Evolution for Everyone

Most people are prepared to admit that we are influenced by our cultures in ways that we don’t understand. As a proverb puts it, the hardest thing for a fish to see is water. Part of the “water” of Victorian culture was an assumption of European superiority. Darwin was progressive for his time but even he was repelled by the “savages” of Tierra del Fuego. When Victorians attempted to view racial and cultural diversity through the new lens of evolutionary theory, some argued that the different races are different species, with Africans closer to the apes. Others argued that we are all one species but that cultural evolution runs along a single track, from savagery to civilization, so that the humane thing to do was make everyone else more like Europeans. Only in retrospect can we look back and see that not only are these theories wrong, but they don’t even follow straightforwardly from evolutionary theory.

What is the water of our culture? I would like to nominate individualism. Individualism is the belief that individuals are somehow a privileged level of the biological hierarchy; that explanations framed in terms of individual action are somehow more “fundamental” than explanations framed in terms of social action; that individual self-interest is a grand explanatory principle that can explain all aspects of humanity. For many people, these beliefs seem like common sense. Water always does.


It wasn’t always that way. Consider the following passage from the social psychologist Daniel Wegner:

Social commentators once found it very useful to analyze the behavior of groups by the same expedient used in analyzing the behavior of individuals. The group, like the person, was assumed to be sentient, to have a form of mental activity that guides action. Rousseau (1767) and Hegel (1807) were the early architects of this form of analysis, and it became so widely used in the 19th and early 20th centuries that almost every early social theorist we now recognize as a contributor to modern social psychology held a similar view.

Even in Darwin’s time, the Russian naturalist and social theorist Peter Kropotkin accused evolutionary theory of being biased by the individualism of British culture, which made competition seem more commonsensical than mutual aid. Even so, Wegner’s passage documents that something happened in the middle of the 20th century that made our culture even more individualistic than it was before. Margaret Thatcher’s notorious quip in 1987 that “There is no such thing as society. There are individual men and women, and there are families.” would have boggled the minds of the Victorians!

Against this background, when evolutionists rejected group selection in favor of “the theory of individual selection” in the 1960’s (see T&R IV), they were just swimming with the other fish. At roughly the same time, a position known as “methodological individualism” became dominant in the social sciences and radical individualism became the dominant position in economics. These parallel events did not take place because scientists were talking to each other across disciplines and changing their views in a coordinated fashion. Much as scientists might like to think otherwise, their formal theories were simply reflecting a larger cultural sea change.

What exactly was this sea change? I would love to know the answer to this question and urge historians of culture and science to study it, or to contact me if they already have. Nazi Germany and the cold war with Communism probably had something to do with it. With Ayn Rand there was a direct connection, since she came from Russia and had a zeal for free-market economics that rivaled religious fundamentalism, as I recount in a chapter of Evolution for Everyone titled “Ayn Rand: Religious Zealot.” Another factor might have been the allure of reductionism; the belief that lower-level explanations are somehow more fundamental than higher-level explanations.

Regardless of the reasons, the hyper-individualism that took hold during the second half of the 20th century became the cultural “water” for the theory of individual selection in evolutionary biology, which portrayed everything that evolves as a variety of self-interest. The zeal associated with hyper-individualism in general might also explain the zeal with which some individual selectionists argued their position, as I documented in T&R V.

Thinking about science as a culturally influenced activity is a tricky business. On one hand, everyone is prepared to admit the abstract possibility and to see it clearly for past examples, such evolutionary theories of racial and cultural diversity in Darwin’s day. On the other hand, most scientists don’t like to admit the possibility for their own theories. To make matters worse, some scholars who study science as a culturally influenced activity conclude that science therefore has no more truth value than any other cultural belief system, such as astrology.

