About Rationally Speaking


Rationally Speaking is a blog maintained by Prof. Massimo Pigliucci, a philosopher at the City University of New York. The blog reflects the Enlightenment figure Marquis de Condorcet's idea of what a public intellectual (yes, we know, that's such a bad word) ought to be: someone who devotes himself to "the tracking down of prejudices in the hiding places where priests, the schools, the government, and all long-established institutions had gathered and protected them." You're welcome. Please notice that the contents of this blog can be reprinted under the standard Creative Commons license.

Friday, May 09, 2008

Chess, psychoanalysis, evolutionary psychology and the nature of pseudoscience

I’m reading a delightful history of chess, The Immortal Game, by David Shenk, and got to the chapter dealing with the dark side of chess: the fact that a small but significant number of top players throughout history have gone off the deep end -- including the famous American world champion Bobby Fisher.

As is usual with correlations (playing chess <=> your brain goes bonk), it is not clear which way the causality goes, if at all. It could be that playing chess at the highest levels affects the mind in negative ways; it may be that abnormal minds are more likely than others to be attracted by the game; or it could simply be that the correlation is spurious, i.e. non-causal.

Shenk does not take sides on this debate, but he does report the pronouncements of a number of Freudian psychoanalysts on the matter. For instance, Ernest Jones (Freud’s biographer and protege), confidently stated that “It is plain that the unconscious motive activating [chess] players is not the mere love of pugnacity characteristic of all competitive games, but the grimmer one of father murder.” What?? That’s right, it’s the good ‘ol Oedipus complex -- itself rooted in the all-encompassing Freudian explanation for human behavior, sex drives -- that pushes players to protect their Queen (=mother) and checkmate the King (=father). Here is some more nonsense from Jones (p. 147 of Shenk’s):

“It will not surprise the psychoanalyst when he learns ... that in attacking the father the most potent assistance is offered by the mother (=Queen). ... It is doubtless [its] anal-sadistic feature that makes the game so well adapted to gratify at the same time both the homosexual and the antagonistic aspects of the father-son contest.” (Never mind that the Queen was introduced relatively late as a chess piece, as Shenk’s history shows.)

Jones wasn’t the only one to psychobabble about chess. Here is American psychologist and Freud disciple Isador Coriat (again, quoted in Shenk, p. 148): “The sole object of the game for these individuals was to render the King (the father) helpless through checkmate, that is, castrate him. The winning of the game produced a feeling of intense pleasure, as a checkmate was unconsciously equated as a castration revenge.” And finally, psychoanalyst (and chess master) Reuben Fine: “[the game] certainly touches upon the conflicts surrounding aggression, homosexuality, masturbation and narcissism. ... [The King] stands for the boy’s penis in the phallic stage, and hence rearouses the castration anxiety characteristic of that period.” Wow, so what does it mean that I enjoy chess only on an occasional basis? That I’ve gotten over my phallic stage, that I’m not sufficiently homosexual, or that I decided that castrating my father wouldn’t be such a pleasure after all?

The point is that these quotes perfectly illustrate why Karl Popper thought that Freudian psychoanalysis is a pseudoscience: it’s no so much that the above is not true (though I very much doubt it is), but that there is not a single shred of evidence that would count for or against such statements. They are, to use Popper’s phrase, unfalsifiable.

The unfalsifiability of psychoanalysis in turn stems from the human facility at telling stories. It is much, much easier to invent an explanation for something than to do the hard work of actually testing whether the alleged explanation actually holds up to the empirical evidence. Heck, often it is even hard to figure out what empirical evidence could possibly bear on the issue to begin with!

Which brings me to evolutionary psychology, a discipline of which I’m about as fond as psychoanalysis, and for similar reasons. Like in the case of psychoanalysis, the problem is not that the basic ideas aren’t sound: sex surely is a fundamental drive of human urges and emotions, and therefore must play an explanatory role in a variety of human behaviors, just like psychoanalysts would have it. Likewise, evolutionary psychologists are certainly correct that natural selection must have played a role in shaping human behaviors and cognitive abilities, as general evolutionary theory would predict. The trouble starts when we get to detailed scenarios aiming at accounting for individual instances. The case of chess briefly described above is an example of how pompously confident psychoanalysts can be of their explanations (“It is plain that...”, “It will not surprise...”, etc.) even though they would be hard pressed to propose an empirical test of what they take to be so self-evident.

