Evolution for Everyone

One reason that I don’t spend a lot of time bashing religion is because there are so many other flagrant departures from factual reality to pick on. Take the patriotic history of nations–the leaders who can do no wrong, the noble “us” and evil “them”–who needs supernatural agents when we can so freely re-arrange the facts of the real world?

Science is supposed to be different. Indeed, science can be idealistically defined as a cultural system designed to hold people accountable for their factual statements. Like religion, however, science as practiced often falls short of science as idealized.

The rejection of group selection and acceptance of the theory of individual selection (see T&R IV) reads disturbingly like a patriotic history. I am aware that this is a serious charge. Basically, I am saying that the theory of individual selection represents a failure of the scientific process and an example of values masquerading as facts, little different than religious, political, and other ideologies. That is why a truth and reconciliation process is needed. Before continuing, however, I want to stress that I remain idealistic about science as a cultural system that–when it works as intended–can indeed hold people accountable for their factual statements. My goal in this series of blogs is to make the scientific process work better for the issues represented by the group selection controversy. Think of me as a scientific reformer.

Consider the following passages written by highly respected evolutionists during the 1970’s and 80’s.

The economy of nature is competitive from beginning to end…the impulses that lead one animal to sacrifice himself for another turn out to have their ultimate rationale in gaining advantage over a third…Where it is in his own interest, every organism may reasonably be expected to aid his fellows…Yet given a full chance to act I his own interest, nothing but expediency will restrain him from brutalizing, from maiming, from murdering–his brother, his mate, his parent, or his child. Scratch and “altruist,” and watch a “hypocrite” bleed (Michael Ghiselin, The Economy of Nature and the Evolution of Sex, 1974 p274).

The intervening years since Darwin have seen an astonishing retreat from his individual-centered stand, a lapse into sloppily unconscious group-selectionism, ably documented by Williams…It is only in recent years, roughly coinciding with the belated rise to fashion of Hamilton’s own ideas, that the stampede has been halted and turned. We painfully struggled back, harassed by sniping from a Jesuitically sophisticated and dedicated neo-group-selectionist rearguard, until we finally regained Darwin’s ground, the position that I am characterizing by the label ‘the selfish organism’, the position which, in its modern form, is dominated by the concept of inclusive fitness (Richard Dawkins, The Extended Phenotype, 1982 p6).

I suspect that nearly all humans believe it is a normal part of the functioning of every human individual now and then to assist someone else in the realization of that person’s own interests to the actual net expense of the altruist. What this greatest intellectual revolution of the century [i.e., the theory of individual selection] tells us is that, despite our intuitions, there is not a shred of evidence to support this view of beneficence, and a great deal of convincing theory suggests that any such view will eventually be judged false (Richard Alexander, The Biology of Moral Systems, 1987 p3).

According to these authors, evolutionary theory ratifies the concept of individual self-interest as a grand explanatory principle. Lest you think that these passages were written for a popular audience, in which case a bit of poetic license might be justified, they are all taken from academic books–scientists writing for other scientists.

Patriotic histories represent conflicts in black-and-white terms and their resolution as definitive. The passages quoted above express complete certainty. Alexander’s phrase “a great deal of convincing theory suggests that any such view will eventually be judged false” does not invite continuing inquiry. Richard Dawkins went even further:

As for group selection itself, my prejudice is that it has soaked up more theoretical ingenuity than its biological interest warrants. I am informed by the editor of a leading mathematics journal that he is continually plagued by ingenious papers purporting to have squared the circle. Something about the fact that this has been proved to be impossible is seen as an irresistible challenge by a certain type of intellectual dilettante. Perpetual motion machines have a similar fascination for some amateur inventors. The case of group selection is hardly analogous: it has never been proved to be impossible, and never could be. Nevertheless, I hope I may be forgiven for wondering whether part of group selection’s romantic appeal stems from the authoritative hammering the theory has received ever since Wynne-Edwards did us the valuable service of bringing it out into the open (The Extended Phenotype, 1982 p115).