The hardest ground to capture, it seems, is the middle ground. Science remains the best cultural system we have for holding each other accountable for our factual statements–vastly better than astrology, for example. But scientists are full of biases, many beneath their conscious awareness, just like everyone else. That’s why a cultural system is required to overcome individual biases. The cultural system does a pretty good job but is especially prone to failure when everyone shares the same biases. Then there is nobody around to propose and defend an alternative hypothesis. The best solution would be to make sure that scientists are as culturally diverse as possible and to employ an army of scholars to scrutinize current scientific theories for cultural bias in a constructive way, sharing the belief that at the end of the day there can be an accumulation of knowledge that deserves to be called factual.

Factual matters are definitely at stake for the issues associated with group selection. What I called “the original problem” in T&R II remains a fact. It is simply the case that “for the good of the group” traits are often locally disadvantageous. If they are to evolve at all, a selective advantage must exist at a larger scale. If group-level selection is sufficiently strong, then “for the good of the group” traits can evolve in the total population, despite their selective disadvantage within groups. Determining the relative importance of within- vs. between-group selection is a straightforward matter of theoretical and empirical research. Even though hard work might be involved, it should be possible to determine the facts of the matter.

What I called The Great Reckoning in T&R IV appeared to deliver a verdict: group-level selection is almost invariably weak compared to individual-level selection. As George C. Williams put it, “group-level adaptations do not, in fact, exist.” Despite the appearances of decades, he was massively wrong.

To be continued.

Comments

  1. #1 Bob O'H
    October 28, 2009

    Factual matters are definitely at stake for the issues associated with group selection.

    Indeed, so could you back this post up with some facts and data, to show that the move towards individualism by the likes of John Maynard Smith were because of political and social considerations, not scientific?

  2. #2 bob koepp
    October 28, 2009

    I agree that individual selection was _over_ emphasized in evolutionary biology, as was adaptationism. But I’m very skeptical of easy explanations in terms of the cultural millieu — shades of the zeitgeist! In fact, given the broadly communitarian social-political ideology espoused by many card carrying individual selectionists, I think you’ve got a serious problem with false consciousness brewing here.

  3. #3 becca
    October 28, 2009

    Yeah, I’m not sure you’ve described the water correctly (current or past).
    I’d argue that throughout the past few hundred years, if not longer, rational individualism was a critical component of the intellectual environment. It’s not new. (“It is vain to talk of the interest of the community, without understanding what is the interest of the individual” Jeremy Bentham)

    Granted, that’s a more nuanced view than the kind of rabid ultra-individualism that has gone in and out of fashion in this last century (e.g. Ayn Rand garbage: “In a capitalist society, all human relationships are voluntary. Men are free to cooperate or not, to deal with one another or not, as their own individual judgments, convictions and interests dictate.”)

    Anyway, the real problem with group level adaptations is that the further you get from the individual, the more difficult it is to provide even indirect evidence. Most biologists can understand easily how seemingly detrimental traits are advantageous to the individual under some circumstances (sickle cell being the cannonical example); so it’s not the complexity of explanation required that bugs us.
    I don’t even think we have a strong (cultural or other) bias against hypotheses that involve detrimental traits being advantageous to the family-group (there’s at least widespread recognition that the grandmother hypothesis of menopause is reasonable enough to test). But it seems to me that the further you get out from the individual, the more explanations sound like ‘just so’ stories. I have no trouble attributing any number of traits that look detrimental to the good of the group- but I haven’t seen a lot of cases where the evidence is compelling that it’s not just a reflection of how the mutational hand was dealt.
    For example, it could easily be that all those crazy CYP P450 polymorphisms evolved as we were co-evolving with plants so that at least *someone* would likely be able to metabolize any given innovation in the plant toxin arena. It’s equally possible some of them are plain deleterious now (e.g. in drug metabolism), but there was a slight positive selective pressure for the individual that is highly dependent on environment (i.e. we have zillions of alleles because we eat zillions of chemicals). These are very similar-looking, how do you tell the difference between individual and group selection? Or does teasing out individual vs. group selection depend on mutual exclusion (i.e. we can only prove group selection exists in cases where we can document that the individual effects were deleterious as they evolved?)

  4. #4 bcoppola
    October 28, 2009

    We are all individuals. Just like everyone else.

    (Sorry. Someone was bound to say it.)