Similarly, when evolutionary psychologists like Randy Thornhill and Craig Palmer claim that rape is an adaptive strategy employed by inferior human males, or when Steven Pinker tells us that all we need to do to understand the human mind is to “reverse engineer it,” they may be correct, but most of their statements are in fact completely unfalsifiable, and therefore -- in the Popperian view (which does have its own problems, to complicate the matter) -- simply not scientific. For instance, to seriously test the rape-as-adaptation hypothesis one would have to know something about the following: a) what was the frequency of rape in ancient human populations; b) the average benefits of the behavior (i.e., number of successfully fathered offspring) vs. the likely costs (e.g., being clubbed to death by the woman’s relatives); c) some details of the social environment of Pleistocene human populations, to assess the possibility of frequency-dependent selection favoring rape, the evolutionary mechanism invoked by Thornhill and Palmer. Failing this, one would want at least to have d) a large sample of species phylogenetically close to humans in which we had data on the frequency and success of rape, for purposes of historical comparison. Needless to say, we have no clue to any of the above. And please note that the other obvious route, studying rape in current human societies won’t do either, because today’s social environment is presumably dramatically different from that of the Pleistocene, so that even if one could show a current adaptive value of rape (which is hard to imagine, by the way) one would still not have made the evo-psych case.

As for Pinker, the case is more complex because he addresses nothing less than the totality of human cognitive abilities, not a single aspect of human behavior. I do think Pinker is likely to be correct that at least some human cognitive abilities are the result of natural selection, but it is easy to make the case that they cannot all be. For instance, one can advance a plausible argument (but, importantly, no more than an argument, i.e., lacking any empirical evidence) that some mathematical ability may have been advantageous to early humans. Perhaps it was necessary to keep track of the group members' hunting scores to insure a fair division of the catch (see how easy it is to come up with a just-so story?), or whatever. But even Pinker, I hope, wouldn’t dream of suggesting that natural selection is responsible for our brain’s ability to solve differential equations -- which is what brought us to the moon and allows much modern technology to work. If I am correct in this, then at the very least the explanatory scope of evolutionary psychology is much more limited than its supporters have been trumpeting for a while.

Allegedly, even Freud once said that sometimes a cigar is just a cigar. Evolutionary psychologists hardly ever admit to the possibility of other explanations for human behaviors, and when they do, they are still clearly confident that they got the big picture right, despite the almost ridiculous paucity of evidence in their favor. OK, enough of this, I need to go back to my chess game; I haven’t managed to castrate my father yet...

16 comments:

  1. I ought to think that you're exaggerating when you say that you are literally just as fond of Evo-Psych as you are of Freudianism. Needless to say, Evo-Psych is not the most rigorous field if Psychology, and is probably on the lower end, in terms of rigor, of the whole scientific spectrum. But really, just as bad as Freudianism? Gimme a break.

    It's worth pointing out why some Evolutionary Psychologists think that the standards of rigor found in other fields need not apply to them. Steven Pinker, the good-looking poster child for Evo-Psych, has made it clear just what he thinks Science is: the application of reason to the understanding of natural world.

    So, in Pinker's mind, you just have to be logical. Look at the evidence, make the best argument you can, and you're doing science.

    It's trivial, then, for someone like Pinker to say that rape is an evolutionary adaptation. Pinker doesn't need evidence from past human epochs because he has a perfectly good bit of reasoning to back up his hypothesis. Consider, for your pleasure, this unpleasurable line by Pinker (Slate, page 347):

    "Many of the sex differences are found widely in other primates, indeed throughout the mammalian class. The males tend to compete more aggressively and to be more polygamous; the female tend to invest more in parenting. In many mammals a greater territorial range is accompanied by en enhanced ability to navigate using the geometry of the spatial layout (as opposed to remembering individual landmarks). More often it is the male who has the greater range, and that is true of human hunter-gatherers. Men's advantage in using mental maps and performing 3-D mental rotation may not be coincidence."

