Aha! I've been hearing about "that German study about facial attractiveness," but I hadn't gone seeking it out because I gathered it was the old study that claimed (at least as people usually seem to present it) that the main thing that determines attractiveness (worldwide, across all cultures) is facial symmetry, and I just don't buy that.
Very asymmetric faces are judged rather unattractive, but very unattractive faces are not necessarily asymmetric. And vice versa: very symmetrical faces [are] not necessarily ... attractive and very attractive faces [are] often [not perfectly symmetrical]. Based on our results, symmetry only seems to be a rather weak indicator for attractiveness. Often it is even difficult to distinguish between the original and the perfectly symmetrical version, because irregularities in shape are rather insignificant. Therefore, the strong influence of symmetry that has been reported in the scientific literature over and over again is questionable.
So of course because it agrees with my intuitive feeling on the subject, I'm much more inclined to take the study seriously than I otherwise would've. Reader agrees with study that supports reader's preconceived biases; film at 11.
The other extremely widespread "everyone knows" idea about attractiveness is that perfectly average faces are (allegedly) universally considered the most attractive. This study suggests that it's true that averaged faces (faces created by averaging real faces) are more attractive than real faces, but also indicates that starting out with more attractive faces results in more attractive average faces. So the most attractive faces in their study were the ones that were the average of several real faces that were themselves judged to be attractive. In fact, the study suggests that skin color and smoothness have a major effect on perceived attractiveness, and perhaps that the smooth-looking skin is the main reason the averaged faces are seen as more attractive.
So on the one hand, the study does support some of my gut feelings. But I still see plenty to argue with.
Primarily, I think, that it doesn't pay enough attention to individual variation. All of these "what is beauty?" studies I've seen fail to seriously address the fact that different people just plain have different tastes. In the "babyface" part of this latest study, for example, the overall numbers support their conclusion, but there's a great deal of variation; nearly 10% of participants found the adult-looking faces more attractive, a number which the study discounts as "[o]nly very few" but which seems fairly significant to me, especially given that there were only six choices (so by pure chance, if I'm understanding the setup correctly, you'd get only 17% picking each one). So even in their study, there was a lot of variation.
I suspect they'd see even more variation if they made more of an attempt at a cross-cultural study—both including more different shapes and structures and colors of faces, and including subjects (observers of photos) from more cultures. They don't even (in the parts I read) talk about age differences among the test subjects; I suspect that if you asked a group of 80-year-old American women and a group of 16-year-old American girls, their favorites wouldn't entirely match. And that's not even considering sexual orientation, which this study appears to ignore entirely.
Also, most of the people I talk with about attractiveness (especially most women I talk with) say that what they find attractive in a person has a lot more to do with other factors than with facial features. Faces have a lot to do with it for me, but various friends of mine weight more heavily such items as voice, laughter, fluidity of motion, body shape, general fitness, friendliness, focus of attention, perceived strength (physical or emotional), etc.
I'm also a little dubious about the study's results given that the results seem to support the philosophical/political ax that the experimenters want to grind. I pretty much agree with what they say, but they make such a strong statement about it that it makes me a little dubious about the objectivity of their research. Their results do tend to corroborate some of the things characters in Ted Chiang's "Liking What You See: A Documentary" say (you knew I'd bring that story up sooner or later, didn't you?):
[T]he results of our studies on social perception suggest that there is a well-defined stereotype of attractiveness: People with more attractive faces were assessed to be more successful, conten[t]ed, pleasant, intelligent, sociable, exciting, creative and diligent than people with less attractive faces. These results particularly show the far-reaching social consequences human facial attractiveness may have.
But they go on to rail against this insidious evil in rather forceful terms. And anyway, short of induced calliagnosia, I'm not sure what can be done about it in the real world.
My personal responses to these faces:
I do find the "sexy" prototype faces more attractive than the "unsexy" ones, but I also find the "unsexy" ones kind of cute (several of them have nice eyes and a sort of solid down-to-earth look, to my eye)—and the "sexy" ones are still a long way from being as attractive to me as various real-life people I've seen in movies (and even a few I've met in person). So I don't buy the notion that the most attractive faces possible are the ones derived by blending real faces; the study indicates that "Natural faces cannot keep up with their artificial 'competitors,'" but to my eye, that's just not true. The study says "Just 3% (!) of the natural faces were rated as 'rather attractive'—the judgements 'quite attractive' and 'very attractive' were never applied to natural faces." But that doesn't say that those judgments couldn't be applied to natural faces; just that the natural faces chosen for the study weren't the most attractive ones possible. (Of course, it may well be that if you averaged, say, River Phoenix and Leonardo DiCaprio's faces, you'd get an average that I'd find even more attractive than the originals, but I still don't think the study's results are as dire as they claim.) (Also, as the study notes, the images we see of famous people are often enhanced in various ways. But I've met real-life people who didn't wear makeup who I found more attractive than these supposedly uber-attractive compound faces from the study.)
Interesting to me: the compound "attractive" woman reminds me of Deanna Troi, but when I actually compare the image with photos of Troi, they're pretty different. Maybe it's the eyes? Not sure.
Speaking of sf, this whole study reminds me of an old sf story, in which a man uses a computer (?) to generate an image of his ultimate ideal female face, and then he meets a woman who has that face. I'm completely blanking on who it's by or what the story's about—it might've been a bar story (possibly even "More than Skin Deep," from Tales from Gavagan's Bar by L. Sprague de Camp and Fletcher Pratt), but then again it might not've. Anyone know offhand?