How (not) to bring psychology and biology together

darwinby Scientia Salon

[This is a new feature here at Scientia Salon: from time to time we will publish very short entries with the abstracts of selected (by our Editor) papers from the primary literature in philosophy, natural science or social science. Links will be provided to the full published article – which will often be behind a paywall – and whenever possible to a downloadable version of the same. The category identifying these entries will be “notable.” For a list and description of all categories used to organize Scientia Salon entries, see our “About” page.]

This paper, by Mark Fedyk, was published in Philosophical Studies, February 2014. Paywall article here. Free version here.

Evolutionary psychologists often try to ‘‘bring together’’ biology and psychology by making predictions about what specific psychological mechanisms exist from theories about what patterns of behaviour would have been adaptive in the EEA for humans. This paper shows that one of the deepest methodological generalities in evolutionary biology—that proximate explanations and ultimate explanations stand in a many-to-many relation—entails that this inferential strategy is unsound. Ultimate explanations almost never entail the truth of any particular proximate hypothesis. But of course it does not follow that there are no other ways of ‘‘bringing together’’ biology and psychology. Accordingly, this paper explores one other strategy for doing just that, the pursuit of a very specific kind of con- silience. However, I argue that inferences reflecting the pursuit of this kind of consilience with the best available theories in contemporary evolutionary biology indicate that psychologists should have a preference for explanations of adaptive behavior in humans that refer to learning and other similarly malleable psycho- logical mechanisms—and not modules or instincts or any other kind of relatively innate and relatively non-malleable psychological mechanism.

56 thoughts on “How (not) to bring psychology and biology together

  1. Am I really the first? Well, fools rush in . .

    Thanks for an interesting, I think important, paper, well worth the trouble to read, if I interpret it correctly. Would I be close to the mark in thinking that you are reinstating Lamarck – without dethroning Darwin — a process already begun in evolutionary biology and, you claim, even more relevant to evolutionary psychology. Since the Darwin/Lamarck dichotomy is taught in highschool, it would be interesting to hear how you would relate your essay to . . . #Darwin/Lamarck.

    I do wonder, however, if the Tomasello hypothesis is ‘scientific’ at all, or not just common sense. Is there a subtle point here – that science should not exclude common sense? (That an elaborate theoretical edifice that produces the same result as common sense is not thereby confirmed?) How would Cimino and Delton control, as you ask them to, for Tomasello’s account?

    Like

  2. Very interesting topic! I’m reading the paper now but I wanted to point out that David Sloan Wilson (Evo biologists) and Steven Hayes (prominent psychologists) among others have recently pointed out the same thing and in many ways trying to “reboot” evolutionary psychology in a more scientifically adequate way. Much of this work is being done under the heading of “contextual behavioral science”, here is a link to the article explaining the approach.

    http://journals.cambridge.org/action/displayAbstract?fromPage=online&aid=9335540&fulltextType=RA&fileId=S0140525X13001593

    I think it’s a good and much better concept that EP as it is now but it still remains to be seen how effective this version would be.

    Like

  3. It does seem a restate of nature, versus nurture, with a preference for nurture, but if I may make a suggestion, that the problem seems to be one of how they view the process of time. The modular thinking is “forward,” i.e. determined past setting conditions for limited selection/determination of the future, while the plasticity view is the evolving, i.e. plastic state of the present, creating the forms and thus past receding.
    Basically, causality yields determination, not the other way around.
    My thumbnail reading, obviously based on the few modules floating around in my mind.

    Like

  4. The Brain Is A Web, Thus So Is Reason

    When one looks at linguistics, biological evolution (in the most general sense, including eco-systems), neurology, civilization, one does see trees, inasmuch as we see webs with reactive (not to say intelligent) strands.

    The latest news, in neurology is, indeed, that the white matter, made of glial cells, axons, oligodendrocytes, etc. itself reacts (not to say “think”).

    Why should not reason itself be the same?

    Should the brain be according to reason, or reason, according to the brain?

    Once reason has become a web, it has a non-trivial topology (in particular with a genus).
    The genius of genus.

    The next natural question is whether reason has a metric, a geometry, a notion of proximity. Indeed look at the brain, namely, the mind. Some nerve impulses go as far as possible, all along a motor neuron, or along a long axon. However, others go short, and are manipulated, not just at the first axon they meet, but even before this, along the axons themselves.

    Some will fear for reason, as they read these lines.

    But a model exists, in physics. Renormalization.
    As a field strength augments, approaching its putative source, the field itself modifies the effect it is supposed to describe. There is no general theory of how this works (although some field theories are known as “renormalizable”).

    To put it in terms non-specialists may understand, the force varies with, and because of, the force itself. In any case, electromagnetism starts with 1/d^2, and ends up very different.

    Another meta-example is graciously offered by mathematics itself. Mathematics is the field depicting reason itself. However, it does not have safe foundations.

    Category Theory became actually strong, because of the mood that underlays it. Namely that foundations, globally, should not be worried about, while, locally, much progress can done by making them richer (that’s what Grothendieck did).

    So, once again, we see that reason is rich, and can get richer, but it is local, not global.

    Evolution, as understood for most of the Twentieth Century, was driven by chance or weird considerations about reproduction (animals were supposed to prefer so and so because it allowed them to reproduce better… as if they cared!) For example, we were told the Peacock’s tail was there to show to females the beholder was healthy, hence would bear them more children.

    (More refined recent studies show the obvious, just as with eyes on butterfly wings, Peacock tails may rather be scary devices, as anybody who has deployed an umbrella for a lions would know.)

    Chance is here to say, but the big Damocles sword over facile explanations is Quantum Physics itself: the Quantum is teleological, and no doubt impact both genetics and epigenetics.

    This means that the most inner machinery is not just potentially teleological, but really teleological.

    That Biological Evolution did not exploit what is, after all, the most fundamental law of the physical world, as it evolved according to physics for 4 billion years, is, frankly, impossible.

    Like

  5. What other work exists that suggests developmental plasticity as a potential area to explore for evolutionary psychologists? After reading this article I felt the impulse to think it was only a matter of time before someone did which lead me to wonder if they already have…

    Like

  6. astro,

    well, someone has to be the first one to comment…

    “Would I be close to the mark in thinking that you are reinstating Lamarck – without dethroning Darwin”

    No, the paper is about phenotypic plasticity, which though it is sometimes confused with a type of neo-Lamarckism, is actually nothing of the sort. Plasticity is simply a measure of the tendency of a given genotype to produce different phenotypes when exposes to different environments.

    “How would Cimino and Delton control, as you ask them to, for Tomasello’s account?”

    This isn’t my paper, so perhaps we should ask the author…

    imzasirf,

    “Much of this work is being done under the heading of “contextual behavioral science””

    I’m actually increasingly skeptical of David Sloan Wilson’s ultra-Darwinism (he sees evolution everywhere). Interestingly, this would be the second time that the original sociobiology gets re-baptized to make it more palatable. I wonder when people will finally admit that it’s not a well thought out idea and just let it go…

    brodix,

    “Basically, causality yields determination, not the other way around”

    While I may agree with this statement, I’m not sure where you saw that in the paper. The terms “backwards: and “forward” don’t refer to causality or determination here, but simply to whether one makes predictions about the past based on the present or vice versa.

    Philip,

    “Since I think that human brains evolved to be more like general-purpose computers than like special-purpose computers … I don’t quite get what ‘evolutionary psychology’ is meant to demonstrate”

    Well, the issue at hand is precisely the extent to which the brain is a general-purpose computer. We know this isn’t entirely true, since there are anatomic areas in the brain that tend to be associated with particular tasks (language, face recognition, etc.). But for the original EP program to get off the ground, as the author of this paper reminds us, one needs to accept the assumption of massive modularity, which is clearly false, given all we know about brain development.

