As a culture, we at least pay lip service to the idea that irrational biases are harmful and should be eliminated. Here are two studies which looked into such biases in classical music auditions: Effects of Physical Attractiveness on Evaluation of Vocal Performance and Effects of Performer Attractiveness, Stage Behavior, and Dress on Violin Performance Evaluation (sorry, abstracts only). They found something they weren't really looking for - that songbirds are pretty.

Both studies were designed similarly: auditions were videotaped and then judged by three groups of people. One group rated the video only, one audio only, and a third group the complete videotape. All the judges were music students or faculty. The studies did find that dressing well and knowing how to behave on stage will get you a better rating than you would receive based on the audio only, and that attractive male singers also benefit from their appearance. However, attractive female singers and attractive violinists did not receive any extra benefit from their good looks. They did receive better ratings on videotape than their uglier competition, but those were not unfair - they also got better ratings on audio only. So, attractive people really do sound better - except, apparently, male singers.

So why would a better-looking violinist sound better? There are many possible reasons, though these studies do not tell us which of them are true or important. One possible explanation is cross-assortative mating - talented male musicians impregnated attractive women so their children inherited both the looks and the talent. Cross-assortative mating is not that difficult to test for - all you would need is to check whether this correlation also holds true within families - if it doesn't it's a telltale sign of assortative mating. Because the "good looks" and "good music" genes get sorted randomly among the children in a family, the better-sounding daughter would not be any more likely to be good-looking than her less musically talented sister.

It could also be cross-assortative mating not involving musical talent itself but something correlated with it such as intelligence. We do know from another study that more symmetrical people are better dancers, and by searching for "Dance" in this huge PDF you can see that a follow-up study suggests that this explained by the fact that they are more intelligent.

A more obvious possibility is that good-looking people are more confident and more confident people may practice more, practice with more enthusiasm, or receive more encouragement to pursue music. Another is that the better-looking musicians were from higher social classes and therefore took better care of their physical appearance (for example were less likely to let themselves get fat), and also started playing music at a much earlier age. After all, few working-class families send their five-year-olds to violin classes.

It would not be difficult to design a study to test which of these possible explanations (and perhaps others) are true. Of course getting funding for such follow-up research would be extremely difficult; not a lot of people in academia are interested in investigating the superiority of the good-looking. The very notion that being attractive is linked to being good at something else violates our culture's deeply held ideas about equality. This is an attack not only on the belief that ugly people have more "inner beauty" but on the entire myth that anyone can succeed if they're just given "fair opportunities". This research isn't nearly as dangerous to the progressive order as genomics and psychometrics but it points in a similar direction - that some of us got lucky and really are more capable than others. It all adds up, though, and at some point defending the fiction may become untenable.

On a personal note, my initial reaction was that this is strange but when I thought about it some I realized that it does seem true. I'm not a good musician (or good-looking) myself, but I've played with a few hundred different people so far in my life - the bulk of those being in their teens and early 20s. The few women I still remember as being exceptionally gorgeous (even if I played with them only once a decade ago) were all definitely above average or better in terms of ability. It does seem to be true. I'm hoping to get in on some of that cross-assortative mating myself.