On Sun, Jun 21, 2009 at 12:16 PM, Don Steiny
<[log in to unmask]> wrote:
I do not want to paint everyone with the same brush either and there is some great work going on, however the idea that networks are a consequence of the interaction of atomic actors is explicitly and publicly criticized in a number of places. Simply discovering that a network is scale free or has various types of centrality seems to be tautological to me and there is too much of that, especially since the term "social network" has come to mean "software for interacting on the Internet." It may be for the type of relational thinking that drove the interest in social networks in the 60's until the mid-80's the term "social network" may have lost its value. I know that I avoid it now because of the misunderstandings that it creates. My point, which was VERY badly made in a public forum is that "social network" has a number of different meanings. Historically, the relational view was a reaction to the functionalism of Talcott Parsons at Harvard. Functionalism is, basically, starting from the current situation, divining the subsystems and creating a plausible story about their relationships. There is no real way to decide if this is true or false. Analyzing an organization as patterns of work flows, communication and so on misses some essential questions and it may be that the ability to describe them as networks is more obfuscating than enlightening. A very good example of something that is similar can be seen in Donald MacKenzie's "An Engine, not a Camera" where he gives some well researched examples of how models can drive behavior in the guise of describing it.
In the "Myth of Network Analysis as a Special Method in Sociology" Granovetter explicitly talks about trying to not be typecast a network guy, and White's Identity and Control both relegates networks to a piece of social formations and explicitly expands his meaning into a phenomenological view and does not consider them a only a measurement construct. He defines "ties" as "stories" and, in general, has a very different view of networks than one would see in many places. So even though I have very badly paraphrased them, careful research into what they say shows that it is somewhat different than the bulk of what is going on in network analysis these days. White's subtly and scope is astounding, so any paraphrase of anything he says will be merely a shadow (and I cannot help but injecting a humorous side note that some might say if anyone says anything comprehensible, he or she is not reflecting White accurately). Granovetter puts incredible care into everything he writes and teaches and is meticulous about presenting his arguments. He does not shoot from the hip.
Don,
Without wanting in any way to detract from your apology, I do want to say that, however accidentally, you have provided a marvelous teaching moment for newcomers to SNA. It may be worth remembering something that David Ogilvy wrote about advertising, the importance of recognizing that you preach to a moving parade. Thus, what seems incredibly old hat to someone who has spent years in the trenches may still be fresh and exciting to newcomers. A case in point: When you write that, "Simply discovering that a network is scale free or has various types of centrality seems to be tautological to me," I recall my excitement when the data that I presented at Sunbelt 09 revealed that the networks formed by memberships in teams that create winning ads had precisely the properties that network analysis predicts: giant components appear, giant bi-components approximate giant components, differences in degree have power-law distributions. I know full well that these results do not explain why particular individuals are central in these networks and that to understand other network details it helps to know the institutional setting and the shifting shares of media in the Japanese ad market over the last three decades. Network analysis is no panacea. Still, those results do turn me on. I find myself recalling the sense of wonder that a thirteen-year old me felt the first time he read Isaac Asimov's Foundation Trilogy. I feel like I've seen a piece of the mathematics of Hari Seldon's psychohistory and now I am looking for the Mule who will confound its predictions.
And, thanks to your indiscretion and this marvelous apology, I am now aware of new places to look, new worlds of scholarship to explore: That Harrison White idea that each link is a story resonates powerfully with my own life experience.
Thanks for this breach of norms that reveals underlying truths and suggests new directions.
John