07 June, 2012

ICWSM: Robin Dunbar - Why Facebook Won't Get You Any More Friends

I've never heard Robin Dunbar speak before but I know his basic arguments pretty well. He doesn't disappoint by repeating them all - monkeys, brain sizes, common functional human grouping size, etc. This talk is about why social network sites aren't going to get you more friends, so Dunbar starts out with a broad statement that "The Internet has promised a greater social connectedness" but apparently that hasn't come to pass. He asks whether technologies really widen our horizons (insert the echo-chamber research) and whether it can really result in more friends. His answer is that it doesn't because of the ... social brain hypothesis...! You knew that was coming, right?

I wonder though - Dunbar seems to take the deterministic argument social technology = more friends as his straw men, but do we really use technologies to try and enlarge our social world or do we use it as part of our habitual social practices which at times can result in an enlargement of our social worlds? I mean, what happened to serendipity and wonder, the chance encounters that can change your life, the discovery of being a part of something that you hadn't realized before. Or is it that once you reach those "about 150" you have to loose one to gain one? Ok, I am over-simplifying and Dunbar as a good scientist is showing lots of studies, graphs and experimental results. He notes that the 150 has high turn-over. It's a clumpy network that pretty much suggests the faceted nature of social life. There is an obvious relationship between frequency of social contact and feelings of intimacy. Inside the 150 is where real relationships live but they can move in and out of that space - becoming more acquaintances at times and maybe sometimes moving back into the circle of close confidants. About half the people you include in your network are family and half are friends (although I suspect this is culturally and individually defined with huge variations across, also what do you define as family - blood ties or perception?) Clearly a huge proportion of our social time is devoted to the "inner circle" and the rest of our attention is spread out unevenly among the ties that are in different states of development (or perhaps deterioration).

Friendship and kinship have different dynamics (though I would argue that some kinds of friendships over time come to resemble certain kinds of kinship). Dunbar points out that the average emotional closeness with kin remains stable after residential mobility (measured over the course of 18 months) while for friendships this plummets. This is my dissertation - crazy! Maybe I ought to publish some stuff from that piece of work finally. Dunbar's data here is exclusively gender-related showing that for women talking and for men doing more activities together results in increases in closeness - very western-normative.

Other interesting bits - laughter influences happiness (this happiness bit is becoming really common in research) - experimental results show that laughter happens more often during F2F encounters and Skype is second-best, while people seem to find everything else - phone, IM, SNS, email - significantly less satisfying. So seeing each other does something to make us laugh more and enjoy the interaction more perhaps regardless of how "useful" or "instrumental" said interaction was (Paul Tomkins, take note - there is a graph proving it!)

The analysis of a cell-phone dataset shows how over time women shift form investing into relational partners to investing into their children. In general, women talk more and invest more into relationships than men (men spike but show shorter spike periods than women). The basic message here is that relationships are costly (Google+ team take note!) they are investments and trade-offs. Romantic relationships means fewer best friends - so falling in love is socially expensive because of the time costs. The more traits we share with someone the closer we feel to them emotionally (homophily) and the more likely we are to act altruistically toward them. Curiously, these traits are NOT personality traits but things like sharing the same language, coming from the same place, sharing the same worldview, politics, religion and shared humor (most important of the six traits). This explains why over time people drift apart.

So what makes relationships work? They relate on emotionally intense grooming and cognitive limits (brain size+cognition).

The cognitive limits are about intentionality and it's relationship to limits of our social circle. By intentionality Dunbar means the ability to keep straight in your mind a hierarchy of intentional action - mary knows that bob knows that jane knows that sharon knows that janet does. Five seems to be the average limit and is correlated with average social clique size.

The emotional part of making the relationship work is grooming, which is all about time investments - grooming is a bonding agent and grooming time is determined by group size. Social grooming in fact triggers endorphin release in the brain which is, presumably, the psyho-pharmacological environment for building trust - mmm... yummy endorphins. Of course, in humans, language allows us to groom more efficiently but we seem to require physical touch to get the endorphin hit from social interaction (ah well, there goes the digital world, we'll have to get our kicks some other way, running?). We bridge the gap between grooming and endorphins through laughter, religion and rituals, dancing or playing music. We can get endorphins simply through physical exertion (see exercise is not just good for you but it's also a drug) but socially shared physical exercise enlarges the effect (tricky, you nature!)

And that's all folks - hug more! Don't expect Facebook to make your social life sparkly and laugh more. I think I like the message here even if I haven't learned anything new - at least for this audience the confounding aspects of emotional investments and costs of relating are important to drive home - repeatedly.


No comments:

Post a Comment