[consilient]

> The basic premise is that the act of science and knowledge acquisition cyclically devolves into ideology and is then disrupted.

This can happen of course, but it's not really what Kuhn is talking about. Kuhnian paradigm shifts are to science what massive refactorings are to a codebase: the problem isn't that the old code was wrong (though it might also be wrong), it's that it can't be extended to meet new requirements. And while some of us are better at writing extensible code than others, no one gets it right every single time.