Evolutionary Psychology Under Fire

Steven Pinker is into evolutionary psychology.
This subject looks as though it has the potential
to be a choppy ride!

Canadian biochemist and evangelical atheist, Larry Moran, has recently posted a blog entry that piqued my interest. The post raised questions about the status of evolutionary psychology as a scientific discipline (if "discipline" it can make claim to!). Neither Moran nor his fellow evangelical atheist PZ Myers have a great deal of respect for evolutionary psychology. I have seen Evo-psych described as "evolutionary just so stories". 

I'm in no position to contradict Moran's and Myers' opinion of evo-pysch. In fact it is quite likely they are right. But are they asking themselves the right questions here? Why is evo-psych, from a formal science angle, so flaky? Given the epistemic distance of the objects it is studying is it ever likely to be anything else? That is, are there fundamental epistemic reasons why it lacks the standards of the physical science? How would Myers and Moran fare as evolutionary psychologists? Would they do any better?  Would they be aware of the epistemic distance of their objects of study? Would they take that into account and with all due allowance cautiously tender tentative conclusions about the natural history of behaviour? Would they be less dogmatic about their conclusions? Would they feel too stultified by the epistemic difficulties and tender no conclusions at all?

Evo-psych is a discipline which mixes biology, natural history, behavioral science, and social science along with a sprinkling of topical political intrigue relevance. Add to that the epistemic distance of its subject matter and you've got an emotionally explosive recipe that's right up the street of left and right wing partisans.

One of the specific criticisms that Moran has of evo-psych is this: 

The field of evolutionary psychology is full of hyper-adaptationist thinking. It's primary task is explaining modern features of human behavior as adaptations that took place in primitive human populations. From an evolutionary perspective, this requires that the behavior has strong enough genetic components to be subject to evolution by natural selection. It requires that primitive populations contained alleles for the modern behavior as well as alleles for a different behavior that reduced fitness. Finally, it requires that selection for the modern behavior is strong enough to lead to fixation in just a few hundred thousand years.

All of these assumptions require supporting evidence that is almost always missing in evolutionary psychology publications. In the absence of evidence, the default assumption should be that the behavior is cultural. If there's evidence of a genetic component then the default assumption should be fixation by drift unless there's evidence of selection

In his second paragraph above I'll bet that Moran is right in saying that the evidence is often lacking for evo-psycho; but then that's measuring it by the standards of the physical sciences. As I have recorded on this blog before Moran appears to have a poor appreciation of the fact that the objects of our world have differing epistemic distances and this will have an impact on the relative rigor of the science that tries to grapple with them. In evo-psych we are not dealing with spring extending and chemical precipitating science.  See the following links for more on this subject:

http://quantumnonlinearity.blogspot.co.uk/2014/01/evidence-not-proof.html

http://quantumnonlinearity.blogspot.co.uk/2013/09/yes-its-all-science-larry-but-not-as.html

http://quantumnonlinearity.blogspot.co.uk/2013/08/epistemic-notes.html

http://quantumnonlinearity.blogspot.co.uk/2012/08/its-science-larry-but-not-as-you-know-it.html

http://quantumnonlinearity.blogspot.co.uk/2015/02/evidence-guide-lines-but-not-tram-lines.html

None of this is to say evolution psychology is right; its just a plea to give a little leeway and understanding to those who grapple with an uncooperative epistemology as they try to hammer out a world view. However, I lose all sympathy with such people if epistemic humility is replaced by dogmatism and an abrasive intimidation of unbelievers; that's something I'm all too used to from Christian fundamentalists. 

***

The process that Moran describes in his first paragraph above is, from a standard evolutionary perspective, quite right; this is precisely what would be required for the propagation of those behavioral alleles that are supposed to offer social advantages in the breeding game. And if the alleles don't provide definite advantages it is likely that DNA, by way of a kind of "linguistic" drift, gets a "local accent".

But I don't go along with Moran when he demands that in the absence of evidence, the default assumption should, repeat should, be that the behavior is cultural. OK, one can adopt the working hypothesis that it might be cultural, but in that case one is thrown into the domain of social anthropology, a discipline whose epistemic standards are not going to be commensurable with that of the physical sciences. Social anthropology will seem just as much lacking in the rigor that a physical scientist expects.  My own heuristic here would be to refrain from a dogmatic position either way and proceed slowly, carefully and tentatively; the extent to which identified human behavior patterns are influenced by genetic factors is a wild card. 