  5. #5 Shane Horan
    October 28, 2009

    I think Wilson’s argument is proceeding in the wrong order. If it turns out individual selection zealots are, in fact, wrong, then it’s certainly interesting to find out why (not being able to see the cultural water, etc.). But if they’re correct, are their cultural prejudices relevant? (Are they correct in spite of their prejudices?) To echo Bob O’H’s call, can we first answer the empirical question: the how of the zealots’ error before the why.

  6. #6 Bob O'H
    October 28, 2009

    Shane – I mentioned John Maynard Smith because he was (a) a big figure in these debates, and (b) a communist. I’m interested to see how the accusation of individualism will stick.

  7. #7 abb3w
    October 28, 2009

    Cultural preconditioning is a reasonable suspect for why individuals might lean in a particular direction. On the other hand, the probability bias of the culture does not have complete control over the probability of the response.

    Within the eukaryotes, the demarcation of “individual” is sharper in terms of space and time membership functions than groups or species. While one may define the latter, the edges tend to be fuzzier than the perimeters given an individual by birth, death, and skin/hair… although that does not mean the perimeters of a group do not exist.

    The notion thus wanders through my mind that emergent traits which help sharpen the perimeter (EG: allow telling who is and is not a member of whatever group) would contribute to selection appearing to operate at that level. I also intuit that there may be some connection to traits governing how big the group can be and continue to function as.

    How do you define existence of a “group”?

  8. #8 Guy
    October 28, 2009

    Responding to bcoppola:

    We are also all coherent, functional groups of lower level agents (e.g., cells, genes), and we are all members of higher level groups that depend in part on the roles we play.

    I take you point to be that we (the ones holding this conversation) are individuals so it is natural that we would be biased toward thinking of individuals as fundamentally important. I suppose that hypothetical conversations at other levels of organization would be similarly biased as to which level is the most fundamental. On the other hand, I am persuaded by the view (I am influenced by Stan Salthe on this issue) that building a hierarchically layered complex system like we see in biology starts at the bottom and the level of primary influence (control?) moves up the hierarchy as new levels or organization emerge on top. My gut feeling is that the individual level of organization was very successful in taming antagonism among lower levels (e.g., among genes or cells). For example, the evolution of Mendelian segregation effectively leveled the playing field among alleles because every allele gets passed into gametes in equal frequency (meiotic drive is the great exception here). Social organization appears to represent the emerging next level, which may have become “primary” in eusocial systems, but which remains relatively unstable for ordinary social systems.

    This is all just thinking out loud on my part, and I’m afraid I went way beyond a reasonable response to a short and possibly cheeky post. It sparked a connection to these ideas for me and I wanted to share them on this blog. 🙂

  9. #9 E.V.
    October 28, 2009

    Wasn’t the Prime Minister’s remark a rhetorical device to dispel the notion of elitism and subliminal caste system? In effect that individuals are worthy and we’re all a part of the human family and not hierarchies of presumed social classes?

    (Incidentally, I’m not a fan of Thatcher)

  10. #10 daedalus2u
    October 28, 2009

    There is a lot of very sloppy thinking here. Where did developmental plasticity come from, if it was not inherited? If it is inherited it can be selected for. If it is not inherited, then it can’t persist.

    Developmental plasticity is a property of a genome, not something that exists independent of a genome. A particular genome may produce multiple different phenotypes depending on the environment but all of those phenotypes derive from the same genome, and the descendents of that genome expressed as the particular reproducing phenotype inherit some of that genome and so inherit some degree of developmental plasticity as that genome encodes it. A genome is not just genes, i.e. DNA that codes for proteins. There is non-coding DNA too, and for most of the genome it is not known what it does. Some of that DNA with unknown function has been inherited with absolute fidelity from fish to mammals. What does it do? No one knows, but it must do something pretty important if there are no extant mutant copies of it.

    When a bee develops into a queen because it is fed royal jelly, that developmental plasticity is inherent in the bee genome. Feeding a grasshopper royal jelly won’t turn a grasshopper into a queen grasshopper because a grasshopper doesn’t have the physiology coded for by its genome to support developing into a queen after being fed royal jelly.

