‘What is Evolutionary Psychology?’ at Skeptics in the Pub Glasgow

A couple of evenings ago I spoke at Skeptics In The Pub Glasgow. Skeptics In The Pub is “an informal social event designed to promote fellowship and social networking among skeptics, critical-thinkers, and other like-minded individuals”. I was delighted when, back in December, I received an invitation to speak on the topic ‘What Is Evolutionary Psychology?’.

I already knew that Evolutionary Psychology (EP) is unfairly maligned: I’ve taken part in more than enough conversations in which gross misunderstandings of EP have been expressed, including those with people who, in my view, should know better. However, I hadn’t appreciated the extent and depth of misunderstanding out there until I started to pull this talk together. As I pointed out at the beginning of my talk, if you type in ‘Evolutionary Psychology is…’ into Google, its first prediction about how you will finish your search term is: ‘..bullshit’. (For comparison, for both ‘Social Psychology is…’ and ‘Evolutionary Biology is…”, the first prediction is: ‘..the study of…’.) It is painfully easy to find statements that simultaneously denounce and misunderstand EP, including, as I say, from people who should know better: high-profile skeptics; academics (including those that use evolutionary thinking in their own research); and so on.

Better people than I have already addressed these criticisms, and at length too. In 2000 a book entitled Alas, Poor Darwin: Arguments Against Evolutionary Psychology brought together a collection of authors united in their distaste for EP. It contained all the standard criticisms: that EPists are genetic determinists; that EP is politically motivated; that EP is unfalsifiable; and others. In an excellent review of that book, Rob Kurzban showed how not only are these claims false, they are “infuriatingly false”, because EPists have frequently said precisely the opposite of what they are criticised for, yet when this is pointed out the critics pay no attention, and repeat the criticisms anyway. Consequently, the idea that ‘EP-is-bullshit’ is, sadly, alive and kicking. Kurzban’s review is both an excellent summary of the major criticisms, and an explanation of why they are misplaced. I urge anyone tempted to conclude against EP to read it before they do so.

Fortunately, I did not have to deal with any EP-is-bullshit merchants either during or after my talk. I’d like to imagine that some of them came along but realised during my talk that they’d been misinformed – but probably they just didn’t come. But most people I spoke to said that they didn’t know much about it, except that it seemed to be controversial, and they were pleased to learn more.

There was a Q&A session after the talk. One question I found very interesting was about looking for the give-away signs of natural selection. The human eye is an exquisitely well-designed thing, with one glaring defect: it is built backwards and upside-down (this is why we have blind spots). Have we, the questioner asked, got any similar examples from the human mind? This query could be interpreted as being about whether there are aspects of the human mind that are suboptimal – and if that was the question, there are many examples. But I think the question is more subtle than that. It’s really: Are there any cases where the sub-optimality that we do see is effectively a signature of the action of natural selection? From an evolutionary perspective the vertebrate eye is interesting precisely because, despite its exquisiteness, no intelligent agent would have designed it with such an obvious and easily-fixable defect. This is why it is often used in arguments with creationism. It would be nice to document a similar case with the human mind. The answer I gave on Monday evening was that no, we don’t yet have such a case, mainly because we simply don’t yet understand the mind nearly well-enough – but perhaps we will find such a case in the future. Having reflected on it a little more since, I think that’s right: our understanding of the mind is very basic relative to our understanding of the eye. If and when we understand the mind as well as we do the eye, I may be able to give my questioner a more satisfactory answer.

In the meantime, if anybody tells you that EP is bullshit, ask them why – and then go and check whether EP actually says the bullshit things its critics claim it does. Chances are it doesn’t.

Advertisements

12 Replies to “‘What is Evolutionary Psychology?’ at Skeptics in the Pub Glasgow”

  1. Part of the reason for the delay on eye type observations is that the mind is not a well defined or easily pointed to thing, unlike the eye. In a sense the question will never be answered as the mind is definitionally opaque I fear. But thinking about behaviour and cognition will lead to answers and the literature on biases in reasoning is especially about blind spots and adaptations. That should be a good source of examples.

