Category Archives: evolutionary psychology

Basketball, Fraud, and the Broken University Incentive System

So, in 2005, Nature published a study by a group of researchers at Rutgers University showing that Jamaican teenagers who exhibited a high degree of body symmetry were judged (by other teenagers) to be better dancers than those with less symmetrical bodies. The study was an exploration of the idea that physical symmetry might be a trait subject to sexual selection — with “good dancers” serving as a proxy for “attractive mates.”

In 2013, Rutgers basketball coach Mike Rice received a $475,000 severance after video footage surfaced showing Rice pelting his players with basketballs, grabbing and shoving them, and shouting obscenities at them.

What do these two stories have in common? The stories reveal how Universities are not motivated by their stated missions (truth, education, etc.), so much as by a drive for money and prestige. As a friend of mine recently put it, “Pursuing truth is great, just don’t let it get in the way of your careerism.”

Now, you may be wondering how you can get a job at a University where you get to abuse students, and if you get caught and fired, you still walk away with nearly half a million dollars. I mean, you probably thought you could only get a job like that in an S & M dungeon, or maybe the CIA.

Here’s what pisses me off about the whole thing. Rice did not lose his job for abusing his players. He lost his job for embarrassing the University. When the video first surfaced, everything was handled in house. Rice was given a three-game suspension and a $50,000 fine — secretly. Then, when the video showed up on line, he was fired, along with an assistant coach and the athletic director (who received a $1.25 million severance package). (Here‘s one version)

Some players have argued that Rice’s behavior was not as bad as it appears on the video, and David Plotz argued in Slate that this type of behavior from his own high-school coach made him “a better player and a better man.” Maybe they’re right. Maybe we would all be better off if we were berated regularly for our shortcomings. There’s a balance in there somewhere between “preventing bullying” and raising a generation of hothouse flowers, and I’m not always convinced that we, as a society don’t veer too far in the direction of kid gloves.

The point is, you can imagine defending having a coach like Mike Rice. Likewise, you can imagine arguing that behavior like Rice’s is absolutely unacceptable in a University setting. What is bullshit is internally saying, yes, Mike Rice is great, he just needs to tone it down a notch, and then, when the public finds out, being all, “Oh my god! I can’t believe this was happening!” Basically, the sense you get is that no one was interested in figuring out what the right thing to do was. Everyone was motivated by protecting their own careers, and deflecting embarrassment away from the University.

So how does this relate to dancing teenagers in Jamaica?

The 2005 paper had seven co-authors, but the three key players are Robert Trivers and Lee Cronk, both from the Rutgers Anthropology Department, and William Brown (a postdoc at the time, now in the Psych department at University of Bedfordshire).

A couple of years after the paper came out, Trivers started to suspect that Brown had manipulated the data to make the results more compelling than they actually were. He went on to detail his analysis in a book published in 2009.

So what happens when one of the authors of a high profile paper comes out and claims that the results presented in the paper were not just wrong, but fraudulent? A news piece just out at Nature notes that Trivers approached Nature in 2008 about retracting the paper, but they were not interested. After all, why would they be? Nature’s business plan hinges on publishing studies that are exciting — studies that will be cited by a lot of other papers, and that will attract a lot of attention from the popular press.

This means that Nature (and other high-impact journals like Science) are particularly prone to a few types of bad papers. One is the shocking result that would lead to a major paradigm shift if it were true — but it is not true. One is the paper that seems to represent a huge advance, but is a modest advance that seems bigger than it is because it fails to cite much of the relevant literature. And one is the type we’re potentially talking about here, where a paper presents some really beautiful results, which are beautiful because the data has been manipulated in some way.

The problem is that this business model works great. The upsides of citations and press coverage apparently far outweigh the downsides of publishing incorrect, or incorrectly presented results. In general, there is not that much effort that goes into replicating results in science (arguably, a lot less than there should be). And even if a study fails to be replicated, and is shown to have been wrong, this typically happens years later, when no one is paying attention to the original study anymore.

That means that Nature and Science tend to have this funny status. In biology, anyway, everyone dismisses them as “magazines” or “tabloids.” At the same time, everyone is desperate to publish there.



So what do you do if you’re one of these journals? Well, if you are driven by a moral principle of promoting truth and spreading the best information possible, you put more effort into vetting papers up front, and when a paper is shown to be wrong, you actively try to correct the misperceptions stemming from the fact that people are saying, “Well, this was published in Nature, so it must be right.” On the other hand, if you are driven by the financial health of your corporation, you continue to publish things that will attract attention, and you retract or correct only when you absolutely have to, and you do it as quietly as possible, so as not to harm your brand.