    Like

  7. An interesting paper, I have often wondered if Ev Psych could get round the Glengyle Castle problem.

    I may have to spend a bit of time to understand the proposed solution in detail though.

    Like

  8. This is the first time I’ve heard of the distinction between the so called ultimate explanations and the proximate explanations. They appear to me as powerful conceptual instruments. I wonder why historians tend to pay little attention to the methods of evolutionary biology.

    SciSal, in regard to: “Plasticity is simply a measure of the tendency of a given genotype to produce different phenotypes when exposes to different environments”, may I ask if it is possible for those expressed phenotypes to have any influence in the genotypes of germ cells and, in so doing, affect the decent of the living being?

    Thank you and Mark Fedyk for sharing with us this paper.

    Like

  9. Massimo,
    ” The terms “backwards: and “forward” don’t refer to causality or determination here, but simply to whether one makes predictions about the past based on the present or vice versa.”
    As I’m seeing it, causality is a process happening in the present, of which determination is the effect. As such the past, as in those events already happened, were a product of their occurrence, not some historical vector, on which the point of the present is merely subjective. So that when we try model this process narratively, i.e., from events already occurred, to the present and future ones, it naturally seems linearly deterministic. Which in this case, would be all those set modules/functions of the brain and thought processes setting the foundation of the current state of the brain.
    While looking at it from a present perspective, the situation is much more fluid and much less linear, so causation doesn’t reduce to sequence, but is much more a dynamic of current energies and their forms interacting. Thus more fluid and plastic.
    Since all we eventually have of past events, is whatever residue they might leave physically, or mentally, it only leaves a very limited and subjective set of information and so trying to place the physical reality of all that is present, on this narrative space, leaves a very truncated view, biased toward the most solid remains, like thinking the calcium in the bottom of a pot is the basis of water.

    Like

  10. Jorge,

    “may I ask if it is possible for those expressed phenotypes to have any influence in the genotypes of germ cells and, in so doing, affect the decent of the living being?”

    Not in standard examples of phenotypic plasticity. However, there is increasing evidence of epigenetic (i.e., non gene based) inheritance, and some epigenetic markers (e.g., patterns of DNA methylation) are indeed influenced by environmental stresses. Again, though, this isn’t “Lamarckism,” since the epigenetic changes are not directly adaptive, i.e. they are not the result of an “active” attempt by the organism to properly adjust to environmental change.

    brodix,

    “causality is a process happening in the present, of which determination is the effect”

    Sure, but, again, the author is talking about hypothesis testing, he is assuming a standard view of causality. You may want to take a look at the full paper, I really don’t think the author would have any problem with your take on causality.

    Like

  11. @Jorge: “This is the first time I’ve heard of the distinction between the so called ultimate explanations and the proximate explanations. They appear to me as powerful conceptual instruments. I wonder why historians tend to pay little attention to the methods of evolutionary biology.”

    This is an old, old distinction. It is essentially the same as Aristotle’s very well known distinction between effective causes and final causes. It is not specific to evolutionary biology, although it often comes into play there.

    Regarding inheritance, as Massimo says, the concepts in the article don’t apply to that. The author is describing the branch of biology known as “evolutionary developmental biology”, sometimes abbreviated “Evo Devo”. You might find it useful to read the Wikipedia article, at https://en.wikipedia.org/wiki/Evolutionary_developmental_biology. The basic principle is to take into account the fact that the direct result of a genetic change is usually to alter the way that an organism develops, not to alter its phenotype directly.

    Like

  12. My personal reactions to this paper:

    1) As frequently happens, I’m irritated by the misuse of the term “module”. But really Jerry Fodor is to blame for that, and the author is just following his usage. But it is annoying anyway to see everything genetically programmed referred to as a “module”. A module, properly, needs to be autonomous — it needs to interact with the other parts of the system via a recognizable interface. Hard-wiring alone is not enough to make something a module.

    2) The bottom line of the paper, expressed in a single sentence, seems to be “Evo devo is telling us that we should prefer blank-slate style explanations over explanations that involve mechanisms genetically programmed in detail.” I don’t think that is actually what Evo devo is telling us. Proponents of plasticity often fail to realize that it doesn’t come for free. In order to get plasticity it is necessary to build a system that is capable of responding to signals from the environment in the right way. In many cases genetically programming such a learning system is quite a bit more difficult than genetically programming a certain type of response. In the end, as the author says in the early parts of the paper, the only way we can know what is going on inside is to actually look inside. We aren’t going to get there using a priori reasoning — not even if it is a different kind of a priori reasoning than people have used in the past.

    Like

  13. Bill, I don’t think the author is arguing for a blank slate. But you are correct that plasticity comes with trade offs and limitations. What I think the author gets right, though, is that the original hypothesis of massive brain modularity (as put forth by Tooby and Cosmides, not Fodor) is definitely out the window.

    Like

  14. Massimo,

    I’m actually increasingly skeptical of David Sloan Wilson’s ultra-Darwinism (he sees evolution everywhere). Interestingly, this would be the second time that the original sociobiology gets re-baptized to make it more palatable. I wonder when people will finally admit that it’s not a well thought out idea and just let it go…

    I agree to a certain extent. When it comes to psychology (as described in the paper I linked), I think Sloan’s approach is more palatable than standard EP as it tries to explain the relationship of evolution to psychology in a way that allows for both inherited traits (modules?) and malleable psychological mechanisms with his darwin machine concept.

    This doesn’t advance psychological research in any direct way and certainly doesn’t change my own research but I think it does provide a more reasonable explanation of how to combine two perspectives that traditionally have been divorced from each other (or in the case of EP, often dismissing the malleability of human traits)

    The paper by Fedyk as I understand after skimming the paper is that he is suggesting something similar.

    However, I know Sloan also suggests that evolution can be applied to humanities as well (and I think it can to some extent) but I’m skeptical of just how it will be applied in any meaningful sense beyond “we evolved to have advance capabilities that allow us to do X”. I enjoyed the commentary on their recent paper as several commentators pointed out how their approach did not address political and moral philosophy, even though such large change efforts should take those fields very seriously.

    Like

  15. Hi Massimo and Bill Skaggs,

    Well, the issue at hand is precisely the extent to which the brain is a general-purpose computer. […] the original hypothesis of massive brain modularity … is definitely out the window.

    And:

    We know this isn’t entirely true, since there are anatomic areas in the brain that tend to be associated with particular tasks

    And:

    A module, properly, needs to be autonomous — it needs to interact with the other parts of the system via a recognizable interface.

    The modularity hypothesis never was about hardware, with distinct and localised hardware regions including interfaces to other regions, it always was about function. When defining modules functionally, the hardware implementation of the function can be non-localised and intertwined.

    The idea of a “general-purpose” computer is problematic since there is no way to write “general purpose” software, all one can do is write multi-functional software with information transfer between the different functions. So, if by “general purpose” one means “multi-functional” then ok, but then the distinction with “massively modular” software is blurred.

    [Aside, one can of course have “general purpose” hardware on which one runs specific-purpose software, and humans create computers like that, but in nature’s neural-network computers there is no hardware/software distinction.]

    Further, one cannot “train” a neural-network device in a “general” way, one can only train it by making specific changes in response to specific actions of the network.

    Similarly, natural selection has no way of evolving a “general purpose intelligence”, all it can do is reward or punish specific decisions and thus specific functions of a brain.

    Of course what matters for natural selection is a whole range of different functions, and thus evolved brains will be highly multi-functional, and they will benefit from sharing information between functions — and a sufficiently multi-functional device with sufficient information sharing ends up being an effectively “general purpose” tool.

    But, the evolutionary heritage will still mean that the brain evolved as a multi-functional device (= massively modular) rather than as a “general purpose” device.

    I also don’t see that developmental plasticity is a refutation of either massive modularity or of “particular conceptions of natural selection” (quoting from the paper from here on).

