Big Structures, Large Processes, Huge Comparisons

9. Conclusions

9 CONCLUSIONS The Tasks at Hand In the light of any formal logic of comparison, most of the inquiries we have been examining are ungainly indeed. On the scale of continents , national states, and regions, the matching of instances with each other only provides the grossest of natural experiments. Therein lie two traps: the trap of refinement and the trap of despair. It is tempting to look for finer and finer comparisons, with larger numbers of cases and more variables controlled. In the present state of our knowledge of big structures and large processes, that would be a serious error. It would be an error because with the multiplication of cases and the standardization of categories for comparison the theoretical return declines more rapidly than the empirical return rises. Only in building better theories by means of comparisons on the scale of a Bendix, a Skocpol, a Moore or a Rokkan will we manage to shift that curve of theoretical return from finer comparison. In a distant future, we can aim to have theories of large-scale social processes sufficiently precise that a well-measured chunk of a single region's experience will provide strong proof of a theory's validity or invalidity. The trap of despair opens up when we decide that such a day will never come-can never come. If we can never get past hesitant generalizations in the style of Stein Rokkan, what's the use? THE TASKS AT HAND 145 The use is this: Historically grounded huge comparisons of big structures and large processes help establish what must be explained, attach the possible explanations to their context in time and space, and sometimes actually improve our understanding of those structures and processes . Rokkan's conceptual maps of Europe, for all their faults, do not simply differ from models of state-by-state political development. They have more explanatory power. They are better models. In the improvement of our understanding, individualizing comparisons , universalizing comparisons, variation-finding comparisons, and encompassing comparisons all have their uses. In fact, they are somewhat different uses. I have described the four as if they were alternative tools for the same task. That helpful simplification will eventually have to give way. The four types of comparison differ, after all, with respect to the sorts of statements they yield rather than with respect to the logic of comparison as such. Their relative value depends on the intellectual task at hand. It also depends on the nature of the social world and the limits to our knowledge of that world. Pragmatic, ontological, and epistemological realities all matter. Pragmatically, there are times when what we need most is a clear understanding of the singularities of a particular historical experience. If people have done a good deal of theorizing, implicitly or explicitly, on the basis of that experience, getting those singularities right will serve immediate theoretical purposes. The English experience in creating parliamentary government and regularized opposition needs scrutiny over and over again because-as the contrasting accounts of Bendix and Moore suggest-that experience appears, transmuted and generalized, in so much argument about the bases of democracy. In that case, individualizing comparison serves quite a general end. Universalizing comparison, if appropriate and well done, has rare clarifying power. To show that the same sequence or conjunction of cause and effect recurs in widely separated settings reduces the intellectual need to erect separate explanatory frameworks for each setting, sharpens our sensitivity to other similarities and differences among settings, and helps identify forms of intervention in those settings that are likely to affect them. Suppose that demographers' long search for a standard sequence of transition, population by population, from high, unstable to low, stable mortality and fertility finally pays off. Knowledge of the sequence will cast intense light on the probable demo- 146 CONCLUSIONS graphic consequences of various programs of investment, employment , agrarian reform, or fertility control. Variation-finding comparison, however, promises to help us make sense of social structures and processes that never recur in the same form, yet express common principles of causality. None ofthe analyses reviewed in this book, for example, provides much assurance that anyone will ever discover a single path leading diverse regions from low income to high income. Yet it remains possible that some correlates of change in income (for example, the tendency of populations to spend smaller shares of their income on food and shelter as income rises...