Notice how Dawkins carefully acknowledges that group selection is a theoretical possibility. The basic logic of multilevel selection is impeccable and was affirmed by Williams and others, as I recount in T&R IV. The question is whether group-level selection can ever prevail over individual-level selection. According to Dawkins, this question had been answered so authoritatively that doubters could be compared to romantic dreamers and intellectual dilettantes searching for perpetual motion machines.

Given the certainty with which group selection was rejected, it was kept alive in articles and textbooks primarily as a cautionary tale for how not to think. It became almost mandatory for authors to inform their readers that group selection was not being invoked. Just as patriots vilify their opponents and make sure that they are counted on the side of the righteous, invoking group selection became a heresy inviting ridicule and exclusion. Here is how Stephen Jay Gould recalls the period in an introduction to Richard Goldschmidt’s The Material Basis of Evolution (p. xv), which also became the subject of ridicule:

I have witnessed widespread dogma only three times in my career as an evolutionist, and nothing in science has disturbed me more than ignorant ridicule based upon a desire or perceived necessity to follow fashion: the hooting dismissal of Wynne-Edwards and group selection in any form during the late 1960’s and most of the 1970’s, the belligerence of many cladists today, and the almost ritualistic ridicule of Goldschmidt by students (and teachers) who had not read him.

In future installments of the T&R series, I will show that the certainty expressed by Ghiselin, Dawkins, and Alexander was sheer bravado. The theoretical and empirical case against group selection was never strong and even what there was began to fall apart immediately. That did not alter the patriotic history, however, which is still dutifully reported in text books and transmitted as an oral tradition among graduate students, who warn each other not to invoke group selection in the presence of their faculty advisors. The patriotic history of individual selection theory is a sorry chapter in the history of science. Why did it occur in the first place?

To be continued.

Comments

  1. #1 David Sloan Wilson
    October 27, 2009

    I will take the opportunity to summarize some of the comments up to this point.

    1) Multiple definitions of major concepts, such as self-interest and fitness.
    2) Altruism and selfishness as behaviorally flexible strategies.
    3) Modeling details, such as continuous vs. discontinuous processes, details of the selfish punishment model, etc.
    4) The need to go beyond models and support group selection with real data.
    5) The gene’s-eye view as an alternative to multilevel selection.
    6) Genes as replicators vs. more holistic conceptions of developmental systems.
    7) Fitting evolution into a crowded biology curriculum
    8) Does group selection assume that the group optimum is achieved?

    I can’t do justice to all of these comments, other than to praise the general level of discourse. Some of the major questions will be dealt with in future installments. Very briefly…

    1) There are indeed multiple definitions, which can be both an asset and liability. Welcome to the wonderful world of pluralism, which is a hot topic in the current multilevel selection literature.
    2) Conditional strategies are indeed an important step beyond early models that assume strict genetic determinism, but they don’t change the basic logic of multilevel selection. For example, the “tit-for-tat” strategy of game theory is behaviorally flexible but it still never beats its social partner; it can only lose or draw. To find the advantage of this strategy, we must look at the level of the interacting pair, or more generally the N of N-person game theory.
    3) Every model makes simplifying assumptions, requiring a large family of models to systematically explore the parameter space.
    4) Of course models by themselves are insufficient. In future installments I will show that there is ample evidence for group selection in both the laboratory and the field. However, it is important to note that group selection was rejected in the 1960’s primarily for being theoretically implausible, not because of a great weight of empirical evidence. Hence, it means something when group selection turns out to be theoretically plausible after all.
    5) Within-group vs. between group selection are genuinely alternative hypotheses; if a trait evolves purely by within-group selection, then between-group selection need not be invoked. The gene’s eye view vs. group selection are not alternative hypotheses, as I will describe in detail in a subsequent installment.
    6) The concept of genes as replicators is problematic in a number of ways, including the fact that they are more like pages being replicated by a copying machine. Even as the concept was developed by Williams and Dawkins, however, it is surprisingly irrelevant to the group selection controversy, as I will describe in a subsequent installment.
    7) The idea that evolution can’t be inserted into a crowded curriculum is profoundly wrong. Evolution needs to be taught first so that it can organize that crowded curriculum! It’s shocking how science education, including higher science education, is still taught as a bunch of unconnected facts that students are expected to memorize. An enlightened science curriculum would begin with evolutionary theory and then teach everything in the light of it.
    8) Group selection does not assume that the group optimum is achieved. Pure within-group selection and pure between-group selection are extremes of a continuum and many traits reflect the action of both levels of selection. Even single organisms are not entirely the product of pure organism-level selection, as the many examples of intragenomic conflict attests. Against this background, the assessment that group selection sometimes occurs is far more reasonable than Williams’ extreme claim that it never occurs.