    Modularity isn’t about particular and localised hardware that is “largely hardwired with specific programs”, it is very much a software/functional view.

    Thus when the author says: “considerations of anticipated consilience with evolutionary biology would now recommend going in for some kind of learning mechanism, not a module”, this is a false distinction. There is no such thing as a “general purpose” learning program, there is only learning specific things (though, again, learning a wide range of specific things can produce an effectively “general purpose” outcome). Thus the “learning program” would be one of the modules.

    Thus the author’s comment: “the mind, according to the massive modularity hypothesis, is like a bone which is unable to remodel itself in response to increases or decreases in its load” is false reasoning. The only way in which a brain could remodel itself in evolutionarily advantageous ways is if that re-modelling were itself a naturally-selected “function” and thus if that remodelling were itself a behavioural function/module of the brain.

    Highly useful traits such as beneficial remodelling of the brain, in response to increases in load, do not come easily or cheaply (as would be obvious if you were to try to design a computer to do that!).

    “The massive modularity hypothesis says that there are developmentally-endogenous computational modules which come pre-programmed with algorithms for those patterns of behaviour that “solved” adaptive problems in the EEA.”

    That is true, *but* the “pre-programmed patterns of behaviour” include the “learning mechanism” and the “remodel under load” behaviour! You don’t get those free with packets of cornflakes, they are themselves highly-evolved functions.

    Like

  16. Massimo,

    “Bill, I don’t think the author is arguing for a blank slate.”

    I can’t see any other way to interpret this sentence: “However, I have also argued that the growing understanding of the evolutionary importance of developmental plasticity has changed this, so that considerations of anticipated consilience now most likely rationalize a preference for strongly non-nativist proximate hypotheses.”

    Like

  17. Massimo,
    “You may want to take a look at the full paper, I really don’t think the author would have any problem with your take on causality.”
    That might be because we are taking the same side of the issue, in terms of how to explain the process. I did read the non-paywall version and it was fairly clear about the point I raised. That treating time narratively naturally places serious constrictions on explaining the physical state of present actions. Obviously narrative is an elemental tool to describe reality, but it is reductionistic and linear, while the state of the present producing these events is not.
    While this might not be a major issue for evolutionary biology, some other disciplines take it to extremes and treat time entirely as single vector/dimension, based on measures of duration between events, then get all bent out of shape, trying to tie up the many loose ends, such as how different clocks run at different rates.
    I can only hope “considerations of anticipated consilience” “can be used to “screen off” “as unlikely-to-be-true those hypotheses which posit” math, determinism, block time, “or some other kind of relatively innate, relatively non-malleable proximate mechanism.”

    Like

  18. Severl brief notes:

    On modularity, yes Massimo is right, and could have written even longer. A lot of the most recent findings on brain structure say that modularity is a dead letter. I referenced this a bit in my essay on free will last year. I mean, “everyday” empirical findings about people recovering from strokes show right there at least a degree of the untruth of massive modularity.

    Bill Skaggs’ point 2 in his second comment is good. Ev Psych done right will look at the tradeoffs involved for having a more plastic brain, in terms of energy expenditure, DNA changes and more.

    On the subject of computer-like thinking, I think that it, while not as rejected as massive modularity of the brain, is falling more and more out of favor with many. Besides “brains in parallel” vs “computers in serial,” I think a number of brain researchers are taking more seriously the “hardware” vs “wetware” issues than do the Kurzweil types, and somewhat more seriously than the Dennett types.

    On epigenetics, back to Massimo. Something still not Lamarckian (and still unknown as to just how broad their effects are), but even less “conventional” than genetics, is prions. They too do not reflect active actions by an organism to adapt to the environment, but are based on changes in response to environmental changes.

    All who have picked up on him. D.S. Wilson is … interesting. I agree with fair chunks of his push to get other evolutionary biologists to be more open-minded about group selection.

    Per Imzasirf On the Wilson et al link, that itself seems less an attempt to remake Ev Psych than it does to go into new territory. Ev Psych, to me, has seemingly focused on un/subconscious human psychology and its evolutionary drivers — ie, what has evolutionarily determined the psychology behind what Kahnemann calls “fast” thinking. This paper, though, is focused on intentionality. Now, intentionality can be subconscious intentionality, tis true, but, per everything else in the abstract, I don’t think that’s the authors’ focus. Intentionality aside, I think he and the other authors are making a good effort to integrate cultural evolution and other things with Ev Psych; how well that will work, I don’t know.

    Now, to the meat of the issue

    On the basic ideas of Ev Psych, or sociobiology, though? I mean, if one is a naturalist, and takes the mind as a naturalistic product, “generated” by the brain via embodied cognition, then the mind evolved just as much as color vision did.

    Whether a twice-rebaptized sociobiology is up to explaining the “hows” of that, or whether, in this case, we should stop rebaptizing and instead throw the baby out with the baptismal font water is a good question, though. (Don’t you love analogies?)

    My answer is yes. Setting aside that the objections of Gould and Lewontin were themselves at times politicized, sociobiology and even Ev Psych when done theoretically right have politicized taints to them.

    Like

  19. Coel

    The idea of a “general-purpose” computer is problematic since there is no way to write “general purpose” software, all one can do is write multi-functional software with information transfer between the different functions. So, if by “general purpose” one means “multi-functional” then ok, but then the distinction with “massively modular” software is blurred.

    The problem with “general purpose” is that it’s largely a straw man that EP folks erected to compare their view of massive modularity. It’s not a view that anyone I have ever come across in psychology, neuroscience or humanities really holds outside of those who strongly reject science to begin with.

    Similarly, natural selection has no way of evolving a “general purpose intelligence”, all it can do is reward or punish specific decisions and thus specific functions of a brain.

    Sure it can, David Sloan Wilson gives examples of it in his paper with what he calls the Darwin Machines (DM), roughly equivalent of modules that have built almost unlimited flexibility. He gives the example of the immune system as one of these DMs and they also go into details about symbolic behavior being another DM. These are evolved modules that have evolved to be malleable or as my neuroscience professors use to say, “the only thing we are hardwired for is to be rewired.”

    But, the evolutionary heritage will still mean that the brain evolved as a multi-functional device (= massively modular) rather than as a “general purpose” device.

    I would just change that to multi-functional device (modular, not massively modular). The general purpose blank slate view is again just a straw man.

    Thus the “learning program” would be one of the modules.

    That is precisely what many proponents of massive modularity deny and precisely what DS Wilson is suggesting as an alternative. I agree with your view for the most part, but it’s not a view traditionally forwarded by EP.

    Bill Skaggs

    “Bill, I don’t think the author is arguing for a blank slate.”
    I can’t see any other way to interpret this sentence: “However, I have also argued that the growing understanding of the evolutionary importance of developmental plasticity has changed this, so that considerations of anticipated consilience now most likely rationalize a preference for strongly non-nativist proximate hypotheses.”

    There are more views than massively modular and blank slate. Notice how the quote says “strong non-nativist” but not completely non-nativist.

    SocraticGadfly

    Per Imzasirf On the Wilson et al link, that itself seems less an attempt to remake Ev Psych than it does to go into new territory.

    The abstract does not get into it but a good portion of the beginning of the paper highlights the flaws of EP and how to reboot it.

    Try this link, I believe it is not behind a paywall but it also does not contain the expert peer commentary:

    Click to access wilson-bbs-d-11-00562_copyedited_final_v2a.pdf

    And the paper really is not about intentionality as a trait but more the idea that we should take cultural and behavioral selection in our own hands and use what we know from applied sciences to societal problems.

    Like

  20. Coel no, you’re wrong. Modular theories of the mind have also been modular theories of the brain, and thus about structure, not just function.