In any case there is a distinction between the claim that a behavior has some kind of genetic influence and the evolutionary "just-so" stories one might seek to tell in an attempt to explain the natural history of this influence. Thus, we have two separate questions here:  a) Is an identified behavioral trait influenced by genes and/or culture? b) If genes are involved and we believe in standard evolutionary mechanisms the next question is how did the behavior evolve?

Evo-pysch is by and large engaged in creating retrospective
 sense making narratives rather than a predictive. science.

I think that Larry Moran is out of luck if he wants to see an improvement in the scientific standards of evo-psych in terms of its evidential base because epistemology is against it. What we might see (although I doubt it) is a less emphatic presentation of evo-psych conclusions once the epistemic issues are acknowledged.  We can but hope.

As I have already said in connection with Moran's concept of science, some objects are just too complex and have too great an epistemic distance to be amenable to the standards of evidence expected in the physical sciences; in particular the expected predict and test methods of "spring extending and test tube precipitating science" aren't necessarily going to be realised. And it's not as if there is a sudden cut-off between the rigors of physical science and the more nebulous objects of the social sciences; for objects of scientific study range from the very epistemically accessible like springs, through planetary motions and what was in yesterday's lunch to, the epistemic barriers found in the study of strings and mysteries like the identity of Jack the Ripper.

As the objects we study get increasingly complex and epistemically inaccessible "predict and test" becomes less effective as a heuristic. History, of course, is a prime example. Statements about the past can often be tested against new found data, but therein lies the rub; the data has to be found and if it doesn't turn up the crucial test of an historical theory is left hanging. But when historical test data isn't extant there is a fall back heuristic; we can then proceed abductively; that is, we seek  a "best fit explanation" given the data we do have. However, what constituents a "best fit explanation" may be subject to some ambiguity: This is a bit like a "joining the dots" game; if there are not enough dots more than one explanatory "sense making" structure fitting the data may be devised. As the quote from Pinker suggests evo-psych has a strong retrospective (as opposed to predictive) sense making role; evo-psych, under the best of circumstances, is not going to be an exact science.

Even when an object's epistemic distance means that its data is sparse there's nothing to stop us proposing a conjectural dot-joining theory, but in doing so being appropriately tentative about our conjecture. Here, however, human nature faces a challenge. There may be many objects for which the available evidence doesn't quiet unambiguously determine the explanatory narrative being proposed. Under this kind of epistemic environment what is being proposed is therefore worldview sensitive. It is then that evangelicals of all flavours tend not to adopt the appropriate level of epistemic humility toward their theories, but instead take on the stance of enthusiastic advocates, champions and partisans for their heroic cause. Evangelicals, whether atheist or otherwise, polarise opinion as they sharply demarcate their field of knowledge and stake out their noetic claims. Try not to meet them in a dark alley if you are less committed than they are!

Ironically it doesn't help to have people like Steven Pinker around. Pinker is a mouthy, suave and, worst of all, an incredibly intellectual guy with a grasp of a wide range of subjects. But I wonder if he's spent hours in the lab separating out precipitates or carefully measuring spring extensions or has spent hours cleaning and measuring fossils? Don't these clever people make you spit! I have a sneaky feeling that interpersonal feelings, as well as world view, might have a bearing on the debates surrounding retrospective sense making science!

****

Larry Moran finishes his post by providing a very interesting list of what one evolutionary psychologist claims are the notable achievements of his subject. Viz:

1. Women alter their preferences for the facial features of men as a function of where they are in their menstrual cycles. When maximally fertile, they prefer men possessing markers of high testosterone.

2. Babies display an immediate instinctual preference for symmetric faces (at an age that precedes the capacity for socialization).

3. Children who suffer from congenital adrenal hyperplasia display a reversal in their toy preferences. Furthermore, using inter-species comparisons, vervet monkeys display the same sex-specific patterns of play/toy preferences as human infants. This suggests that contrary to the argument made by social constructivists, play has an evolved biological basis.

4. Individuals who score high on an empathy scale are more likely to succumb to the contagion effects of yawning. This is indicative that this particular contagion might be linked to mimicry and/or Theory of Mind.

5. How provocatively a woman dresses is highly correlated to her menstrual cycle (a form of sexual signaling found across countless Mammalian species).

6. Culinary traditions are adaptations to local niches. For example, the extent to which a culture utilizes meat versus vegetables, spices, or salt is a cultural adaptation (this is what behavioral ecologists study).

