Saturday, 19 May 2012

'You've Got a Friend in Me, When the Road Looks Rough Ahead': On Patterns of Friendship

Introduction: 'I get by with a little help from my friends'
Humans are deeply unusual creatures- we are the only species to form 'long-standing, non-reproductive unions'- that is, we have friends! From C.S. Lewis and J.R.R Tolkien to Boswell and Samuel Johnson to Gertrude Stein and Ernest Hemingway to even fictional friendships like that of Achilles and Patroclus- friendships are some of the most important relationships we have. Indeed, a decline in friendships in the United States (an American Sociological Review study found the number of people with at least one close confidant has dropped from 80% to 57% from 1985 to 2004) has been linked to an increase in psychological disorders. But why do we have friends at all? And perhaps more interestingly: who are we likely to be friends with?

I will trace evidence that cooperation is important in human societies and that this likely explains the psychological rewards of friendship. I will also explore new evidence that even in tribal societies we tend to befriend people who cooperate similar amounts to us, have similar genes to us (even among non-relations) and are physically and socially similar.

'Lean on Me, When You're Not Strong': The Evolution of Human Cooperation
There is strong evidence from chimps on the antecedents of friendships- for chimps non-reproductive connections provide a form of direct reciprocity- support in a fight, borrowing valuable tools, food in time of scarcity (this has been particularly documented by Pruetz and Lindshield). While these aren't exactly friendships as we'd categorise them- they are based too much in reciprocal giving and taking- they do provide clues on why friendships make evolutionary sense.

Further, it has been documented in primates that those who have a better ability to form coalitions have an evolutionary advantage over their competitors- which has been posed as a possible explanation- the logic being that many of the same characteristics (a giving nature etc.) are the same as we prize in friends and potential members of an alliance.

Baboons who form strong non-reproductive bonds also have better immune function and energy savings, which have been explained as being relieved of the burden of being continuously vigilant of potential challenges and attacks and the potential reduced sense of vulnerability.

As Bowles and Gintis (who on a side note wrote papers for MLK Jr.'s Poor People's March back in the day) document in The Cooperative Species, the relatively warlike nature of the hunter-gatherer existence and the rapid extinction of many groups precipitated the genetic and cultural evolution of social emotions such as shame and guilt because they conferred an advantage on any member of a relatively cooperative group. It is theorised that these emotions provided the jump from so called 'contingent cooperation' (think: if you buy coffee for your co-workers, then you expect them to buy you coffee back at a relatively fixed point in the future) and true friendship.

But can we thus shed any light on who we become friends with?

'Don't walk in front of me; I may not follow. Just walk beside me and be my friend': Who are we more likely to be friends with?
Friendship is obviously a culturally contingent phenomenon- witness the breakdown in affectionate male friendships in particularly Anglo-American society that occurred after the Oscar Wilde trial (and from which the Anglo-American world has never really recovered- men used to walk arm in arm in Hyde Park- would many straight men ever do that again?). However, studies have shown amongst groups as diverse as Americans and the Hadza people of north-central Tanzania that there a few common threads amongst those who we choose to be friends with. Broadly speaking, interpersonal similarity is the strongest predictor: we are rarely friends with those who are completely dissimilar to us (except in the case that through repeated interaction we grow to like them).

Much like Erving Goffman's 'matching hypothesis' for couples, there is evidence that people often pick people of similar 'worth' as defined by different cultural characteristics e.g. looks, intelligence, interests etc. Apicella et al found that the strongest predictors of what they call 'social assortativity' (a measure of the regularity of interactions based on the idea that we tend to interact more with our friends) is highest amongst those who cooperate in similar amounts (this is unsurprising- we like friends who are friendly!). A similar result has also been found for US students and Honduran adult villagers- meaning it is likely to be robust to cultural variation. Physical similarities are also prized amongst the Hadza -- after all foraging is labour intensive and if you've got friends who can physically help more, they are going to be contributing more to your life or group. This may also explain why it has been observed that even in modern society we tend to group with people of reasonably similar physical attractiveness to us- although this is obviously also socially attuned- more attractive people are also more popular. Similar positions in a social group are also a strong predictor of friendship- they both bring people together more often and increase the desire for continued social interaction. 

There is also interesting new evidence that people may befriend those with similar genotypes- in particular a study by James Fowler found that whether a person carries DRD2 (which has been linked to alcoholism) and CYP2A6 (which has been linked to openness) is strongly linked to whether they befriend another person with or without those genes, even accounting for social proximity. This of course is particularly bad news for alcoholics, it turns out that not only are they more likely to be genetically predisposed to drink to excess, they may be genetically predisposed to be friends with others who are also predisposed as such. But it provides an interesting broader point- is friendship also for the benefit of the genes? If we follow a Dawkins logic, some of the purpose of friendship may actually be to benefit our genes. It should also be noted that the Fowler study found that 4 other genes were not linked to friendship- so this question needs further exploration. 

Some Further Questions
Obviously this is an area where many new discoveries are being made- studies of the evolution of cooperation more broadly are on the frontier of science after having been largely ignored by evolutionary biology for so long. But there is interesting evidence that far from just being social constructs, friendships were evolutionary advantageous to humans as a form of reciprocity, social association and possibly even genetic association. None of this of course is to downplay how important and varied friendships really are- it just asks an interesting question: how was I able to feel this way towards others in the first place?

Dan Gibbons is a third year Bachelor of Commerce (Economics) student at the University of Melbourne. He has a forthcoming publication in Intergraph: A Journal of Dialogic Anthropology (about memory and nationalism) and is currently submitting papers on the rise of modern consumerism, the role of criminology theory in literary criticism and the institutional theory of nationalism. Dan is a keen debater and public speaker.


  1. Great article - I especially like the reference to pre-Wilde male relationships. It's fascinating to consider that many other societies have dramatically different attitudes not only towards sexual relationships (pederasty in Pashtun societies being one example), but also towards non-sexual relationships. The intimacy between heterosexual men that you describe is still very much present in parts of India - and is viewed with derision by many in the 'liberal, tolerant' West. This assumption - that intimacy between men is necessarily sexual in nature - is surely testament to the lack of imagination of most people in perceiving how much their behavior is defined by social norms.