6. Never, ever pay any attention to Sat/oshi Kan/azawa

So, those are all great and all, and certainly a good first pass to see if somebody writing a given evolutionary psychology paper is an idiot or not. But they are certainly all issues that people doing evolutionary psychology (even shady evolutionary psychology) think about. You'll note that in one of the cases where she mentions a specific study, not only is it twenty years old, but it was subject to lots of devastating critiques in the journal issue in which it was published..

It's also sort of funny that the one paper she cites approvingly is just as potentially culturally contingent as any other ev psych paper, but has an anthropologist as first author.

Never underestimate the laziness of journalists summarizing scientific findings, Sify. This is good for them to read.

Yeah, I mean that's fine. I'm just pushing back against using this as support for the unfogged "aha, evolutionary psychology is all junk!" consensus, because most non-shitty ev psych (that means, at the very least, research that has nothing to do with any economists mentioned in comment 1 in this thread) is perfectly aware of (and attentive to) these concerns.

1.The trait is variable. The number of fingers on a human hand is not significantly variable since most everybody has five. Hair length is variable.

I don't get this at all. If a new mutation is very favorable then pretty soon everyone (or nearly everyone) will have it. So if everyone has a trait that is evidence that it has been selected for. If a trait varies then that is evidence it is not under strong selection pressure (within the range of common variation). Or else a recent favorable mutation has not had time to become universal.

There are some exceptional cases like sickle cell anemia where one copy of a gene is good but two are bad where you don't expect the frequency to be driven to zero or one but I think they are a bit unusual.

6: the idea with variation of personality traits is that there are multiple possible evolutionarily stable strategies.

7

6: the idea with variation of personality traits is that there are multiple possible evolutionarily stable strategies.

Sure but she says:

There are ways to be able to be more confident about whether a trait you're interested in has been selected (or rather, not eliminated) ...

which seems to say that a trait like babies crying when they are hungry which is universal (or so I assume) hasn't been selected for which makes no sense to me.

When Prof. Clancy writes "There are ways to be able to be more confident about whether a trait you're interested in has been selected (or rather, not eliminated). You can see if it conforms to these three principles..." the implication goes only one way. That is, she's saying that if a trait conforms to those three principles then that's evidence that it has been selected for. She's not saying that every trait that has been selected for conforms to those three principles.

The point of the "variation" principle is that if a trait has no variation it's hard to carry out a study that shows that it has a reproductive advantage (because what are you going to compare it to?). Of course, maybe you can discover an ancestral population in which there was variation in the trait. But then it's going to be much harder to produce convincing evidence. There's always the possibility that the trait reached 100% through some other means than selection. For example, a bottleneck. (In the case of fingers: there used to be tetrapods with various numbers of digits, but all extant tetrapods are thought to be descended from a species with five digits. It seems rather unlikely that there's anything special about the number five.)

It seems rather unlikely that there's anything special about the number five

What! Why do you think our numerical systems are base 10?

I saw the figure five in gold.

9.2 is totally wrong.

9

The point of the "variation" principle is that if a trait has no variation it's hard to carry out a study that shows that it has a reproductive advantage (because what are you going to compare it to?). ...

First if a trait has variation this is evidence that it isn't (within the range of variation) important for reproductive success or the variation would have been selected away. Second carrying out a study today makes no sense if the current environment differs (as I would expect it usually does) from the environment over which we evolved. Third any such study would be hard to do well. For example the linked article claims:

Voice pitch, however, is a good example of a trait that is variable, heritable, and has been shown to be correlated with the number of children a man has ...

Now the abstract she links says nothing about number of children. But suppose there was a gene that gave you a deep voice which attracted women but also made you more likely to die in childhood. Then a study might show deep voiced men had more children but that would not mean the deep voiced gene was advantageous.

Follow-up to comment 5: hey, look.

Except now Martin Wisse has expanded the premise to five ways psychology (not just ev psych!) needs to evolve. I'm not saying the original article was an anthropologist writing a hit piece about psychology as part of an ongoing turf war over funding for field work vs. lab work, or anything, but it is serving that purpose well.