7.  Maternal grandmothers and paternal grandfathers invest the most and the least respectively in their grandchildren. Whereas all four grandparents have a genetic relatedness coefficient of 0.25 with their grandchildren, they do not all carry the same level of "parental uncertainty." In the case of maternal grandmothers, there is no uncertainty whereas in the case of the paternal grandfather, there are two sources of uncertainty. This last fact drives the differential pattern of investment in the grandchildren.

8. Good male dancers are symmetric (paper published in Nature). One would expect that some behavioral traits might correlate with phenotypic quality as honest signals of an individual's desirability on the mating market.

9. Self-preference for perfumes is linked to one's immunogenetic profile (Major Histocompatibility Complex).

10. When a baby is born, most family members (especially those of the mother) are likely to state that the baby looks like the father. This phenomenon is found in countless cultures despite the fact that it is objectively impossible to make such a claim of resemblance. The reason for this universally found cultural tradition lies in the need to assuage the fears of paternity uncertainty.

11. Environmental stressors (e.g., father absence) and the onset of menarche (first menses) have been shown to be highly linked. In numerous species, the likelihood of a female becoming reproductively viable is affected by environmental contingencies.

12. Women are less receptive to mandatory hospital DNA paternity testing (for obvious reasons). In other words, their willingness to adopt a new product/service is fully driven by an evolutionary-based calculus.

13. Women can smell the most symmetric men. In other words, women have the capacity to identify men who possess the best phenotypic quality simply via their nose. This is what I have referred to as sensorial convergence.

14, Using fMRI, the exposure to ecologically-relevant stimuli (e.g., beautiful faces) yields distinct neural activation patterns in men and women.

15. In choosing a mate, humans tend to prefer the smell of others that are maximally dissimilar to them along the MHC. This ensures that offspring possess a greater "defensive coverage" in terms of their immunological system.

That's certainly an interesting list of theories most of which I've haven't heard before! I will look at human behaviour differently from now on! Moran suggests that this list should be approached with these questions in mind:

a) Is there evidence for genes (alleles) that are responsible for this trait?

b) Is there evidence that in primitive societies this trait improved fitness more than the original, presumably deleterious, trait?

c) Is there evidence that this is a universal trait present in all human populations?

Good questions and Moran probably thinks the evidence for the listed achievements is wanting!  Going back to comments I have already made:. There are in fact three distinct questions here: Firstly,  the human behavior patterns claimed by the above list have the potential to be tested and perhaps settle the debate about their existence (Moran's question c).  But, secondly, even if they should be verified as human behavior patterns, there is then the question of whether they can be traced back to a genetic influence. (Moran's question a). Thirdly, if there is a genetic factor, does the evolutionary "just so story" add up? (Moran's question b)

Grasping the nettle of human nature theory. 

Another reason why I think people are so partisan about this subject is that it raises the contentious questions of the nature-nurture debate, a debate which is bound up with political interests. If such a thing as "human nature" is denied and human beings are thought of as "blank slates", that is tantamount to saying that they are very flexible inter changeable components which can be pressed into shape to fit the society of our dreams. I submit that both the Marxist and libertarian social dreams come unstuck on the question of human nature of which their economic reductionism fails to sufficiently take into account.

The  denial that there is even such a thing as "human nature" is certainly wrong. Human minds are highly proactive surfaces that select and categorize the data which falls on to them according to interest, motive and a set of predetermined neural constants that probably vary from person to person. Pinker is no doubt right to criticize "Blank Slate" theory. However, needless to say, separating out the nature and nurture factors has proved difficult, difficult enough to keep the genes vs culture debate fueled. And that's before we even get onto the question of the natural history of those behavioral patterns and genes.

***

Finally I must add a disclaimer. I think it likely that standard evolutionary mechanisms are insufficient to explain natural history. So, I have been exploring what I call "intelligent creation". I distinguish this from paradigm of the god-of-gaps Intelligent Design movement which believes that natural history is awash with information discontinuities. In intelligent creation the processes of physics are regarded as part and parcel with the intelligent process of information generation & information selection. Just as the third person perspective only ever sees the first person as behavior patterns and (on closer look) a system of molecular and neural signalling, then when I look closely at physics I see the mind of God expressed as a quantum signalling process spread across the cosmos.  But all this is a highly speculative blue skies project and so I don't make claim, unlike the young earthists and IDists, to being God's gift science. This is an enjoyable private hobby which keeps me out of trouble!.