    I look forward to the next round of comments.

  2. #2 Jesse
    October 27, 2009

    I can think of a pretty simple reason that individual selection appealed to people — especially Western-educated scientists – as a theory. A rather large hunk of our culture is based on the premise that we act as individual actors. It’s not even something we think about consciously. We just assume that we are individuals acting in our own interest. It’s one reason why we see capitalism as a good idea.

    Put another way, we have a lot invested in the idea of individual merit. Reading the passages you quote above, it seems to me that these folks have a tough time with the idea that a collective of any kind is even feasible as a way of organizing — even though group selection can be demonstrated by any group of people that forms a unit bigger than two and survives long enough to create a village-sized community.

  3. #3 Justin Wagner
    October 27, 2009

    If I remember correctly, in a more recent book (A Devil’s Chaplain, 2003) Dawkins briefly responds to the ideas in Gould’s The Structure of Evolutionary Theory regarding selection acting at different levels in the hierarchy. He acknowledges that such selection may be possible (without commenting on relative frequency), but stresses that he believes selection at any level must always funnel consequences down to the gene level.

  4. #4 InfuriatedSciTeacher
    October 27, 2009

    Not entirely sure if the discussion between Mike and me regarding the teaching of evolution is the source of #7, but for clarification I was referring to an entire undergraduate biology program and a fair number of master’s classes that didn’t address group selection beyond its dismissal, not the teaching of evolution in general.
    A more interesting and disappointing take on that conversation is that there are teacher-candidates that want to teach biology and openly declare that they will not teach evolution because it conflicts with their religious views. NC’s science standards directly address teaching evolutionary theory in the the biology class, but it definitely isn’t an interwoven theme as it needs to be when implemented. (What to present is not a real choice, how to present it certainly is. The curriculum is heavy on cell bio and weak on evo/eco).

    5) seems quite plausible, given a multi-level examination of the process. The issue with reductionism is that it might miss emergent properties, but not attempting to reduce is failing to address fundamental questions about whatever system is being studied.

    8) Right, just as selection of any kind doesn’t assume that an optimum is acheived… thus the constant competition and changes… running just to stay in place, as it were (Ridley’s Red Queen analogy, in essence)

    6) I’ll wait on the evidence I don’t have to comment on that one… the view to which I was exposed was essentially Dawkins’.

    Jesse> interesting idea… collectivism isn’t taken well some aspects of Western society, I agree. Do you think this truly extends into the examination of non-human organisms as well? Postmodernist psych and Ed psych are full of examples of collectivism (distributed cognition for example) that are purely theoretical or fail to emerge when examined experimentally. We might want to consider basing our theories on observations of reality rather than the writings of Smith and Rand on one side and Marx and Engels on the other.

  5. #5 InfuriatedSciTeacher
    October 27, 2009

    One reason that I don’t spend a lot of time bashing religion is because there are so many other flagrant departures from factual reality to pick on. Take the patriotic history of nations–the leaders who can do no wrong, the noble “us” and evil “them”–who needs supernatural agents when we can so freely re-arrange the facts of the real world?

    Good point. The conflict is between rationality and those flagrant departures from factual reality. The reason I do tend to bash religion is that it makes such a handy tool for those who want to whip up a furor of nationalist frenzy (references to godless communists, anyone?) and it’s a source of broader anti-science ignorance than nationalism in general. Rampant nationalism, religious bent or not, is probably responsible for more death and suffering than religion in isolation, however. I’m not even going to attempt to document the numbers, as shifting definitions of both nationalism and religion are going to cloud the issue.
    Think there might be/have been a selection pressure (cultural or biological) for forming in-group/out-group divides, regardless of the ideological framework?

