I have thought that the physical models were interesting ways of describing cities. I had some nice results in 1968, results that have taken me the subsequent 35 years to understand (two of my books). I did further work in this area (another book, Marginalism and Discontinuity) over the years, and more since then.

So most of what I see in the book Critical Mass is not unfamiliar. It is fairly presented, for the most part. It is a good survey.

I have never been a "believer." It's interesting, and if you can develop modes of empirically verifying the mechanisms of the physical models, that would be significant. And the physical models do provide phenomenologies of various social phenomena, albeit most of the ideas are already available outside of physics or formal mathematics.

To me, when I read the various applications in papers, of network theory, of game theory, of complexity and agent based modeling, and even of much of statistical regression modeling, and even of what is called rational choice--it's all cute, usually handled a bit roughly and sloppily, and perhaps a bit illuminating. Of course, you can improve the models and so publish another paper. Etc.

But what I want is rich understanding of society, and that richness is rarely encompassed by what people do.

NN may well be terrific, but when he showed what he did, all I could think of is that you need to think more about what you know about individual firms and divide up your sets of time series first.

Most of these people using network theory, use an elementary idea, when in fact it is the more advanced ideas that make the theory interesting.

Game theory is nice, and so are market mechanisms, and I love to read the stuff that is done by the masters--eg Modigliani-Miller or Fischer Black's musings, and sometimes Schelling. But in the hands of the less than masters, it's not so interesting to me. There is something predictable about law and economics for the most part. I want to know more about legal compexities, and the law and history people in our Law School have the edge.

Critical Mass is about smart people with powerful tools, but they are not as smart as they need to be.

I'm being brutally honest. When I go to a seminar, most of them are obvious once you see what's up, and then I ask, how good are the empirics, the fieldwork or the archival work. If it does not take 2-4 years to gather the empirical data, how can you make an advance?

Real physicists or mathematicians don't act like Einstein or Grothendieck. But lots of the social scientists do. How many Einsteins can a field have? It's embarrassing.

I know this is not very kind to social science as a discipline, but I have always been taken by theory that has large and rich data that needs to be matched to is. It's usually the other way around. I have learned a great deal from the great thinkers in politics, society, etc.

By the way, if you look at Richard Easterlin's most recent book of essays, much of what I am saying is mirrored from a different point of view. I am not against the prevailing social science, but it rarely makes me want to know more.