>This sounds to me as though Habermas believes one can write a history of
>ideas -- that by analyzing Set 1 of ideas (worldview, etc) and Set 2,
>one can discover an explanation of Set 2.
I don't think this is what is happening here. Habermas is writing a social theoretical analysis and historical account of rational processes, not reason in itself, but how people use their reason to begin with.
Habermas points out that a person who can take the position of an other in dialogue is more rational that a person that cannot. This has nothing to do with logic or ideas, per se, but has to do with the capacity for reflexivity in the reasoning process. For instance, an adult that can reflexively participate in an argument, adopting first, second, and third person perspectives has superior reasoning skills in relation to a person who can only see things from their perspective. Likewise, a society that has in place political mechanisms that provide feedback between the public sphere and the economy, for instance, is superior - in terms of rationalization, to a bureaucratic economy that is only able to operative from the top down. I don't exactly see how this is comparable to writing a history of one set of ideas and explaining that history with a second set. This avoids the relativist fallacy that you seem to be hinting at if the stages of development can be demonstrated to be unavoidable (i.e. you can't take a post-conventional perspective until after you have already acquired knowledge regarding pre-conventional and conventional perspectives).
Now one can say that this is ethnocentric, or that the stages of development are not accurate in one way or another, but one must then demonstrate why the critique itself is superior / true, which entails demonstrating that one understands what one is criticizing and has moved beyond it into greater insight, i.e. one must prove a greater degree of insight relative to the one being criticized, either through internal contradiction or a separate scale of measurement.
ken