  6. #6 piker
    October 27, 2009

    “2) Conditional strategies are indeed an important step beyond early models that assume strict genetic determinism, but they don’t change the basic logic of multilevel selection. For example, the “tit-for-tat” strategy of game theory is behaviorally flexible but it still never beats its social partner; it can only lose or draw. To find the advantage of this strategy, we must look at the level of the interacting pair, or more generally the N of N-person game theory. ”
    The problem with game theory, even when combined with sophisticated computer modeling, is that the computer cannot apply any forms of intelligence to the strategies that were not initially part of its programming. The “intelligence” used by life forms contains inferential premises that we as yet have too little knowledge or understanding of.

  7. #7 Bob O'H
    October 27, 2009

    The theoretical and empirical case against group selection was never strong and even what there was began to fall apart immediately.

    Doesn’t this comment conflate two different types of group selection – the naïve version, as exemplified by Wynne-Edwards (poor bugger), and trait group selection (which is about to appear…)?

  8. #8 Jesse
    October 27, 2009

    InfuriatedSciTeacher >

    Well, I was just thinking about Gould. One of the things he points out numerous times in his essays is that the idea of a “chain of being” was deeply embedded in Western culture from the get-go, so the fact that non-human organisms would be set up that way was just logical, from a 19th century point of view.

    Or, the idea of time as a separate entity from space, and God as an “outside” observer from the universe, I think deeply affected people and is one reason why Relativity seemed so odd at first.

    It’s not about saying “I am a collectivist” or “I am an individualist” and tailoring your theory to fit, it’s more about unconscious biases that make a theory (or way of thinking) “feel” right.

    Note that initially, Darwin’s (misunderstood) theory was the basis for a lot of social darwinist thinking because the observation (that some people do less well in society) seemed to fit a general framework so well (that some organisms are better “fit” for their environment). You can have observations fit theories that aren’t necessarily correct, even if you are a meticulous observer.

    After all, prior to a lot of modern technology (say, electrical devices) there was simply no way to test a lot of things such as aether theory. And the observation that light traveled in waves was pretty ironclad. The fact that it fit preconceived, pre-scientific classical (Greek, mostly) philosophical notions as well (like the indivisibility of space) was a bonus.

  9. #9 Allen MacNeill
    October 27, 2009

    Having thought about these issues for a very long time, it seems to me that the underlying reason for the “patriotic” adherence to individual selection is the historical fact that it is grounded in R. A. Fisher’s simplification of evolution by natural selection to changes in allele frequencies as the “ultimate” basis for evolution. Dawkins’ “selfish gene” theory is based on this concept, as is most of modern evolutionary theory: what really matters is changes in genes (or, more precisely, alleles), which cause changes in phenotypes.

    In The Genetical Theory of Natural Selection (the founding document of modern population genetics, and therefore of modern evolutionary theory), Fisher is very clear on this point: that Mendelian genes correspond (in essentially a one-to-one fashion) to phenotypes. Change the frequency of the genes and you change the frequency of the corresponding phenotypes. By contrast, Darwinian natural selection is about changes in relative frequencies of phenotypes, with no direct reference to the underlying genetic “causes” of those phenotypes.

    What has happened over the past fifty or so years is that the supposedly necessary connection between gene (i.e. allele) frequencies and phenotypes has been undermined to the point that no less a respected historian of the “modern evolutionary synthesis” than William Provine has pronounced that synthesis “dead”. And, indeed, it is; evolutionary changes in phenotypes have been shown to be only very loosely related to changes in the underlying genes. The work of Mary Jane West-Eberhard (following in the tradition of C. H. Waddington) and Eva Jablonka (following in the tradition of Jean-Baptiste Lamarck) has shown that much of heritable phenotypic variation is developmental (i.e. acquired), rather than genetic (i.e. genetically inherited). Furthermore, the work of Motoo Kimura, Tomoko Ohta, and others has shown that most changes in genes and/or proteins do not necessarily correspond to changes in phenotypes. In other words, any process that results in phenotypic variation (no matter what its genetic or developmental basis) can provide the “fuel” for evolution, and the “patriotic” stranglehold of the “change the gene to change the trait” model of biological causation has been decisively superceded.

