Optimal Matching Methods for Historical Sequences

Abstract

common script is standard historical and sociological fare. In the passage from which this quote is drawn, Rude describes a script proceeding from general grievances to triggering events and on to a fixed repertory of rick-burning, machine-breaking, and so on. Throughout his earlier chapters he presents stories of individual riots, on the evidence of which this generalized rests. But one may well ask when we can really speak of such a fixed of events. How would we identify such a quantitatively? How could we discover if the characteristic in rural riots differed sharply from that of urban disturbances? How might we analyze the effects of the occupations of participants on such sequences of events? The answers to such questions are not easily found. The problem of event in riots is typical of a much larger category of problems in history. In each case the data are lists of events. In each case the analyst wants to separate the common pattern from its particular realizations. Yet in each case the analyst wants also to understand why each particular realization took the form it did. The only practical approach to these tasks has been to generate a common pattern using an ideal type or comparative analysis, and then to consider the variations from it on an individual basis. There have been no effective quantitative methods for analyzing such sequence data, a fact that goes far