Search

This article proposes a theory to explain how actors in social exchange networks reach agreements on division of resources. Social exchange theory grew. from the application of the economic theory of exchange to social relationships. Sociology focuses on a problematic area for economic theory, the exchange of valued objects in, relatively small groups, where actors seek to settle on one optimal outcome out of a range of possibilities. Central to theoretical development today is a class of networks in which subtle power differences occur. This phenomenon is known as weak power. In these weak power networks, some positions may have advantages over others in acquiring resources through exchange. However, unlike the advantages in strong power networks, advantages in weak power networks are not progressive. Over a series of exchanges, a strong power advantage eventually results in one, exchange partner receiving nearly all available resources. Weak power is limited in range and magnitude. This necessitates theoretical refinement because adequate assessment of power differences between positions in weak power structures requires more, precise predictions of exchange rates at equilibrium.

Production systems are ordered sets of if-then rules linking knowledge and action. The authors show how these systems, developed and applied in artificial intelligence and cognitive science, can be used to construct empirically based symbolic interactional models of situated human interaction. The authors argue that production system models enhance the conceptual clarity and formal analytic capabilities of key ideas in Herben Blumer's methodological directives for sociology's symbolic interactionist perspective. The authors present both a general discussion and outline of production system model building and a case study in which they evaluate a model for the detailed analysis of customer-waitperson interactions in a restaurant setting. [ABSTRACT FROM AUTHOR]

Narrative analysis is applied to episodes of interaction generated by coupled production systems con- trolling the activity of different processing units. Formal analysis of such coupled systems, conceived of as models of institutionalized social action, has shown that they generate context-free languages of interaction. This note suggests that narrative analysis of interaction strings in such languages has limited analytical utility: it offers neither a method to discriminate between well-formed and malformed strings of interaction nor a procedure to induce the underlying generative rule structure from a corpus of well-formed interaction strings. Therefore, the narrative method holds hale promise as a technique to examine the routinized activity composing a social institution. [ABSTRACT FROM AUTHOR]