However, Jerry Coyne tries to (partially) defend evolutionary psychology and he offers the following paper as evidence that the field is not entirely worthless.
Confer, J.C., Easton, J.A., Fleischman, D.S., Goetz, C.D., Lewis, D.M.G., Perilloux, C., and Buss, D.M. (2010) Evolutionary Psychology: Controversies, Questions, Prospects, and Limitations. American Psychologist 65:110–126. [doi: 10.1037/a0018413]The main problem with the adaptationist approach is identifying adaptations. Is something really an adaptation or are there other explanations? Part of the answer involves providing evidence that the behavior has a specific genetic component.
I don't think those problems are satisfactorily addressed in this review in spite of Jerry's recommendation. If this is the best the field has to offer then it's in real trouble.
Here's an example from that paper of the "best" kind of science.
The science of confirming and falsifying hypotheses, of course, is typically more complex than these examples indicate. Often a hypothesis is embedded within a larger theoretical network. For example, one evolutionary prediction is that women will prefer men as potential mates who express a willingness to invest in them and their offspring (Buss, 1995). This is derived from the hypothesis that in paternally investing species, females will use cues to a man’s willingness to invest as a criterion for mate selection. In turn, this hypothesis is derived from parental investment theory, which posits that the sex that invests more in its offspring will be the choosier sex when selecting mates (Trivers, 1972). Finally, the logic behind parental investment theory is derived from inclusive fitness theory (Hamilton, 1964a, 1964b), the modern formulation of evolution by natural selection. As in all realms of psychological science, the evaluation of each evolutionary psychological hypothesis, as well as the broader theories within which they are embedded, rests with the cumulative weight of the empirical evidence.I've always been puzzled about the evolutionary significance of mate selection in humans. Whenever I look at genealogical records I'm impressed by the fact that almost all men and women who reach maturity seem to find a mate. This jibes well with my personal experience since all of my friends who wanted to live with a partner succeeded in finding one. I wonder how powerful these mate selection criteria can be if there aren't significant numbers of people don't succeed in mating. Where are the ugly mean and women who couldn't find a mate?
I understand that there are studies showing that men and women in different cultures will prefer certain physical characteristics in their mates. What I don't know is whether this actually translates into mate selection when the time comes to form a stable partnership (e.g. get married). It doesn't seem like it to me otherwise almost all wives would look like Marilyn Monroe.
If women are the choosers then why aren't there lots of single men who will never have children even though they want to? Where are the losers in our societies? How common were they in the past?
What about hypothesizing that we all dream about the appearance of the ideal mate but that when it comes time to make a choice we put our emphasis on other characteristics (e.g. availability)? Maybe our preferences aren't really adaptations at all? Was that one of the hypotheses that was considered or do evolutionary psychologists jump immediately to adaptive story-telling?
38 comments :
For the sake of playing devil's advocate: just because most everyone can find a mate doesn't mean that the male partners are the fathers. (This effect is not significant enough, based on what numbers I've seen; life seems to be driven by "good enough" rather than any maximizing efforts.)
Larry wrote "Whenever I look at genealogical records I'm impressed by the fact that almost all men and women who reach maturity seem to find a mate. This jibes well with my personal experience since all of my friends who wanted to live with a partner succeeded in finding one."
Geneological records are a lousy way to assess reproductive success or failure. Many such records fail to record individuals who fail to mate and produce offspring. Much more reliable is empirical demographic data which does not rely on geneological information. According to such data, it is NOT the case that "all of my friends who wanted to live with a partner succeeded in finding one". This is pure anecdotal bullshit. In fact, there is a significant asymmetry between male and female reproductive success: while almost all females do successfully mate (and most go on to produce children), the same is not the case for males. In my own research I found that less than 3 percent (2.88% to be exact) of women failed to mate and reproduce, while more than three times as many men failed to do so. This asymmetry is mostly due to vertical polygamy. A significant fraction of males mate and reproduce with more than one female (usually in sequence in western society, but simultaneously in others). By contrast, almost no females produce offspring with more than one male (for the complete data set and statistical analysis, see http://evolutionlist.blogspot.com/2006/04/polygamy-de-facto-and-de-jure.html).
