Under the circumstances, then, the most likely approach would combine two widely neglected techniques in the sociologist's repertoire: observation and archival data analysis.46 All of the Propositions call for a comparative analysis of qualitatively differentiated categories of organizations. While space limitations preclude detailed consideration of each Proposition, perhaps a few brief illustrations will suffice. Proposition 3, for example, would entail a comparison of societies qualitatively classified according to the extent of observed regimentation and political oppression. Similar approaches would be appropriate for Propositions 5, 7, and 8. The first Proposition, on the other hand, could be tested by comparing the developmental histories of several organizations roughly matched on all characteristics except in their approximation to Simmel's constructed type. According to the Proposition, the analysis should reveal that organizations whose ideologies, goals, and activities are closely dependent on restricted distributions of information develop into full-fledged secret societies more frequently than other organizations. Finally, tests of Propositions 2, 4, 6, and 9 would require comparative content analyses of official documents, personal letters and diaries (where appropriate), and other recorded descriptions and accounts of life in the secret society. In conclusion, the Nine Propositions presented above are not exhaustive of Simmel's essay. They extract only a part, and probably the most obvious part, of hiis argument. Moreover, when the insights of this essay are combined with those of hlis otlher investigations, still more Propositions are evident. Coser's reformulation of Simmel's classic work on conflict, for instance, readily suggests, among many others, a continuation of our Proposition 4:47 the greater the tendency of the secret society toward total inclusion of its members' activities, sentiments, ideas, and objects, (1) the greater the intensity of its internal conflicts, and (2) the more forcefully it responds to acts of renegadism and heresy committed by its members. Nevertheless, while not exhaustive of their source, the Nine Propositions do illustrate the kinds of testable statements that the essay will yield. If Simmel's analysis is to be used as the basis for generating explanations of the origin and development of secret societies, or, more generally, of the role of secrecy in social organization, the necessary first step is a conversion of imaginative insight into imaginative but manageable Propositions. lation between the scientist's quest for knowledge and the individual's right to privacy? See Edward A. Shils, Inquiry and the Autonomy of the Individual, in D. Lerner (ed.), The Human Meaning of the Social Sciences (Cleveland: Meridian Books, 1959), pp. 114-157. 46 For an excellent discussion of the uses of archival data, observation and related techniques, see Eugene J. Webb, et al., Utobtrusive Measutres: Nonreactive Research in the Social Scientces (Chicago: Rand McNally & Co., 1966). 47 Lewis Coser, The Functions of Social Conflict (New York: The Free Press, 1956), pp. 6772.