48 SOCIAL SCIENTIST The following analysis of this framework's initial advances proceeds in two steps. First, social structures of accumulation (SSA) theory is presented in its own terms, with emphasis on tensions and problems which, as I argue, the effort to translate social relationships into variables defined in power terms produces. In a second stage of criticism, the framework will te briefly examined from the stand-point of an explicitly suggested alternative conception of Marxian theory. This second step will question the basic conceptions of power and agent behavior on which the framework builds. A SOCIAL APPROACH TO ACCUMULATION THEORY For a thorough presentation of the concept of a social structure of accumulation, Bowles, Gordon, and Weisskopf (1988, 9)2 refer their readers to an earlier work, Gordon, Edwards, and Reich's Segmented Work, Divided Workers (1982). The general definition offered there bears quotation at length, for it will help to orient the analysis of more recent developments which follows. By social structure of accumulation we mean the specific institutional environment within which the capitalist accumulation process is organized. Such accumulation occurs within concrete historical structures: in firms buying inputs in one set of markets, producing goods and services, and selling those outputs in other markets. These structures are surrounded by others that impinge upon the capitalist accumulation process: the monetary and credit system, the pattern of state involvement in the economy, the character of class conflict, and so forth. We call this collective set of institutions the social structure of accumulation (Gordon, Edwards, and Reich 1982, 9). A social structure of accumulation is then defined by the set of institutions which either directly 'organize* the making and reinvestment of profits, or more generally 'impinge upon' or modulate those processes. Gordon, Edwards, and Reich make one motivation for taking this as their starting point quite clear: it reacts against problems thought to characterize the abstractly fixed 'inner logics' constructed by hitherto dominant interpretations of Marxian economic theory. Indeed Gordon, Edwards, and Reich criticize their own prior use of monopoly capitalism theory's stage-theoretic approach to capitalist development. By tying its argument so closely to product market structure, they now contend, that approach operates on too abstract a level and fails to 'capture the breadth and complexity of the capital accumulation process' (Gordon, et al. 1982, 22). Bound up with this drawback is a problem common to Marxian and orthodox traditions. While both recognize that investment is dependent upon expectations, they have nevertheless 'tended either to elide the importance of the