As an evolutionary psychologist myself (see http://www.amazon.com/The-Modern-Scholar-Evolutionary-Psychology/dp/B00435HBGO/ref=sr_1_1?ie=UTF8&qid=1355165268&sr=8-1&keywords=allen+macneill and http://www.amazon.com/The-Modern-Scholar-Evolutionary-Psychology/dp/B005KR5J8G/ref=sr_1_3?ie=UTF8&qid=1355165268&sr=8-3&keywords=allen+macneill), I find it interesting that Larry's critique of Jerry Coyne's much more balanced analysis relies entirely on anecdote and personal experience (dare we also include heresay?), which he deplores in evolutionary psychologist. Gander, meet sauce.
Ever heard the phrase "S/he's out of your league"? Even if everyone ends up with a mate, there's still a massive amount of sorting going on - the Brad Pitts end up with the Angelina Jolies and the losers with the losers. Not very adaptive when the losers' kids mostly still make it to adulthood, but presumably we are talking about a time when that was far from the case. So the losers still end up with a mate, but they may not be able to afford to raise children, or their children may have reduced chances of survival. Why is this not a plausible hypothesis?
And maybe you should read a textbook on evolutionary psychology, rather than the crap that gets written up by soi-dissant "science" journalists. I'd start here: http://www.amazon.com/Evolutionary-Psychology-Science-Mind-Edition/dp/020501562X/ref=sr_1_1?ie=UTF8&qid=1355165518&sr=8-1&keywords=evolutionary+psychology. I realize that relying on personal experience is fun and much more relaxing, but if I were to do that on the subject of biochemistry (which I teach at Cornell as part of biology), you would think me pretty shoddy, right?
Far be it from me to defend evolutionary psychology, which has a lot of methodological and overreach issues, but surely if our preferences were not adaptations but merely random then we could expect a greater number of us to actually prefer old, sick, insane, obese or anorexic partners? Does not the fact that the ideal is generally young, healthy and fit tell us something already?
Surely you would not also, for example, say that it is unwarranted adaptationism to call the fact that plants grow their green leaves above the ground instead of below it an adaptation to where the light is, even before I have formally demonstrated the genetic basis and selective pressure?
I'm well aware of the problems with anecdotes and personal experience but there are times when it serves as a reality check.
It's good to see that you have real data. Of the 2.88% of women who don't mate, how many fail to do so because nobody selected them as partners? Is it true that most of these women fail some of the obvious tests for attractiveness? Or is it possible that some percentage of women are infertile or just don't want children? How many?
In order for the adaptive scenario to be true, the ones who fail to find a mate must have some feature that makes them less likely to bear healthy children. Did you try to find out what features the 2.88% had in common that makes them less desirable as potential mothers?
What about the roughly 9% of men who don't have children? What feature(s) did they have in common to make them less likely to be selected as fathers? Did they all have flabby stomachs or bald heads?
BTW, I just picked out one thing from the article that interested me. Do you think that on the whole it's a good defense of evolutionary psychology? Can you point me to the part that discusses how you decide whether something is an adaptation or not? Can you point me to the discussion of how certain maladaptive or neutral traits can be fixed in a population? Can you highlight the part that explains the significance of selective coefficients and population size?
How about the section explaining why we are so confident that we know how small groups of hominids must have behaved on the African savannah 200,000 years ago? I missed that part.
Ian says,
... presumably we are talking about a time when that was far from the case.
Exactly. Was there ever such a time? Do we have evidence?
So the losers still end up with a mate, but they may not be able to afford to raise children, or their children may have reduced chances of survival. Why is this not a plausible hypothesis?
It's only plausible if there was a time in the past when everyone got to mate. Then there was a mutation followed by adaptation for being picky about who you mate with. It's possible, but when you work through the details it seems unlikely that it could work in a small hunter-gatherer social group.
There are nonadaptive scenarios that seem equally plausible but they aren't given much consideration. Why?