  10. #10 piker
    October 27, 2009

    Now, that’s how to get right to the point and make it. Brilliantly done.

  11. #11 James Taby
    October 28, 2009

    I love that opening line about delusions in the real world too. Do you think that Dawkins ever actually reads any of your blogs? You don’t seem to have the cult of personality he does. I think you own him pretty bad.

  12. #12 InfuriatedSciTeacher
    October 28, 2009

    By contrast, Darwinian natural selection is about changes in relative frequencies of phenotypes, with no direct reference to the underlying genetic “causes” of those phenotypes.

    Right, because Darwin hadn’t the slightest idea how traits were inherited…this was considered one of the weak points in his theory at the time, but let’s not pretend that the original theory is being ignored because someone followed an established mechanism for heredity.
    You are correct in the sense that the phenotype is what is directly experiencing the selection pressure, regardless of however that trait is passed down. Do you honestly think that the lack of one-to-one, gene-to-phenotype mapping discredit the idea that phenotypic traits are inherited via genes?

  13. #13 Mike
    October 28, 2009

    @InfuriatedSciTeacher

    Do you honestly think that the lack of one-to-one, gene-to-phenotype mapping discredit the idea that phenotypic traits are inherited via genes?

    For my part, the answer to that particular question would be “no”. But seeing as the phenotype is determined by developmental plasticity, epigenetic, behavioral and cultural inheritance as well as genes, it is certainly true to say that the idea that the phenotype is determined solely (or even for the utmost part) by genes is discredited, as well as the (of course related) idea that that the inherited ‘parts’ of the specific phenotype are inherited solely through the genes.

    It appears the genes provide the rough building plan, and the other forms of inheritance (including niche construction) as well as the (non niche-constructed) environmental idiosyncrasies determine the specific way this rough building plan is realized and individualized.

    Frankly, I don’t see how that can be denied in light of the evidence.

  14. #14 InfuriatedSciTeacher
    October 28, 2009

    Mike>
    For your part, we agree… nor was I clinging to the idea of gene X causes trait Y… even in microbes it isn’t always that simple. “Rough building plan” seems a valid way of explaining that. After actually looking at who Allen McNeill is, it becomes fairly obvious that he probably wasn’t making the all or nothing statement I took it to be… (I’ve spent too much time with people who term things in black and white). That being said, credentials or not I don’t buy the statement that there is no genetic effect whatsoever. The traits are inherited via genes, and the expression of those is impacted by a number of other factors. This statement is not intended to include behavioural factors in anything that can reasonably be assumed to think on some level. Dawkins framed the same idea as “recipe, not a blueprint” in TGSOE, so I’m somewhat wondering where the disagreement on that really lies, since it would appear that all sides accept that there are both environmental and genetic factors that influence development. Perhaps I’m giving RD more credit for open interpretation than I should be on that?

    In other words, any process that results in phenotypic variation (no matter what its genetic or developmental basis) can provide the “fuel” for evolution,

    is about as fair a summation as I’ve seen.

  15. #15 becca
    October 28, 2009

    Well, group selection theories are often romantic. That shouldn’t be a dirty word.
    It seems to me like an individual selection fetishist would argue we want to have group selection simply because there are so many more warm ‘n fuzzy explanations and we are biased toward wanting to believe nice things about others to support cooperation (i.e. it’s romantic).
    And a group selection fetishist (not that you are, Dr. Wilson, but it’s something to be aware of) could argue that we want to have individual selection simply because there are so many more ways to ‘justify’ selfish behavior and we are biased toward excusing such behavior in ourselves.

    Bias in both directions is possible, depending on whether one is an optimist or a pessimist about ‘human nature’ (or even alternative whether one is a forest- or a leaf- focused sort of person…). But just because a rationalization is useful doesn’t mean it isn’t true. Both types of selection are obviously going on, and both types of understanding are obviously useful to people with their own biases or agendas.