    Of course, those reflex EP-phobes probably exhibit a blindspot too but whether it is an adaptation to deny adaptationism is an open empirical question… I did once see a paper at a conference reporting on reproductive success within EP compared to postmodern scholars. Goodness knows where the data came from, but I am sure you can guess the predicted and supported hypothesis…

  2. Hi Thom,
    I tend to start with the “EP is bullshit” assumption, but thankfully nobody expects me to know better.

    Re. the point about suboptimal aspects of the human mind as a signature of their evolution: it’s potentially an interesting question. Isn’t the mundane answer that most of the standard explanations for suboptimality that EP people evoke to defend themselves against the charge of adaptationism (e.g. aspects of the mind as spandrels, as products of stochastic evolutionary processes, as adaptations to earlier environments) fall into this category, since they’re based on the idea that you’re dealing with a stochastic, local, slow, conservative evolutionary process?

    Alternatively, maybe you couldn’t think of any examples equivalent to the eye because, for the eye but not for the mind, we know what the function is, what problem it solves, and therefore it’s fairly easy to see that a blind spot is suboptimal, at which point you try to figure out why it’s suboptimal. My impression for the mind is that, unlike for the eye, we often don’t know what the function of a particular ‘component’ is, and it is therefore tempting to adopt the approach of saying “I know the mind does X: what is the function such that X is the solution?”. If that’s the way you tend to think, even if it’s only your starting hypothesis, it’ll be hard to identify suboptimal aspects of the design of components of the mind, unless they’re flagrantly suboptimal, and in fact if there are suboptimal aspects of the design of the mind you’ll tend to infer the wrong function – imagine asking “what is/was the function of the eye, such that having a blind spot is/was the best solution?”.

    Kenny

  3. Thanks both for the comments.

    Tom – Yes, I agree. Indeed, the lack of clear definition of what we even mean by the mind is symptomatic of how much further advanced is our understanding of the eye, in comparison to the mind. As for EP-phobia itself being an adaptation – that sounds like a bonkers presentation that I would quite like to have seen.

    Kenny – Did you read the Kurzban review I linked to? I’d be genuinely interested to hear your thoughts about it.

    I’m not sure I follow your first paragraph. Could you give me an example or two?

    As for function – yes, I think that’s right. Indeed, I made the point in my presentation that all vision scientists are effectively EPists. I used a nice quote from a professor of vision and perception at Yale saying just that. And so yes, we should not be drawn into questions analogous to “what is/was the function of the eye, such that having a blind spot is/was the best solution?”. That’s why EP emphasises the importance of identifying the task first (a task that is itself not straightforward).

  4. Hi Thom,

    > Kenny – Did you read the Kurzban review I linked to? I’d be genuinely interested to hear your thoughts about it.

    I did – I think I’ve read it before, but I had a quick skim before I wrote my comment, mainly to check I wasn’t just about to repeat one of those criticisms. I don’t know how accurately he’s depicting the criticisms – presumably he is, in which case he may well be right in saying they’re unfounded. Two thoughts though. Firstly, coming up with a couple of counter-examples to a particular criticism isn’t very convincing – it’s possible to think that the EP approach is often badly done without thinking it’s necessarily always badly done. Second, I think some of the criticisms have validity expressed in a weaker form. I guess what I was shooting for in my original comment was a weakened version of the “panadaptationism” criticism – I don’t think that all EP people think everything is an adaptation, but I think that for at many of them it’s their null hypothesis. Similarly, with the genetic determinism stuff: I don’t think EP people are stupid, but I think there’s possibly an over-emphasis on biological evolution. For instance, it’s a bit of a mystery to me why natural selection of cultural variants (still natural selection, just operating on traits inherited by a different route) isn’t more main-stream in EP.

    > I’m not sure I follow your first paragraph. Could you give me an example or two?

    The whole Environment of Evolutionary Adaptedness explaination for people’s suboptimal behaviours in the modern world (I think eating too much fat is everyone’s favourite example) hinges on the idea that there is some timelag in the process that leads to adaptation – evolution by natural selection, especially biological evolution, would predict such lags. As another example, Kurzban in that review talks about “random processes”, and on consulting the Tooby & Cosmides article he was citing I see this refers in fact to linkage, where selection on one trait can drag other non-functional but linked traits to high frequency – again, that’s part and parcel of an evolutionary process where traits aren’t inherited seperately.