    Sure, Pinker, it all makes sense. Science is just making sense and nothing more, so he's being scientific as all Hell in this paragraph.

    When Pinker starts throwing garbage like that around, I do get annoyed. I also got irritated reading his article about how Jews are, on average, a good bit smarter than non-Jews. And I got irritated reading about how men have, on average, more scientific talent than women.

    But it seems to me that, in this article, you are mistaking the ravings of a popular psychologist with the study of evolutionary psychology. I don't, for example, think that there is anything particularly unrigorous about the work of David Buss on evolutionary sexual psychology. Suppose that a scientist gathers evidence on which waist-to-hip ratio is best for childbirth, and, lo-and-behold, that is the one that is most attractive to men. How unrigorous would it be for such a scientist to claim that this correlation is caused by natural selection? Not at all unrigorous, in my mind.

    ReplyDelete
  2. Criticizing psychoanalysis as pseudoscience is shooting fish in a barrel. I guess it could be called sporting. I find psychoanalytic theory like fortune telling, entertaining and capable of interesting insights, but sacrosanct only to believers.

    The conclusion seemed off target compared to the rest of the post.

    "Evolutionary psychologists hardly ever admit to the possibility of other explanations for human behaviors."

    I doubt that evolutionary psychologists as a group are that absolute, except to claim that the explanations for human behavior have natural origins (rather than imbued by one or more dieties). I can't claim to have read any of their works, but Massimo writes "Pinker is likely to be correct that at least some human cognitive abilities are the result of natural selection."

    "... and when they do, they are still clearly confident that they got the big picture right, despite the almost ridiculous paucity of evidence in their favor."

    Yet Massimo seems to agree that they do have the big picture right. "The problem is not that the basic ideas aren’t sound... evolutionary psychologists are certainly correct that natural selection must have played a role in shaping human behaviors and cognitive abilities. The trouble starts when we get to detailed scenarios aiming at accounting for individual instances."

    ReplyDelete
  3. in the Popperian view (which does have its own problems, to complicate the matter)

    Popper himself makes some mistakes, and some of those mistakes are severe. The mistakes, however, seem correctable and falsifiability as a demarcation criterion seems rock solid.

    (I also think naming whole schools of thought after specific philosophers is a mistake. Scientists do not usually privilege individuals to quite the same degree. I'd prefer to call it Falsifiability rather than the "Popperian view.")

    ReplyDelete
  4. Your way of thinking about human cognitive abilities seems to misplace the focus. Does anyone really think we have an "ability to solve differential equations" as an independent cognitive capacity, emphasis on the independent? More plausibly, we have flexible cognitive capacities which can be trained for a large variety of cognitive tasks. The appropriate question is "why did evolution give us cognitive faculties flexible enough to be trained for a wide range completely novel mental tasks?"

    When you look at cognitive faculties at the right level of analysis, it's hard to see any reason to resist the conclusion that a solid majority of them formed because they were adaptive. Our mental faculties aren't minor features of our species. On a crude physical level, our brains are big enough to frequently kill mothers giving birth. Yes, some features of organisms are byproducts of something else, or products entirely of contingency, or the result of parasites messing with them. However, are any of these explanations plausible in the case of our big brains and the flexible minds they support?

    ReplyDelete
  5. When you look at cognitive faculties at the right level of analysis, it's hard to see any reason to resist the conclusion that a solid majority of them formed because they were adaptive.

    Regardless of the level at which you look at any feature, that the feature is adaptive is a positive hypothesis that must be specifically supported by data. That's the fundamental lesson Gould was trying to teach us in his criticism of the adaptationist fallacy.

    It is simply not sufficient to say, "it's hard to see any reason to resist the conclusion that a solid majority of them formed because they were adaptive." That's not science. We shouldn't "resist" conclusions, we should always try to substantiate them.

