Friday, August 17, 2007

Physicists Will Save Us?

I've praised Mark Buchanan's The Social Atom here in the past, a very good analysis of complex adaptive systems, networking and nonlinearity, that shows how human interactions are not as special as humans (surprise) think they are. Buchanan has a blog and has discovered not only crime but "routine activity" theory in criminology. A nice post here on where it might go and a promise to do more analysis soon. (There's also a good comment citing the incomparable Jane Jacobs as support for Buchanan's perspective.)

What's important about this is the fresh set of eyes that Buchanan and his fellow physical science people bring to our situation. The application of new methodology alone, which almost has to be done by the mathematicians and physicist types, will advance us in our study of criminal behavior far beyond what the simplistic, linear, fragmented studies in social science have done in the past. But they will also literally be "outside the box," not trapped by the decades and decades of structured thinking that afflicts every aged discipline. They won't come up with all the answers for the hows and whys of crime and how to stop it, but they will bring some fresh thought and enliven the old truths. It will be fun to watch. Go give him some hits so he knows we're looking forward to it.


JSN said...

It is true there is no unified theory of crime causation nor is there likely to be one in the near future because there are too many variables that are uncontrolled and the data quality is not very good.

A good exercise would be to attempt to devise a theory of pick-pocketing because all pockets have a guardian one of the variable is constrained and the intuitively obvious prediction is that risk is proportional to the product of the pocket density and the offender density. One problem is to devise a self adaptive theory that allows the offenders and guardians alter their tactics. A second problem is figuring out how to test such a theory if one can be devised.

Gritsforbreakfast said...

I haven't read The Social Atom, but Buchanan commented on Grits recently to mention that the central thesis in it was similar to an hypothesis I suggested in this post on Voronoi cells and privacy. (He didn't discuss Voronoi cells, he said, but used some of the same math concepts to apply to the social sciences in similar ways. I'm very interested in how computer-aided higher math, fractal analysis and chaos theory can help reinvent criminal justice theory and practice. That and the implications of new brain science findings show how the hard sciences could drive certain criminal justice reforms in the near future, at least, perhaps, after the science is nailed down a little more concretely. best,