    By the way, in the process of checking an EP textbook for an example of a maladaptive trait explained by timelag (I looked in Barrett, Dunbar & Lycett, “Human Evolutionary Psychology”), I came across this nice quote which I think exemplifies the point I was trying to make re. not spotting maladaptive aspects of an organisms’s mind because you assume they aren’t there:

    “We should beware of concluding that a trait is maladaptive simply because we cannot see an obvious advantage to it… The claim that a trait has no function or is maladaptive may simply be a statement of ignorance. As a general rule of thumb, therefore, the conclusion that a trait is maladaptive should be an explanation of last resort after all other possible adaptive explanations have been excluded (Dunbar, 1982)” (p 14 of Barrett et al).

    That sounds like the opposite of what you are saying is best practice, i.e. knowing the function first. Again, it’s not the case that everyone does the wrong thing all the time, but it’s not hard to imagine that applying that sort of rule of thumb might shape the conclusions you draw, and sometimes in the wrong direction. Mischevilously, the following opposite heuristic looks like a safer one to me: “We should beware of concluding that a trait is adaptive simply because we think we can see an obvious advantage to it… The claim that a trait evolved to fulfil a certain function may simply be a statement of ignorance. As a general rule of thumb, therefore, the conclusion that a trait is adaptive should be an explanation of last resort after all other possible non-adaptive explanations have been excluded”.

  5. Hi Kenny.

    > Coming up with a couple of counter-examples to a particular criticism isn’t very convincing – it’s possible to think that the EP approach is often badly done without thinking it’s necessarily always badly done.

    Of course it’s possible to think that. But: I don’t know of any good evidence that bad science is any more prevalent in EP than elsewhere. There’s plenty of rubbish in, say, differential psychology, but there is no ‘differential psychology is bullshit’ meme. No sign of one either. Same for social psychology, neuroscience, psycholinguistics, whatever.

    Either way, I don’t think the criticisms that Kurzban is addressing – and the ones that I think are behind the EP-is-bullshit meme – are about lax practice. They are more fundamental than that: for many critics, the very idea of EP is taken to be bullshit. Kurzban illustrates very clearly that this is simply wrong: EP’s foundational texts just do not say the things that critics of this sort claim they do.

    > Second, I think some of the criticisms have validity expressed in a weaker form. I guess what I was shooting for in my original comment was a weakened version of the “panadaptationism” criticism – I don’t think that all EP people think everything is an adaptation, but I think that for at many of them it’s their null hypothesis.

    Well, whether that is true of EPists or not is an empirical question, and again, I don’t see the data on that to justify the EP-is-bullshit meme. But either way, I’m not sure I see why your alternative null hypothesis is any better. What the null hypothesis is depends on the specifics of each case. My belly-button is the perfect size and shape to store hazelnuts, but common sense tells us that this shouldn’t be the null hypothesis. At the same time, theory-of-mind, say, is sufficiently cognitive demanding and clearly important to human life that it would be odd to assume as a default hypothesis that it’s not an adaptation of some sort.

    Following on from that, I wouldn’t say the Dunbar quote is the opposite of what I’m saying. The function-first approach (which EP shares with evolutionary biology) says: “Work out what the task is, and then predict what the mechanism should look like”. Dunbar is adding a coda: “..but if your predictions don’t match up the first time, don’t be too quick to assume that there’s no function” (but I do grant that he’s making that point in a particularly strong way).

  6. > I don’t know of any good evidence that bad science is any more prevalent in EP than elsewhere. There’s plenty of rubbish in, say, differential psychology, but there is no ‘differential psychology is bullshit’ meme. No sign of one either. Same for social psychology, neuroscience, psycholinguistics, whatever.

    On the contrary, in the past couple of weeks I’ve seen an anthropologist worrying about systemic problems with social anthropology, and a neuroscientist worrying about the prevalence of locationalist bullshit in neuroscience, concerns that I’ve come across before. It’s not necessarily a problem unique to EP, or even worse in EP than elsewhere – the key thing is to figure out if there is a problem (which I guess is what we’re arguing about now), and if so what to do about it.

    > My belly-button is the perfect size and shape to store hazelnuts, but common sense tells us that this shouldn’t be the null hypothesis.

    Well indeed, but if we ignored your very sensible not-an-adaptation null hypothesis and did what the textbook seems (to me) to be saying, we’d have to assume that the shape of your belly-button is an adaptation to *something*, and we just didn’t figure out what yet. I think there’s an interesting question here about what our null hypothesis should be when trying to explain the behaviour of a complex, learning, cultural animal like a human. But coming back to the topic of your original post: if sub-optimal or maladaptive aspects of an organism’s design are particularly revealing about the evolutionary processes that shaped that organism, assuming that such features don’t exist might be a bad idea, since it cuts you off from the kind of evidence you actually really want to find.