One place variation does indicate selection is variation between populations (rather than within them). If the sickle cell gene is much more common inside Africa than outside it that is strong evidence that the African environment has selected for it.

Then a study might show deep voiced men had more children but that would not mean the deep voiced gene was advantageous.

It certainly could, if there were more surviving grandchildren with than without the gene. `Advantageous' isn't measured for individuals.

What does human evo-devo call traits that are advantageous in ensemble? E.g., a population that's a third each A, B, C is more successful than a population of any single type?

I also don't understand lack of variability as being evidance that a trait has not been selected. If all pigeons fly, would that be evidance that flying was not evolved.

Evolution can not work if there was no variation ever, but if all humans have some trait that no other primate has, at some point there may have been some variation on that trait.

9: if a trait has variation this is evidence that it isn't (within the range of variation) important for reproductive success or the variation would have been selected away.

Similarly, there can never be a $10 bill on the street because someone would have picked it up.

But let me try again: Prof. Clancy is not claiming that every trait that has been selected for exhibits variation. No doubt most traits that have been selected for have reached fixation. But knowing that's true in general doesn't help you in any particular case. Do five fingers give a reproductive advantage over other numbers of fingers? Or did dactlyly reach fixation by some other means (e.g. bottleneck or genetic drift)? What kind of evidence can you bring to bear on this question?

The point of the three principles in Clancy's article is that these are the cases in which you have a chance of being able to come up with some scientific evidence.

17: I also don't understand lack of variability as being evidance that a trait has not been selected

Now you're just making me sad.

16

It certainly could, if there were more surviving grandchildren with than without the gene. ...

How many of these studies are looking at the number of surviving grandchildren?

16: I think you're looking for the term "group selection."

Glad to see she cites to the great paper The Weirdest People In The World" , a fantastic piece of work that deserves its own thread. (The discussion pieces after the paper are also excellent). It's a comprehensive strike back at a particular piece of stupidity which is a key feeder of bad ev psych, namely the belief that complex behavioral traits are physically hard-wired through evolution instead of culturally adapted. It's funny that this should have become such an article of faith among pop ev psych types when it's rather obvious that if any human trait is important for reproduction, it's flexibility to cultural environment.

18

The point of the three principles in Clancy's article is that these are the cases in which you have a chance of being able to come up with some scientific evidence.

I think it is bizarre to restrict yourself to cases where selection is currently occurring particularly since (as I said above) recent drastic environmental changes for most human populations mean that the present situation is not representative of the conditions under which we evolved.

And there are certainly other ways of obtaining evidence. For example comparing with chinpanzees. The five finger example is sort of silly since (as I understand it) it was established long ago and path dependence can explain why it has continued even absent current reproductive advantage.

This dude thinks that there is a developmental maximum of 5 fingers:

http://dev.biologists.org/content/116/2/289.full.pdf

I forgot that Gould thought that language was a spandrel.

21

... when it's rather obvious that if any human trait is important for reproduction, it's flexibility to cultural environment

Really? Was culture rapidly changing during most of our evolutionary history?

20: mocked and recrudescent alternately.

The Wikipedia summary is particularly entertaining for me because all my thinking about evolution is community- or niche-level, in which we assume -- even observe! -- a lot of things like the haystack model. Thinking about one species at a time feels like cheating, although for humans it's just possibly true; too soon to tell, as was said of another triumph of instrumental reason.

WEIRD people make strong inferences from computations of similarity, whereas populations with greater familiarity with the natural world, despite their capacity for similarity-based inductions, prefer to make strong inferences from folkbiological knowledge that takes into account ecological context and relationships among species (Atran, Medin, & Ross, 2005).

I'll be getting back to my folkbiology, then.

(High time, too. In the annals of I Suck, after tossing my life into the air for eight month to get more fieldwork done, I have been indistinguishably sick or depressed for a month and haven't done much of anything. The panic and weather would be getting me out, except that I am finally definitely sick. I;m lucky that the weather has been neutral, but unlucky in that I wanted rain instead.)

re: 24

I can't speak for our whole evolutionary history, but books like After the Ice: A Global Human History, 20,000 - 5000 BC * give pretty compelling examples of massive cultural variation in the palaeolithic, mesolithic and neolithic periods.

