In a rather grammatically- and stylistically-challenged article, the Atlantic talks about the latest study:

Viren Swami and Martin Tovée at the Universities of Manchester and Newcastle, respectively, look into the intricate world of why physical ideals are ideals, and in turn why they drive people beyond reason and morality in the current Archives of Sexual Behavior.

Stylistic note: this lead make it sounds like the study is unique. But I’m guessing that the Archives of Sexual Behavior have published dozens if not hundreds of articles on why physical ideals are ideals. Indeed, the abstract says as much. So why are we talking about this one in particular? Is it the best done so far? I’m going to make the case below that it isn’t even close. What we’re about see is what I call the Scientific Peter Principle: poorly designed studies usually have the most attention-getting results.

(Also, do ideals drive people beyond reason and morality? That’s an awfully loaded statement.)

The problem is primal, so the research methods are not to be outdone. 361 white British men were “taken to a quiet private location” to look at women. Not real women; 3D computer renderings. The men were allowed to rotate them 360 degrees. The only difference among the women was breast size.

The men were then asked to “make their ratings on a paper-and-pencil survey.”

Emphasis mine. “Not to be outdone”? I can think of about a dozen ways I could outdo this study.

Swami and Tovée compared the results with the men’s preferences in breast size, which showed that “men who more strongly endorsed benevolently sexist attitudes toward women, who more strongly objectified women, and who were more hostile toward women idealized a large female breast size.”

The study’s abstract, which is all I have access to, is rather stunning in its lack of humility. After noting that previous studies have been ambiguous, they boldly proclaim their results and then say:

These results were discussed in relation to feminist theories, which postulate that beauty ideals and practices in contemporary societies serve to maintain the domination of one sex over the other.

Even if we were to accept the conclusions of this article — and I don’t — it’s a long way from there to beauty ideals maintaining the domination of one sex over the other. Would you like some science with your ideology? Actually, we don’t even need to go to the abstract to see the boldly stated ideological bias. The title is: “Men’s Oppressive Beliefs Predict Their Breast Size Preferences in Women”.

So, yeah.

You probably know that I’m not going to be sympathetic to this and not just because of my distaste for ideology. In my previous post, I stated my hypothesis that the breast fetish is just like any other fetish — something that the male mind has latched onto as a way of identifying potential mates. It’s commonality is simply because of its obviousness — visible breasts are the easiest way to identify the female of our species. It’s not a social construct, per se. It is a preference that arises within a social construct. If it weren’t breasts, it would be something else (and almost always is). But the key point here is that fetishes are not really chosen. They just happen. It’s just something that, on a very primal level, the human sexual id locks onto.

Still, even without my prior assumptions and biases, we can easily see that this study, which has now been widely cited by various mainstream sites (and not just because they like to talk about breasts), has some big problems.

First, the study was of 361 men. 361 men who were willing to be taken to a “private, quiet location”. 361 whose age, employment and marital status is not exactly clear. That’s an awfully small and demographically narrow number to be drawing conclusions from.

Second, if the 3D drawing in the Atlantic article is an accurate reproduction of what they were shown, this wasn’t a reasonable test at all. I hate to break this to the authors, but the average bust size in the Western World is quite large and increasing: at least a 36C by old standards and probably larger if the lamentations of bra fitters are to be believed. This is partly rising obesity, marginally because of implants and mostly for reasons that aren’t really clear. This has had a significant effect on the landscape in that men’s perception of what constitutes a big bust has changed. Looking at the figures, even the last one didn’t really cross me as “very large”. Were these informed by some statistical survey of women’s breasts sizes? That’s one way you could improve this “not to be outdone” study.

There’s a related issue of body type. Critics of male sexuality often claim that men want big breasts on skinny bodies. Certainly, there is a subset of men who like that but most men who prefer busty women actually prefer curvy women. They like big hips and curvy backsides just as much as they like big breasts. Asking these men to look at 3-D computer models — frankly, none of which look like a real woman — is problematic at best.

(Aside: as I argued in my previous blog, male preferences are not monopolar. All things being equal, a man may prefer a woman with bigger breasts. But in the real world, things are rarely equal. He may be fine with a woman with smaller breasts if she has other features he finds attractive — enchanting eyes, a warm smile, a slender frame, beautiful hair. And — this is a critical point — if a man likes a woman, finds her interesting, enjoys her company — he will begin to see her as attractive. She will become beautiful to him. He will see the beauty in her even if there really isn’t that much to see on an objective level.

