Smell the glove

GRIP firmly, maintain eye contact. What you do during a handshake is clear. But after? Research published this week suggests that humans, like other animals, use smell when they greet each other.

Noam Sobel of the Weizmann Institute of Science, in Israel, took note of an observation made by epidemiologists in 2008 about just how often people touch their faces. Dr Sobel had a hunch that it might have something to do with the gathering of scents—and could thus bear on the question of whether human beings secrete odiferous signalling molecules, sometimes known as pheromones.

To test this idea his team invited 280 volunteers into their laboratory for an undisclosed experiment. The volunteers were first asked to wait alone for a few minutes. Then an experimenter (taken from a pool of 20 men and women) greeted them with or without a handshake, explained that the trial would start soon, and left. Little did the volunteers know that the experiment, details of which appear this week in eLife, had long since begun. They were being filmed the whole time.

Dr Sobel and his colleagues measured how often volunteers placed their hands near their faces before and after the greeting, and whether this differed between those who had shaken hands and those who had not. They found that people who had been greeted with a handshake touched their faces more often than those who had not been, and also that such face-touching tended to be closer to their noses.

To find out whether face-touching was accompanied by sniffing, as Dr Sobel’s hypothesis would predict, the team fitted another group of volunteers with gizmos that measured nasal airflow, and repeated the experiment. They discovered that the rate of flow more than doubled when volunteers touched their faces: a clear sign that they were smelling something.

Not all of Dr Sobel’s observations make immediate sense. If the purpose of sniffing is to learn something about the person with whom you have just shaken hands then it would be expected that the hand sniffed was the one which did the shaking. That, though, was true only when the shakee was of the same sex as the shaker. Those who shook hands with someone of the opposite sex more often smelled the hand that had not done the shaking.

Dr Sobel does not, at the moment, have any idea why that might be the case. Nor does he know which chemicals, carrying what information, are being transferred. He has though done a few preliminary experiments to show that molecules such as squalene, hexadecanoic acid and geranylacetone—known to be chemical signals in some other species—can indeed be passed on by a simple handshake.

Such back-to-basics experiments are precisely what Tristram Wyatt, a pheromone researcher at Oxford University, argues for in another paper, just published in the Proceedings of the Royal Society. The question of what might constitute a human pheromone is complicated by the fact that the concept was originally developed for insects. Many species in this group of animals really do have what are, in essence, external hormones. These molecules induce a precise and predictable effect—often sexual attraction, but also, in social species, things like raising the alarm and directing nest-mates to food.

Among mammals the pheromone concept is hazier. Mammals certainly use scent to communicate. A dog cocking its leg is leaving a message about its identity, and which bits of the landscape it thinks belong to it. But what a second dog does with that information will depend on that dog’s sex, social status, friendship and familiarity with the marker, rather than being the automatic reaction an insect would have. Moreover, a decades-long hunt for pheromones in humans has turned up nothing. Dr Wyatt says that a few early candidates have garnered too much attention, and his paper calls for researchers to devise more rigorous experiments that treat humans as if they were a newly discovered mammal.

As such, he is intrigued by Dr Sobel’s paper, which more or less does just that. Dr Sobel’s results suggest handshakes may be a socially acceptable means of acquiring another’s scent, much as a dog might sniff a backside in order to match the individual to its scent marks in the environment.

Nor need handshakes be unique in this function. There are many greetings in which it is easy to imagine how scent might play a role—from one common in India, in which a young person touches the feet of an elder, who then rubs the younger’s head, to one employed by some Australian aboriginals, among whom greeters place their hands in their armpits and thence onto the chests of the greeted. Dr Sobel’s findings will no doubt spark a range of further study, from the chemical to the ethnographic. A rich line of inquiry is at hand.