    Per both SEP and Wiki, Fodor, who originated the modern concept, included points that must be about hardware, not software, or at a minimum, about hardware as well as software:
    1. Domain specificity, modules only operate on certain kinds of inputs—they are specialised
    2. Informational encapsulation, modules need not refer to other psychological systems in order to operate
    3. Obligatory firing, modules process in a mandatory manner
    4. Fast speed, probably due to the fact that they are encapsulated (thereby needing only to consult a restricted database) and mandatory (time need not be wasted in determining whether or not to process incoming input)
    5. Shallow outputs, the output of modules is very simple
    6. Limited accessibility
    7. Characteristic ontogeny, there is a regularity of development
    8. Fixed neural architecture.
    Certainly, 7 and 8 are about “hardware” as much as anything. Even if not totally about localized hardware, it is about hardware.

    It’s true this isn’t the same as functional specialization of the brain. However, that idea did appear to have some influence on Fodor, and probably on others.

    I link to Wikipedia because it’s got good “additional reading” links at the bottom. http://en.wikipedia.org/wiki/Modularity_of_mind

    And, per the likes of Ev Psychers bête noire, David Buller, even if we look at this from a “software” angle, there’s no evidence for it.

    That said, this is the big problem with massive modularity ideas. They analogize wayyyyyyyyyy too much from computer models to the human mind.

    Then this:

    Similarly, natural selection has no way of evolving a “general purpose intelligence”, all it can do is reward or punish specific decisions and thus specific functions of a brain.

    First, and you know this how? The neural net of a hydra doesn’t really make a lot of specific decisions, but, nonetheless it or something even simpler was “selected for” in the past. You may or may not be right about intelligence at the level of human consciousness, but, if “intelligence” is considered part of what brains in general evolve to do, I don’t think you’re so right.

    That said, the real issue is that you’re assuming two things to create an extended syllogism, neither of which may be warranted. You’re assuming specific decisions are always associated with specific functions, first. Second, you’ve already indicated that specific functions are modular but in a software-only way. Again, there’s no guarantee either is warranted.

    Imzasirf has otherwise addressed this nicely. Speaking of, thanks for the response on Wilson. Per your take, it’s like I’ve said about neuroscience: We’re in the Early Bronze Age as for having any clue about consciously, intentionally influencing cultural evolution. I have no use for something that sounds one-third like Wilson’s version of Kurzweil’s singularity.

    On the “blank slate” issue, I might be halfway between you and Bill. No, it’s not quite “blank,” but the strongly tilts near that.

    Like

  21. Thank you, Mark Fedyk, for providing us the opportunity to read this well-written and intriguing article. This is a topic that is of great interest to me, but another of which I have little understanding and with which I struggled. Bill Skaggs helped to clarify some of my confusion about proximate explanations/mechanisms and ultimate explanations/mechanisms. But the principle of anticipated consilence is still a matter of some confusion. I read a bit about Whewell’s notion of consilience and took in what I could. What is a source of confusion is the use of “anticipated.”

    You write:

    “What has emerged in the last decade is a near consensus view amongst evolutionary biologists that developmental plasticity occurs in all areas of the biological world and that developmental plasticity is fundamental to evolution. . . . Developmental plasticity is fundamental to evolution, and learning is an ‘unusually important’ kind of develop mental plasticity. . . . This demonstrates that evolutionary psychology is not, well, evolutionary psychology.”

    What you seem to be saying is that evo psyc is at a methodological crossroads and that the principle of anticipated consilience may be fruitfully employed to reorient it. What is not clear to me is whether the “near consensus view amonst evolutionary biologists that developmental plasticity occurs in all areas of the biological world and that developmental plasticity is fundamental to evolution” employed the same principle of anticipated consilience to reach that consensus that you recommend for evo psyc. In which case, there seems to be some unavoidable piggy-backing going on here, and I don’t mean this to sound in anyway derogatory. But any further clarification of my confusion here from either you or the other readers would be helpful to me.

    Like

  22. Coel: “The idea of a “general-purpose” computer is problematic since there is no way to write “general purpose” software, all one can do is write multi-functional software with information transfer between the different functions.”

    But humans are more general-purpose computers than special-purpose (or perhaps “limited-purpose” is better) ones, which defines other animals. Since we can imagine and create just about any culture (software), we can “evolve” our cultures. No other animals can do that. (Unless you watch The Planet of the Apes movies.)

    Like

  23. The following is hopefully a reasonably accurate and more ‘main street’ version of the paper.

    Traditional evolutionary psychology:

    1) Posit an adaptation, a pattern of human behavior (e.g. distrust of newcomers), that could have been favored by selection, thereby becoming a brain module (or function) in the ‘The Environment of the Evolutionary Adaptation’. The ultimate explanation.

    2) Make a list of possible reasons why this could have been a good point to select on. For example to help avoid free riders, less trustworthy, less entitled to coalition benefits, who deserves more punishment, who are judged less competent, i.e. who will be judged as less likeable (Cimino and Delton, 2010).

    3) Test if people judge ‘outsiders’ in the ways just described (proximate explanation), and if so you can claim (tentatively) that those responses mean a brain module (or brain function), that evolved as an adaptation a long time ago, is what is producing them.

    But:

    “proximate explanations and ultimate explanations stand in a many-to-many relationship [and] entails that this inferential strategy is unsound. Ultimate explanations almost never entail the truth of any particular proximate hypothesis.”

    Evolutionary theory has advanced a lot since the 90’s and so to realign the field the author suggests we start with the many possible proximate causes and filter them in the light of the best and current (anticipated concilience) cognitive science and evolutionary biology theory.

    “set the initial plausibility of any given novel proximate hypothesis proportional to the plausibility of significantly similar (kinds of) proximate hypotheses as can be found in any of the biological sciences which have as their subject adaptive patterns of behaviour.”

    Conclusion:

    “I argue that inferences reflecting the pursuit of this kind of consilience […] indicate that psychologists should have a preference for explanations of adaptive behavior in humans that refer to learning and other similarly malleable psychological mechanisms – and not modules or instincts or any other kind of relatively innate and relatively non-malleable psychological mechanism.”

    Like

  24. Sometimes I think that we pretend to know more about the brain than we do.

    There is not much point in making ad hoc models of the brain based on analogies with what we happen to know about computing right now, because our inability at the moment to understand things like intelligence and consciousness pretty much demonstrates that we don’t know enough about these things to make those kinds of models.

    For my part I am more interested in things that I observe that I am able to do with my mind rather than the things that some theory says I ought to be able to. Unless of course that theory predicts I can do something which I don’t know of yet, and that prediction checks out.

    More study needed. That is not so much a comment about the paper under discussion, which may be regarded as part of the study that is needed. It is more a comment about the rather confident assertions that some are wont to make about the mind, even if the mind (or brain or computing) is not their particular field.

    Like

  25. Of course human beings have an enormous amount of developmental plasticity — far more than other animals, which is why in their short existence they wound up in Australia, otherwise cut off by tens of millions of years from the mainstream of mammalian evolution, as well as the Pacific islands and the Americas.

    I don’t think we got that way without evolving a tremendous amount of unique mental machinery though — specifically I have to say that the essence of Chomskian linguistics seemed to me a necessary explanation of a deep mystery from the time I 1st ran across it. The alternative – that every individual must improvise an understanding of language without a great deal of structure, and that language yet exhibits such similarity everywhere, among strands of the human race separated by tens of millennia makes no sense.

    Nor did the common wisdom of the 50s/60s when I was growing up, that we became human by achieving a certain critical brain size, and having opposible thumbs seem at all reasonable. Given all the complex machinery that we now know makes up vision, I would expect as much or more complexity to go into language.

    The really important problems, like language, and aspects of how we thing that we probably haven’t even identified yet, are extremely difficult, while what is tempting is relatively trivial: low hanging fruit; naturally anything to do with reproductive choices, or choices of what effort to put into childrearing (the “Cinderella effect” referred to in the paper) are both relatively easy (to come up with *plausible* theories at least) and make for popular science lite literature.

