31/01/2012

Echo Chambers Reconsidered

Steve Sailer remembers
the kind of problems I used to run into when socializing before the Internet gave me a better outlet for the corrosive side of my intellect. At a party, I'd start out popular because I knew enough about most things to be able to ask other guests intelligent questions about their personal field of interest. But I tended to get really interested in what they told me and kept asking questions, which at first they found flattering. But I'd eventually get carried away thinking about the topic and get to the point where I'd ask some extremely unsettling question that the other guest didn't want to think about at all (e.g., So, if the Efficient Markets Hypothesis that you studied at B-School is correct, how do you people in the stocks and mutual funds business add value?).
Yeah. I, too, needed a surprisingly long time to realize that, as a blogger more popular than myself would put it, conversation is not about exchanging ideas. In particular, it was only fairly recently that I realized that when someone tells me how X is a wanker because he did y, it is not my job to contribute my assessment of the situation, let alone go and specifically try to see the other side of the story. It is to agree that X is totally a wanker - for having done y, but more generally, too, sure, everyone knows that! (Wish I could heed my own advice.)

More generally, truth is a valuable value, but so is being agreeable. Always giving the former the upper hand over the latter is the wrong strategy.

Ever since the internet became a mass medium, people have worried that it would cause their fellow citizens to communicate less and less with others unlike themselves. Let's call it the Hyperhomophily Hypotesis (HyHy). (I think I developed that hypothesis myself, only to find out from blogs that lots of people had thought of the same idea.) And indeed; most people who spend substantial amounts of time on the internet have probably come across blogs the comments sections of which are best described as echo chambers. I believe there was some economic study a few months back which claimed to have found that the HyHy was overrated, but let's, for the sake of the argument, assume that study was wrong. Would that really be so bad? Three counterarguments.

One, some of the concern about HyHy seems to be based on comparing societal outcomes against an overly harmonious ideal of Let's All Agree on Everything. Me, I like harmony, too, but I also recognize that disagreement between people is a natural ingredient of social relationships, in part because many games just happen to be zero-sum or mixed motive. It's inherent in social life. Any collection of people had better find ways of dealing with that, but using Let's All Agree on Everything as the leading ideal for a society seems to lead to poor results. We tried, here in Germany. Twice. Didn't work out so well.

Two, people tend to like spending time with others like themselves and dislike spending time with others unlike themselves.

Three, perhaps Sailer is right that congregating with like-minded people on the web serves as a valuable outlet for tendencies that don't go down so well in the offline world (where you don't have as much choice regarding who you'll communicate with). Underlying this idea is a steam boiler model of the psyche: once you've released some of that steam, there's less pressure left. That model has been thoroughly discredited for aggressive behaviour, but is quite appropriate for others (e.g., sexual, at least in the short run). What about an overemphasis of truth vs. agreeableness? More research is needed, but for the time being, I'm inclined to agree with Sailer.

P.S.: I would like to point out that this post is not, although it would have been the obvious choice, called "Beyond Universal Agreement: Rethinking Echo Chambers".

No comments: