When it came to solving the riddle of the peacock's tail, Charles Darwin's powers of evolutionary deduction were second to none — the more extravagant their feathered displays, he reasoned, the greater their chances of attracting a peahen. But when he tried to account for the human propensity to weep, Darwin found himself at a loss. "We must look at weeping as an incidental result, as purposeless as the secretion of tears from a blow outside the eye," he wrote in 1872.

In this Darwin was almost certainly wrong. In recent decades, scientists have offered several accounts of how the capacity for tears may have given early hominids an adaptive advantage. These range from the aquatic ape theory, according to which tears were an adaptation to saltwater living, to the notion that by blurring our vision tears may serve as a "white flag" to potential aggressors — a signal that the crier is incapable of harm. Then there are the straightforward biological theories, such as the claim that tears evolved to keep the eye moist and free of harmful bacteria.

But perhaps the theory enjoying the widest currency at the moment is the notion that tears are a form of social signaling that evolved from mammalian distress calls — a clear visual signal in other words that someone is in pain or danger and needs help.