There are lots of excellent biochemistry textbooks but that doesn't prevent biochemists from making fools of themselves from time to time.
We try to weed out those biochemists who display a lack of knowledge of basic principles in their field. We're not perfect but the percentage of bad apples is quite low.
Why are there so many evolutionary pychologists who don't seem to have a good textbook level of understanding of evolution? As you point out on your blog [Evolutionary Psychology: The Science of Human Nature], you assign the spandrels paper to your students. This is good. How many of your fellow evolutionary psychologists have read the paper and what percentage of those who have read it understand it?
When an ethologist observes that there is a statistically significant asymmetry between female and male mating rates and reproductive success, it is rarely the case that s/he can ask the elephant seal (or langur or whatever) what was going through her/his mind when they were mating and reproducing. Does this lack of anecdotal evidence therefore invalidate the observations? If so, then there is no science of ethology.
Furthermore, I thought I had proposed a testable (and mathematically plausible) hypothesis for the asymmetry between female and male mating rates and reproductive success: vertical polygyny, whereby some males have children more than one female, while others have none. Standard stuff in ethology, but not in biochemistry I guess.
As for what distinguishes an adaptation from an accident, I propose the following criteria:
1) an evolutionary adaptation should be correlated with increased relative reproductive success, either at present or in the empirically verifiable past;
2) an evolutionary adaptation should be correlated with an ecological context within which differences in relative reproductive success can be consistently correlated with specific environmental variables; and
3) an evolutionary adaptation should be correlated with an underlying genetic and developmental program that brings about the expression of the adaptive phenotype.
If the adaptation is behavioral, one more correlation should be demonstrable:
4) a behavioral evolutionary adaptation should be correlated with an underlying neurological/neurochemical mechanism that elicits the behavior in question.
It is, of course, not accidental that these criteria conform fairly closely to Tinbergen's "four questions" in ethology. Since "evolutionary psychology" seems to get your dander up, would you be more comfortable with the term "human ethology" being applied to the empirical analysis of behaviors that meet these criteria? There is an International Society for Human Ethology, founded by a pioneering ethologist, Irenhaus Eibl-Eiblesfeldt (look it up; Google is your friend). The subject matter of human ethology is nearly isomorphic with the subject matter of evolutionary psychology, but gets less publicity (probably because of the technical term "ethology").
As for inferring the Pleistocene EEA for most human behavioral adaptations, ethologists don't generally speculate on the Paleocene (or Triassic) origin of other animal behaviors. Lorenz, Tinbergen, and von Frisch got their Nobel for watching animals behave here and now, and for formulating and empirically testing hypotheses to explain the origin and evolution of those behaviors. That's generally what evolutionary psychologists/human ethologists do as well. Yes, correlation isn't necessarily causation, but without correlation all you have is meaningless data.
This assumes that hunter/gatherers were exclusively monogamous. However, the overwhelming evidence points to the conclusion that humans have been (and still are) polygynous, with some men having proportionately more mates and offspring than others (and some none). The heritable characteristics of those males would be disproportionately represented in the populations to which they contributed, while the characteristics of the males who were less successful were not. This is not controversial when one is discussing elephant seals, African lions, or prairie voles. Why is it so much less plausible when one refers to humans?
Almost all of the evolutionary psychologists I know well have read the spandrels paper, and most have read Gould and Vrba/exaptations as well. The same could not be said for the "science journalists" who write about evolutionary psychology, nor the editors that preferentially publish the crap because it's easier for non-scientists to "get". Evolutionary psychology, like any empirical science (and especially evolutionary biology), is not easy for non-specialists to fully understand, and so a lot of misunderstandings get promulgated. However, unlike biochemistry, people think they know what evolutionary psychology is about without having to do the hard work of really learning about it from, say, a textbook on the subject, published by an academic press. Have you done that yet, Larry, or have you already decided that since a lot of nonsense about evolutionary psychology gets published in the popular press, it must therefore all be crap?
