Philosophy represents an interesting case, because the Anglo-American branches of Analytic philosophy treat philosophical history and its contexts as secondary to generating new problems and solutions.  The Continental branch is much more open to reexamining and reworking older problems and positions, and in this respect is much closer to literary studies.  This is one reason why literary scholars in America discuss Derrida much more often than philosophers.

The hard sciences, however, seem to have more difficulty using "superseded" frameworks for generating new understandings.   I would expect these frameworks to have a pedagogical use, but not a really generative use for new research and emergent problems. 

This, at any rate, is how I interpret Paul's response to Anne about the situatedness of the scientific observer.  If I understand Paul correctly, this kind of observation about situatedness is nice to have, but doesn't seem crucial to making new discoveries in his field, a very different attitude than that found in the humanities or even the "soft" sciences.

 And Kuhn's notion of paradigms and paradigm-shifts seems to demand that scientists themselves remain incapable of switching their perspectives from one paradigm to another at will, or of translating the results of an analysis in one paradigm into terms recognizable for another paradigm.  This is what Malhotra calls "incommensurability":

The thesis of incommensurability implies that rival theories are radically incommensurable. The impossibility of full translation between rival paradigms is further exacerbated by the fact that the advocates of different paradigms often subscribe to different methodological standards and have nonidentical sets of cognitive values (Kuhn, 1977).

So what does the incommensurability thesis do to our hope for "advancements in knowledge"?