    Furthermore, the idea that the majority of our cognitive abilities are adaptive is not very interesting, even if it were definitely true. Which abilities are adaptive? Adapted for what? What were the specific selection pressures? Why this ability and not some other?

    Answering such questions requires more than just armchair speculation and just-so stories.

    ReplyDelete
  6. Joseph,

    "I ought to think that you're exaggerating when you say that you are literally just as fond of Evo-Psych as you are of Freudianism. "

    Yes, but only slightly, and your example from Pinker, I think, makes my case rather nicely...

    "I don't, for example, think that there is anything particularly unrigorous about the work of David Buss on evolutionary sexual psychology."

    Actually, his study of the waist-to-hip ratio in women has been severely criticized, as it turns out that there is more variation among cultures than he reported, and that such variation is correlated with the degree of Western influence on those cultures. No, I really can't think of much serious stuff that has come out of the evo-psych literature, honestly.

    Hallq,

    "Does anyone really think we have an "ability to solve differential equations" as an independent cognitive capacity, emphasis on the independent? More plausibly, we have flexible cognitive capacities which can be trained for a large variety of cognitive tasks."

    Of course, but that isn't what evo-psychs are saying. They reject the model of a general-purpose brain and insist on naturally selected specific "modules." To some extent the brain does have localized structures and functions, but not to the degree implied by evo-psych, and it is of course an entirely different matter whether specific human abilities were selected or not.

    As for solving differential equations, of course nobody seriously argues that that ability evolved by natural selection. But that is my point: evolution can explain only the embryonic versions of some human cognitive capacities. The rest, it would seem, is a matter of cultural evolution, which is anathema to evo-psychs.

    ReplyDelete
  7. Hmmm... maybe I'm just not reading the flakier evo psych people. Pinker has his line about the differences between Einstein and a high school drop-out being trivial compared to the differences between a high school drop-out and a chimpanzee, and has suggested some of what people claim about cultural differences is a matter of taking official propaganda at face value (say, it's not that romantic love doesn't exist in some cultures, it's just that it's official disapproved of and hidden away). But is there anywhere Pinker is so dismissive of cultural evolution, that it's actually fair to say calculus is a problem for his views?

    ReplyDelete
  8. "Actually, his study of the waist-to-hip ratio in women has been severely criticized, as it turns out that there is more variation among cultures than he reported, and that such variation is correlated with the degree of Western influence on those cultures. No, I really can't think of much serious stuff that has come out of the evo-psych literature, honestly."

    Do you mean that there were allegations that he fudged his numbers, or that there were simply new studies performed which showed greater variance?

    If he actually fudhed his numbers, then he's a downright shmuck, needless to say.

    But if he merely got different numbers the first time than other scientists did, I don't see how we can charge that he was being pseudoscientific. If he was refuted, that must mean that his hypotheses were subject to experimental refutation...

    ReplyDelete
  9. Joseph,

    no, no allegations of fraud, just that he did not consider any measure of cultural relatedness among groups. In other words, sloppy science. And the line between sloppy science and pseudoscience is sometimes not very sharp, as I'm arguing in the book I'm writing now (but that will come out toward the end of next year, so we've got time... :)

    ReplyDelete
  10. I've been thinking about this comment by Joseph:

    "Steven Pinker, the good-looking poster child for Evo-Psych, has made it clear just what he thinks Science is: the application of reason to the understanding of natural world. ... So, in Pinker's mind, you just have to be logical. Look at the evidence, make the best argument you can, and you're doing science."

    First off, I'd like to know if anyone has an actual quote from Pinker to that or similar effect.

    In general, however, it seems interesting to me that Pinker is describing what philosophers do, not scientists (this used to be called natural philosophy, appropriately enough, and it was the forerunner of science).

    This is important because -- as readers of this blog know -- I certainly don't discount philosophy, on the contrary, I find the borderlands between philosophy and science (sometimes referred to as "sci-phi") fascinating.

    But Pinker, like Dawkins in his God Delusion book, is making an argument from plausibility, meanwhile ignoring that he is doing philosophy, not science. Plausibility does have a role to play, and a plausible argument is certainly better than an implausible one. But as Joseph points out, science is something different.