    Similarly, epigenetic programming of DNA only happens because there is physiology coded for by DNA that selectively programs DNA epigenetically. That is mostly what happens during differentiation, cells program their DNA, so that only some of the genes will be expressed. That is why brain cells look like brain cells and liver cells look like liver cells even though all types of cells all have the same DNA. The developmental control mechanisms that control proliferation and differentiation during embryo development are mostly not understood. That the physiology to do that exists and is coded for by DNA is undisputable.

    No becca, multiple P450 polymorphisms didn’t evolve so that “someone” would be resistant to random plant toxins. P450 polymorphisms arose just like all other mutations, they happened. If the polymorphism was “bad”, so that it didn’t make or metabolize an essential compound, then that phenotype died and had no ancestors with that P450 polymorphism. Unless that phenotype had a redundant P450 polymorphism that did make or metabolize the essential compound, then the new mutant didn’t need to. A gene doesn’t need positive selection to persist, it needs to not have negative selection (i.e. it needs to not kill its phenotype) to persist.

  11. #11 David Sloan Wilson
    October 29, 2009

    I enjoyed the discussion of his post. Here are some quick replies before moving on.

    Several of you commented that what motivates a given idea does not bear upon its factual validity. I couldn’t agree more and never meant to imply otherwise. Even so, the fact that science is a culturally influenced process needs to be taken seriously. There must be a better way of incorporating it into the scientific process than ignoring it until the errors become obvious fifty or a hundred years later.

    Bob and Becka caution against easy explanation, but so do I! In my post, I say that it will require an army of scholars who are sympathetic with scientific objectives, in contrast to many people who identify with the field of science studies. Deep scholarship is required. Scientists need to be very careful about what was said and meant in the past, in addition to the present. I am trying to provide that service for the group selection controversy, in this series of blogs and in my academic writing. I am held accountable by other historians, sociologists, and philosophers of science such as Elliott Sober, Samir Okasha, Mark Borello, and others. When this kind of deep scholarship replaces the patriotic history and the “tower of Babel” problem that I describe in T&R VII, the truth and reconciliation process will have succeeded.

    I disagree with Becka that the further away you get from the individual, the more indirect the evidence becomes. Especially for single traits such as altruism, aggression, sex ratio, feeding rate, etc., showing that they are selectively disadvantageous within groups and evolve only by virtue of the differential contribution of groups to the total gene pool is part of the same exercise. Keep an eye out for a lovely demonstration of this for sexual conflict in water striders that is about to appear in Science magazine, by my most recently minted PhD student Omar Eldakar.

    abb3w asks how groups are defined. The short answer is that they are the sets of individuals that influence each other’s fitness with respect to the particular trait whose evolution is being considered. All evolutionary models of social behavior must define groups in this way, regardless of what they are called. If you want to know if social behavior X evolves, you need to determine the fitness of a X individual vs. a not-X individual. Since X is a social behavior, the fitness of an individual will be determined by its own phenotype and the phenotypes of the other individuals with which it interacts. Those other individuals define the group. Therefore, all groups must be defined in reference to particular traits. I coined the term “trait group” in 1975 to emphasize this fact, but it is implicit in all models of social behavior. For example, it defines N in N-person game theory. See my book with Elliott Sober, Unto Others, for a fuller discussion.

    Guy anticipates the subject of major transitions, which shows that the individuals of today were the groups of past ages. This will be the subject of a future installment.

    Daedalus2u raises issues about complex genotype-phenotype relationships, which go beyond simple gene-thinking in some respects but still qualify as standard heritable phenotype variation in other respects–the way that Darwin thought about it, knowing nothing about genes. Lots of stuff to discuss, not but in the space of a comment on comments!

    Thanks once again for the intelligent, vigorous, and respectful discourse.

  12. #12 Mike
    October 29, 2009

    @daedalus2u

    It seems to me you’re missing the point(s) (Some sloppy thinking perhaps?).