    At any rate, mere explanations don’t make for very good science, unless they, like the explanations of chemical combinations lead to lead to substantial new knowledge and capabilities.

    Some sciences though might take a long way to get to that point. Ev Psych seems to still be largely in the “collecting specimens phase, like biology in the 18th/19th century, with here and there a stab at a grand theory. Linguistics looks to much like 19th century (esp the 1st 2/3 century) chemistry, with a lot of tantalizing relationships and almost-relationships, and incremental and debatable progress for decades, but based on the example of chemistry, I wouldn’t dismiss it.

    Language is like a radical new platform in the software sense, that facilitates an orders of magnitude increase in our level of cultural plasticity.

    I just have to laugh at the idea of cooperative hunting as something for beings to whom it has never occurred to to just observe and think “Hmmm, that looks like I good idea”. It just seems like too many levels to jump over at once, In fact more than very trivial cooperation calls for language, and with language (at a sufficient level), it seems inevitable.

    Like

  26. I find it very hard to escape the impression that there is an awful lot of turgid pseudo-profundity in this paper. Maybe I am missing something, but it seems to me to be about 10 times as hard to read as it needs to be.

    As an example, I think the 377 words of footnote (I can’t include it without going over 500 words) could be paraphrased in the following 165 words (more like 140 if you take out a few observations I inserted, and the numbers (i)..(ii)… For anyone who wants to tell me where I’ve gone wrong, I’ve made a two column table of the original footnote and my paraphrase, matched phrase by phrase at http://jmisc.net/fedyk-paraphrase.html

    (i) Sometimes an ultimate explanation might, after all, imply a proximate explanation.
    (ii) If an organization ever had some adaptive behavior(B), it must have some mechanism capable of causing B.
    (iii)Whatever enables B might be able to cause other behaviors,
    (iv) ???
    (v) ditto (ii).
    (vi) Call the function/mechanism responsible for the behavior T.
    (vii) T might have other functions.
    (viii) T might be a “module”; if so, producing B *may* be its only function.
    + T is relatively innate and non-malleable. (non sequitur?)
    (ix) T might be a learned capability, which would make it very context dependent
    (x) T might be a domain-general faculty. [faculty/system of faculties seems a matter of convention at this level of abstraction]
    (xi) T would be able to cause B, as well a a bunch of other things; maybe an infinite number of things.
    (xii) Either a module, something learned, or a domain-general faculty might be able to produce B.
    (xiii) Hence observation (ii) doesn’t imply much of anything.

    Since everything under discussion here is behavioral/psychological, it seems unnecessary to use the word “psychological” not to mention so many times, so I’ve left it out.

    Like

  27. Ev Psych seems at present to be largely descriptive and suggestive, like paleontology some decades ago. Many speculations will remain in limbo until and unless actual mechanisms are discovered that mediate, say, differential reciprocal altruism at the various levels of kinship, or particular treatments of newcomers.

    One thing that seems to me off about this paper is the constant refrain of ultimate causes “explaining” proximate causes. The few proximate causes we have for anything in human psychology are painfully extracted from neuro-anatomy, and I doubt that even regular evolutionary biologists often get to proximate causes. At the top of p5, he makes a point that makes me wonder why he continues to mention proximate causes: “these various debates about the ultimate explanations for different traits can proceed even in complete absence of any understanding of which proximate explanations are plausible. We simply do not have the resources to determine the molecular and physiological processes that fix the coloration of every insect species; recent estimates of the number of insecta put this number at around 9 million ( Mora et al. 2011).”. I.e. we are working at the level of saying “the moths’ wings changed from brown to black during the industrial revolution as the trees around London became darkened by soot”. Figuring out the genetics behind brown or black wings seems like another discipline. The mention of 9 million species of insects and the citation to go with it seem gratuitous.

    Like

  28. A number of additional thoughts.

    First, on the main paper? With a more thorough reading, I think that Fedyk risks throwing out babies with bathwater. One can critique, and strongly critique, evolutionary psychology, enough to call for a new model of the evolutionary development of human psychological traits, without insisting that such a new model should begin with a non-nativist stance. We may need more non-nativist ideas than current EP, but that’s not as much of a spectrum change as he proposes.

    Wilson’s piece? I’ll give him a kudo at the start for calling for “depolarization” between ev psych and the “standard social sciences model.” I still disagree with the idea that we as a species can actively manage changing our social and cultural practices though, at least not short of Huxley’s soma or Clockwork Orange. Neoliberal social pundits like Cass Sunstein talk about their “nudges,” but, I don’t think that’s at the scale nor scope of what Wilson is getting at.

    That said, Wilson does tie in with the main paper. Tooby et al analogized human minds to computers because computers were limited then, as well as sexy; they may also have been thinking along Dennett’s lines of evolution being algorithmic, and thus thinking, in a somewhat parallel line of thought, that not only was such an analogy sensible, it was semi-required.

    At the same time, I don’t really like Wilson’s “symbotype,” analogizing units of symbolic thought to genes. Especially in creatures with social learning and other such types of factors, he’s pushing the evolutionary angle too much.

    He also gives a have-your-cake-and-eat-it-too add-on to his definition:

    In any case, the term symbotype refers not to a single cultural trait but rather to a given set of symbolic relations, which results in an entire suite of phenotypic traits (the phenotype).

    Well, D.S., if you’re first analogizing “symbotype” to genotype, but now to phenotype, well, Houston, we have a problem.

    The next main part of the paper is where the rubber hits the road, though. His “applied science of change” ultimately looks a fair bit like his version of E.O. Wilson’s consilience, if not ramped up even more.

    That said, again, there are also ideas to like, such as his bringing group/multilevel selection back. Interestingly, earlier this evening, I was reading about some of the psychological therapies he mentions. Some of them, Massimo, even more than their father or stepfather, CBT, would fit with your Stoic meditation ideas quite nicely.

    However, back to Wilson’s main idea. These psychology schools all operate on individuals, not groups, and surely don’t have “throw weight” for what Wilson envisions. He next gets to small groups, but I don’t find him convincing.

    Beyond that, even with advances in computing power, we don’t have computers that have the throw weight to model his ideas. Again, we’re probably in the Early Bronze Age there.

    We also have no idea whether or not some things from individual behavior, like tit-for-tat, cheater detectors, etc., scale well to groups.

    Like

  29. Massimo,

    I’m not sure how you are able to conclude that massive modularity is false given what we know about brain development. The modular framework deployed by evolutionary psychologists (such as Tooby and Cosmides) is quite different than the one Fodor originally conceived. Many still don’t understand this, and that’s unfortunate. I suspect (but only suspect) that a key reason why there is still this kind of misunderstanding is because it has been repeated a number of times in other sources, and so perhaps people have just imbibed it in this manner.

    For those who still think modularity in evolutionary psychology is tantamount to Fodorian modularity, here are a couple of correctives (via philosopher Peter Carruthers and evolutionary psychologists Clark Barrett and Rob Kurzban, respectively):

    Click to access AoM%20-%20Precis.pdf

    Click to access Barrett%20Kurzban%202006.pdf

    imzasirf,

    D.S. Wilson et al.’s example of the immune system as a type of ‘Darwin Machine’ is actually a modular-like system that is nonetheless functionally specialized for guarding against disease. It just so happens that its specialized design is highly generative and thus capable of producing a wide range of outputs. In this sense, it is exactly what evolutionary psychologists mean when they speak of modular psychological adaptations, regardless of how flexible or generative they might be. Functional specialization needs to be underscored, and it is no way incompatible with flexible cognitive architectures that develop in highly contingent ways.

    SocraticGadfly,

    “And, per the likes of Ev Psychers bête noire, David Buller, even if we look at this from a “software” angle, there’s no evidence for it.”

