"Ein Freund, ein guter Freund, ist das Schönste was es gibt auf der Welt!" —Die Drei von der Tankstelle
In the 1930 UFA movie Die Drei von der Tankstelle (The Three from the Gas Station), Willy Fritsch, Heinz Rühmann, and Oskar Karlweis fall in love with the same girl, Lillian Harvey, who ultimately chooses the debonair macho Willy. It does not go well romantically, but the friendship among the three males survives. Much of the music featured in the film has become iconic (auditonic?), but the theme song (see epigraph) block-busts everything else. Even today, middle-aged Germans will break into song if you, after a glass of Riesling, intone Ein Freund. The mood will be sentimental, and the notion of same-gender friendship will mentally manifest in its fullest glory. Granted, Ein Freund struck homoerotic undertones, but a Freudian would not be surprised: Isn’t all friendship among men a thinly sublimated form of same-sex lust?
Perhaps. I am not here to refute this argument. But there is an interesting biological/game-theoretical alternative: Forming and maintaining friendships is at bottom about forming and maintaining alliances. Allies have each others’ back. They have made a commitment – whether explicit or not – to side with one another. NATO is an example. All partner nations say that an attack on one of them is an attack on all. Other non-NATO nations know this, and this is an important deterrent. Alliances are common among mammalian species. Even one of Freud’s theories was about the alliances the hungry and undersexed males in a group form to challenge the alpha.
Game theory offers formal models of alliance building, which predict how rational beings use alliances to further their own interests. Such models can look cold and unemotional (they are), and thus far removed from the deep experiences of loyalty, obligation, and quasi-kinship we can experience with close friends. But that is the point. Biologists and game theorists might say that these feelings are a trick of nature, as it were, to get us to make and honor the kinds of alliance commitments that get us through life.
Peter DeScioli, Rob Kurzban, and their colleagues (allies?) have built and tested a theory that casts friends as allies. With that, they can make precise and testable predictions about how we will respond to certain challenges arising in the social world. In their latest set of studies, they (Shaw et al., 2017), show that friends are sensitive to a particular kind of betrayal. Seeing you as an ally, they expect you to support them in a conflict with a third party. If this support is not forthcoming, they will downgrade you as a friend (your rating may drop from AAA to A). If John gets into an argument with George, and his friend Paul is around, he expects Paul to back him up. If Paul remains neutral or disinterested (he doesn’t even need to side with George), John will think that an implicit but powerful code has been violated.
When Kurzban and colleagues find that friends who remain neutral in a conflict are disliked there is no sign of trouble yet. Perhaps this rarely happens because friends know their obligations, and they too have a motive to maintain the alliance for the sake of future benefits. Perhaps the most troubling finding is that 81% of the respondents said they would remain neutral in a conflict scenario presented to them. Although many friendships can be more durable than romantic of marital pairings, they have their own kind of fragility. Another powerful fragility, of course, comes from competition over women. When these socio-sexual needs are pitted against each other, which would you choose? The Three from the Gas Station gave musical expression to this conflict, and perhaps that is one source of its continued popularity.
Arguably, the ally theory of friendship is incomplete. Perhaps all friends are allies, but not all allies are friends. NATO nations need not feel a warm glow about one another. Friendship may claim evolutionary primacy. Modern contract-based alliances seek to replicate what friendships in the wild do naturally. If so, Kurzban and colleagues may have gotten one thing backward: Friendships model alliances more than vice versa.
Shaw, A., DeScioli, P., Barakzai, A., & Kurzban, R. (2017). Whoever is not with me is against me: The costs of neutrality among friends. Journal of Experimental Social Psychology, 71, 96-104.