Similarly, what do you do if you are a University associated with publishing a fraudulent paper? Again, if your guiding principles are truth and honest scholarship . . . oh, who am I kidding, those have not been the guiding principles of Universities for decades. As a University, your guiding principle is to maximize your money and prestige (to the extent that prestige begets money). That means that you will promote truth to the extent that you are required to do so, in order to maintain your brand as an important “research” and “education” institution. But what you really want is to keep the accusations of fraud as quiet as possible.

According to Trivers’s account of events, his going public with his book detailing the case for fraud, Rutgers was forced to undertake an investigation, since the work was supported in part by a grant from the NSF. Had they failed to do so, they might have become ineligible for NSF funds. When they completed their investigation in 2012, the University refused to make the results public. Trivers has posted the report on his own site (here). In brief, the report says that, yes, there is evidence for fraud in the paper (although the Nature news piece notes that Brown still denies having manipulated the data).

There’s another part of the story, though, that is much more disturbing. Trivers tells of an incident where he went to confront co-author Lee Cronk in his office about the situation. Trivers felt that Cronk had backed Brown’s version of events, and that the Rutgers report had vindicated his position over theirs, and he called Cronk a “punk.” Cronk then reported the incident to the campus police, and Trivers wound up getting banned from campus for alleged violation of the University’s anti-violence policy.

Now, clearly, I was not there, and so I have no independent source of knowledge about what happened that day. However, the version of event related by Trivers (read the whole thing here), sounds completely consistent with my own interactions with him. His version of the story suggests motivations and actions on the part of others:

It suggests that Cronk overreacted / acted punitively when Trivers confronted him. Maybe he was embarrassed and angry, and pretended to have felt physically threatened by Trivers in order to get back at him. Maybe he really did feel physically threatened, because, maybe, like most academics, he’s sort of a wuss. Maybe he overreacted in the heat of the moment, later wishing that he had not gotten the police involved. I don’t know. I’ve never met him, but this does pass the sniff test as the type of thing I can imagine a lot of Professor types doing.

It suggests that Rutgers sided against Trivers in the incident, perhaps out of retaliation for his original whistleblowing about the paper and the embarrassment and financial cost his actions imposed on the University. After all, if he had kept his mouth shut, they would not have had to undertake the whole fraud investigation, which I’m sure cost a bundle, and they would not have had the embarrassment of national press coverage. Again, clearly, I have no privileged details here, but this sounds like something you would do if you were a University that

I don’t want to pick on Rutgers in particular here. I think that this is the course of action that would have been taken by any University. That does not make it right, though.

I also don’t want to pick on Brown and Cronk, both of whom I suspect are perfectly bright and competent researchers. Even if Brown did actively manipulate the data, I see that as a pathology of the system, where getting a Nature paper can mean the difference between getting a job and not getting a job. If we didn’t hand out jobs and funding and promotions based on numbers of publications and citations — but rather on the quality and rigor of people’s research — there would not be incentives to manipulate or fabricate data. Even if Cronk maliciously got the Rutgers police to ban Trivers from campus, he did so in the context of a system that almost never rewards standing up and saying, “You know what, I was wrong.”

I know a lot of great scientists, people who are their own harshest critics, who are reluctant to publish results until they are 100% certain, who view even critical reviews of their manuscripts as good opportunities to make their work better. I also know a lot of scientists who are happy to cut corners, exaggerate the significance and importance of their results, and promote themselves, even, sometimes, at the expense of the truth.

Unfortunately, the way the current incentive system works, the latter group tend to have much better jobs than the former.

Until we can fix that, everything is going to conspire to encourage researchers to exaggerate, manipulate, and even fabricate their data. And everything is going to conspire to discourage Universities and journals from addressing and correcting fraudulent and erroneous results.

We have to reward honest, serious work that is not necessarily flashy. And we have to reward people who are willing to admit and correct mistakes.

Most of all, we have to reassert the principle of “truth” as our highest value in academia, and fight against its erosion by the secondary values of fame, prestige, and, most importantly, funding.


Disclosure statements.