    Of course, evolutionary psychologists would disagree with you – as do I. Language would be one example of a functionally specialized modular system that evinces a high degree of apparent design, hence the hallmarks of natural selection. One must look at the messy empirical details in this kind of debate as well; theory and philosophy alone aren’t enough to arbitrate these debates.

    Also, it is possible to use both the forward- and backward-looking methodological heuristics of evolutionary psychology in reciprocally-illuminating ways. One isn’t necessarily merely handcuffed with one of them in isolation with the other.

    Like

  30. This is, of course, yet another attempt to resurrect the Blank Slate. The author leaves us in no doubt on that score, cutting to the chase in the abstract, where he writes that we should prefer, “explanations for adaptive behavior in humans that refer to learning and other similarly malleable psychological mechanisms.” The futility of such efforts was amusingly demonstrated by the media response to Hawking’s recent comments “On Aggression,” as if there had never been anything the least bit controversial about them.

    There is really nothing original here. It’s just another instance of the hoary strawman argument that EP is equivalent to “genetic determinism.” The author merely substitutes the rather clumsy “relatively innate, relatively non-malleable proximate mechanisms” for that somewhat too obvious term. As usual, the author pretends that there is a stark “nature vs. nurture” divide, with EP representing the former, and continues with a string of misrepresentations that have been refuted literally hundreds of times in the EP literature. For example, EPers are all supposed to be in lockstep with Fodor’s rigid definition of “modularity,” they are all supposed to embrace “nativist” versus “non-nativist” arguments, they are all supposed to believe that human behavior is “hardwired,” excluding “learned adaptive behavior,” they are all supposed to reject the notion of “developmental plasticity,” etc. All this is only possible if one completely ignores the abundant refutations to all these spurious claims that have appeared in the EP literature over the last two decades. A good example is “Modularity in Cogniton: Framing the Debate,” by Barrett and Kurzban, published in 2006 and available free online. I can only suggest that anyone who seriously believes there is anything novel about Fedyk’s arguments read that paper.

    Like

  31. Hal Morris: “Language is like a radical new platform in the software sense, that facilitates an orders of magnitude increase in our level of cultural plasticity.”

    That’s a great quote I’d like to quote!
    (saying as a programming languages geek)

    Like

  32. Helian et al., since I’ve read the full paper, I can categorically say that the author is not proposing a blank slate model, and I find it fascinating that so many are so ready to jump to that conclusion. The was always the issue whenever I was trying to explain phenotype plasticity to others, they would immediately take it to be a rejection of genetic influences. It is not, since plasticity, by definition, is a property of the genotype. What it does, though, is to make simplistic genetic deterministic accounts untenable. Those happen to be the sort of accounts often relied upon by evopsych, which is why this paper presents an issue for that research program.

    Like

  33. Thomas, unfortunately the author likely doesn’t even know his paper is being featured at SciSal. In the future I will make a point of alerting him/her. This new feature works like the old “picks” at Rationally Speaking, with the difference that I highlight one paper at a time, instead of several, and that the paper in question is from the primary literature rather than from secondary or popular sources. In alignment with the different purpose of SciSal. This is why I have been the one answering questions, and not the author. Sorry for the confusion.

    Like

  34. This was a really interesting paper – glad I read it. It’s possible that evolutionary psychology is benefitting, in a sense, from the lack of clarity about modularity that’s been argued over in this thread. The paper does a good job of trying to pin some of that down.

    To those arguing about hardware vs. software, I’d say that it’s important to keep in mind the limitations of the metaphor. If you think of cortical remapping vs. more local changes to neurons, would you say that one has to do with “physical structure” and one doesn’t? I’d hope that you wouldn’t — both clearly involve physical changes. A view of plasticity that doesn’t involve remapping as “non-structural” privileges “what’s connected to what” in a way that isn’t warranted. And again, if you were to compare synaptic vs. non-synaptic plasticity, you might be tempted to make the same sorts of (unwarranted) distinctions, because one involves more obvious morphological changes to cells. You might say, for example, that a neuron whose potential has changed is “really the same neuron but just acting differently”. This way of looking at it misses the fact that physical changes to the neuron are necessary for the change in potential to happen, and that physical changes can be more and less morphological (or morphological at smaller and larger levels of organization).

    In the hardware/software metaphor, we’re working with a machine that doesn’t undergo a lot of morphological changes — and so the distinction has a much more pronounced meaning than it can be made to have when talking about the brain. Plus — the whole idea of a “set of instructions” that are “executing” in the brain leads to a lot of the conceptual errors that lead us to think incorrectly about modularity.

    TLDR; it’s really all “structural” changes — just structural changes at different scales and extents.

    Like

  35. Very interesting article, but I’ve only been able to skim it and need time to read it closely. For now, it seems pretty much right on the money.

    I couldn’t help noticing one passage in particular:

    “‘Rather, we should expect parental feeling to vary as a function of the prospective fitness value of the child in question to the parent… When people are called upon to fill parental roles towards unrelated children, we may anticipate an elevated risk of lapses of parental solicitude. (Daly and Wilson 1985, 197)’
    This prediction about parental feeling was tested by analyzing rates of child abuse in the Hamilton-Wentworth area in Southern Ontario, where Daly and Wilson found significantly higher rates of abuse for children living with at least one step-parent (…).”

    Are we supposed to take such ‘predictions’ and ‘testing’ seriously? Coming from a dysfunctional family, and knowing others who also have, it is quite clear to us that favoritism and abuse have more to do with the psychological problems of the parents, or inability to learn parenting skills, than some meta-inheritance of selection for ‘fitness.’

    One of the glaring deficiencies of EvPsych is that their models and studies both derive from, and return to, the cultural norms of their own society. They show virtually no deep reading in anthropology, or real understanding of the issues raised in observing patterns of behavior in other cultures.

    We know there have been tribal cultures where children were seen as progeny of the clan, or even the whole tribe, rather than the ‘nuclear’ family, and where it was presumed that all members of the clan or tribe would effectively parent the child. We also know that there were, and still are, more developed cultures where lineage and/or gender determined favoritism – often by law. George the Third was not fit to be the King of England, but on the throne he sat. And I don’t see how the dumping a girl baby in the river because the parents want a male for their first born (as happened in ancient Rome, as happens still today in China, occasionally) evidences “a function of the prospective fitness value of the child,” since the infant hasn’t been child long enough to determine its fitness. The Daly and Wilson study only tells us something about the culture of Hamilton-Wentworth, Ontario, surely that’s obvious. EvPsych has a fundamental research problem – its testable predictions all seem to derive from already existent observations – ‘From evolutionary theory, we can predict that, during socialization rituals where alcohol is consumed, someone will get angry when called a cross-eyed baboon’ – as if there could be any doubt of that.

    EvPsych is raising interesting questions; but that’s not enough to save it. Its theorists really need to recognize their own cultural up-bringing and biases. If they do, perhaps they can incorporate the criticisms and suggestions of Fedyk’s essay. If they don’t, their project seems doomed to the same waste-basket as phrenology and eugenics.

    Like

  36. Jedi Master You’re behind the times, at a minimum. The modularity of language has also been more and more rejected, at least in its more robust forms, along with other ideas about massive modularity.

    Aeon has a good article on this, including noting that Chomsky himself has pulled in his horns a fair bit from his original claims:http://aeon.co/magazine/culture/there-is-no-language-instinct/

    I’ll take the rest of your comments in light of your claims about the modularity of language.

    (I also recommend that link to Hal and others on this thread talking about language issues.)

