Why do mothers kill their children? Over at Scientific American, guest blogger Eric Michael Johnson, of Primate Diaries fame, has fashioned a nicely turned essay considering one answer  to this question— or at least a partial answer — offered by researcher Dario Maestripieri: When mothers kill their children, they are reacting to a particularly toxic combination of stress, powerlessness, and social disadvantage.

Maestripieri, Johnson writes,

has spent most of his career studying maternal behavior in primates. In particular, he’s focused on the factors that influence a mother’s motivation towards her young. As a professor of Comparative Human Development, Evolutionary Biology, Neurobiology, and Psychiatry at the University of Chicago he has enjoyed the kind of cross-disciplinary success that most scientists only dream of. His 153 academic papers and six books have been cited more than a thousand times by scholars (including this one) in many of the world’s top scientific journals. His latest paper is scheduled to be published in early 2011 by the American Journal of Primatology. In it Maestripieri lays out the argument he’s built over the last two decades showing how one of the most serious impacts on maternal behavior, one with potentially lethal results, is so common in modern life as to be nearly invisible: stress.

Johnson notes right off that stress is usually a good thing, for it’s really a way to ramp up mind and body to meet a challenge, whether fighting off an attack, calming a screaming toddler, or giving a talk before a big crowd. Most of the time our “stress response” — a heightened energy and vigilance driven by (among other things) a rise in cortisol levels — serves us well. Extended and or intense stress, however, does not. It erodes the body and tends to lead to behaviors that aren’t adaptive.

In this case the maladaptive behavior is the murder of one’s children, and to examine it, Johnson looks at Maestripieri’s work on how stress affects mothering in rhesus macaque monkeys — and how that effect differs in monkeys of different social strata. This area remains ripe despite being heavily harvested. Both Maestripieri and Stephen Suomi, among others, have shown that within the high-stakes hierarchy of rhesus society, status plays a huge role in a monkey’s health, lifespan, and genetic success (that is, whether and how robustly the monkey passes on his or genes). Monkeys with secure positions in higher-status family groups  do better on all these counts because they  get more food and mating opportunities and have more allies in the sometimes violent power struggles through which rhesus work out their social positions.

Being low in this hierarchy has serious consequences, as Maestripieri found when he studied monkeys in a large, wild colony on an island in Puerto Rico:

[T]he team analyzed the colony’s mortality records covering a period of ten years and found that infants born to low-ranking females were much more likely to die in their first year than those born to high-ranking ones. As a result, low-ranking mothers were living in a state of constant panic. They would watch as their offspring were confronted by dangerous group members but they were powerless to do anything about it. Unable to act while their innate warning system screamed at high alert, their anxiety simply grew, expanding out of proportion as a result of the social inequality.

Average change in cortisol levels for pregnant/lactating females in three social ranks. Image reproduced from Hoffman et al. (2010).

I’m not sure the anxiety expanded out of proportion; you could argue that an anxious response is proportional in the sense that it recognizes that the offspring face grave odds. On the other hand, you could consider it out of proportion because it can drive responses that make matters worse. Stressed-out rhesus mothers, for instance, tend to  be overcontrolling or overly harsh with their offspring. This generates anxiety in the young monkey and compromises its social skills. This in turn continues the cycle of low status, social isolation, and stress. As often happens in life,  poor responses help create a self-reinforcing loop.

In any case, Johnson, apparently echoing the researchers, emphasizes that poverty, low social status, and a feeling of helplessness can combine with other problems to create a deadly mix — including heightening the chance that a mother will kill her children.

I don’t doubt it. Johnson cites this as good reason to address root causes of poverty, and I can’t disagree with that either; it is a good reason to address poverty more seriously, and in particular to create economic policies that prevent long stretches of unemployment. A society as wealthy as ours should be able to do these things. Instead we seem to be headed the other way.

So I agree with what we might call the social lessons Johnson draws from these studies. But I wonder if stopping there pulls us up short of some of the more interesting scientific implications of these studies. I wonder if stopping there mires us in a vision of environment that gives it too much primacy. It suggests that environment is deterministic.