Since 2011, I have had an unpaid adjunct / visiting scholar position at Rutgers in the Genetics Department. During that time, I have gotten to know Trivers a bit, and we, in fact, have plans to work on a project together. In my experience, he is brilliant, boisterous, foul-mouthed, politically incorrect, and honest to a fault — exactly the sort of person all academics should aspire to be. Unfortunately, in their current incarnation, Universities much favor mediocre minds and personalities who will toe the corporate line, and who will bring funding to the University without inconvenient things like “truth” get in the way.

In elementary school, I was once on the losing side of a full-length, YMCA-league basketball game that ended 6 to 5. In junior high, I once attended a week-long basketball camp at the end of which I received the Orwellian “Most Improved Player” award.

Scocca on Toy Gender Apartheid

So, here’s an awesome piece to make you feel shitty about whatever you bought your kids for Christmas. Tom Scocca has written an excellent screed about “Toy Apartheid” and the cultural enforcement of gender norms on young children. Here are a couple of highlights:

Christina Hoff Sommers—who has made a nice career in the Boys’ Toys section of the opinion-having business by arguing over and over that men have been victimized by feminism—explained to the readers of The Atlantic’s website how dangerous this intervention in the toy-marketing business really is. “[N]othing short of radical and sustained behavior modification” can change children’s “elemental play preferences,” she wrote. And:

The Swedes are treating gender-conforming children the way we once treated gender-variant children.

They are calling them special epithets and beating them up and sometimes killing them? (Also: “Once”? Do tell.) But no, sorry, what she means is that these scheming Nordic elites are trying to modify the poor children’s natural behavior. Only they aren’t even doing that, really. They’re just putting some different photos in the toy ads.

Here, he responds to an anecdote Sommers tells about the time that her granddaughter, when given a toy train, placed it in a baby carriage and covered it with a blanket:

But as the sociologists say: so fucking what? What’s the damage if little Eliza wants to rock Baby Train to sleep? Baby Train is snug and warm, and is also an inanimate object. Little Eliza is enjoying herself. Why does her grandmother have to be an asshole about it?

And here is, I think the correct response to anyone who uses anecdotes about children to justify gender conformity as “natural”:

Children are stupid and rotten and conformist, and elevating their weakness to a point of pride is insane. Trying to make them to stop being that way isn’t confusing or cruel; it’s one of the basic duties of being a parent and adult. Any pundit who starts holding forth on the superior behavioral wisdom of small children deserves to be bitten by one.

I miss when he had a regular blog at Slate sooo much.

Epigenetics and Homosexuality

So, last week featured a lot of news about a paper that came out in the Quarterly Review of Biology titled “Homsexuality as a Consequence of Epigenetically Canalized Sexual Development.” The authors were Bill Rice (UCSB), Urban Friberg (Uppsala U), and Sergey Gavrilets (U Tennessee). The paper got quite a bit of press. Unfortunately, most of that press was of pretty poor quality, badly misrepresenting the actual contents of the paper. (PDF available here.)

I’m going to walk through the paper’s argument, but if you don’t want to read the whole thing, here’s the tl;dr:

This paper presents a model. It is a theory paper. Any journalist who writes that the paper “shows” that homosexuality is caused by epigenetic inheritance from the opposite sex parent either 1) is invoking a very non-standard usage of the word “shows,” or 2) was too lazy to read the actual paper, and based their report on the press release put out by the National Institute for Mathematical and Biological Synthesis.

That’s not to say that this is a bad paper. In fact, it’s a very good paper. The authors integrate a lot of different information to come up with a plausible biological mechanism for epigenetic modifications to exert influence on sexual preference. They demonstrate that such a mechanism could be favored by natural selection under what seem to be biologically realistic conditions. Most importantly, they formulate their model into with clear predictions that can be empirically tested.

But those empirical tests have not been carried out yet. And, in biology, when we say that a paper shows that X causes Y, we generally mean that we have found an empirical correlation between X and Y, and that we have a mechanistic model that is well enough supported that we can infer causation from that correlation. This paper does not even show a correlation. It shows that it would probably be worth someone’s time to look for a particular correlation.

As a friend wrote to me in an e-mail,

I found it a much more interesting read than I thought I would from the press it’s getting, which now rivals the bullshit surrounding the ENCODE project for the most bullshitty bullshit spin of biology for the popular press. A long-winded-but-moderately-well-grounded-in-real-biology mathematical model does not proof make.


Okay, now the long version.

