May 18, 2007

Getting the science of communication backwards

A key technique for "finding facts in a world of disinformation", according to unSpun, the recent book by Brooks Jackson and Kathleen Hall Jamieson, is "RULE #4: Check Primary Sources" (p. 160). And as they point out, the internet makes it easy to do this, especially if you have a university affiliation that gives you access to journal subscriptions online.

I warn you, though, experience with the process may make it hard for you to follow their excellent "FINAL RULE: Be Skeptical, but Not Cynical" (p. 175).

Here's a case in point. A few days ago, I took a look at the media's (uncharacteristically minimal) reaction to some research by Camelia Suleiman and Daniel O'Connell on alleged gender differences in interviews with Bill and Hillary Clinton ("Women and men again, you know?", 5/13/2007). And one of the most interesting things that turned up in a Google News search for {Suleiman Clinton} was Marilia Duffles & Jeffrey Lord, "Why Hillary Talks Like Bill", The American Spectator, 5/7/2007.

Duffles and Lord present a perspective that has its roots among progressive-minded feminists, but has recently become popular on the American right. And in support of this viewpoint, they cite some recent research on the neuroendocrinology of communication that turns out to -- well, you'll see.

Here's their theme:

Women are biologically predisposed to behave in a certain way with the aid of hormones like estrogen, which has dictated their softer physical characteristics throughout our prehistoric lineage. Hope as anyone may for uniform conduct, the incontestable fact is that our stone-age biology still holds sway against the cultural tides of modern society, manifesting as behavior we still anticipate: women as care-givers, provider of bosom, as the more empathetic, the more socially connected of the two genders. And this also happens to have biochemical proof.

When women chit-chat, their oxytocin level -- a feel-good hormone that elicits feelings of trust, bonding and love -- rises, according to a recent study by Dr. Shelley Taylor, psychology professor at UCLA. This means they experience pleasure and feel connected with others. And it lines up with Harvard psychologist Carol Gilligan's work on the difference in moral thinking between the sexes that concludes men generally rely more on a universal set of rules that determine obligations and rights (justice-based reasoning), whereas women are oriented towards care-based reasoning that focuses attention on the needs of others. [emphasis added]

For more on this bio-political trend, see "David Brooks, Neuroendocrinologist" (9/17/2006). Our focus here is not the broader ideological movement, it's the relationship between specific assertions and the scientific sources that they cite. So we're zeroing in on the sentence in bold, about how oxytocin levels rise when women chit-chat, reflecting the experience of pleasure and a feeling of being connnected with others; and we're going to check the "recent study by Dr. Shelley Taylor". That would be Shelley E. Taylor "Tend and Befriend: Biobehavioral Bases of Affiliation Under Stress", Current Directions in Psychological Science 15 (6), 273–277 (2006):

... we examined the relation of plasma oxytocin levels to reports of relationship distress in adult women (Taylor et al., 2006). We found that women who were experiencing gaps in their social relationships had elevated levels of oxytocin. Specifically, women with high levels of oxytocin were more likely to report reduced contact with their mothers, their best friends, their pets, and social groups to which they belonged. In addition, those with significant others were more likely to report that their partners were not supportive, did not understand the way they felt about things, and did not care for them. Poor quality of the marital relationship and infrequent display of affection by the partner were also associated with higher levels of plasma oxytocin. Thus, oxytocin appears to signal relationship distress, at least in women. [emphasis added]


Additional details, tables and graphs can be found in Taylor, S.E., Gonzaga, G., Klein, L.C., Hu, P., Greendale, G.A., & Seeman S. E., "Relation of oxytocin to psychological and biological stress responses in older women". Psychosomatic Medicine, 68, 238–245 (2006).

Now as R. Bowen explains in Pathophysiology of the Endocrine System, the physiological functions of oxytocin are many and complex:

In years past, oxytocin had the reputation of being an "uncomplicated" hormone, with only a few well-defined activities related to birth and lactation. As has been the case with so many hormones, further research has demonstrated many subtle but profound influences of this little peptide. For example, administration of oxytocin to species ranging from mice to humans has revealed a number of effects on social behavior.

Well-documented effects of oxytocin in humans include promotion of cervical dilation and uterine contraction during childbirth, and the "letdown reflex" in lactating mothers. Injecting oxytocin into the cerebro-spinal fluid causes erections in male rats, and vaginocervical stimulation releases oxytocin within the spinal cord in female rats. Oxytocin has been implicated in pair-bonding in monogamous prairie voles, maternal behavior in ewes, protective inhibition of fetal brain activity during childbirth, and so on.