We are intelligent beings. Simple application of that intelligence will tell us that some partners aren't going to be very good at reproduction (e.g. insane people). That kind of mate choice doesn't have to be a specific adaptation that requires specific alleles.
The tick is to identify those components of mate choice (if any) that are specific adaptations whose alleles confer on us the ability to recognize good mates wherein the past, before the mutations arose, we couldn't tell a good mate from a bad one.
"Was there ever such a time? Do we have evidence?"
You might want to take a look at "Farewell to Alms" by Gregory Clark. Lots of data from pre- and post-industrial England. Except for recent times, humanity mostly lived at the Malthusian limit: even 250 years ago in England (arguably the most "advanced" country in the world) average expected life-span of a newborn was ~20 years: on average 4 births needed to maintain the population stable.
http://www.amazon.com/Farewell-Alms-Economic-History-Princeton/dp/0691141282
(sorry, don't know how to make a link clicky)
Finally, you wrote "There are nonadaptive scenarios that seem equally plausible but they aren't given much consideration."
Mostly because non-adaptive scenarios don't result in two things:
1) consistent, significant, repeated asymmetries/patterns within phylogenetic lines (e.g. patterns of infanticide in primates), and
2) consistent, significant, repeated asymmetries/patterns between unrelated phylogenetic lines (e.g. patterns of infanticide in widely separated vertebrate and invertebrate taxa).
Yes, drift can fix particular genotypic/phenotypic characters within particular lineages, but it does so randomly, not over and over again within and between widely separated phylogenetic lineages. That's convergence, and it's the result of natural selection in similar ecological contexts.
You ask "Why are there so many evolutionary pychologists who don't seem to have a good textbook level of understanding of evolution?"
How many are we talking about here, and do you mean practicing evolutionary psychologists (i.e. people doing actual empirical research in accredited academic institutions), or are you talking about the "science journalists" and other fiction writers who have no actual background in the science? Just curious...
I understand the scenario. What you're proposing is that at some time in the past specific alleles arose in the population that made women more selective in their choice of mates. Those alleles caused the women to eliminate certain men from consideration even if it meant sharing the winners with other women.
Meanwhile, the women who didn't have the "picky" allele(s) were free to mate and produce babies with any man in the tribe.
Explain to me again how the "picky" allele increases in the population? Aren't there a lot of other assumptions required?
Is there any direct evidence that the preferences of women for certain mates has a genetic component other than intelligence?
I'm often alerted to silly evolution psychology papers by bad science reporting but when I go and look at the actual papers I'm not impressed.
I think we agree that there's a lot of crap out there and it's not just the science journalists who are at fault.
However, unlike biochemistry, people think they know what evolutionary psychology is about without having to do the hard work of really learning about it from, say, a textbook on the subject, published by an academic press. Have you done that yet, Larry, or have you already decided that since a lot of nonsense about evolutionary psychology gets published in the popular press, it must therefore all be crap?
I've read some of what's in the textbooks. You could really be helpful if you would take the time to look in the very latest issues of the top evolutionary psychology journals and pick out the very best papers you can find. Choose three, one from each of the three best journals. We could look at them to see if there's any evidence that the authors have read and understood the spandrels paper.
This shouldn't be too hard, should it?
Divalent says,
... average expected life-span of a newborn was ~20 years: on average 4 births needed to maintain the population stable
What's that got to do with the subject we're discussing?
I don't find it surprising that there are different levels of selection in male humans vs female humans (See http://www.pnas.org/content/early/2012/04/24/1118174109.abstract for a study of this in an ostensibly monogamous human population). Presumably there is some level of sexual selection, which could take place at the physical or behavioral level. It's still up to people to prove that a characteristic has a substantial selection pressure associated with it.
Well, I find it rather more convincing that we have these preferences instinctively. When the average man sees a woman he likes, does he think "she has got wide hips, so she will be good very specifically at birthing my children", or does he think "wow, she's pretty"? And simple application of our intelligence seems to lead to people having fewer children later in their lives wherever you give them the choice to do so!