    I kinda get the feeling I did sitting in Carl Woese’s class. He’s an older gentleman, and has been fighting so hard for so long that he didn’t even realize when the dominant paradigm swung in his direction. I think I might have been the first generation of student he taught to grow up with the three domain model at least alongside, and sometimes superseding, the five kingdom model. I couldn’t understand why he was so vehement about the microbe oriented world. Of course most of the biochemical and numerical species diversity is microbial and any evolutionist who ignores that does so at their peril. I mean, duh.
    😉

    Dr. Wilson, I wonder if the same thing has happened to you. Why wouldn’t we assume group selection occurs?

  16. #16 albert rogers
    October 28, 2009

    Am I the first in this thread to say that Dawkins is right?
    Genes work through the phenotype, but the phenotype is evanescent. Groups are somewhat less so, which is what I think E.O.Wilson’s Sociobiology is about. But the hereditary material in a group, other than what Dawkins calls memes, is gene frequency. So although a gene for altruism, if you have it, can get you killed, a preponderance of such genes in an extended family group can give it advantages over other groups.
    It is still the gene, not the group, that is behaving as if it only cares about its own multiplication. But a group that has a system of behavior patterns (perhaps not genetically inherited) that restrain group-destructive activity might very well in the long run be changing its gene pool. This is in no way incompatible with Dawkins’s position.

  17. #17 albert rogers
    October 28, 2009

    Piker says that “the computer cannot apply any forms of intelligence to the strategies that were not initially part of its programming.”

    I think he’s wrong.

    For a start, you could say the same thing about a new born baby. Yet identical twins turn into significantly different persons, their bodies and minds being molded by mechanisms just as physical as the working of a computer. The computer is merely vastly less complex in its electronic structure.

    It of course raises the question “What do you mean, intelligence.”

    I know perfectly well that computers are not intelligent.
    They only think that they are.

  18. #18 piker
    October 28, 2009

    “So although a gene for altruism, if you have it, can get you killed, a preponderance of such genes in an extended family group can give it advantages over other groups.
    It is still the gene, not the group, that is behaving as if it only cares about its own multiplication.”

    Did or does Dawkins then, along with the Wilsons, propose that there are actually genes for altruism, and that some will have them while others don’t? Or that those that don’t have genes for selfishness in their stead?
    Even Dr. Wilson, as noted in his comment above, has started to speak of these traits as conditional strategies. (Even though he still seems to see them as mutually exclusive.)

  19. #19 piker
    October 28, 2009

    And also Albert, computers can’t draw inferences from any data not at some point input by one or another of their operators. Babies, however, do so incessantly.

  20. #20 Mike
    October 28, 2009

    @albert rogers

    Well, Dawkins’s concept of memes is quite unrealistic as opposed to other models of behavioral and cultural inheritance, because what is transmitted are phenotypic traits, namely behavioral dispositions and the like. Memes are thought of as discrete, and having their own fitness-values. But it’s just not realistic to individuate something as a unit of cultural transmission. In some cases (emulation learning) you have inheritance of complex, holistic behavioral schemata, in others (imitation learning) you have rather modular aspects of a complex behavior that can be inherited from diverse sources to create the behavioral phenotypic trait in the organism. And if that organisms becomes the source of imitation or emulation learning for the next generation, it is passing on (holistically or in a modular fashion) a behavior that can be composed of modular aspects from multiple cultural ‘teachers’, yet can be taught as a whole. That’s where the concept of an individualized meme breaks down. Furthermore, multiple neurocomputational, neurophysiological and behavioral realizations of a single culturally/behaviorally inheritable trait make the concept of “a meme” completely unrealistic.

    If you’re interested in behavioral and cultural inheritance, I can suggest the works of Boyd & Richerson, Henrich, Jablonka and others. Far more scientific, and in the case of Boyd & Richerson complete with mathematical models.

