Howard S. Becker and Blanche Geer
MANY people think of the work of science as the production of general propositions stating the relation between two or more variables under a specified set of conditions. Such propositions take, in the simplest case, this form: if A, then B -- provided that conditions D,E, and F obtain. These kinds of propositions avoid taking account of the unique characteristics of any given case and attempt to abstract from the particular case only those variables contained in the proposition, while controlling all others. Students of small groups, for instance, work with propositions relating such variables as cohesion, communication, and deviance in ways that are purposely independent of those qualities unique to the groups on which their studies are done.
Generating such propositions is an important part of scientific work, but sometimes the study of human organization requires a different approach. For many practical and theoretical problems, we want to take account of as much of an organization's complexity as our theory will allow. We may be interested in applying existing theory to a particular case in order to understand and possibly control it, or in developing theories about social systems and their relations with the environment, or in understanding the nature of a particular social problem. When our____________________
Some of the material presented here originally appeared in Howard S. Becker ( 1958) and is reprinted with permission of the Editor. Alvin Gouldner's comments on an earlier version of that paper were of great value. Another paper on participant observation is by Howard S. Becker and Blanche Geer ( 1957).