ion. * Part of a paper presented at the annual meeting of the American Sociological Society held in Denver, Sept. 7-9, 1950. That paper was a substantive report on a research described in William J. Goode, Some Problems in Postdivorce Adjustment,. American Sociological Review, 14 (1949), 394-401. Thanks are due to Melvin Tumin of Princeton University for aid in formulating the present treatment. Such an assertion does not deny the analytic differences between the economic and the sociological levels of abstraction. Economic analysis does represent a different level of emergence.1 A study of chairs within the economic framework of supply and demand is different from a sociological investigation which treats them as social objects, defined by social custom, expressive of social status, and imbued with meaning (the empty chair which suggests an absent friend, a throne, the bench outside the vil1 For a more extended analysis, see Talcott Parsons, The Structure of Social Action, New York: McGraw-Hill, 1937, pp. 737 ff., 768-772, etc. Cf. also William J. Goode, Religion Among the Primitives, Glencoe: Free Press, Chaps. 2-3. This content downloaded from 157.55.39.215 on Tue, 30 Aug 2016 06:03:46 UTC All use subject to http://about.jstor.org/terms ECONOMIC FACTORS AND MARITAL STABILITY 803 lage barbershop, etc.). Similarly, the analysis of the expenditure patterns of a family, to determine how it allocates its scarce resources to alternative ends, is different from a sociological study of its status-seeking and status-maintaining activities. We can discriminate between the abstract levels of economics and sociology, as we can between these and a biological framework or abstract level. Indeed, we ordinarily justify the departmental separation of disciplines on some such basis. However, to identify the differentiated abstract level on which a given analysis proceeds does not deny the theoretical and concrete importance of factors at that level which are usually dealt with at other abstract levels. We do differentiate a biological level at which we might analyze glandular activity, but we must be prepared in some cases to analyze such activities on the sociological level, in terms of social meaning. The case is similar for economic factors. Just as the economist can (and does) analyze the impact, upon economic processes, of factors ordinarily dealt with sociologically so must we take into account some of these economic processes-but within a sociological framework, i.e. at the sociological level. It is in such terms, then, that factors usually considered economic may be observed and analyzed in the divorce process. We can measure their importance, and must do so within a framework of social meaning. It is an interesting commentary on recent sociological history that while we have tended to reject economic factors partly because they are on another emergent level, we have not done the same with so-called personality factors. Rather, we have used the latter as standard tools in our sociological analyses of the family, although they are as clearly on a different emergent level as economic or political variables. This has led us to stress the causative character of personality and psycho-dynamic variables in the shaping of economic processes, and to gloss over the real possibility that economic variables may equally well be important in the shaping of personality