Tuesday, May 11, 2010

The Social Contagion of Disease :: The Flows Within Social Networks


The above TED talk by Nicholas Christakis, MD/Ph.D./MPH, discusses his research on social networks and the contagion of behaviours and emotions within them. He discusses the likelihood of being overweight and obese and one's social ties. If you have a friend whose obese, you are more likely to have a friend who is also obese and vice versa. There is this effect, albeit diminished, with friends of friends {second order ties} and so on::

What explains this?

He found evidence for all of the following 3 factors::

  1. Induction :: Adopting behaviours of your friends leading to obesity/thinness
  2. Homophily :: Sharing norms about obesity/thinness with your friends
  3. Confounding {spurious correlation}:: share common cause for obesity/thinness

He also found that emotions can be clustered, e.g., happy people cluster together, and that propensity to be gregarious is heriditary. More interesting findings is what I call compartmentalization. There are genetic predispositions to broker friendships between friend clusters with some, while others like to keep their friend domains distinct.

He uses the carbon molecule analogy {carbon can be diamonds or graphite, depending on its structure, i.e., how atoms are configured} to emphasize the importance of structure in social networks. So, he states that humans act within a larger "superorganism" akin to a beehive, so that the study of individuals becomes problematic. First, I'm often wary of biological metaphors applied in social contexts, as the mechanisms that make them work are often reduced to reductive logics. I'm not so sure about this superorganism hypothesis, as I think there are often tensions between micro-social interactions and the macro-social order, so the actions of the individual often do matter and not always subsumed in a larger macro-social order.

While I agree with him when he echoes what sociologist have been saying for years in that social networks equal social capital and that they form the conduits of things, resources, actions, ideas, emotions, etc., I disagree with his take that they are "good" entities.

Brian Uzzi has found that networks can be "overembedded" or "underembedded". Overembedded organizations are too insular and subject to groupthink among other things, while underembedded organizations are too thin and don't have enough access to various forms of capital, ideas, knowledge, etc. Networks can be dysfunctional and they can be harmful, so it's important to examine them with respect to the content that flows within them. We also don't always choose our social networks. Some are determined by accident of factors like geography. While we may control if we maintain ties in a network, e.g., the old gang from high school or college, this may be subject to other factors that override our negative feelings of being within a given social structure. On social networking sites like Facebook, when boundaries collide with etiquette and real-life connections becomes readily evident when "nana" or your boss wants to be your Facebook friend. Sure, you can set privacy settings, but my point is that social ties are ties wrapped up in history and context.

Now, go look at your Twitter and Facebook walls and figure out if your network is good or evil.

Song:: Au Revoir Simone-'Through the Backyards'

Twitterversion:: Interesting TED talk {N. Christakis} puts social networks on a pedestal, but doesn't who & what flows within matter most?  @Prof_K

2 comments:

Anonymous said...

Hmm. You got me to thinking... Are social network sites like Facebook truly social networks? Outside of cyber life -- in the real world -- do they count as social capital? Hunh. According to your definition, yes: they are conduits of resources, actions, ideas, emotions, etc. Weird though: on FB you can ostensibly choose your connections, but natural affinity is probably only a fraction of the basis for connection in most cases. Otherwise, it's etiquette, professional or social pressure, status. Just like other pre-ordained social organizations, I guess, like family or small community. Okay, but like I'm just so indignant that there is corporate profit from these technology-based social networks.

But me, I'm a mystic at heart (er, a wanna-be mystic...). I believe that connection is inherently good, because it is life-serving in an emotional and spiritual way. Plenty of touchy-feely New-Agey evidence of that. Well, shit, there's plenty of scientific evidence of that.

And, for the record, I surely don't believe that tecnology-based social networks are inherently good! They're a marketable product, fer crying out loud!

Hungrily chewing on thoughts,
shannon

Kenneth M. Kambara said...

Hi Shannon,
Thanks for stopping by.

Great points. I'm not sure what Facebook is and I'm definitely not convinced it counts as social capital. I say that because of the structure and the content of the communications, which are related to your issues about natural affinity. What I mean by structure is that these social networks on Facebook can consist of ties that aren't really meaningful. I have zero identity with my high school, so while there is a social network there that I could be a part of, it would be devoid of meaning. Moreover, when I think of social capital, I think of a true relational quality to the ties and at least some reciprocity. I'm not convinced that FB really taps into this. I think it can, but not necessarily.

What frustrates me the most about Facebook is that it's not easy to develop new communities around affinities and content.

I like the ideas of creating social networks and linking people, but I see Facebook as a cash grab. They sold everyone on their privacy settings and now are lifting them because they need a more open system to make money. Will people leave in droves? I don't think so, since it offers a "commodified" social networking that's too irresistible for many. Nevertheless, I think there are models for social interactivity and networking {social portals} that can be more affinity-based and hold great potential for offering value without being such an overt and ham-handed cash grab.