It’s well known that people trust, and are more cooperative with, those who they rate as looking similar to themselves. This makes evolutionary sense in terms of kin selection, the idea that we’re wired to cooperate with genetic relatives who share our genes and, to an extent, our looks.Now a team of psychologists have found that the reverse is true too: people who are more trustworthy are rated as looking more similar to ourselves. Harry Farmer, Ryan McKay and Manos Tsakiris of Royal Holloway, University of London, had 59 people play a trust game, in which they were initially given £2.50, some portion of which they could transfer to another player (the trustee). The transferred amount was then tripled by the experimenters, after which the trustee could split the rewards with the original player (the trustor). Each trustor was told before they decided how much to transfer that the trustee they were playing with had already decided how much they would return for every amount they could potentially receive — so a trustor might expect that if they transferred a large share of the original £2.50, the trustee would generously split the total amount once multiplied.In reality, there was no trustee to receive the transfer. Instead Farmer and colleagues had made pre-recorded videos in which the trustee either gave back 70% of the tripled transfer, or a measly 10%. So some trustors transferred a big share of their money, expecting a decent return, only to be betrayed by the trustee.Before and after playing this trust game, trustors were shown a series of photos in which their own face was morphed to varying degrees with an image of the trustee they were supposedly playing with. Trustors had to say at which point the image showed a 50/50 blend of their own and the other player’s face, known as the point of subjective equality (PSE). The idea behind using PSE is that if it is judged to have been reached when the morphed image contains more than 50% of the other person’s face, then that person must be regarded as looking like us — and vice versa.Farmer and colleagues found that when people played the trust game and were betrayed by an untrustworthy player, their estimate of PSE was roughly the same before and after the game. By contrast, people playing against a trustworthy player (who returned what they said they would) increased the percentage of the other person’s face in the composite required to reach PSE. The results are described in a paper published online in the journal Psychological Science.The authors argue that this means that players perceived their fictitious partners as looking more like themselves once trust had been established. Harvey Whitehouse, an anthropologist at the University of Oxford who was not directly involved with this research*, says that this effect may be an example of phenotype matching – using the behaviours or other traits of individuals as a cue to determine whether they are kin or non-kin (a process that has been observed in non-human species).In the case of humans, it’s possible that when people cooperate with us, treat us kindly or show us altruism, we tend to see them as more like family, and this influences the way we perceive their physical features. What’s more, the factors beyond kin relations that promote cooperation and cohesion within groups, such as ritual practices, also create a sense of living within an extended family with whom we have close, trustworthy ties. On this view, trust itself becomes a cue as to who should be considered as part of this extended, non-kin family. And so cultural practices that engender a sense of family, including rituals but other shared markers of social identity, could have played a crucial role in the evolution of altruism and cooperation directed to people beyond our immediate kin group.At the very least, the authors argue that their results show that the way we perceive faces is not simply based on objective physical characteristics, like relative position of eyes, mouth and so on, but also by social factors. “Our study highlights the strong links between physical and social representations of closeness between one’s self and others,” says Farmer.*Although Whitehouse did not carry out these studies, he does work with Farmer, McKay and Manos on the Ritual, Community and Conflict project, of which this research was a part.Dan Jones is a freelance writer based in Brighton, UK. You can follow him on Twitter @MultipleDraftz.