Really, I see your point that for any given allele random drift must be the null hypothesis, but surely there is no way around the observation that most interesting features must logically be the result of adaptation. For example, it appears reasonable to assume that us not having our brain tissue on the lower side of the feet is an adaptation to some kind of environmental pressures. In fact, considerably more than 99.9% of all conceivable genotypes and phenotypes would be just as instantly lethal, so by that reasoning alone we could consider the remaining few viable ones as being the result of adaptation to an environment in which they don't die.
I agree with Mr moran's stats.
It does seem now and in the past everyone got married who wanted to and choosing was not very choosy.
It was very much who was close.
Everyone wants the better people as they score it but easily settle.
Darwin emphasized breeding selection in humans had made a big deal but in reality it never did.
Larry said
Explain to me again how the "picky" allele increases in the population? Aren't there a lot of other assumptions required?
This scenario has actually been shown to work: It is Fisher's runaway sexual selection, also known as "sexy sons". What makes the picky allele increase its frequency, is that it soon is in gametic phase equilibrium with the alleles that make males attractive to the picky females. Hence, the allele creates its own selective advantage. Of course, at the population level there is no adaptive benefit.
Allen, thank you. That makes a lot of sense. I like a lot of Professor Moran's earlier work, but he loses me here. Personal anecdote, use of geneological records, plus a failure to understand that until very recently a significant portion of the population did not survive to reproduce, leads me to believe that his conclusions might be based on allegiances or political bias. It's still a fascinating discussion though.
I couldn't find a mate and didn't reproduce. I have almost no family that I'm aware of. This is the end of the line. What does that prove? Nothing.
I really like your four criteria points above: they constitute the most damning critique of "evolutionary psychology" that I have yet encountered.
But of course, if you could just point me to a clear and uncontestable example of your point 4), you could make me into a convinced supporter. You have several handy for just such a challenge, no ?
allen
I think Darwin's 'Any variation which is not inherited is unimportant for us', quite well summarizes your four criteria: Evolution is in our genes.
So, are e.g. mate preferences genetically fixed within particular lineages, f.e. as compared to other preferences? Or to sexual 'regrets', for that matter. (eg http://www.ncbi.nlm.nih.gov/pubmed/23179233) I think it's hard to sort that out using juist 'free responses, written scenarios, detailed checklists, and Internet sampling'.
ps
I mean, Obviously there isn’t a gene controlling how people answer questions about their sexual preferences or 'regrets'.
For the record ....
I am using real data (genealogical records) and personal observations (anecdote) to question the implicit assumptions of a key part of evolutionary psychology. There's nothing wrong with that. Instead of focusing on whether my data and observations are definitive (they are not), why not focus on whether mate selection actually exists and whether it has a specific genetic basis?
And why in the world would you ever think that I don't understand that lots on infants and children died in the past? The genealogical records show that very clearly and so do cemetery records.
As for your personal mating history, it's only relative if you put it into the context of current evolutionary psychology thinking. The standard explanation is that people who fail to find a mate have some physical characteristic that makes them unacceptable to potential suitors. Those suitors have been programmed by their genes to only choose mates of a certain type. Is your face asymmetric? :-)
Well, using anecdote, I can tell you that mate selection certainly exists and appears to be driven by women. We cull from available suitors, since the offer of sex is *always* there (or at least it was for me and all of my friends/ acquaintances when we were younger). The trick is find someone who's appealing enough and wants to stay to help raise your child or support whatever else you might be doing with your life. Those who aren't appealing are "creepy." Appeal for women doesn't appear to be based on looks, but on compatibility and status. Anyway, my personal history would lead be to believe that some of the EP hypotheses are spot on -- that is, to the extent we have choice over anything, a thought that inevitably sends me spiraling down the ultimate causation tunnel. At any rate, I see no reason to place limitations on scientific speculation, and I don't see modern EP as being substantially different from other areas of psychology when it comes to standards for publication. It's merely more controversial due to the politicizing nature of the issues addressed.