As far as I've been able to determine, though, no one has ever measured oxytocin levels during female chit-chat. So where did Duffles and Lord come up with their howler about how oxytocin rises when women interact with one another, and "means they experience pleasure and feel connected with others"?

One possibility is that they got it from a 2003 popular book by Shelley E. Taylor, The Tending Instinct, which says things like this:

...oxytocin ... appears to be both a cause and a consequence of social support. In the face of at least some threats, oxytocin is rapidly released. From what we know about its effects, oxytocin may be one of the biological factors that propels people and animals into one another's company in stressful situations: as an "affiliative" hormone, oxytocin leads people to seek contact with others. It is released in the company of others, so it may also be implicated in the emotional experiences of bonding that are seen in the aftermath of tragedy.

I haven't been able to find any human studies to support the claim that oxytocin "is released in the company of others" -- as far as I know, this generalization comes from what happens when you put a bunch of rats together in a cage, in circumstances that are at least as likely to cause the rodent version of "relationship distress" as "trust, bonding and love". In any case, Duffles and Lord may have been led astray by Taylor's theory of oxytocin as an "affiliative hormone", which mediates what she calls the "tend and befriend" response to stress, as opposed to the better-known "fight or flight" response. This theory has been deservedly influential, though it's by no means uncontroversial.

However, you'd have to read Taylor very carelessly in order to characterize oxytocin the way that Duffles and Lord do. I suspect that they were influenced by another source -- Louann Brizendine's book The Female Brain. From p. 36-37:

And why do girls go to the bathroom to talk? Why do they spend so much time on the phone with the door closed? [...]

There is a biological reason for this behavior. Connecting through talking activates the pleasure centers in a girl's brain. Sharing secrets that have romantic and sexual implications activates those centers even more. We're not talking about a small amount of pleasure. This is huge. It's a major dopamine and oxytocin rush, which the beiggest, fattest neurological reward you can get outside of an orgasm. Dopamine is a neurochemical that stimultes the motivation and pleasure circuits in the brain. Estrogen at puberty increases dopamine and oxytocin production in girls. Oxytocin is a neurohormone that triggers and is triggered by intimacy. When estrogen is on the rise, a teen girl's brain is pushed to make even more oxytocin -- and to get even more reinforcement for social bonding. At midcycle, during peak estrogen production, the girl's dopamine and oxytocin level is likely at its highest, too. Not only her verbal output is at its maximum but her urge for intimacy is also peaking. Intimacy releases more oxytocin, which reinforces the desire to connect, and connecting then brings a sense of pleasure and well-being.

Now again, as far as I can determine, no one has ever measured oxytocin in teen girls on the telephone, or reported any other direct evidence of the effects of conversational interaction on oxytocin levels in men or women of any age. (If I'm wrong, please let me know.) So there's no direct evidence to show that this picture of female neuroendocrinology is false. On the other hand, there's no evidence that it's true, either -- it's a sort of neuroendocrinological science fantasy. And the evidence from Taylor 2006, that oxytocin levels in adult women rise with "relationship distress" and general social disconnection, certainly seems to point in entirely the opposite direction.

I don't want to disagree with Jackson and Jamieson's FINAL RULE: it's true, you should be skeptical, not cynical. But I'd suggest a codicil: if you read something about the science of communication in the popular press or in a popular book, these days, it's probably incomplete and misleading, and it may well be 100% backwards.

It would be easy to sneer at the American Spectator, which is not exactly famous for concern about whether stuff is actually, you know, like true or anything. But the fact is, the rest of the media, from the New York Times to ABC's 20/20 to the BBC, has not been any better in covering the neuroscience of sex differences in general, and sex differences in communication in particular.

There are several effects that combine here to create a perfect storm of misinformation. One is the modern tendency to treat science as a source of morally instructive tales, not to be taken literally. Another is the generally abysmal level of scientific understanding among journalists and editors. And then there's the special ignorance of language and communication among our society's intellectuals in general, who may have forgotten the physics and biology that they took in college, but who never learned any linguistics to start with.

As I said, it can be hard not to let skepticism turn into cynicism. My own method, as I've explained elsewhere, is to join H.L. Mencken in viewing the mass media as a "daily panorama ... of private and communal folly inordinately gross and preposterous, so perfectly brought up to the highest conceivable amperage, so steadily enriched with an almost fabulous daring and originality, that only the man who was born with a petrified diaphragm can fail to laugh himself to sleep every night, and to awake every morning with all the eager, unflagging expectation of a Sunday-school superintendent touring the Paris peep-shows".

Posted by Mark Liberman at May 18, 2007 06:50 AM