    Developmental Plasticity comes in two forms. The first is what you describe – the activation of a developmental program specified in the genome. This is called (with Sandra J. Smith-Gill) ‘developmental conversion’. Here, the information for the phenotypic traits is indeed encoded fully in the genome and only takes a cue to activate it.

    But there’s a second type – called (again with Smith-Gill) ‘phenoytypic modulation’. Phenotypic Modulation is when the environment modifies rates/degrees of expression of (parts of) the developmental program. You’re overall problem (including with epigenetics) seems to be that you’re thinking the fact that information in the genome being necessary is sufficient for dismissing all these forms as genomic effects. It is not, because we need to think in terms of the information that is necessary and sufficient for producing the phenotypic trait. And this is not specified in the genome in this case. Of course the genome is necessary – but the information contained in it is not sufficient.

    Furthermore, developmental plasticity need not necessarily be encoded in the genome in order to persist. That is quite simplistic thinking. It can also be due to the fact that the transcription and construction mechanisms cannot – physically – specify every single aspect of the product. This is a rather simple truth because the physical information contained in the genome is less than the information in the products of genetic expression. You have molecular bonds, interactions between products, between products and the environment, spatial orientation, chemical relations etc. Developmental plasticity means simply that the effects of the expression of the genome can be influenced by the environment. This is simply a fact of complex biological systems, not something that needs to be encoded and inherited. It is ‘inherited’ in virtue of organisms being complex biological systems participating in complex interactions with the environment. What is encoded in the genome pertaining to developmental plasticity is – concerning developmental conversion – the alternative developmental programs.

    As for epigenetics – again, the point is not that the genome is necessary, or that the mechanisms responsible for epigenetic encoding are themselves constructed from information in the genome. The point is that information that specifies the phenotypic traits (of cells or organisms – there are epigenetic effects on all levels within the organism) is not entirely genomic, but that crucial information is epigenetic, i.e. present in chromatine markings, RNA-interference and spatial structures with specific physical and chemical properties that guide template-copying processes.

    I welcome the criticism – but I think it’s a little uncalled for to claim that “there’s a lot of sloppy thinking” in what you criticize.

  13. #13 Guy
    October 29, 2009

    Responding to deadalus2u:

    I disagree with your fundamental point.

    “Developmental plasticity is a property of a genome, not something that exists independent of a genome.”

    I recognize that this assumption has underpinned most of the literature on the evolution of phenotypic plasticity, but I hope there will be a critical evaluation of this assumption soon. My personal view is exactly the opposite of your premise. As I see it, developmental plasticity (not limited to adaptive plasticity) is an inherent aspect of any dynamical complex system (dissipative system). I see adaptive evolution of the genome as usually constraining plasticity to improve the chance that a functional phenotype will develop (developmental canalization). I acknowledge that there are instances in which selection molds the genome to yield mechanisms guiding adaptive plasticity in the face of particular predictable aspects of environmental variation (e.g., getting a sun tan). However, that is in my opinion a very small piece of the phenotypic plasticity story.

  14. #14 beccas
    October 29, 2009

    @daedalus2u
    Obviously developmental plasticity is part of the genome. Which raises the first-derivative question- what is the rate of change of the gene/genome/organism/species/group? What are the selective pressures on the rate?
    That’s partly what the P450 example is about, but it’s also partly about what the selective pressure is *on*.
    There is *genomic* plasticity in the P450s- could it be because there is selective pressure on the species to have some diversity in these genes? You’re still thinking selection only applies to the individual, but I suspect the P450s are as heterogenous as they are in part to allow us *as a species* to be able to adopt to different environments.
    (obviously regardless of whether selection can occur on the species level, mutations do not generally [possible microbial exceptions aside, perhaps] arise *for* any purpose. Sloppy writing, yes, but I do understand the distinction)

    My problem is that I don’t see how you *prove it*. Dr. Wilson, I think those examples assume that any persistent trait that is individually deleterious must be beneficial to the group. That’s indirect.
    My real trouble (illustrated by the P450 example) is if a trait is advantageous to the individual and advantageous to the group, how do you figure out where the selective pressure is coming from?

