At 07:03 PM 9/24/99 -0400, CHarles Brown wrote: > > >Charles: I must say , I don't see how you avoid an abitrariness problem at the level of selecting your unit. I recall well this problem in studying anthropology when we were trying to decide how to draw boundaries around cultures or groups. Selecting the part can be just as arbitrary as selecting the whole. > >I mean the point on a production unit is that , today, production is so social, the division of labor so complex, calling it global is almost not an exaggeration. So, a factory, a production unit, has connections all across the globe, etc. > Charles, of course every unit of analysis is to a significant degree arbitrary. What I was suggesting was to: - construct units of analyis that closely correspond to the percpetions of social actors, rather than those of the researchers; it is quite frequent that social actors act on behalf of organizations, much less common than they act on behalf of a social class, bur acting on behalf of a 'world system?? ? - gimme a break! - construct units of analysis that allow for empirical comparisons, that is, examining the effects of a hypothesized factors under the 'ceteris paribus' conditions; such comparisons are nuts and bolts of empirical science (without that we enter the realm of religion and ideology); the less aggregated (i.e. simpler) the unit of analyis, the easier it is to find cases that are empirically similar in certain respects (ceteris paribus) and diffretent in the hypothesized effects (counterfactual); thus it is possible to find organizations that are similar in many respects and differ only in a few respectes, but it is very difficult to find nation-states that are similar in many respects but differ only in a few; as far as a 'world system' is concerned, however, it is a single sui generis case - so what is the basis for a comparison? what is the counterfactual? by selecting such a large unit of analysis we effectively abandon the realm of empirical science and enter that of story-telling and mythology i.e. we are confined to finding only corroborating evidence and apriori precluding any effective falsification of our stories. Consequently, the thirdworldist stories of some this list's contributors are as good as, say, the "white man's burden" - they are accepted based on their emotional appeal rather than empirical evidence. wojtek