The Problem of Homosexuality

The first thing to remember is that when an evolutionary biologist talks about the “problem of homosexuality,” this does not imply that homosexuality is problematic. All it is saying is that a straightforward, naive application of evolutionary thinking would lead one to predict that homosexuality would not exist, or would be vanishingly rare. The fact that it does exist, and at appreciable frequency, poses a problem for the theory.

In fact, this is a good thing to keep in mind in general. The primary goal of evolutionary biology is to understand how things in the world came to be the way they are. If there is a disconnect between theory and the world, it is ALWAYS the theory that is wrong. (Actually, this is equally true for any science / social science.)

Simply put, heterosexual sex leads to children in a way that homosexual sex does not. So, all else being equal, people who are more attracted to the opposite sex will have more offspring than will people who are less attracted to the opposite sex.

[For rhetorical simplicity, I will refer specifically to “homosexuality” here, although the arguments described in the paper and in this post are intended to apply to the full spectrum of sexual orientation, and assume more of a Kinsey-scale type of continuum.]

The fact that a substantial fraction of people seem not at all to be attracted to the opposite sex suggests that all else is not equal.

Evolutionary explanations for homosexuality are basically efforts to discover what that “all else” is, and why it is not equal.

There are two broad classes of possible explanation.

One possibility is that there is no biological variation in the population for a predisposition towards homosexuality. Then, there would be nothing for selection to act on. Maybe the potential for sexual human brain simply has an inherent and uniform disposition. Variation in sexual preference would then be the result of environmental (including cultural) factors and/or random developmental variation.

This first class of explanation seems unlikely because there is, in fact, a substantial heritability to sexual orientation. For example, considering identical twins who were raised separately, if one twin is gay, there is a 20% chance that the other will be as well.

Evidence suggests that sexual orientation has a substantial heritable component. Image: Comic Blasphemy.

This points us towards the second class of explanation, which assumes that there is some sort of heritable genetic variation that influences sexual orientation. Given the presumably substantial reduction in reproductive output associated with a same-sex preference, these explanations typically aim to identify some direct or indirect benefit somehow associated with homosexuality that compensates for the reduced reproductive output.

One popular variant is the idea that homosexuals somehow increase the reproductive output of their siblings (e.g., by helping to raise their children). Or that homosexuality represents a deleterious side effect of selection for something else that is beneficial, like how getting one copy of the sickle-cell hemoglobin allele protects you from malaria, but getting two copies gives you sickle cell anemia.

It was some variant of this sort of idea that drove much of the research searching for “the gay gene” over the past couple of decades.  The things is, though, those searches have failed to come up with any reproducible candidate genes. This suggests that there must be something more complicated going on.

The Testosterone Epigenetic Canalization Theory

Sex-specific development depends on fetal exposure to androgens, like Testosterone and related compounds. This is simply illustrated by Figure 1A of the paper:

Figure 1A from the paper: a simplified picture of the “classical” view of sex differentiation. T represents testosterone, and E represent Estrogen.

SRY is the critical genetic element on the Y chromosome that triggers the fetus to go down the male developmental pathway, rather than the default female developmental pathway. They note that in the classical model of sex differentiation, androgen levels differ substantially between male and female fetuses.

The problem with the classical view, they rightly argue, is that androgen levels are not sufficient in and of themselves to account for sex differentiation. In fact, there is some overlap between the androgen levels between XX and XY fetuses. Yet, in the vast majority of cases, the XX fetuses with the highest androgen levels develop normal female genitalia, while the XY fetuses with the lowest androgen levels develop normal male genitalia. Thus, there must be at least one more part of the puzzle.

The key, they argue, is that tissues in XX and XY fetuses also show differential response to androgens. So, XX fetuses become female because they have lower androgen levels and they respond only weakly to those androgens. XY fetuses become male because they have higher androgen levels and they respond more strongly to those androgens.

This is illustrated in their Figure 1B:

Sex-specific development is thus canalized by some sort of mechanism that they refer to generically as “epi-marks.” That is, they imagine that there must be some epigenetic differences between XX and XY fetuses that encode differential sensitivity to Testosterone.

All of this seems well reasoned, and is supported by the review of a number of studies. It is worth noting, however, that we don’t, at the moment, know exactly which sex-specific epigenetic modifications these would be. One could come up with a reasonable list of candidate genes, and look for differential marks (such as DNA methylation or various histone modifications) in the vicinity of those genes. However, this forms part of the not-yet-done empirical work required to test this hypothesis, or, in the journalistic vernacular, “show” that this happens.