    “I take you point to be that we (the ones holding this conversation) are individuals so it is natural that we would be biased toward thinking of individuals as fundamentally important.”
    We disagree.
    😉

  15. #15 daedalus2u
    October 29, 2009

    How are you defining “information” in a completely developed phenotype? There is never sufficient information from the genome and from the environment to completely specify the information in the phenotype. I think there is (or should be) general agreement on this. If identical genomes were developed into phenotypes under “identical” environmental conditions, each and every one of those phenotypes would be different. They would not contain “the same” information.

    If you take a coin and flip it a million times, the resulting sequence has more information than the coin had. It takes a million-bit long binary string to encode the information from the million flip sequence. Where did that “information” come from? It came from the same place that much of the “information” in a fully developed phenotype comes from. A single genome can support the development of zillions of phenotypes, many more phenotypes than there are bases in the DNA of that genome. We know this is correct because there are more neuronal connections than there are DNA bases. Just as a coin can support a very large number of flip sequences, so a genome can support a very large number of phenotypes. It is unreasonable to expect the genome to specify each phenotype to an arbitrary degree just as it is unreasonable to expect a coin to specify which coin flip sequences it can be used to generate.

    Development is a process of many coupled non-linear parameters. It is inherently chaotic. Differential effects early in the process become chaotically amplified and produce macroscopic effects later, sometimes known as the “butterfly effect”. Identical twins are not identical.

    I read the 1983 SJSG paper in Amer Zool. I think her attempt to distinguish between what she calls “developmental conversion” and “phenotypic modulation” is arbitrary and non-physiologic (and I think wrong). I think a lot of her problem stems from her assumption of homeostasis (an assumption which is wrong). There is no such thing as homeostasis. No physiological parameter is regulated to be “constant” or “static”, physiological parameters are only regulated according to the control paradigms that are coded for by the genome.

    She appeals to the idea of a “development vector” as if a certain and specific path of development to reach a specific phenotype was a goal of the process of development which is not correct. There is no goal, there is only the path. I think the idea of goal oriented development is akin to the idea of homeostasis. I think the appearance of “developmental goals” and of “homeostasis” are both similar artifacts of human perceptions. If something appears to happen as a consequence of a process, then humans have a very strong tendency to anthropomorphize that what “happened” was the “goal” of the process. The path may be complex, and there may be downstream parts of the path that can “fix” upstream missteps, but there is no goal.

    The genome doesn’t “know” anything about a developmental goal, all it “knows” is the physiological processes it is coded to perform. It goes about those chemical and physiological processes and what results is the developed phenotype. To the extent that there is what is called developmental plasticity, that plasticity is coded for in the genome because the genome doesn’t specify a different level of developmental specificity (higher or lower plasticity). In other words, much of the complexity of the final developed phenotype cannot be coded for in the genome because there is insufficient information present in the genome to do so. The genome specifies a path, and the path leads to multiple phenotypes.

  16. #16 piker
    October 29, 2009

    daedalus2u,
    Biological systems “learn” to take advantage of chaos, in that there is a predictable consistency to chaos, informally known as the law of averages. The flipped coin, however, has no ability to learn from its experiences. So your analogy there basically fails.

  17. #17 Bob O'H
    October 29, 2009

    Bob and Becka caution against easy explanation, but so do I! In my post, I say that it will require an army of scholars who are sympathetic with scientific objectives, in contrast to many people who identify with the field of science studies. Deep scholarship is required. Scientists need to be very careful about what was said and meant in the past, in addition to the present. I am trying to provide that service for the group selection controversy, in this series of blogs and in my academic writing.< (blockquote>
    I also asked for some evidence for your position: without it, it looks like you’re just poisoning the well. So how about providing the evidence? You claim to be doing the scholarly work, but it isn’t evident in this post.

  18. #18 Mike
    October 29, 2009

    @daedalus2u

    It’s not about the totality of information in the developed phenotype – it’s about the information necessary to produce that phenotype. Of course the genome is never strictly enough. But for normal development according to a developmental program specified in the genome, the crucial information is in the genome. But concerning epigenetic modulation of genomic programs, the information necessary to produce the product is not genomic.