    As for a related issue, analogizing from machines to humans, Massimo’s new essay refudiates that idea quite nicely. I’m shocked that, on it, Coel committed a non sequitur.

    https://scientiasalon.wordpress.com/2015/02/27/the-danger-of-artificial-stupidity/

    MassimoThis all reminds me of something you’ve written about before: embodied cognition. I say that in part because I’m reading a new book about it. Any theories on evolutionary effects on the development of human psychology needs to do so in light of evolutionary development of the human mind as an embodied system. Not just a system, but an embodied system.

    That’s an additional bit of work, of course, but, so be it.

    Per whether the paper pushes the “blank slate” or not? I go back to my previous comment and expand. Fedyk is not pushing a blank slate for individual human nature, I don’t think, but he is proposing that the contents of that slate are fairly circumscribed.

    You know I like to discuss things in terms of continua, so, let’s say 0 is totally nativist and 100 is totally non-nativist. I’d put Fedyk at around, say, 70. That said, this is why I second the thoughts of others and say that an essay’s author should be “tagged” in order to allow for his or her comment.

    Whether he’s right or not remains to be seen, of course. This is one area where I think EP needed some pushback, but his “strongly” and other things, I think, mean that he may be pushing back too far. Again, he may not be, but I stand by the idea that he may be risking that.

    EJ Yes, the stepparent abuse studies in general have been found “wanting” for the reasons you mention and similar ones. I do agree, as noted previously, and on my blog, that there’s a fair amount of bias — usually some sort of political libertarianism — behind the curtain.

    Like

  37. Jedi Master

    D.S. Wilson et al.’s example of the immune system as a type of ‘Darwin Machine’ is actually a modular-like system that is nonetheless functionally specialized for guarding against disease. It just so happens that its specialized design is highly generative and thus capable of producing a wide range of outputs. In this sense, it is exactly what evolutionary psychologists mean when they speak of modular psychological adaptations, regardless of how flexible or generative they might be. Functional specialization needs to be underscored, and it is no way incompatible with flexible cognitive architectures that develop in highly contingent ways.

    I agree with the first part of your paragraph but not the second half.

    D.S. Wilson’s argument is that EP researchers put too much emphasis on the module and downplay the highly flexible and generative systems (as in the immune system example).

    You yourself state it as we must “underscore” the functional specialization but my question is why? It’s certainly an important part but not the only or most important part. The DM that Wilson talks about is not underscoring either, just showing you can have both in the same system.

    What the EP researchers (some not all obviously) have difficulty with is that they want the module to be more important and downplay symbolic human behavior and human culture but I think this is unjustified. Once you get DMs, it’s hard to justify only studying one part and not the other and studying the module only will leave out some of the most essential aspects of the machine.

    That is why people criticize EP research as making strong conjectures regarding what we evolved to do and why we do it because if we evolved to be highly flexible, it means very little to say we evolved to X when X can have seemingly limitless combination with other factors (of course we evolved to do this).

    In other words, the argument made in the article that Massimo linked to still holds, that very little can be said about the psychology by looking at the modules alone and that studying the generative part of the module is a more productive line of research to pursue in psychology.

    Like

  38. Not to step too far outside the box, but isn’t basing the evolution of psychology as starting with the Pleistocene a bit like remodeling an old house, to make it seem like your own? There was quite a lot of evolution, both biological and psychological, leading up to that stage.
    No, we don’t have a blank slate, in terms of the primate form, but what about going to a fairly blank slate and considering how that form arises from it and what functions are emphasized?
    Much has been made of opposable thumbs, big brains, walking, etc. What I think is one of our more notable physiological features is that of all species of animal, primates have their eyes placed closest together. As our species originated as creatures swinging around in trees, being able to judge distances and locations very accurately would be of primary importance. So when these creatures started throwing rocks and sharped sticks at food and enemies, the underlaying computational tools for such factors as trajectories, angles, weights, etc. would have been quite instinctual.
    Yet do these factors create bias? Most animals have eyes on the corners of their heads and many on the sides. Might there be a linear, point/object oriented focus on our part, whereas other animals might have a more developed spatial/contextual awareness? Even now, we treat space as though it were simply an effect of its own measurements, I suppose as it can’t be seen or touched. Considering Smolin’s argument, some even think the linear process of time is more fundamental.
    Obviously this linear impulsion was benefited us enormously, as we have pushed out in every possible direction; consciously, spatially, technologically, communicatively, etc. Yet where do we go from here? Safe to say, few of us are leaving this planet and the resources on it are getting seriously taxed.
    While we have evolved to a high level of intellectual complexity, it is as Socraticgadfly points out, early bronze age, when compared to our biological complexity. One significant factor to consider, given everything from the periodic table, to the history books, is that complexity can grow unstable and tends to increase and decrease.
    Given that significant contraction appears possible in our future, what lessons do we learn? What will be the science of our age’s contribution to the evolution of our species and viability of the planet?
    Will it be iPhones, nuclear weapons and multiverses?
    Is there some, not quite yet “anticipated consilience” in our future? If there were to be some Santa Fe Institute type meeting of the inter-disciplinary minds, what lessons could it draw?
    Do we have that degree of plasticity, or would everyone insist that no one else is informed enough about their particular field to pass any judgments pertaining to it and we all crawl back in our particular modules?
    I suspect the likely outcome would be more heat, than light. Missing the forest for the trees.

    Like

  39. @SciSal

    “I can categorically say that the author is not proposing a blank slate model, and I find it fascinating that so many are so ready to jump to that conclusion.”

    Forgive me for my naïve assumption that, if something walks like a duck, quacks like a duck, and flaps its wings like a duck, it actually is a duck. Whether you want to call him a Blank Slater or not, the point of my comment was that Fedyk’s paper is a collection of strawman arguments that completely mischaracterize the “common ground” that EPers are supposed to stand on. I congratulate you on reading Fedyk’s entire paper. I suggest you also read the entire paper by Barrett and Kurzban I referred to in my first comment. Among other things, I suspect you’ll notice that Fedyk’s strawmen have been around for decades, and that EPers have convincingly refuted them over and over again. Later day Blank Slaters like Fedyk simply ignore these responses, and continue to trot out the same old hackneyed arguments, occasionally repackaging the strawmen by, for example, serving up the old “genetic determinism” canard clothed in new jargon as a belief in “relatively innate and relatively non-malleable psychological mechanisms,” by way of keeping up appearances.

    Like

  40. The Cimino and Delton paper (also freely available) is cautious in its conclusions, and I don’t think the authors suggest they have excluded alternative explanations. WRT to Tomasello’s work, I would think that C&D would argue that they are looking for evidence that similarities across cultures in “culturally specific social norms” eg hazing and initiation, arise from genetic behavioural predispositions rather than just being a widely applicable organisational solution. This is a pretty broad hypothesis, and testing it will be hard.

    The behaviour genetical approach to this would be to look for genetic variants associated with variation in the “Tenure Categorization” trait in modern populations and see if they are in plausible pathways. The EP model, by contrast, essentially imagines that stereotypic characteristics of humans arise from fixed genes (fixed back in the good old days), so possibly one could look for genomic signatures of selection (which genes???).

    A study by the same authors (along with Cosmides and Tooby)

    http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3365621/

    looks at the related concept of the free rider (again a universal across diverse societies) and its moral psychology. The discussion of alternative hypotheses and study limitations for that paper are a bit stronger.

    With respect to Daly and Wilson (1985), again the alternative “proximate explanations” are partly subsumed as mechanisms eg less time bonding or annoyance with stepchildren is confounded by the concept that bonding with the neonate is the selected trait.

    Like

  41. Hi imzasirf,

    The problem with “general purpose” is that it’s largely a straw man that EP folks erected to compare their view of massive modularity.

    I was responding to Massimo’s use of the phrase, but if you don’t think that it is a prevalent view then ok.