Leaky Epigenetics and Sex-Discordant Traits

Assuming for the moment that there exist various epigenetic marks that 1) differ between and XX and XY fetuses and 2) modulate androgen sensitivity. These marks would need to be established at some point early on in development, perhaps concurrent with the massive, genome-wide epigenetic reprogramming that occurs shortly after fertilization.

The theory formulated in the paper relies on two additional suppositions, both of which can be tested empirically (but, to reiterate, have not yet been).

The first supposition is that there are many of these canalizing epigenetic marks, and that they vary with respect to which sex-typical traits they canalize. So, some epigenetic marks would canalize gonad development. Other marks would canalize sexual orientation. (Others, they note, might canalize other traits, like gender identity, but this is not a critical part of the argument.)

The model presented in this paper suggests that various traits that are associated with sex differences may be controlled by distinct genetic elements, and that sex-typical expression of those traits may rely on epigenetic modifications of those genes. Image:

The second supposition is that the epigenetic reprogramming of these marks that normally happens every generation is somewhat leaky.

There are two large-scale rounds of epigenetic reprogramming that happen every generation. One occurs during gametogenesis (the production of eggs or sperm). The second happens shortly after fertilization. What we would expect is that any sex-specifc epigenetic marks would be removed during one of these phases (although it could happen at other times).

For example, a gene in a male might have male-typical epigenetic marks. But what happens if that male has a daughter? Well, normally, those marks would be removed during one of the reprogramming phases, and then female-typical epigenetic marks would be established at the site early in his daughter’s development.

The idea here is that sometimes this reprogramming does not happen. So, maybe the daughter inherits an allele with male-typical epigenetic marks. If the gene influences sexual orientation by modulating androgen sensitivity, then maybe the daughter develops the (male-typical) sexual preference for females. Similarly, a mother might pass on female-typical epigenetic marks to her son, and these might lead to his developing a (female-typical) sexual preference for males.

So, basically, in this model, homosexuality is a side effect of the epigenetic canalization of sex differences. Homosexuality itself is assumed to impose a fitness cost, but this cost is outweighed by the benefit of locking in sexual preference in those cases where reprogramming is successful (or unnecessary).

Sociological Concerns

Okay, if you ever took a gender-studies class, or anything like that, this study may be raising a red flag for you. After all, the model here is basically that some men are super manly, and sometimes their manliness leaks over into their daughters. This masculinizes them, which makes them lesbians. Likewise, gay men are gay because they were feminized by their mothers.

That might sound a bit fishy, like it is invoking stereotype-based reasoning, but I don’t think that would be a fair criticism. Nor do I think it raises any substantial concerns about the paper in terms of its methodology or its interpretation. (Of course, I could be wrong. If you have specific concerns, I would love to hear about them in the comments.) The whole idea behind the paper is to treat chromosomal sex, gonadal sex, and sexual orientation as separate traits that are empirically highly (but not perfectly) correlated. The aim is to understand the magnitude and nature of that empirical correlation.

The other issue that this raises is the possibility of determining the sexual orientation of your children, either by selecting gametes based on their epigenetics, or by reprogramming the epigenetic state of gametes or early embryos. This technology does not exist at the moment, but it is not unreasonable to imagine that it might exist within a generation. If this model is correct in its strongest form (in that the proposed mechanism fully accounts for variation in sexual preference), you could effectively choose the sexual orientation of each of your children.

Image: Brainless Tales.

This, of course, is not a criticism of the paper. The biology is what it is. It does raise certain ethical questions that we will have to grapple with at some point. (Programming of sexual orientation will be the subject of the next installment of the Genetical Book Review.)

Plausibility/Testability Check

The question one wants to ask of a paper like this is whether it is 1) biologically plausible, and 2) empirically testable. Basically, my read is yes and yes. The case for the existence of mechanisms of epigenetic canalization of sex differentiation seems quite strong. We know that some epigenetic marks seem to propagate across generations, evading the broad epigenetic reprogramming. We don’t understand this escape very well at the moment, but the assumptions here are certainly consistent with the current state of our knowledge. And, assuming some rate of escape, the model seems to work for plausible-sounding parameter values.

Testing is actually pretty straightforward (conceptually, if not technically). Ideally, empirical studies would look for sex-specific epigenetic modifications, and for variation in these modifications that correlate with variation in sexual preference. The authors note that one test that could be done in the short term would be to do comparative epigenetic profiling of the sperm of men with and without homosexual daughters.

