I am quite impressed by Lakoff's work - I think he is right on target claiming that political differences are in fact differences in cognitive styles. I was saying that for long time, albeit not as eloquently as Lakoff. I think that his two cognitive frames (or schemata) of strict father and nurturing parents morality are quite appealing, but I am not sure how that would work outside the US. My impression is that these family models have still deeper roots in personality traits, especially different styles of organizing categories.
My own thinking, albeit I have not developed it into a more or less coherent argument yet, is that categories represent, metaphorically speaking, "islands of meaning" or perceptual chunks cognitively glued together by some perceived similarity and cognitively separated from the rest of the world. Each "island" needs two elements to exist: the organizing core or principle i.e. what is believed to be the common or most salient feature of the island, and the boundaries that separate members of the island from the rest of the world.
Albeit both "cores" and "boundaries" are needed to construct a cognitive category, I conjecture different people may use these two elements differently. Some of them may depend more heavily on towers while others - on walls.
The core-centric way of category construction essentially involves constant comparing a given element with the core to ascertain its similarity. It is not that much concerned with boundaries as long as the core is clearly perceived - it can tolerate porous or poorly defined boundaries as long as it is cognitively capable of comparing elements with the "core." This is a very dynamic way that requires constant mental assessments and adjustments.
The boundary-centric way of category construction essentially involves drawing a thick line that separate members and non-members. As long as members are clearly on one or another side, they are classified accordingly, often regardless how they compare to the core . The boundary-centric approach does not tolerate poorly defined boundaries and ambiguities, because it is lost in such situations, especially if the person in not well versed in the core-centric approach. This is very static approach that requires fair amount of rigidity and permanent markers to function properly.
It is easy to see that progressives tend to be more core-centric in their approach to category construction, while conservatives tend to be boundary-centric. I also conjecture that these approaches are determined to a large extent by bio-chemical factors in a similar ways as our affective types and disorders are. Some people may be "chemically" predisposed toward one or the other, or perhaps have both.
I believe that this approach, if true, can go beyond the US experience and explain facts that Lakoff's archetypes cannot - e.g. why foreigners who have not been exposed to the strict morality/nurturance models of family recognize "their" own type in the US. I saw that a lot among E. European immigrants - ex CP apparatchiks and functionaries became staunch GOP supporters.
Another thing that Lakoff aptly observes is that while there are many different types of progressives/liberals based on their own modes of thinking, they all share the same core principles based on the nurturing family archetype. I totally agree - I think that we should concentrate on what we have in common, rather than what separates us. Especially before Novemeber.
Wojtek