Since the 1930s, the idea that creatures might advertise their " fitness " as mates through symmetry has been around. It rests on the notion that factors ranging from bad genes to coming off worse in too many fights will lead to asymmetries others can see. And over the past few years, experiments with animals and humans seemed to support the idea that symmetry in features such as plumage and facial characteristics influences mate selection.

But no one had actually checked that the animals in these experiments can actually see the subtle differences in symmetry, typically of between 1 and 2 percent. Now the first such experiment has been carried out. The results are worrying for advocates of symmetry detection as a powerful factor in mate selection.

John Swaddle, an ethnologist at the University of Bristol, performed the experiment using wild starlings, which have excellent eyesight. He trained them to hit keys marked with bar patterns with varying levels of asymmetry in order to receive food rewards. This showed the starlings could easily detect asymmetries of between 5 and 10 percent. But at between 1 and 2 percent, their performance plummeted to no better than random guessing.

" This suggests that the levels of asymmetry that birds encounter in nature will often be just too small to be detected, " says Swaddle. He says experiments that linked small asymmetries to fitness may have produced misleading correlations, and says the only direct evidence that birds use asymmetry as a cue to mate fitness involves levels of asymmetry of at least 10 percent. " I think signaling by asymmetry will probably only occur when species show such very large asymmetries—and this doesn’t occur that often. "

This raises questions about research suggesting that humans are influenced by visual asymmetry in their choice of partner. Michael Butt of the Perception Laboratory at the University of St. Andrews, who has carried out such experiments, concedes that no one has ever checked if humans have a threshold to asymmetry detection. He thinks a conclusive test of this would probably involve manipulating complex three-dimensional images: " It would be a very difficult experiment to do "

Burt adds that the abilities of animals to detect asymmetry in bar patterns may not reflect their talent for spotting asymmetries in body shape. " The visibility of an asymmetry may well depend on its type, " he says. Swaddle agrees, and is planning further experiments. " But I suspect that asymmetry is used as a visual cue less often than most people appear to presume, " he warns.

According to the text, in mate selection human beings（）

A. need symmetry detection

B. need complex three-dimensional images

C. begin to realize their dependence on symmetry

D. cannot be proved whether to rely on asymmetry