12 December 2007
I have been too busy catching up on things to blog about the conference last week on computational social science (scroll down to see the program). David Allen prodded me with a thoughtful e-mail about the event, which I reproduce below. We had a substantial attendance-- filling a large auditorium-- so I invite attendees, if so inclined, to add their comments (small or large).
One presenter at the conference on Computational Social Science (Fri Dec 7) pointed back to a ‘Top 40 list,’ a progressive account of ever-mounting achievements, tallied earlier in the day. The entire thing, this half-day event, was a Super Bowl, a gathering that may not happen again so soon – we were fortunate. Here is just one proposition stimulated by the discussion:
The lesson from Tycho Brahe emphasizes the importance of data, to advance knowledge. Naturally, that highlights a computational science. However, as important as was data to the Copernican revolution, its sea change in worldview turned at least as much on overturning an entrenched ideology. If I remember my history (and I may not …), even the Inquisition threw around its perfidious muscle, trying to prevent humankind falling from the center of cosmology.
Interestingly, ‘stickiness’ in the conduct of human affairs – we might identify defenders of an ideological faith as ‘sticking’ to their guns – was prominently on display in results reported at the conference. First Laszlo Barabasi and then Sandy Pentland took time to detail how their results quickly portrayed habitual behavior (and so, perhaps surprisingly, predictability of their subjects).
It is not a far step from well-worn paths, in those two presentations, to well-worn mental paths. (Of course Thomas Kuhn was the modern expositor on this subject. Earlier, Max Planck made it succinct with, "Science advances one funeral at a time.")
Was there evidence, at the conference?
Across the course of the day, there was evidence of change in some prior views. Particularly, Mark Granovetter’s notion of ‘the strength of weak ties’ came in for inspection. Again it was Laszlo Barabasi who presented results that support strong medium ties, instead. The point may also have been touched in another presentation, but I have lost reference to it if so.
But, we also heard, in comment from the floor, the difficulties finding publications that accept these papers. That belies wider resistance to a ‘new approach.’
Almost ironically, in the concluding session on IRB, what I take to be the underlying tension proved to be crisply on display.
Marshall Van Alstyne offered an elegant construction, a solution that takes advantage of the prevailing notions from the currently accepted worldview, at least in neoclassical economic thinking. In response, Allan Friedman raised questions about their realistic applicability.
Those two will speak for their views. Here, I will suggest where lie the deeper tensions, between the currently predominant paradigm and the social network view busily being developed.
Social network analysis would vary from the neoclassical view particularly in two fateful ways, I suggest: Rather than begin from static equilibria (borrowed of course from physics, earlier), dynamics are ‘natural’ to social network analysis. More, neoclassicism takes off from the individual, or individual firm; there is no place, really, for connections among the atoms. Social network analysis comes at phenomena, of course, from exactly the opposite direction.
The implications boil up all the ideological struggles – quite beyond intellectual quarrels – that roil standard politics, in the US and elsewhere. No wonder there is resistance.
At least there was, in this event, serious effort on display, to move from data to conception and theory. New intellectual lands are colonized only with such landscaping.
Posted by David Lazer at December 12, 2007 10:40 AM