* Chris Y's recommendation.

27: that's 1/200th of the period since the last common ancestor with chimps.

... fairly conservatively.

(On the other hand early hominids left Africa way the hell early, the 200,000 years or whatever since modern humans left Africa is plenty of time for evolutionary change even without thinking about things like interbreeding with other hominids, even within Africa there's no reason to think that early hominids lived in a single environment or had any kind of monoculture, etc. etc.)

re: 28

Sure, but it's a significant chunk of our recent evolutionary past. Obviously the evidence gets sparse the further back we go, but it's entirely plausible that you'd have some variation, possibly quite a bit of variation, pretty much as far back as you have recognisably modern humans.* Maybe not. It's hard to infer much from lithics alone. It doesn't seem obvious to me that the cultural uniformity hypothesis is by any means a certain one.

[Typed before reading 30 on preview]

* dates depending whether you count that as anatomical or behavioural modernity and whether you believe they differ.

WEIRD people make strong inferences from computations of similarity, whereas populations with greater familiarity with the natural world, despite their capacity for similarity-based inductions, prefer to make strong inferences from folkbiological knowledge that takes into account ecological context and relationships among species

It's kindalike this, and that, and it's all the same really.

Human reproduction is very clearly culturally embedded and mediated through cultural institutions. We don't reproduce by going into heat and raping the neighbor girl (or boy).
It seems logical that human cultures did vary a fair amount over evolutionary time; certainly hunter-gatherer societies vary a lot culturally. But if they didn't is the argument that culture is biologically hard-wired too, so there is no distinction between the instinctive and the cultural? Bad ev psych often does argue exactly this, e.g. all reproductive institutions look like marriage in 1950s Long Island because that's nature's way.

Clew, I also find the existence of some amount of group selection to be very intuitive and there's plenty of scientific support for the idea -- this piece by DS and EO Wilson is a good and pretty clear introduction.

32 -- not the argument at all. It's not that western societies classify and non-western societies don't, it's that they classify differently based on their experiences and what is emphasized in their learning process. It's very common sense actually.

31

... It doesn't seem obvious to me that the cultural uniformity hypothesis is by any means a certain one.

Sure but the claim I was doubting was much stronger:

... when it's rather obvious that if any human trait is important for reproduction, it's flexibility to cultural environment

Perhaps flexibility is very important but I don't think it is obvious. And how often were early humans raised in one culture (which might vary by geographical area) and then required to function in a vastly different culture as adults (flexibility)? Modern humans often have some trouble with this but of course the chances of exposure to greatly different cultures under modern conditions seems (intuitively at least) higher than was likely typical for early humans.

33

Clew, I also find the existence of some amount of group selection to be very intuitive and there's plenty of scientific support for the idea -- this piece by DS and EO Wilson is a good and pretty clear introduction.

Nevertheless the conventional wisdom (which I share) is that selection has mostly acted on individuals not groups.

Remind me, clew, have you moved?

33: Bad ev psych often does argue exactly this,

Bad ev psych which is mostly found in the strawmen erected by social scientists, or the stupid attempts to use it to make political points by popular press/political pundits/asshole economists.

Did an ev psych researcher shoot your dog, PGD?

the stupid attempts to use it to make political points by popular press/political pundits/asshole economists.

The thing is, though, those stupid attempts are the only reason anyone who isn't an academic is particularly aware of the field. It's not less interesting than any other arcane field of study, but unless it can be used to make political/social points, it's not more interesting. And the political/social points (as opposed to the underlying research) do tend toward the stupid side.

I honestly fail to see how the article in the OP was aimed at non-scientists making stupid political/social points as opposed to strawman psychologists. Do pundits really do a bad job operationalizing variables?