I sincerely appreciate your work, and enjoyed reading what you had to say on evolution and chance. You helped me realize that I was thinking about a legitimate issue debated by scientists, and not merely misunderstanding something simple in evolutionary theory. This is what Myers is using to completely dismiss EP at the moment, and I'm more persuaded by Dawkins and Wilkins. Myers' dismissal of EP on these grounds should lead to dismissal of many other evolutionary sciences as well, but it doesn't. Why not?
Facial symmetry? Are you making fun of EP again? ;-)
4) a behavioral evolutionary adaptation should be correlated with an underlying neurological/neurochemical mechanism that elicits the behavior in question.
This seems a laudable goal. Clearly there's a connection between neurological/neurochemical mechanisms and behavior, and many studies exist to show this. If this weren't the case, we'd be the only species with "souls."
Corneel,
You meant to write gametic phase disequilibrium, I'm sure.
It's not so easy to get Fisher's process to work. Assuming there's natural selection against the male trait (think peacock's tail and a tiger) and against female choosiness (it takes time = $$$ to choose), selection has to be quite weak and the preference has to be quite strong for it to work in mathematical models. It may be easier to evolve if the male traits "advertize" some level of heritable superiority wrt other males.
Here are just a few:
http://www.richardronay.com/CBS/Research_files/The%20presence%20of%20an%20attractive%20woman%20elevates%20testo.pdf
http://primate.uchicago.edu/2009PNAS.pdf
http://www.psych.ucsb.edu/research/cep/papers/TOMbroadnarrow.pdf
Here are three excellent evo psych papers, all of which explore byproduct hypotheses:
http://rspb.royalsocietypublishing.org/content/270/1517/819.full.pdf
http://www.pnas.org/content/98/26/15387.long
http://www.psych.ucsb.edu/research/cep/papers/PNASCosmidesBarrettTooby2010.pdf
troy, the first does not seem to be a warranted assumption. There's no reason why the attractiveness need confer a selective disadvantage, those are just more interesting cases for geneticists (and easier to notice, since they provide a convenient adaptative explanation for the prevalence of an otherwise clearly nonadaptive trait).
And, as your last sentence recognizes, such mate preference is not at all surprising in cases where the attractiveness is associated with some other measure of fitness. This is routinely, trivially seen all the time, in female animals' preference for healthy males by various measures (courtship rituals in birds requiring the male to be a sufficiently strong flier, for instance). (I'm using females' selection of males here, but of course it can go either way)
Not a bad review here:
http://biology.anu.edu.au/hosted_sites/kokko/Publ/Matechoicereview.pdf
Troy said:
You meant to write gametic phase disequilibrium, I'm sure
Absolutely. Thanks for the correction.
Re. advertising male fitness. I am not a big fan of Zahavi's handicap principle. I just think it makes no sense to advertise superior fitness by sporting traits that completely nullify whatever small fitness advantage there was. Also, the heritability of fitness is typically very low, so there is little variation to fuel the evolution of handicap traits.
Hi Gary,
troy, the first does not seem to be a warranted assumption. There's no reason why the attractiveness need confer a selective disadvantage, those are just more interesting cases for geneticists (and easier to notice, since they provide a convenient adaptative explanation for the prevalence of an otherwise clearly nonadaptive trait).
That's an interesting point. There could be 'boring' traits that go largely unnoticed by us but which are the target of female mating preferences. Can you think of a good example? It occurred to me that it might be hard to detect such 'preferred' traits in simultaneous hermaphrodites, where sex-specific traits are absent by definition, but where (mutual?) mate choice might be important.
On the other hand, it has been argued (handicap principle, as noted by Corneel) that traits advertizing 'good genes' must be costly to express - otherwise such traits could too easily be faked by 'inferior' males and they would cease to be correlated with good genes.
Thanks for the link to the review paper!
Hi Corneel,
Re. advertising male fitness. I am not a big fan of Zahavi's handicap principle. I just think it makes no sense to advertise superior fitness by sporting traits that completely nullify whatever small fitness advantage there was.
That's probably why it took a long time to be, sort of, accepted as a potential explanation. Until some modelers (Alan Grafen I think) showed that it actually can work.
Post a Comment