As Opposed to What?

The conclusions reached by models in evolution are most strongly shaped by the set of alternatives that are considered in the model. That is, a model might find that a particular trait will be selectively favored, but this always needs to be interpreted in the context of that set of alternatives. Most importantly, one needs to ask if there are likely to be other evolutionarily accessible traits that have been excluded from the model, but would have changed the conclusions of the model if they had been included.

The big question here is the inherent leakiness of epigenetic reprogramming. A back-of-the-envelope calculation in the paper suggests that for this model to fully explain the occurrence of homosexuality (with a single gene controlling sexual preference), the rate of leakage would have to be quite high.

An apparent implication of the model is that there would then be strong selection to reduce the rate at which these epigenetic marks are passed from one generation to the next. In order for the model to work in its present form, there would need to be something preventing natural selection from finding this solution.

Possibilities for this something include some sort of mechanistic constraint (it’s just hard to build something that reprograms more efficiently than what we have) or some sort of time constraint (evolution has not had enough time to fix this). The authors seem to favor this second possibility, as they argue that the basis of sexual orientation in humans may be quite different from that in our closest relatives.

On the other hand this explanation could form a part of the explanation for homosexuality with much lower leakage rates.

What Happened with the Press?

So, how do we go from what was a really good paper to a slew of really bad articles? Well, I suspect that the culprit was this paragraph from the press release from NIMBios:

The study solves the evolutionary riddle of homosexuality, finding that “sexually antagonistic” epi-marks, which normally protect parents from natural variation in sex hormone levels during fetal development, sometimes carryover across generations and cause homosexuality in opposite-sex offspring. The mathematical modeling demonstrates that genes coding for these epi-marks can easily spread in the population because they always increase the fitness of the parent but only rarely escape erasure and reduce fitness in offspring.

If you know that this is a pure theory paper, this is maybe not misleading. Maybe. But phrases like “solves the evolutionary riddle of homosexuality” and “finding that . . . epi-marks . . . cause homosexuality in opposite-sex offspring,” when interpreted in the standard way that I think an English speaker would interpret them, pretty strongly imply things about the paper that are just not true.

Rice, W., Friberg, U., & Gavrilets, S. (2012). Homosexuality as a Consequence of Epigenetically Canalized Sexual Development The Quarterly Review of Biology, 87 (4), 343-368 DOI: 10.1086/668167

Update: Also see this excellent post on the subject by Jeremy Yoder over at Nothing in Biology Makes Sense.

Legitimate rape, seminal priming, and preeclampsia

So, as you are well aware, a couple of days ago, human-shaped pile of garbage Todd Akin articulated his belief that “legitimate rape” rarely leads to pregnancy, due to the magical uterine “shutting that whole thing down” properties of the uterus. Here at Lost in Transcription, we discussed the fact that there are some species that do, in fact, exhibit the capacity for “post-copulatory female choice.” However, humans are not one of these species, unless you count the set of medical interventions that Akin is trying to outlaw (along with Romney, Ryan, and the official Republican party platform).

If you’re interested, Kate Clancy wrote up an excellent summary of the actual science on the topic of pregnancy and rape.

Jesse Bering (an evolutionary psychologist who has been featured on this blog previously) also weighed in on the science, using twitter to point to an article that had recently written on “Darwin’s Morning After Pill.” In the article, Bering outlines an argument for the adaptive value of preeclampsia. The argument features “seminal priming theory,” which Bering calls “criminally unread,” and which has been promoted by Gordon Gallup (who, like a certain goat I know, has an adaptationist story for just about everything).

Roughly, the argument is this. Women don’t want to let a man get them pregnant unless they are certain that the man is going to stick around for the long haul. So, you want to have a biological mechanism that prevents pregnancy from one-night stands, but encourages pregnancy when you are in a committed relationship. Preeclampsia is a convenient (if life threatening) way for mother nature to terminate your pregnancy when it would be better not to have a baby. Therefore, preeclampsia should be more common for pregnancies resulting from sex with an unfamiliar male. Preeclampsia should be less common when it is the product of a long-term sexual relationship.

The proposed mechanism is that exposure to a male’s semen sort of habituates the female to the biochemistry of that particular male. Preeclampsia is associated with certain inflammatory features that share some similarities with an immune response. In that sense, preeclampsia is sometimes thought of as the mother rejecting the foreign body of the fetus, sort of like how one might reject a transplanted organ. The idea, then, is that through exposure to the male’s semen, the female ratchets down this response, thereby allowing the pregnancy to move forward.