I was talking generally about ev psych grousing -- you're right that the original post is more about research. Is it really all strawmen, though? I thought your first comment said that the criticisms were reasonable but not universally applicable.

re: 35.last

You seem to be confusing flexibility as a trait of individual adult humans, with flexibility of humans in general. Sure, Ugg might be born in a particular type of society and live and die without experiencing a huge amount of cultural variation. But Gugg, living a few hundred miles away, and 2000 years later, might live and die in a radically different society.

re: 35.last

It's true, of course, that selection of individual humans (or genes, or whatever) may not select specifically for an ability to adapt to multiple radically different cultural environments, since specific instances of selection aren't going to operate over time and space in that way. Even if you believe in group selection [rather than just individual selection, or gene selection], it isn't going to apply across multiple generations and in multiple places.

But the existence of huge cultural variation is at least prima facie evidence that universal recent or relatively recent selection _for_ adaptation to a very specific social and cultural environment hasn't taken place.

41: the criticisms are reasonable, widely understood, and broadly accounted for (or at least acknowledged) in all the research I've looked at primary sources for (excluding things like K/an/azawa, who is a hack and an economist besides). The way the post is written makes it sound like these criticisms will come as a surprise to ev psych researchers -- or that ev psych researchers simply don't address these things -- which is really not the case. Something like the WEIRD paper (written by psychologists, incidentally) points to a real problem, potentially, but it's more of a problem in social psychology than evolutionary psychology (which tends to be very concerned with looking at things cross-culturally, for obvious reasons) but the idea that you can solve that issue just by heading out and doing some fieldwork is simplistic to the point of being ludicrous.

Since I somehow always find myself in the position of defending an approach I don't actually think is terribly promising (for reasons related to what ttaM has been saying) I should point out that there actually are very cogent criticisms that can be made of the ev psych approach, but she hasn't made them, and the fact that she approvingly linked to a paper that strikes me as perfectly problematically confounded (but which has an anthropologist as first author) makes me wonder.

But the existence of huge cultural variation is at least prima facie evidence that universal recent or relatively recent selection _for_ adaptation to a very specific social and cultural environment hasn't taken place.

Right. This is where I always get stuck on the possible explanatory power of very specific mental modules for dealing with situations familiar from the veldt. They could perfectly well exist. But so much of what people do in day to day life is wildly variable cultural stuff, more like chess than like responding affectionately to an infant, that don't seem like something that such very specific mental modules would have much of an effect on.

Adaptive (or adapted) or not, we demonstrably have a huge, huge capacity for cultural flexibility.

so much of what people do in day to day life is wildly variable cultural stuff, more like chess than like responding affectionately to an infant

I actually think the latter describes much more of our day-to-day behavior than the former, but I agree with the premise that uniquely human capabilities are much more likely to be devoted to flexibly compensating for variable environments, so never mind.

Comity!

Sure, close enough.

The WEIRD paper really is interesting. Everyone should read it!

49: Everyone should read it!

Sure. So they can learn how little it has to do with the critique in the article linked in the post per Tweety's 44.2.

46: but I agree with the premise that uniquely human capabilities are much more likely to be devoted to flexibly compensating for variable environments

Of course that begs the question of the extent to which uniquely human capabilities are controlling a particular behavioral pattern.

51: let alone how important uniquely human capabilities actually are to any evolutionary explanation of behavior, but, you know, I swept that under the comity.

50: ummm, no it has to do with the linked article since point 2 in the article is entirely devoted to talking about it. But in any case that's not why you should read it, the linked article is just OK and not really anything anyone here needs to read, while the WEIRD article is really interesting and important in its own right.

Anyway, I'm not sure why people are taking off on the linked article. I don't think it's that good, these are things you should already know and some are so vague as to be pointless, but the various provisos in it are things social scientists would do well to pay attention to in all of their work. Particularly on the first two points. If you don't think peer-reviewed evolutionary psychology articles run afoul of these things, I recommend you check out the collected works of David Buss' lab at UT Austin.

Bad ev psych which is mostly found in the strawmen erected by social scientists, or the stupid attempts to use it to make political points by popular press/political pundits/asshole economists.

I specified bad because I didn't want to rule out good and because I specifically wanted to include the (massive) impact bad ev psych has had in the popular culture. I don't think you get to rule that out, it's been an important influence on the broader intellectual climate. But some of the issues she points out -- particularly the first two points -- are relevant to psychology research in general.