    But apart from that, what is inherited is not just genes, but also epigenetic markers, morphological traits through ecological inheritance (when an environment shapes the organism in its developmental plasticity in a certain way), as well as certain environmental features produced (or modified) by the former generation that radically change the fitness-profiles of the offspring (like termite-mounds etc.. the concept is called ‘niche construction’).

    Furthermore, I don’t think its clear that the view that ‘genes “caring” only about their own multiplication’ is in all cases an adequate view of what’s going on in evolution, much less the only adequate perspective. That’s partly because of epigenetics, partly because of behavioral and cultural inheritance, and partly for reasons of group selection. There is, e.g. a good point to be made about taking the perspective of the individual as being interested in the persistence through time of its traits, which can explain horizontal transfer of cultural/behavioral traits as well as genetic inheritance.

    @InfuriatedSciTeacher

    That being said, credentials or not I don’t buy the statement that there is no genetic effect whatsoever. The traits are inherited via genes, and the expression of those is impacted by a number of other factors.

    Well, if you’re excluding cultural/behavioral inheritance (which happens also in species that I would not judge to be able to anything we would recognize as ‘thinking’), then I can agree about 98%. I think it’s all about how you individuate phenotypic traits. Naturally, if we take behavioral schemata as phenotypic traits (which makes absolute sense), then they can arise purely by behavioral/cultural inheritance, without the trait itself (as opposed to the physiological requirements for being able to acquire the trait) being influences by genes.

    I say 98% because shaping of the course of developmental plasticity can lead to traits that are not genetically encoded depending on how you individuate traits. Of course, since it’s developmental plasticity, it can be said that this is controlled by the genes – but if one is prepared to call something a phenotypic trait that arises through developmental plasticity in certain environments but not in others, given the same genes, then it seems to me one can say that this is a trait that is defined without a genetic component. But I guess that’s more or less semantics (not entirely unimportant, though).

  21. #21 piker
    October 28, 2009

    So what are the semantic parameters of the concept of mutual exclusivity as applied to behavioral schemata?

  22. #22 David Sloan Wilson
    October 29, 2009

    Here are brief replies to the comments on this post before moving on.

    Jesse, InfuriatedScienceTeacher, and others anticipate my discussion of individualism and cultural influences in T&R VI.

    Justin correctly states that all levels of selection (as conceptualized by multilevel selection theory) result in evolution at the level of genes and are therefore examples of “gene selfishness” (as conceptualized by selfish gene theory). This means that the concept of genes as replicators is no argument at all against group selection. Dawkins can say that now in The Devil’s Chaplain (and also The Extended Phenotype), but in the Selfish Gene he described the replicator concept as a drop-dead argument against group selection. I call this “naive gene selectionism” in a future installment and it is still widespread. Even Dawkins still clings to it, as you will see.

    Piker points out that game theory models can’t go beyond the conditional strategies that are entered into the models. There are models in which the strategies are created in a more open-ended fashion (e.g., through genetic algorithms), but this does not alter the basic dynamic of within- vs. between-group selection.

    Bob asks if there is a new and legitimate group selection that differs from an old group selection that deserves to be rejected. The short answer is no. Wynne-Edward’s thesis that populations might evolve to regulate their size to avoid overexploiting their resources is perfectly reasonable. His error was to assume that it happens all the time and to interpret a vast array of social behaviors as examples. Today we can say more modestly that it happens in some cases but not others. To see a beautiful demonstration of Wynne-Edward’s thesis in an experimental microbial population, read the article by B. Kerr et al titled “Local migration promotes competitive restraint in a host-pathogen “tragedy of the commons” (Nature Vol 442|6 July 2006|doi:10.1038). Characteristically, they describe exactly what Wynne-Edwards was talking about but don’t cite him or use the word group selection, even though they are explicit about selection within and among groups in every other respect–a widespread problem, as I will document in future installments.

    Allen MacNeill (Hello, Allen!) makes some good points about simplistic models in which genes code directly for phenotypic traits, vs. more complex genotype-phenotype relationships. The addition of complex gene-gene and gene-environment interactions introduces a whole new world of possibilities for multilevel selection that weren’t even remotely on the radar screen when group selection was rejected in the 1960’s. Laboratory experiments on group selection show this especially well, as I will describe in a future installment.