One way of thinking about preeclampsia is as the rejection of an alien body by the mother

Jeremy Yoder has written a nice piece detailing how, even if we accept all of this, it is ridiculous to think of this mechanism as an adaptation. In particular, even under the most generous set of assumptions, natural selection acting on such a mechanism would be vanishingly small. And, of course, even to get there, you have to buy the typical evolutionary psychology assumption of an “environment of evolutionary adaptation” that looks an awful lot like the normative middle-class, suburban values of 1950s television America.

There are a few lines of evidence that are cited (by Bering, and in general) in support of the idea. The bulk of the evidence hinges on the observation that changing partners increases a woman’s probability of preeclampsia. For example, if your second pregnancy has the same father as your first pregnancy, you are less likely to develop preeclampsia than if the two pregnancies have different fathers. This is a finding that has been replicated a number of times, and with very large samples, so that’s pretty solid, right?

Actually, no. While the association between paternity switch and preeclampsia is true, it probably doesn’t mean what Bering and Gallup think it means, and the relevant data doesn’t actually support the seminal priming theory.

The problem is that a change in paternity correlates with time between pregnancies. So, if your two kids have different fathers, it is more likely that the two pregnancies were spaced farther apart. My reading of the literature is this: in every case where there is an association between paternity switching and preeclampsia, the study has not separately controlled for time between pregnancies. In each study where time between pregnancy is explicitly controlled for, the association with paternity switching vanishes. (See, e.g., this or this.)

In fact, controlling for time between pregnancies, if you have preeclampsia in your first pregnancy, switching partners actually makes you less likely to have preeclampsia in your second one. Don’t get too excited though. The converse is also true. If you don’t get preeclampsia in one pregnancy, switching partners makes you more likely to get it in the next one.

What that suggests (to me, anyway), is that some fathers are more likely to produce preeclampsia than others (or, alternatively, that the probability of preeclampsia depends on some interaction between the maternal and paternal genotypes). According to this explanation, if you don’t get preeclampsia, it means that you and your partner are at low risk. If you switch partners, though, you go back into the standard risk pool. (This interpretation is also consistent with this study, which followed fathers.)

There are a few other lines of evidence, which are cleaner in their implications. One study on artificial insemination finds preeclampsia more often in cases where the woman was inseminated with a stranger’s sperm than in cases where she was inseminated with her partner’s. There is also a study that finds that frequent oral sex correlates with a reduced risk of preeclampsia. (That’s her performing oral sex on him, not the other way round.) Does the frequency of oral sex correlate with the spacing between kids? I don’t know. I’m hoping that some of you will weigh in on that in the comments.

Males ingesting female gametes also has well documented health benefits.

The problem with these studies is that, unlike the partner-switching studies, we’re looking at small numbers. Whether or not they will hold up under more extensive analysis it not yet clear.

My read on the whole thing? At the moment, the data just isn’t there. All that exists in support of the seminal priming theory is an adaptationist fairy tale and a couple of small studies that have yet to be reproduced.

Oh, and also a whole bunch of studies that, if you cherry pick from among them, and ignore all of the studies that contradict them, support the theory. Of course, that’s pretty much true of any theory, which is exactly why evolutionary psychology continues to be such a booming field.

Toxoplasmosis Extravaganza: Ride Complete!

So, this week at Darwin Eats Cake, we celebrated our one-year anniversary with a series of nine strips on the zooparasite Toxoplasma gondii. This parasite, which causes Toxoplasmosis, is the reason why pregnant women are encouraged to avoid cat litter.

Here’s the full series, presented for your one-stop-shopping viewing pleasure. The strips do not, I think, assume any expert biological knowledge, so you don’t need to be a parasitologist to enjoy them. However, a dorky and juvenile sense of humor will help a lot. Alternatively, you can read them on the Darwin Eats Cake website, where they look a little better, I think. The series starts at

At this point, Darwin Eats Cake will return to its regular programming schedule, with twice-a-week updates, usually on Mondays and Thursdays, except for those days that have been recognized as official holidays by the Darwin Eats Cake Council of Freeholders and its chairwoman, the duly elected Queen of Naboo.