I think there's a general issue in the social sciences with abstracting away from cultural context. This goes beyond 'evolutionary psychology'. I also wonder if 'evolutionary psychology' makes sense as a separate discipline.

53: I specifically wanted to include the (massive) impact bad ev psych has had in the popular culture.

OK, this is probably some of the difference in our response to ev psych, I see no "massive" impact on popular culture. I see a relatively marginal academic discipline (sure with some fairly well-known proponents) that has some results which get trumpeted and taken out-of-context* by people with political axes to grind or a need for content (pundit/journalists).

*Much like economics.

I wish I lived in your world where economics was a relatively marginal discipline.

Even if you believe in group selection [rather than just individual selection, or gene selection], it isn't going to apply across multiple generations and in multiple places.

I do believe in group selection - to be precise, in multi-level selection - and I don't think that gene selection, except at the level of "selfish DNA" is any kind of alternative. Genes are what are copied, phenotypes and groups of phenotypes are what are selected, or winnowed.

The Gintis/Bowles line on this is that there could be, and in fact has been, selection for individual traits that make groups more coherent (including the policing of individualism). In this way, belonging to a successful group leads to reproductive success, and this continues to be true across multiple generations and in multiple places. Obviously it's not the only factor in determining the success of groups - things like smallpox can be just as important. But it is silly to deny that this actually exists.

Note that Gintis and Bowles are in fact economists, so feel free to hate. But I could quote E O Wilson., DS Wilson and others

55: Poorly written and footnoted by me. Only meant that economics was likewise taken out of context, not that economics shared the "relatively marginal" characteristic.

Honestly, the findings in that WEIRD article are so interesting -- it's a litany of differences in human cultures that certainly surprised me, and really does create some extremely interesting issues for generalizing psychological theories of human nature, whether or not they are "evolutionary." Here's a typical passage:

By the mid‐1990s researchers were arguing that a set of robust experimental findings from behavioral economics were evidence for a set of evolved universal motivations (Fehr & Gächter, 1998; Hoffman, McCabe, & Smith, 1998). Foremost among these experiments, the Ultimatum Game, provides a pair of anonymous subjects with a sum of real money for a one‐shot interaction. One of the pair--the proposer--can offer a portion of this sum to a second subject, the responder. Responders must decide whether to accept or reject the offer. If a responder accepts, she gets the amount of the offer and the proposer takes the remainder; if she rejects both players get zero. If subjects are motivated purely by self‐interest, responders should always accept any positive offer; knowing this, a self‐interested proposer should offer the smallest non‐zero amount. Among subjects from industrialized populations--mostly undergraduates from the U.S., Europe, and Asia--proposers typically offer an amount between 40% and 50% of the total, with a modal offer of 50% (Camerer, 2003). Offers below about 30% are often rejected. With this seemingly robust empirical finding in their sights, Nowak, Page and Sigmund (2000) constructed an evolutionary analysis of the Ultimatum Game. When they modeled the Ultimatum Game exactly as played, they did not get results matching the undergraduate findings. However, if they added reputational information, such that players could know what their partners did with others on previous rounds of play, the analysis predicted offers and rejections in the range of typical undergraduate responses. They concluded that the Ultimatum Game reveals humans' species‐specific evolved capacity for fair and punishing behavior in situations with substantial reputational influence. But, since the Ultimatum Game is typically done one‐shot without reputational information, they argued that people make fair offers and reject unfair offers because their motivations evolved in a world where such interactions were 10 Weird People not fitness relevant--thus, we are not evolved to fully incorporate the possibility of nonreputational action in our decision‐making, at least in such artificial experimental contexts. Recent comparative work has dramatically altered this initial picture. Two unified projects (which we call Phase 1 and Phase 2) have deployed the Ultimatum Game and other related experimental tools across thousands of subjects randomly sampled from 23 small‐scale human societies, including foragers, horticulturalists, pastoralists, and subsistence farmers, drawn from Africa, Amazonia, Oceania, Siberia and New Guinea (Henrich et al., 2005; Henrich et al., 2006). Three different experimental measures show that the people in industrialized societies consistently occupy the extreme end of the human distribution. Notably, some of the smallest scale societies, where real life is principally face‐to‐face, behaved in a manner reminiscent of Nowak et. al.'s analysis before they added the reputational information. That is, these populations made low offers and did not reject.

