Big Structures, Large Processes, Huge Comparisons

4. Comparing

4 COMPARING Eradicating Pernicious Postulates How can we eradicate the perniCIOUS postulates? Two approaches , one direct and the other indirect, promise to do the job. Directly, we should track the beasts to their dens, and battle them on their own grounds. We should look hard at the logical and evidential bases for generalizations about social change, about the use of illegitimate force, about differentiation as a master process. We should confront them with real historical cases and alternative descriptions of what actually went on. They cannot resist these weapons. The indirect approach makes it easier to discover appropriate historical cases and to devise alternative explanations. It consists of fixing accounts of change to historically grounded generalizations. I do not mean universal statements confirmed by a wide variety of instances in different eras and parts of the world; at that level of generality, we have so far framed no statements that are at once convincing, rich, and important. I do mean statements attached to specific eras and pa~t~ of the world, specifying causes, involving variation from one instance to another within their time-place limits, and remaining consistent with the available evidence from the times and places claimed. Big structures, large processes, and huge comparisons enter the analysis at precisely this point. They provide the stanchions to which ERADICATING PERNICIOUS POSTULATES 61 we lash our historically contingent statements. Analyses of structures and processes operate at four historical levels, all of them involving comparison. At the world-historical level, we are attempting to fix the special properties of an era and to place it in the ebb and flow of human history. Schemes of human evolution, of the rise and fall of empires, and of successive modes of production, operate at a worldhistorical level. At the world-systemic level, we are trying to discern the essential connections and variations within the largest sets of strongly interdependent social structures. World-system analyses, strictly speaking, certainly qualify, but so do Toynbee-style studies of civilizations. At the macrohistorical level, we seek to account for particular big structures and large processes and to chart their alternate forms. At the microhistorical level, we trace the encounters of individuals and groups with those structures and processes, with the hope of explaining how people actually experienced them. Need I warn that the distinction of exactly four levels, rather than three, five, or some other number, leaves great room for debate? Unless we have compelling evidence that some kinds of large structures persist, cohere, and constrain all the rest, the number of levels between the history of a particular social relationship and the history of the world remains arbitrary. We should resist the temptation to reify the levels. I place the number at four on the wager that through most of history the world divided into at least two largely independent networks of production, distribution, and coercion. Our own singlenetwork era began when the network of production, distribution, and coercion centered in China became inseparable from its counterpart centered in Europe. If so, we can reasonably distinguish among analyses of (I) variation from network to network, (2) the operation of particular networks, (3) variation among structures and processes within particular networks, and (4) clusters of experience that people within particular networks treated as having common properties. Those define four levels: worldhistorical , world-systemic, macrohistorical, and microhistorical. If the world forms but a single coherent network, then the first two levels collapse into one. If the only significant uniformities and variations among structures and processes are those identified by the participants themselves, the distinction between the last two levels dissolves. 62 COMPARING How many levels exist and what units define them are partly empirical questions. Within limits, we can amass evidence for or against Toynbee 's claim that great civilizations, defined by people's interdependent involvement in a distinctive system of cultural premises, constitute the largest intelligible units of historical analysis. Within limits, we can also bring evidence to bear on the claim that at a certain point in time-including our own time-the entire world formed but one such system. Adjudication of the evidence, however, requires agreement on the practical definitions of difficult terms such as "coherence" and "interdependence ." If any connection counts, we will most likely discover that with trivial exceptions the world has always formed a single system . If only the sort of coherence nineteenth-century analysts attributed to societies counts, we will...