So, stop by on Monday for a new strip, or any time to trawl the archive:

It’s Toxoplasmosis week at Darwin Eats Cake

So, tomorrow (March 13) marks the one-year anniversary of the launch of my webcomic Darwin Eats Cake on its very own website (here). Normally, Darwin Eats Cake updates approximately twice a week (hemicircaseptanally), on approximately Monday and Thursday (circa-Mondarily and circa-Thursdarily, I assume). However, to mark this special anniversary occasion, we are rolling out a daily series of strips on Toxoplasma gondii, the parasite responsible for Toxoplasmosis. This bug was recently in the news thanks to a profile of Jaroslav Flegr published recently in the Atlantic (here).

Here are the first two of this week’s six strips:

Best URL for sharing:
Permanent image URL for hotlinking or embedding:
Best URL for sharing:
Permanent image URL for hotlinking or embedding:

And remember: Sharing is Caring!

Evolutionary Psychology Freudian slip

So, here’s something that’s a little bit awesome.  At least it’s awesome if you’re the type of Evolutionary Biologist who likes to poke fun at Evolutionary Psychology. Which is to say, if you’re an Evolutionary Biologist.

If you go to John Tooby’s webpage, and click on the link labeled “Advanced Theory and Method in Evolutionary Psychology,” you get this:

Hat tip to someone whose name I won’t post here, lest it should negatively impact his and/or her job prospects.

On the nocturnal penile tumescence

So, here’s the latest Darwin Eats Cake. Once again, Guillaume is gracing us with his adaptationist explanations. This time, he is answering a question from Bastian Greshake (@gedankenstuecke), champion of evolution, creative commons, and all sorts of other good stuff.

If you’re not familiar with Creative Commons, it is an alternative to traditional copyrights. It’s a great option if you’re committed to an open culture, where creations can be shared, but want to protect yourself against having your creations exploited.

For instance, all of the Darwin Eats Cake strips are published under a creative commons license explicitly stating that you are free to share them. You can e-mail them, copy them into your own blog, print them out, pretty much anything you want, just so long as you provide attribution. The only thing you can’t do is sell them.

Best URL for sharing:
Permanent image URL for hotlinking or embedding:

Why do we make odd faces when we orgasm? A romance in three parts

So, Guillaume’s Mailbag has continued on its mission to provide an adaptive explanation for every existing trait. The most recent trait Guillaume has been tackling was submitted by John Wilkins, who asked, “Why do we make odd faces when we orgasm?”

In case you missed when I’ve plugged him before, JoHn Wilkins (no recent relation) is a philosopher of science in Australia. His most recent book is Species: A History of the Idea, and he runs an excellent blog called Evolving Thoughts. He recently concluded an excellent series of posts on “Atheism, agnosticism and theism” in which he discusses, among other things, what it means to have a belief. You can find the start of that series here.

But back to the face of orgasm. Guillaume took three full strips to answer this one, so I’ve waited until he was done to post them here. I think I’ve finally figured out how to make these full-page versions more readable on the blog, but it involved lowering the resolution of the JPEG, so, for higher-res versions of these three comics, head on over to Darwin Eats Cake. The first of the series of three can be found here.

Best URL for sharing:
Permanent image URL for hotlinking or embedding:
Best URL for sharing:
Permanent image URL for hotlinking or embedding:
Best URL for sharing:
Permanent image URL for hotlinking or embedding:
For those who are interested, a couple of vole and oxytocin citations are provided below to get you started. The vole literature is actually quite extensive and all interesting. I’ve included a relatively recent paper, which will contain citations to a lot of the other work. No peer-reviewed publications are yet available on the eating and mating habits of Ursus philorgasmii.

Ross HE, Cole CD, Smith Y, Neumann ID, Landgraf R, Murphy AZ, & Young LJ (2009). Characterization of the oxytocin system regulating affiliative behavior in female prairie voles. Neuroscience, 162 (4), 892-903 PMID: 19482070

Carmichael MS, Warburton VL, Dixen J, & Davidson JM (1994). Relationships among cardiovascular, muscular, and oxytocin responses during human sexual activity. Archives of sexual behavior, 23 (1), 59-79 PMID: 8135652

Although at least one study suggests that, in men, prolactin is actually more strongly correlated with orgasm than oxytocin is:

Krüger TH, Haake P, Chereath D, Knapp W, Janssen OE, Exton MS, Schedlowski M, & Hartmann U (2003). Specificity of the neuroendocrine response to orgasm during sexual arousal in men. The Journal of endocrinology, 177 (1), 57-64 PMID: 12697037