Fascinating! And, did you know that people in different cultures perceive the Mueller‐Lyer illusion substantially differently, with Western subjects extreme visual outliers? I for one did not.

I liked this passage, too, which is perhaps more directly relevant to the discussion above. Just because Apes and Western Children do something doesn't actually mean it's a human universal:

Working with children and non‐human primates is essential for understanding human psychology. However, it is important to note that despite its great utility and intuitive appeal, such research does not fully obviate these challenges. In the case of primate research, discovering parallel results in great apes and in one human population is an important step, but it doesn't tell us how reliably developing a particular aspect of psychology is. As the spatial cognition work indicates, since language and cultural practices can--but need not--influence the cognition humans acquired from their phylogenetic history as apes, establishing the same patterns of cognition in apes and Westerners is insufficient to make any strong claims about universality. Suppose most psychologists were Hai\\om speakers (instead of Indo‐European speakers), they might have studied only Hai\\om‐speaking children and adults, as well as nonhuman apes, and concluded (incorrectly) that allocentric spatial reasoning was universal. Similarly, imagine if Tsimane economists compared Ultimatum Game results for Tsimane adults to chimpanzees (Gurven, 2004; Henrich & Smith, 2001; Jensen, Call, & Tomasello, 2007). These researchers would have found the same results for both species, and concluded that standard game theoretic models (assuming pure self‐interest) and evolutionary analyses (Nowak et al., 2000) were a fairly accurate predictors in Ultimatum Game behavior for both chimpanzees and humans--a very tidy finding. In both of these cases, the conclusions would be opposite to those drawn from studies with WEIRD populations.14

Studying children is crucial for developing universal theories. However, evidence suggests that psychological differences among populations can emerge relatively early in children (as with folkbiological reasoning), and sometimes differences are even larger in children than in adults, as with the Mueller‐Lyer Illusion. Moreover, developmental patterns may be different in different populations, as with sex differences in spatial cognition between low income vs. middle and high income subpopulations in the U.S., or with performance in the false belief task. This suggests a need for converging lines of research. The most compelling conclusions regarding universality would derive from comparative work among diverse human populations with both adults and with children, including infants if possible. Human work can then be properly compared with work among non‐human species (including but not limited to primates), based on a combination of field and laboratory work.

Gintis is my favorite Amazon book reviewer:

http://www.amazon.com/gp/cdp/member-reviews/A2U0XHQB7MMH0E?ie=UTF8&display=public&page=1&sort_by=MostRecentReview

"Schooling in capitalist America" is good too.

Minivet, yes. Town, charming; solitude, more to my taste than it is to my benefit.

How long have humans been living in a wide array of environments? I don't know if we spread along all the nice littorals first, only moving into the deserts & jungles and mountains and all 2e5ya, or in the penultimate interglacial, or what. The patchiness of resource distribution is a powerful predictor of sociality in subterranean rodents (near my piece of the elephant), and I know people who argue that it's strongly explanatory of sociality and types of institutions in post-agriculture and industrial humans, but I don't know how long we've been exploiting distinctly different environments.

re: 56

FWIW, I do to, and was fairly convinced by Wilson and Sober's book. I had in mind that the sorts of examples of group selection that are likely to persuade 'selfish gene' sceptics aren't much like the sort of Gintis/Bowles claim. Which, fwiw, I'm not familiar with the specifics of, and am now going to find out more.

62: An explanation designed to convince a Dawkins believer can take many forms, but it should always be printed in VERY LARGE TYPE and wrapped around the end of a two by four before delivery. This increases penetration of the arguments.

63: I get it! Because Dawkins is always so rude and aggressive, right?

No, actually: because they simply parrot arguments about rationality and evidence without having thought them through very carefully.