    For the concept of information in biology, I recommend two papers by Jablonka:

    “Information: Its Interpretation, Its Inheritance, and Its Sharing” in Philosophy of Science, 69 (December 2002) pp. 578-605

    and

    “The evolution of information and the major transitions”
    in Journal of Theoretical Biology, 239 (2006) 236-246

    Concerning developmental plasticity, I don’t see how what you write establishes that the distinction between the activation of an alternative developmental program and the modulation of the expression of a program is arbitrary. Actually, I think that distinction makes a lot of sense. Of course there’s no teleology, but the concept doesn’t necessitate any.
    Furthermore, I don’t see how the treatment of homeostasis changes the validity of that distinction… (?)

    (On a side note – yes, no (non-trivially specifiable) parameters in an organism are maintained at exactly one level. But a vast number of parameters need to be kept in a certain range of values for the organism to function properly, and a lot of the actions of the organism and its organs have the function to maintain the range of values. Metabolism and endocrinology are primarily about that.)

    In any case, my main point was that we can think of straightforward expression of a program in the genome, and in comparison to that, epigenetic modification of programs adds information that is absolutely crucial for the specification of the phenotype or phenotypic trait in question. At least that was what I meant to express in my last post.

    Best,
    -Mike

  19. #19 daedalus2u
    October 29, 2009

    Mike, I think I understand what you are trying to say, but I think that you have no way to precisely define what you mean by a “phenotype” that has been produced by a “genetic program” and how much information of what type is necessary and sufficient.

    If we do a thought experiment, and grow a million clones of a single genome into a million different individuals and then grow those million different individuals in a million more/or less different environments; how many “phenotypes” are present? I would say a million, I think that you would have to look at each one and use some definition of a “standard phenotype” and then count how many there are of each different “standard phenotype”.

    The defining properties of those “standard phenotypes” are arbitrary (to some extent).

    I think you are using an ambiguous definition of “information”. MarkCC has a number of posts on information vs. meaning over at Good math bad math.

    http://scienceblogs.com/goodmath/2009/09/information_vs_meaning.php

    The “information” that results in the output of an epigenetically programmed bit of DNA which is epigenetically programmed in response to an environmental stimulus is not only in the stimulus. It is in the DNA, the stimulus, and also (and most importantly) in the relationship between the stimulus and the DNA change as transduced by the physiology that produces the epigenetic programming. The details of how physiology does this are mostly unknown, but that it is done, and done controllably is not in doubt. If it is done controllably, there is physiology that does it. That physiology came from somewhere, and is instantiated in the physiology produced by the genome.

    I was only able to see the abstracts of the articles you linked to. I think he makes the mistake in thinking that the “information” is in the environmental trigger that initiates the epigenetic programming of DNA. I think that is not correct. I think the “information” is already in the genome that produced the “phenotype” that produces gametes with a certain type of epigenetically programmed DNA in response to certain environmental stimuli. And I am not trying to be cute.

    Epigenetic programming of DNA is not simple. It obviously requires multiple genes to do it right (even if we don’t know what those genes are or how they are controlled). The right enzymes need to be expressed at the right time, in the right place in response to the right environmental stimulus which then need to methylate the right genes and not methylate the wrong ones. How many different epigenetic programming subroutines are there? At least a few for each different type of differentiated cell. A process that uses multiple genes in concert is not a process that can “just happen”.

    I don’t have a problem with thinking about development as a “program”, but it is a “program” of which we have near complete ignorance of the details. We have insufficient information to know if any particular phenotypic outcome is a “feature or is a “bug”. I think in the sense that SJSG is using the two terms, “developmental conversion” results in “features” and “phenotypic modulation” results in “bugs” (mostly).

