Once more, without the straw man

Personhead pantryslut gave me the link to the article that was raggedly and incorrectly referenced by various people I have talked to in the last year or so.  It turns out -- after I had written for hours about stranger anxiety and (apparent) race and how babies don't, in my experience, tend to express stranger anxiety about skin color in environments where more than one skin color is present normally (did I ever actually manage to say exactly that?  I was tanked on vicodin all day yesterday, so nothing is guaranteed -- I'm afraid to go back and read the thing) -- it turns out, after all, that the study concluded two things about babies' preference (not babies' anxieties):

Newborn babies don't exhibit preference based on "own-race:"
Three-month babies do.
They think that this is because they have been building up an idea of what people ought to look like and that they use cues that also differentiate between races.  They don't think they're talking about skin color only, but also about facial features in general.

How did they determine this (very tiny and possibly shaky) conclusion?

First they took a pile of white newborns and exposed them to a black and white photos of unsmiling men.  They were quite rigorous about how they did this.  They recorded the babies' eye movements and found no difference in how long they looked at the pictures.

Then they took a new pile of white three-month babies who had only been exposed to white people and exposed them to the same set of pictures in the same conditions.  They recorded the babies' eye movements and discovered that the babies gazed longer at the white guy.

Go look at the article and check if I have represented it correctly.

Look at the pictures they used (I gather there were more of them but they were all about like that). Personally, I think you can find more facial features variation than that within a "race."  I hate the word, by the way, and it's not because of squeamishness: it's because of something it implies about genetics and evolution that simply isn't true in the human species.  Our genetic variation simply isn't great enough to warrant the label, and even more to the point, the variation within groups that we call races is greater than the variation across races. As they exist in our minds, races only exist in our minds.

Of course, there are things we can point to and say -- those eyes, that nose, that mouth, those eyebrows, that skin color, that hair texture -- but if you really, really pay close attention, you will find that (except --historically --for the very palest and the darkest skin colors) all those features show up all over the human family. But they only mean something because we say they do.

The article references a mixture of other baby-preference studies, some of which I have read in the past and have found dubious either in their conception, methodology, or implication.  I'm wary of baby-preference studies anyway.  Sixty-four babies  gazing at photographs in a laboratory just doesn't impress me as much as babies on buses making eyes at real people. There's apparently at least one study sayign that babies raised by a female caregiver prefer female faces and those raised by a male caregiver prefer male faces.  But in real life, with real babies, when we see young babies raised in an all-female household, some of them gravititate towards the men in our program (staff, fathers, uncles, big brothers, etc).  There's a difference between a photograph and a real person with an animated, smiling face, saying gooey things and wiggling fingers.

One gripe I have is the term "own-race" which implies, though the authors would probably admit it isn't warranted by the design or results of the study, that the babies are actually forming a theory of race.  They apparently haven't done a study with babies raised in a homogeneous environment of people whose apparent race is not the same as theirs, nor a study with babies raised in a diverse environment.  Or if hey have, I haven't found it. 

My prediction would be that, if the study is reproducible at all, they would find that the first set of babies would prefer faces like their caregivers' and the second set of babies would have idiosyncratic tastes.  My prediction too is that these photograph tests would be utterly unreproducible with real human beings instead of photographs.