    That is in line with the idea that, in a nutshell, I consider the general idea of evo-psych plausible, but not scientific. This is also why it wouldn't strictly be a pseudoscience, but a non-science, or a border-science.

    ReplyDelete
  11. By the way, if anyone is interested, the paper debunking evopsych's rake on the waste-to-hip ratio in women is:

    Losey, J. E., L. S. Rayor and M. E. Carter (1999). Is beauty in the eye of the beholder? Nature 396: 321-216.

    ReplyDelete
  12. "Likewise, evolutionary psychologists are certainly correct that natural selection must have played a role in shaping human behaviors and cognitive abilities,.... The trouble starts when we get to detailed scenarios aiming at accounting for individual instances."

    Interesting discussion! This statement above by Massimo segues into what I think is one of the problems of evol-psych. That evolution shaped human behaviour is a truism, the problem is of specific behaviours. I propose the opposite, evolution equiped humans to have highly variable behaviours, and adaptation based on malleable cultural behaviours.

    For example, for most animal and plant species, they are adapted to a fairly narrow range of ecological circumstances. Humans via culture were able to adapt to just about any ecological circumstance on earth, from hunting/gathering in the arctic, to the San desert, to tropical forests, to alpine, to maritime ad naseum. Then humans developed agriculture, through which they shaped whole new ecological niches to their needs. All these adaptations did not have much to do with bio-genetic and morphological changes.

    Then there is the issue of sex for human evolution. While certainly sex drive is the result of evolution, that doesn't stop quite a few crazy Catholic priests and other wierdos from deciding to not have sex or offspring, resisting the greatest evolutionary imperative, to have sex and reproduce! Plus there are many people in modern society who, while they may have sex, actively choose not to reproduce due to various cultural factors.

    ReplyDelete
  13. Dear Massimo,

    No offense, but I think you're strawmanning ev-psych a bit, pulling out the most ridiculous cases like Dawkins supposedly does with religion. Your favorite is the rape case; I've seen you mention it in almost all discussions you've had of ev-psych. Of course the rape case is ridiculous but that shouldn't damn the whole field. Take an ev-psych claim that was, in fact, testable, tested, and falsified: homosexuality is advantageous because homosexuals will care for their nieces and nephews. This was tested and homosexuals do not, in fact, care more for their nieces and nephews than heterosexuals do.

    I think you have to assess ev-psych on a case by case basis. Some claims are more ridiculous than others. What do you think?

    Sincerely,
    Mark

    ReplyDelete
  14. Mark,

    yes, some cases are more ridiculous than others, though the rape one is a classic, by two major exponents of evo-psych, so I consider it perfectly fair game.

    As for homosexuality, you picked a good one, but it is good precisely for the reasons that Jonathan Kaplan and I think some evo-psych research can actually be done: homosexuality as a kin or group care strategy can be tested because it isn't just human: one can observe it and experiment on it in a variety of other species, including several birds.

    Jonathan and I do not argue that evo-psych is doomed as a matter of principle, only that they insist in picking the worst types of human behaviors to study, the so-called "universals" that are human-specific (and they do so because those traits are "sexy").

    If a trait is a universal among humans, then there is little or no variation within the species to make comparisons for tests; and if it is truly a human-only trait then one can't make comparisons among species, leading to the worst case scenario of just-so stories.

    ReplyDelete
  15. Dear Massimo,

    Thank you for your response. You make good points and I finally understand your beef with evo-psych and your comparison with psychoanalysis. So, basically, if evo-psychs want to be doing better science, they need to be putting the EVO into evo-psych instead of being anthropocentric.

    Sincerely,
    Mark

    ReplyDelete
  16. Massimo, on the topic of mathematical ability. Would it be fair to say that there may have been a selection pressure for biological hardware that allowed for the abstract abilities needed to do maths? I'm finding it difficult to draw the line between evolved ability and non evolved ability. Or is your main point that there isn't enough data to support the hypotheses and that hypothesising without the science is what is being done in evolutionary psychology.

    ReplyDelete

Note: Only a member of this blog may post a comment.