    A lot of what I am working on now is in stress responses as triggered and regulated by low nitric oxide. Low NO is a stress response and many stress responses are triggered by low NO. Many stress responses that are triggered by NO are not appreciated as adaptive responses because when they are continued for a long time there are adverse effects. An example is something like post traumatic stress disorder, PTSD. It is characterized by insomnia, anxiety, hyper vigilance, avoidance of stimuli associated with the trauma, and a hair-trigger violent response. If you were living in a war-zone, those would be “features” and not “bugs”; features that could save your life in a war-zone but are maladaptive in times of peace. Is PTSD a “developmental conversion” or a “phenotypic modulation”?

    I think that essentially all of the “complex genetic disorders”, such as obesity, heart disease, diabetes, autism, hypertension, Alzheimer’s, other neurodegenerative diseases are not “bugs”, they are “features”. The reason those doing genetic studies are having such difficulty finding “disease genes” is that these are not diseases. They are “features”, that have become maladaptive when continued for too long.

    You are right that ”a vast number of parameters need to be kept in a certain range of values for the organism to function properly” the problem is, we don’t know what those values are and if they are constants or variables. The idea of homeostasis posits that they are all the value “at rest”, but we know that is not the case.

    When Professor Wilson gets the group selection mess cleaned up, I hope he will address the homeostasis mess (which I think is even worse).

  20. #20 Mike
    October 30, 2009

    @daedalus2u

    re: standard phenotype
    I basically agree – but all those clones will have many things (phenotypic traits) in common, and will have gone through essentially the same developmental program (given the same epigenetic information), but have gotten different environmental inputs, and thus we get different phenotypes. I think that much can be agreed upon

    re: information
    Thanks, I’ve read the articles. But since I’ve taken courses in and/or otherwise studied studied logic, information theory (from Shannon information to Kolmogorov complexity etc), they didn’t contain anything really new to me.

    I also agree that the relationship between the environment and the genome via the cellular and organismic mechanisms that determine how the environmental idiosyncrasies will result in epigentic programming is very important. But again, the chromatine markers (taking one epigentic inheritance system as an example) are the bottleneck, the information that is basically added to the information in the DNA affecting transcription/expression. If you kept the chromatine markers but took away the environmental cues and the actions of the mechanisms determining the chromatine markings, you’d get the same change in transcription/expression. And that is the sense I was talking about. That information in chromatine markings was not encoded in the genome (or the parental genome of the mother-cell) before.

    re: articles
    If you drop me a mail at MBauer.MPhil@googlemail.com, I can send you the articles. Btw, the author is a she – Eva Jablonka.

    I guess we just have to agree to disagree where the essential information is ‘located’. But I’m okay with that.

    The stuff you work on sounds interesting – I’ll make sure to take a lot a blog!

  21. #21 daedalus2u
    October 30, 2009

    I don’t disagree that there is information coded on the cromatin after the epigenetic programming has happened. I am saying that the “information” came from other parts of the genome and not from the environment. The epigenetic programming information was stored in the genome in another form (as yet unknown, but likely distributed over enzymes, transcription factors, regulating DNA, and other things, (that is distributed and not encoded in “closed form”, i.e. simple genes)), and was converted by the epigenetic programming mechanism (which is also unknown) from that other form to a form that would be read out as changes in DNA expression (i.e. methylation) following future cell divisions. The environmental stimulus simply provided the trigger that lead to the different epigenetic programming, not the data that is the content of that programming. Different cells (with different genomes) with the same environmental stimulus would exhibit different epigenetic programming.

    To make an analogy, suppose you had a self-replicating robot that made copies of itself. When the robot anticipated a change in the environment its descendents would exist in, it modified their design, for example if it anticipated an iron-poor environment, it used aluminum as a structural material instead of iron. When those aluminum utilizing descendents were transferred to an iron rich environment, they switched back to using iron as a structural material in their descendents. Switching between iron and aluminum may never happen in an existing robot because aluminum and iron can’t be welded together. Would we surmise that the information necessary to switch from a design using iron to a design using aluminum and back came from the environment? Not at all.

    When using aluminum, there is less reliance on the iron utilizing subroutines, so there is less selective pressure to maintain them in high readiness. There will be “genetic drift” away from using iron and toward using aluminum.