• Choose a category
• All
• Classic-Fiction

By Root 904 0

can also develop “concatenated” theories by dividing a complex causal process into its specific component theories, or sequential stages, focusing on particular policy instruments or the views of designated actors. For example, Alexander George and Richard Smoke divided deterrence theory into a number of more specific theories which deterrence comprises: commitment theory, initiation theory, and response theory.221 Similarly, Bruce Jentleson, Ariel Levite, and Larry Berman broke down protracted military interventions into sequential stages and the differing dynamics of getting in, staying in, and getting out.222 Such designations help identify subtypes of undertakings and phenomena that occur repeatedly throughout history which can be grouped together and studied as a class or subclass of similar events. This can be done through statistical analysis when a sufficiently large number of cases of a particular phenomenon is available, or through qualitative analysis of a small number of instances.

Where should one draw the line in developing ever more finely grained types and subtypes? As Sidney Verba put it many years ago:

To be comparative, we are told, we must look for generalizations or covering laws that apply to all cases of a particular type. But where are the general laws? Generalizations fade when we look at particular cases. We add intervening variable after intervening variable. Since the cases are few in number, we end up with an explanation tailored to each case. The result begins to sound quite idiographic or configurative… In a sense we have come full circle… . As we bring more and more variables back into our analysis in order to arrive at any generalizations that hold up across a series of political systems, we bring back so much that we have a “unique” case in its configurative whole.223

Yet Verba did not conclude that the quest for theory and generalization is infeasible. Rather, the solution to this apparent impasse is to formulate the idiosyncratic aspects of the explanation for each case in terms of general variables. “The ‘unique historical event’ cannot be ignored,” Verba notes, “but it must be considered as one of a class of events even if it happened only once.”224

One criterion that helps determine where to draw the line in the proliferation of subtypes is the notion of “leverage”—the desirability of having theories that explain as many dependent variables as possible with as few simple independent variables as possible. This is not the same as parsimony, or simplicity of theories. We agree with Verba and his co-authors Gary King and Robert Keohane that parsimony is “an assumption … about the nature of the world: it is assumed to be simple … but we believe [parsimony] is only occasionally appropriate … theory should be just as complicated as all our evidence suggest.”225

The recognition that even unique cases can contribute to theory development strengthens the linkage between history and political science. Some of the particular qualities of each case are inevitably lost in the process of moving from a specific to a more general explanation. The critical question, however, is whether the loss of information and simplification jeopardizes the validity and utility of the theory. This question cannot be answered abstractly or a priori. Much depends upon the sensibility and judgment of the investigator in choosing and conceptualizing variables and also in deciding how best to describe the variance in each of the variables. The latter task in particular—the way in which variations for each variable are formulated—may be critical for capturing the essential features of “uniqueness.” For this reason, investigators should develop the categories for describing the variance in each of their variables inductively, via detailed examination of how the value of a particular variable differs across many different cases.

THEORY DEVELOPMENT AND GENERALIZING ACROSS TYPES

The most general kind of finding from a deviant case is the specification of a new concept, variable, or theory regarding a causal