Now, I know that Johnson doesn’t think the environment is deterministic; he’s a smart guy schooled in evolutionary biology, and he knows very well that not only do genes play a role in how we respond to our environments, but that our response comes out of a sort of black box — a highly complicated set mechanisms still hidden from us — in which genes and environment experience produce a behavior that in turn determines outcome more directly than environment does. In this case, for instance, the behavior is murder, and the outcome is dead children. Clearly the social environment, the tremendous stress of poverty and social isolation, contributes. But that’s not all that’s going on in the black box. And the problem with stopping with environment as a cause — as an endpoint to a social argument, if not a scientific argument — is that it leaves the black box closed. And I think this is a time when we should be trying to crack open that black box

Unfortunately, the black box here — the one in which genes and experience mysteriously mix to create behavior that determines our fates — is a Pandora’s box politically. We have been long stuck in an argument about nature versus nurture,  genes versus environment, in which the nature/gene argument is associated with conservative and racist views, which emphasize individual agency, and the environment/nurture argument is associated with liberal views that seek to legislate an egalitarian, fair, decent society. I’m all for the latter. I understand the political dangers of overemphasizing agency: of pushing the idea that since behavior shapes our fates, people who want their lives to go well should simply behave differently — as if it’s just as easy for the poor kid to act with confidence, consistency, and strength, to say nothing of skillsets, as it is for his richer, more fortunate peers. That’s a warped view. It  ignores that the poor kid who achieves great success is notable precisely because she defies the odds.

So yes, putting behavior on the table makes for a trickier game. But I think when we end discussions of the scientific links between environment and poor outcomes with a simple call for better environments, we trap the conversation in a place from which we need to move on — and can move on now that we’re getting a better handle on how genes and environment mix to produce behavior. We can begin, for instance, to  reframe how we view genetic variants that clearly affect our responses to environment. We can get a richer view inside the black box and see dynamics that can simultaneously emphasize the environment, one’s genes, and one’s agency. We can recognize that behavior occurs not as the endpoint of a set of prior experiences — a domino that is struck and falls — but as part of a circular dynamic in which behavior is both cause and effect.

This may be an arbitrary caveat to raise about a strong, intelligent, insightful essay I largely agree with. And I don’t know exactly where to go with this discussion, or how Johnson might have left me feeling more satisfied with his treatment. Yet as much I enjoyed the piece, I did finish it feeling slightly unsatisfied. And I think this is because Johnson, in trying to emphasize a link between poor environments and a certain outcome–in this case, poverty and the murder of children — left closed a black box that I think we should take every chance to open.

I would love to see, for instance, those low-ranking, stressed out monkey mothers in the Maestripieri study split by genotype. For as I noted in an article about stressed-out rhesus moms that Suomi studied, neurotic rhesus mothers in low social positions are more likely to raise neurotic, low-status offpsring if the mothers themselves were a)  raised in such circumstances themselves and — a huge and — b) they carry the S-allele of the serotonin transporter gene, which is associated with greater sensitivity to environment in general and social experience in particular. It’s this combination that produces the risk — that creates a disproportionate number of the overanxious, undernurturing moms, the high cortisol levels and low social skills, the self-perpetuating loop of low status and inherited and taught dysfunction.

Figuring out how that works is a lot harder than just pointing to environment. But it seems to me a more interesting problem. And in a world where the “He’s a product of his environment” argument is not exactly carrying the day anyway, perhaps this more complex problem is a more fruitful one to explore politically as well as scientifically.

I say this not knowing where it will lead —except out of a dead-end argument we’ve been having for too long anyway.

[Ed. note, 25 Nov 2010: fixed a mess of typos and confusing diction (some due to voice-recognition software booboos) and a few sloppy constructions. No substantive changes.]

Image: Eugene Delacroix, Medea prepares to kill her children. Courtesy Wikimedia Commons.