    To James: Dawkins knows about me but thinks that he can settle the issue by throwing a tantrum, as you will see. He refuses to answer people who try to organize a debate.

    Mike and InfuriatedScienceTeacher discuss the point made by Allen. It’s true that complex phenotype-genotype relationships can still result in standard heritable phenotypic variation–evolution as Darwin thought about it, knowing nothing about genes–but they can still make a big difference for multilevel selection. For example, when genes code directly for behaviors, phenotypic variation among groups is directly proportional to genetic variation among groups. With complex genotype-phenotype relationships, a small genetic difference among groups can lead to large phenotypic differences. I and others have written numerous academic articles on this subject.

    Becka raises two points: First, that biases can go in both directions but there is still a fact of the matter, regardless of biases. This anticipates the conversation following T&R VI. Second, that group selection has already become accepted and that I might be fighting a war that is already over. I do know that a sizable proportion of the evolution community has fully accepted multilevel selection and that the trajectory is strongly in the direction of increasing acceptance. However, there is still a big mess to clean up, which is the subject of T&R VII. I’d be delighted if Becka was right and I was wrong!

    Thanks to Albert for defending the gene’s eye view. As I briefly state in my reply to Justin, the status of genes as replicators, or the “fundamental unit of selection” was treated as a definitive argument against group selection. If it it turns out to be nothing of the sort–if group selection results in gene frequency change just like any other level of selection–then the entire “group selection doesn’t work because the gene is the unit of selection” argument collapses. Furthermore, what I call the original problem doesn’t require groups to be tight-knit units like individuals. If a behavior such as altruism is selectively disadvantageous within groups, but nevertheless evolves in the total population, then there must be a selective advantage at a larger scale regardless how loose or fuzzy the groups are. These points will be made in more detail in future installments.

    InfuriatedScienceTeacher and Mike introduce the subject of cultural evolution, including memes vs. other models. There’s much to discuss, but this comment on comments has grown long enough!

  23. #23 piker
    October 29, 2009

    Dr. Wilson,
    But you haven’t commented on whether an organism using altruism as a conditional strategy also has the option of using selfishness as a conditional strategy, and if so, at what point does an organism use one of these strategies so exclusively that it becomes, in effect, unconditional and measurable or identifiable as that individual’s dominant trait – so that for purposes of your theories, the individuals with that trait achieve dominance in a group at one level, and their progeny, who purportedly inherit that trait, will as individual trait carriers, lose that dominance at another level where the particular strategy has (according to the models) lost its effectiveness.
    In other words, why is it posited that the groups are seemingly selecting for individual organisms that carry the particular trait, rather than selecting for the traits that will then come to the fore in the new alignment of individuals? So that in the latter case, we would not have so much of a change in the phenotype as more simply one of the phenotype adapting its behavior to the group culture?
    Put another way, there’s clearly a mutual exclusivity paradox involved when you tie the effectiveness of the trait to the individual that carries it, rather than tie the effectiveness of the individual to the choice of that trait in a particular group dynamic.

  24. #24 Bob O'H
    October 29, 2009

    Bob asks if there is a new and legitimate group selection that differs from an old group selection that deserves to be rejected. The short answer is no.

    1. That’s not what I asked. I’m amazed that you twist my question around like that – it makes me uncertain about your ability to summarise other peoples’ views correctly. My question was predicated on there being a difference between the old and new forms of group selection: this seems to be accepted by almost everybody in the field.
    2. My books are in Helsinki waiting to be shipped here, so I can’t check this up, but my recollection is that Wynne-Edwards argued that individuals would give up reproductive success “for the good of the group”, but there was no accounting for conflict between the different levels at which selection could act: the assumption was that the good of the group won out.

    As I briefly state in my reply to Justin, the status of genes as replicators, or the “fundamental unit of selection” was treated as a definitive argument against group selection.

    Can you clarify what you mean here? Are you equating the levels at which selection and replication act?

New comments have been disabled.