ReplyDelete
  18. To anonymous,

    Points well taken, especially for odd behaving human mutations like mitochondrial t8993G (a NARP mutation) and a8344G (a MERRF mutation) where these specific mutation vary wildly in pedigrees. Something odd and interesting is going on here, compared to A3243G (common MELAS mutation), which seems to recur, but segregate neutrally in the germline.

    My criticism was directly against the energy balance / energy optimization stories at the organsimal level, that often used.

    ReplyDelete
  19. Here's an interesting thought. An adaptationist has to explain the presence of religion in man as being something extremely beneficial to the species.
    Since even tribes of humans who have been seperated from extratribal members of their species since before the Upper Pleistocene still exhibit religious practices and superstitions, and since the lack of religion in humans is only exemplified by a fractional margin in any given modern human population, it must follow that "the religious gene" evolved in humans via natural selection long before civilization.

    Religion cannot be seen as a "cultural phenomena" because all races and varieties of humans have it and some of those populations have not been in contact with the rest of human culture since soon after the species evolved.

    All of the members of mitochondrial Eve's clan were most likely religious,(possibly excluding the usual 5%-10% agnostic/atheist), and she most definitely was....but the question is, why had religion (or why WOULD religion eventually)enabled her to outcompete so many other hominids, if natural selection is the impetus?

    ReplyDelete
  20. "why had religion (or why WOULD religion eventually)enabled her to outcompete so many other hominids, if natural selection is the impetus?"

    Natural selection is not competition between species.

    ReplyDelete
  21. OK but isn't natural selection a way that a species can be seperated into other species as the mother species spreads out across different environments?
    If we look at the point I was attempting to make in the microcosm of HSS, then I can rephrase the question by asking why only the HSS who developed religion survived.

    ReplyDelete
  22. "OK but isn't natural selection a way that a species can be seperated into other species as the mother species spreads out across different environments?"

    In other words, natural selection occured differently for pre-homo erectus in Africa (yielding ergaster), Asia (yielding Homo Erectus) and Europe (yielding antessessor).

    ReplyDelete
  23. "OK but isn't natural selection a way that a species can be seperated into other species as the mother species spreads out across different environments?"

    Yes, but this speciation, not (out)competition.

    ReplyDelete