Speaking of attachment…
You guys, there are things I don’t know. I bump into brand new stuff – even stuff I think ought to have known before – pretty regularly. Example: Clelia Mosher‘s work was unknown to me until a couple months ago.
So I’m absolutely certain that there are plenty of things that I’ll learn in the future and plenty of things that I’ll feel like I ought to have known before.
But here is something I’ve never heard before, from our old friend the Evolutionary Agony Aunt that just sounds like… well… total bullshit:
Human behaviour seems to be under the control of two evolutionary programs: one that results in fertilisation, disillusionment and a series of partners, and the other that enables humans to develop the lasting relationships that lead to long, happy and healthy lives.
The two evolutionary programs, as she describes them, are controlled by dopamine and oxytocin respectively.
So (1) I’ve never heard this idea before and (2) it seems a little nutsy and confused.
I am not a neuroendocrinologist (neither is Ms Jahme), but I know enough to be able to say that it’s, like, waaaaaaaaay more complicated than that.
Prolactin does damp down sexual responsiveness after orgasm, counteracting the effects of dopamine, but “post-orgasmic increases of prolactin probably do not represent a direct negative feedback mechanism to the CNS but are likely to be one signal among a complex interaction of neuropeptides, monoamines and neurotransmitters controlling sexual behavior.” In other words, it’s not even REMOTELY that simple.
Furthermore, oxytocin can both increase and decrease dopamine, AND dopamine in turn can increase oxytocin (I don’t know if it can also decrease OT). So it’s not like they’re separate systems – indeed dopamine and oxytocin are reciprocally interacting bedmates in the hormonal dance of attachment and a vast range of socio-affiliative behaviors.
Furthermore, oxytocin isn’t just the happy feel-good hormone of the securely attached; under stress, oxytocin levels increase in order (it’s hypothesized) to motivate a person to seek the safety of a social group. Google “tend and befriend.” In other words, OT is not really predictive of a “long, healthy, happy life.”
This kind of nonsense gives evolutionary psychology a bad name among real scientists.
The broader short-term/long-term system dichotomy is something Ms Jahme and her cohort of evolutionary psychologists (David Buss, et al) have been proposing for years, and as a theory it doesn’t work. It’s a cheap attempt to explain the defeatingly varied array of sexual, social, and romantic behaviors people exhibit. It ignores the true complexity of human behavioral neuroendocrinology.
Maybe I’ll do a post on why the dichotomy doesn’t work. It’s complicated and technical and probably fairly boring.
I would add here something I’ve mentioned before, which is that the best (by which I mean my favorite) relationship research is John Gottman’s observational studies of couples’ communication patterns and stress responses. He can predict with staggering accuracy whether a couple will divorce within the next 5 years by watching them fight. He’s not looking at their brain chemicals, he doesn’t need to know their plasma levels of prolactin or oxytocin. He just needs to see how critical they are, if they roll their eyes, if they listen to each other, if they allow time for stress levels to go down.
And furthermore these are teachable skills, not biological programs. (Oh god, I dread the day when someone says that learning these relationship skills “transitions a person from the short-term evolutionary strategy to the long-term evolutionary strategy.” Utter. Bullshit.)







Gottman’s done some really interesting work. But his methodology on the astronomically accurate marriage success prediction rate is very very flawed.
http://www.stat.columbia.edu/~cook/movabletype/archives/2010/06/a_wikipedia_whi.html
Meh. I should probably give more of a shit than I do about the details of the statistical analyses, but I just can’t see my way to it, because the work is so incredibly good and useful and consistent. A model that accounts for 80-90% of ANY human behavior is a really fucking good model, and it’s the factors in that model that are consistent.
http://www.blog.sethroberts.net/2010/06/06/can-john-gottman-predict-divorce-with-great-accuracy/
It’s always possible to fit a model to data after the fact, but it’s irrelevant if it doesn’t accurately predict the future. There’s no evidence that Gottman’s models do.
Except that teaching couples to behave like the successful couples makes them more successful. Why is that not evidence?
The problem is that his model accounts for only the human behaviour *in* his sample. He has not presented any evidence to show that the model does well in predicting *out* of sample. You can always write down a model that accounts for 100% of the behaviour in your sample (its called over-fitting in statistics), but that says nothing about how well it will predict the behaviour of people in general. If a theory is to have any bite, it must be able predict and not just explain ex post.
In my experience, 80-90% prediction accuracy belongs more to the realm of the natural sciences. In the social sciences, the margin of uncertain is much much larger, which makes Gottman’s results very suspicious.
Well then you should do some science then. G’ahead. I’m all for replication of studies and advancing our understanding of how people work. The statistical stuff – again, I should care more than I do, but I just don’t – doesn’t impact the remarkableness of the observational method or the power of results in helping people make their relationships better. The predictive power of his model is less interesting to me than the outcome studies of relationship education using Gottman’s work.
So the model predicts in his sample, AND teaching couples to behave like the couples that stay married makes couples stay married longer. That’s compelling to me.
Pingback: a terrible system | Emily Nagoski :: sex nerd ::