Why are we drawn to some faces more than others?

Partly because of how faces look, and partly because of what we have seen. Two features keep turning up in the research: faces close to the average, and faces that are symmetrical, both tend to be rated as more attractive. And an averaged face is not dull. It is, in fact, quite good looking.

The discovery was an accident. Francis Galton blended photographs of vegetarians and of criminals, hoping each group would show a typical look. It did not. But he unexpectedly noticed that the blended face was more attractive than any of the individual faces. In 1990, psychologists Langlois and Roggman repeated the idea with computers: 300 judges rated single faces and composites of up to 32 faces, and the 32-face composite won.

Is it built in? Newborns are a clue. At 15 minutes old they show no preference, but 72 hours later they already stare longer at faces adults find attractive. The preference shows up within days. Why is debated: one idea is that an average face is attractive simply because it is prototypical, another that average features hint at fewer harmful mutations. The Hadza of Tanzania add a twist: they liked averaged Hadza faces but had no preference for average European ones.

Beauty is not only average, though. A face averaged from attractive faces appealed more than one averaged from all women aged 20 to 30. Standards of beauty have changed with culture, and a beautiful face carries an unearned bonus: in a 1972 study, most people expected attractive people to have better personalities, happier marriages and more career success. That is the halo effect.