  7. > On the contrary, in the past couple of weeks I’ve seen an anthropologist worrying about systemic problems with social anthropology, and a neuroscientist worrying about the prevalence of locationalist bullshit in neuroscience, concerns that I’ve come across before.

    Oh, sure, within disciplines there are certainly these worries. What I meant my saying the meme isn’t out there is that if you type “social anthropology is” into Google, its first prediction is not “bullshit” (I just checked).

    Just to be clear: I’m perfectly willing to accept that there is poor work in EP, just as in any discipline. What I’m resisting is any suggestion that the agenda of EP (in a broad sense) is fundamentally flawed.

    > if we ignored your very sensible not-an-adaptation null hypothesis and did what the textbook seems (to me) to be saying, we’d have to assume that the shape of your belly-button is an adaptation to *something*, and we just didn’t figure out what yet

    No. The textbook says: determine what the tasks are that humans need / needed to solve, and based on that work out what the optimal sort of mechanism would be (within sensible limits of what sorts of mechanisms the organism might be expected to evolve – the best way for humans to travel large distances is to fly, but we shouldn’t then predict that we should have wings). Then go and test that prediction. And if the results don’t match your predictions, then reassess. What form that reassessment will take depends on the specifics, of course.

  8. Brilliantly, though, if you type “neuroscience is …” into google, it guesses “bullshit”.

    Re. what the textbook says: I guess we’ll have to disagree what that particular textbook says – it seems quite clear to me. Re. what it *should* say: I’d also suggest you need to add to the end of your procedure “If the results *do* match, consider whether there are alternative explanations for the apparent fit between form and function”. You should probably also do a ton of other stuff about figuring out the evolutionary history of the trait (e.g. is it a recent innovation in humans, or shared more widely among primates?), working out whether the trait is actually evolvable (is there plausibly a relatively smooth path of ever-increasing fitness from absence to presence of the trait? is the trait actually inherited by the mechanism you think, e.g. genetically?), and maybe doing some comparative stuff to see if we can identify other species that face similar challenges in their environment and find similar solutions (e.g. cases of convergent evolution, as has been done for the eye). Obviously not every paper is going to contain all that stuff, but I think spotting a potential fit between form and function is only part of the story – personally, I’m rather keen at looking at alternative explanations, since I’ve had some mileage from that in my own work.

  9. Poor old linguistics fails the google test miserably: “linguistics is … a science”. Smacks of insecurity to me.

  10. Wow – I would not have guessed that neuroscience has enough critics to fail/pass* the Google-bullshit-test. Indeed, I didn’t even bother to check – I assumed that if either of the disciplines you mentioned would fail/pass*, it would be social anthropology. That will learn me.

    On doing the other research: it goes without saying that that is all good. But note that those are mostly phylogenetic questions you’re pointing to. All interesting and valuable, of course, but EP’s focus, like psychology itself, is on mechanism. It uses function to inform/constrain that, and phylogeny can be used for the same purposes. But EP is not principally about evolutionary *history*; it’s about using evolution as a way to understand mechanism.

    On the textbooks: considering function first doesn’t mean that one can’t coherently do things the other way around, as least heuristically i.e. take a trait (especially one that seems to carry costs), assume that it has a function, and try to work out what that function is. There’s nothing a priori wrong with this, so long as we keep in mind that our assumption that there is a function may be flawed. In this case, Dunbar’s coda does not apply nearly as much.

    * Delete as appropriate – I’m not sure what counts as a pass and what as a fail.

  11. Rather than look for suboptimal design in cognitive adaptations that are comparable to the structure of the eye, it could be useful instead to consider the variety of documented biases we see in cognition and perception that also reveal the signature of selection. There are many empirically solid options, including auditory looming (Neuhoff), height and distance perception (Jackson), and cross-sex mind reading (Haselton).

    The systematic biases in decisions under uncertainty across these domains are not suboptimal in a klugey sort of way (at least not obviously), but rather are responses to different constraints presented by the respective adaptive problems (each with their own specific design). The beauty of these examples is that they generate quite predictable effects, reveal the hand of selection, and the general principle (i.e., error management) has been formalized.

    Just a thought.

    More google data: Sociology is “a joke,” psychology is “bullshit,” and cognitive science is “the downfall of psychology”

Leave a reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s