    Hi Socratic,

    Modular theories of the mind have also been modular theories of the brain, and thus about structure, not just function. … Fodor, who originated the modern concept, …

    I’m approaching this from the EP angle (rather than from philosophy), and certainly in EP “modularity” is defined in functional terms. There is no requirement that a “functional module” needs to be implemented with localised hardware, nor indeed that the hardware needs to be single-function. In practice the relationship between function and hardware is going to be messy and complex (as expected for a device cobbled together by evolution).

    The basic EP argument for functional modules is then that evolutionary selection will have been for specific functions and behaviours.

    First, and you know this how? The neural net of a hydra doesn’t really make a lot of specific decisions, but, nonetheless it or something even simpler was “selected for” in the past.

    I’m puzzled by your argument. Any *simple* neural network can only have a rather limited set of output states, and that is exactly what I mean by “specific function”. E.g., the algorithm “Given input states X, Y and Z, set output state to 3″, is a specific function, and will be selected for if that behaviour is beneficial.

    The fact that the human brain must have evolved out of a collection of such specific functional modules, which then developed into multi-functionality and entwining and information-sharing between functional modules, is the starting point for the modularity hypothesis.

    Of course if you do have sufficient entwining of multiple functions then it ends up pretty much “general purpose”, but that’s not the evolutionary origin of the device.

    You’re assuming specific decisions are always associated with specific functions, first.

    If you define the “function” as “making that decision” then it is!

    Second, you’ve already indicated that specific functions are modular but in a software-only way.

    Again, the term “modular” here is just a reference to the specific function, so again that follows by definition. As for the “software only”, it’s rather that I’m making no claim about the hardware implementation, I’m simply approaching this from the evolutionary point of view that it is *function* that is selected for.

    Like

  42. I’d like to add a comment on the aspect of this that is most important to me, although it isn’t really a central issue in the article. The term “modularity” has come to be misused to a sinful degree in cognitive science, and generates more obscurity than clarity. I’d like to explain how that came about, and what we should do about it.

    Jerry Fodor was the first to write seriously about modularity, in his 1983 book *The Modularity of Mind*. His definition of modularity was rather complex, though. As Peter Carruthers summarizes it, in Fodor’s account “modules are held to be mandatory in their operation, swift in their processing, isolated from and inaccessible to the rest of cognition, associated with particular neural structures, liable to specific and characteristic patterns of breakdown, and to develop according to a paced and distinctively-arranged sequence of growth.” (http://faculty.philosophy.umd.edu/pcarruthers/Moderate-modularity.htm) But this is too complicated a definition for such a basic concept, and as Carruthers explains, later authors relaxed it in ways that made it even fuzzier. Currently, it seems, modularity means hardly anything more than hard-wiring.

    In the meantime, the meaning of modularity in computer science has simplified and clarified. In modern computer science (and other branches of engineering), modules are simply subsystems whose interactions are channelled entirely through interfaces. An interface is a restricted interaction scheme that uses a defined set of signals and obeys a definite specification. Modules can be either functional or hardware-based — subroutines in a program are functional modules; devices in a computer are hardware modules.

    The result is that what cognitive scientists mean by modularity has come to be quite different from what engineers and computer scientists mean by modularity. The solution is to throw away the useless hash that cognitive science has developed, and use the term in the simple way computer scientists do. In particular, stop using it as a synonym for “hard-wired”.

    Like

  43. HelianUnbound,

    The author is not proposing that modularity be discarded, and he is definetly not proposing a return to the blank slate.

    I think he is simply saying that when evolutionary psychology is proposing the appearance through adaptation of ‘higher leveled’ modules one should focus more on ‘psychological’ mechanisms, in tune with the best of contemporary theory in evolutionary biology, and other relevant fields.

    In other words, proposing that when we look at higher level behavior we focus more on functions like ‘social learning’ and less on things like a module for ‘distrust of others’ to explain current human behavior.

    And I agree, moreover I think evolutionary psychology has been used way too often, intentionally or not, to explain in inevitable biological terms, with an aura on scientific respectability, all kinds of cultural stereotypes and dysfunctional social behavior.

    Like

  44. I’m sorry, but when an author suggests that learning must be preferred unless one can perform a double back flip over his “anticipated consilience” barrier, that’s blank slatism at its best. I’m stunned that all of you have swallowed Fedyk’s implausible yarn about Lorenz’ hydraulic theory being the only game in town in the very heyday of the Blank Slate without even blinking. After all, some of that orthodoxy’s very arch-priests, such as Richard Lewontin and Stephen Jay Gould are/were evolutionary biologists, the “mature discipline” EP is supposed to emulate. They, too, had a “positive model in biology of the proximate causes of adaptive behavior,” in the form of the Blank Slate. Fedyk is speaking of a time in which the Blank Slate dogmas were virtually unchallenged in the behavioral sciences, and anyone who got out of line was shouted down as a fascist, or worse. And yet we are supposed to swallow the ludicrous imposture that Lorenz’ hydraulic theory not only overshadowed the Blank Slate dogmas, but that those dogmas never even existed! I can swallow a little judicious historical revisionism, but doing a full Trotsky on the Blank Slate is a bit rich.

    Like

  45. First, Coel.

    Sorry, no, you don’t get to define the terms. While modularity in EP may be used more in functional than strucutural terms, it references both. We’ve talked about language, and certainly a Pinker, who has written about language, was doing so based on a structural module for language.

    But, it doesn’t matter anyway. That Aeon piece, which I link again to make sure you can’t avoid it, notes that, in language at least, new thought rejects both structural and functional massive modularity. The MM gig is up. Kaput.

    (P)erhaps the uniqueness of language arises not from a ‘where’ but a ‘how’. What if there is a type of neurological processing that is unique to language, no matter where it is processed in the brain? This is the idea of ‘functional’ rather than ‘physical’ modularity. …

    In The Language Instinct (1994), Pinker examined various suggestive language pathologies in order to make the case for just such a dissociation. For example, some children suffer from what is known as Specific Language Impairment (SLI) … That seems like a convincing smoking gun – or it would, if it hadn’t turned out that SLI is really just an inability to process fine auditory details. It is a consequence of a motor deficit rather than a specifically linguistic one. Similar stories can be told about each of Pinker’s other alleged dissociations.

    http://aeon.co/magazine/culture/there-is-no-language-instinct/

    Brodix then gets to the heart of the problem with Ev Psych, with his unnamed reference to the EEA, the environment of evolutionary adaptedness.

    Yes, that’s the core problem with ev psych. The sexism and other biases a lot of people see in it stem from that. Beyond that, it’s just unproven in general.

    How do we know that that era, whether in rapidity of genotypic changes, or degree of “quality” of phenotypic expression of those changes, was more important than anything else? How do we know that humans are, as alleged, most adapted to that era?

    As for why the EEA was so elevated? Along with massive modularity, man the (noble) hunter-gatherer has also pretty much been knocked into a cocked hat.

    Homo sapiens was a less noble scavenger gatherer long before becoming a hunter-gatherer. That’s not so sexy, though, either metaphorically or literally. Many ev psychers seemed to focus on man the hunter-gatherer in the other sense, of the two stupidly overlapping ones in English, namely “males the hunter gatherer.” Of course, there’s a pretty straight line from that to Randy Thornhill’s and Craig Palmer’s “rape is adaptive” claims.

    Non-nativism does come indeed to the fore here. If they had ever looked at a small tribal society, beyond knowing that rape tends to have a lower reproductive rate than consensual sex, they’d have seen how quickly our ancestors would have made it non-adaptive.

    All the issues associated with the EEA’s problems are why ev psych in generally just can’t be taken that seriously. Thornhill and Palmer sound like they’re writing a “Saturday Night Live” skit.

    Like

  46. Helian, you can keep saying that until the end of time, but the author is *not* advocating blankslatism, and we have been explained why. And to call Lewontin and Gould “priests” is a gross mischaracterization of what those people (a *geneticist* and an evolutionary biologist) were writing.

    Like

Comments are closed.