Citation

Perry RW. Int. J. Mass Emerg. Disasters 1989; 7(3): 305-327.

Copyright

(Copyright © 1989, International Sociological Association, International Research Committee on Disasters)

DOI

unavailable

PMID

unavailable

Abstract

Recent years have seen the rise of a consensus among disaster researchers that a pressing need exists for the development of both classification systems (typologies) and theoretical models in the study of disaster phenomena. Given the importance and urgency assigned these tasks, the question of how they fit together arises. Indeed, two issues of such importance probably merit joint pursuit by the community of scholars. Yet, there is some evidence that the relationship between taxonomy and theory building is not clear to many social scientists. They are often treated as independent activities which should be pursued separately. The purpose of this paper is to examine the ways in which taxonomic activity meshes with theory construction, particularly with the construction and testing of causal models. The substantive focus for the discussion will be the problem of individual response to warning messages. It has been argued that of all disaster phenomena, warning response is among the "best studied." In fact, there is a very long history of research on warning response and a large backlog of field studies. Given all these data, a question that has long challenged researchers is: how can it be organized theoretically? One answer is that theoretical organization can be achieved both through taxonomy and model construction. From a meta-theoretical perspective, taxonomy and modeling are complementary activities, not any more distinct or separate than research and theory. The remainder of this paper is built around discussions of three issues. First, we will examine the meta-theoretical context of taxonomy and modeling. This task involves examining recent issues in causal modeling practice and specifying the relationship of classification schemes to models. Second, we will develop -- drawing upon the warning response literature -- a model which takes into account the basic tenets of the recent modeling literature. In this context, primary theoretical and methodological issues will also be discussed. Finally, the model will be examined in terms of a comparative dimension -- across events -- using classification (or typology) as a means of providing interpretative context.