I would posit that there are very few men who date or are attracted to women entirely on bust size. Their preference in models and pornography — situations in which there is no interaction — may reflect a preference (although even then there is probably a broad range). But their behavior in real life can be wildly at variance with this. I would bet you that a significant fraction of the men who preferred “very large” breasts are dating or married to skinny women. And I would bet that some of the men who preferred “very small” breasts are dating or married to busty women. And I would further bet that they find the women in their lives attractive despite not conforming to their preferences in zombie-like computer models.)

Third, the questions. I don’t have access to the study, but here are the sample questions they provided:

Attitudes Toward Women Scale (sample prompt: ”Intoxication among women is worse than intoxication among men.”)

Hostility Towards Women Scale (sample prompt: ”I feel that many times women flirt with men just to tease them or hurt them.”)

Benevolent Sexism subscale of the Ambivalent Sexism Inventory (sample prompt: “Women, as compared to men, tend to have a more refined sense of culture and good taste.”)

Whoa, really? Those are your sample prompts? Those three prompts are all judgements. You would probably find lots of women who would agree with at least a couple of those. You would probably find that a man would agree or disagree based on his emotional state (if he’s just had a bad break-up, for example). And prompt three (and many of the questions on Ambivolet Sexism Inventory from which they are taken) aren’t clearly sexist. Many of even the most blatant ones probably probe misanthropy far more than they probe misogyny specifically.**

(Another aside: the Atlantic author illustrates sexism by quoting a lawsuit in which a boss constantly commented on a co-worker’s breasts and once shook her breast as a substitute for shaking her hand. This is not the behavior of a man who likes big breasts or thinks women have a more refined sense of cultural taste. This is the behavior of a sociopath.)

But I think the real flaw is highlighted by Ann Althouse:

They were taking a science-y survey, so deference to authority and desire to be socially acceptable would be an influence along with real-world sexual preference.

…

The scientists found “men who more strongly endorsed benevolently sexist attitudes toward women, who more strongly objectified women, and who were more hostile toward women idealized a large female breast size.” Were these men really the ones who “idealized a large female breast size,” or were they simply the ones who didn’t feel as strongly compelled to moderate their opinions to conform to the perceived demands of polite society?

Exactly. I keep harping on this in the social sciences: there is a huge difference between what people think and do and what they tell a group of leering scientists that they think and do. Most people do not want to be perceived as abnormal (or sexist). This is a big problem with this study since, if I read it correctly, the men were shown all five images at the same time. This creates a very obvious social pressure that is different from if five groups of men were shown five different images separately. Hell, if I were put in a room and asked which image I liked, I might say 3 or 4 even though I would prefer 4 or 5 (and would actually prefer a real women with real physical proportions).

How would I improve this “not to be outdone” survey? First of all, I would have a lot more than 361 white British men. Second, I would show each man only one image and ask him to rate her on a scale of 1-10. Second, I would get images of real women and digitally alter them, using some statistical model based on women’s actual bust sizes. Third, I would make a second axis by having some women altered to have both bigger hips and bigger breasts and others to just have bigger breasts. Breast size and hip size are correlated, as anyone who has seen real women instead of 3-D models knows. Fourth, I would use something a little less ambiguous than these prompts. For example, I might give the men two different job applications and just change the gender and see how they rated the applicant. Or have some people enact a job situation and ask them what they thought of the woman’s behavior. Something a little more direct, at any rate.

Or I might go to the gigantic database compiled by the authors of A Billion Wicked Thoughts who gathered data from Google when men didn’t know they were being studied. The only problem is that I would probably find — as those researchers did — that men actually prefer curvy women, not just just busty ones. And that would ruin my thesis that a preference for big boobs is a results of sexism.

So, let’s sum up: a small and poorly designed study asked men to look at unrealistic images of women. They were then asked leading questions of dubious utility. And from this, we conclude that men who like big boobs are more likely to be hostile to women and that feminist theory is vindicated.

That makes me feel some hostility all right. But it’s not directed against women.

**Update: Michael Talarski alerted me that there are links to the questions in the Atlantic article. Here is the attitude toward women quiz. The other triggers a download. The questions are mostly reasonable probes of attitudes toward women (although a few are bit ambiguous). But I would be curious to see how women score on that test. And I would be especially curious to see if these attitudes correlate with actual behavior.