Frank Kogan

In regard to Balkanization: Kuhn thought that, in the hard sciences, specialization was an inevitable outcome of scientific evolution: after a paradigm shift, there would likely be more offspring than there'd been parents, and the offspring would be technically more difficult and elaborate, and each would have a narrower range.

Based on something Richard Rorty once wrote back in 1980 or so, I got the impression that philosophy's Balkanization is of a different character: it isn't so much the creation of ever more subspecialties but the creation of competing schools, both figuratively and in some cases literally, a department at a particular college having one emphasis, a department at another having a different one, this being the result of analytic philosophy's no longer having a common idea of what it is to do an analysis.

the philosophy of science that i do know a bit about -- Karl Popper, for example -- is also primarily concerned with border patrol: what counts as science, what doesn't, why this might matter.

I'd hope that the word "primarily" is an exaggeration here. I haven't read Popper, but I'd had the impression that he was interested in induction and falsification for their own sake, not merely so that he could say, "Oh, what you guys are doing isn't science." My guess is that the emphasis on falsification made science seem very heroic, constantly challenging itself. Whereas for Kuhn, paradigms (in the broad sense) allow the practitioners of a science not to spend time testing basic assumptions but to take a whole shitload for granted, rather dogmatically, and to move the discipline along until the vessel cracks, at which point either the vessel repairs itself or alternative shitloads of assumptions are put forth and new paradigms are born.

Kuhn wasn't at all interested in either patrolling borders or breaking them down. I got the impression he thought the sciences could take care of themselves, and I think he was surprised by the impact Structure had (and thought a lot of the impact was based on misunderstanding). Structure presented a different view of how science worked (and how scientists worked), one different from empiricism, and, intertwined with this, developed a model for how sciences evolve.

I'd say that Structure gets rid of the ontological division between science and nonscience, so the criterion stops being "You're a science if you do things scientifically," i.e. use a particular method; but Structure actually increases the sense (or my sense) of sociological distance between the sciences and the nonsciences, since the latter can't simply start acting more "scientific"; they've got to get everybody on the same page, which turns out to be very difficult, since concepts and assumptions worth agreeing on prove hard to come by.