flects ideas as old as chaos itself û the idea that chaos is not the absence of any order at all but rather the presence of two many competing orders. This idea is a fundamental element in a number of ancient religions, including Hinduism.
Chaos theory allows scholars to see patterns in the midst of disorder. In many areas of the hard sciences, these patterns are often expressed in terms of fractals, or complicated shapes that are structurally detailed at any level of magnification. There are no perfect analogies for fractals in the social sciences or indeed in the world of human behavior, but social science researchers do search for analogues of fractals, elements of a larger picture that will help them see order in the midst of change as well as potentially helping them to predict the outcome of that change.
The second question proposed above û what is it that makes a good theory û is a harder question to answer, for there is no universally agreed-upon model of what constitutes a good social science theory. This lack of agreement results from at least two things. There is first of all the fact that the various social sciences are themselves so different from each other; it is hardly surprising that economists and ethnographers would differ in their definitions of a good theory. Secondly, there is the fact that the model of what constitutes a good (social) theory is historically and in large measure epistemologically based upon norms derived from the hard sciences. In many cases these norms suffer rather badly in the transition to areas of research in which variables are much slipperier and harder both to operationalize and to quantify than they are in scientific research. This is an important historical point, for a good scientific theory is not the same thing as a good social scientific theory û despite numerous, voluble and often quite vociferous arguments to the contrary. For the purposes of this discussion as a descript...