Search

This article aims to present some of the initial work of developing a social science grounded game theory-as a clear alternative to classical game theory. Two distinct independent initiatives in Sociology are presented: One, a systems approach, social systems game theory (SGT), and the other, Erving Goffman's interactionist approach (IGT). These approaches are presented and contrasted with classical theory. They focus on the social rules, norms, roles, role relationships, and institutional arrangements, which structure and regulate human behavior. While strategic judgment and instrumental rationality play an important part in the sociological approaches, they are not a universal or dominant modality of social action determination. Rule following is considered, generally speaking, more characteristic and more general. Sociological approaches, such as those outlined in this article provide a language and conceptual tools to more adequately and effectively than the classical theory describe, model, and analyze the diversity and complexity of human interaction conditions and processes: (1) complex cognitive rule based models of the interaction situation with which actors understand and analyze their situations; (2) value complex(es) with which actors operate, often with multiple values and norms applying in interaction situations; (3) action repertoires (rule complexes) with simple and complex action alternatives-plans, programs, established (sometimes highly elaborated) algorithms, and rituals; (4) a rule complex of action determination modalities for actors to generate and/or select actions in game situations; three action modalities are considered here; each modality consists of one or more procedures or algorithms for action determination: (I) following or implementing a rule or rule complex, norm, role, ritual, or social relation; (II) selecting or choosing among given or institutionalized alternatives according to a rule or principle; and (III) constructing or adopting one or more alternatives according to a value, guideline, or set of criteria. Such determinations are often carried out collectively. The paper identifies and illustrates in a concluding table several of the key differences between classical theory and the sociological approaches on a number of dimensions relating to human agency; social structure, norms, institutions, and cultural forms; patterns of game interaction and outcomes, the conditions of cooperation and conflict, game restructuring and transformation, and empirical relevance. Sociologically based game theory, such as the contributions outlined in this article suggest a language and conceptual tools to more adequately and effectively than the classical theory describe, model, and analyze the diversity, complexity, and dynamics of human interaction conditions and processes and, therefore, promises greater empirical relevance and scientific power. An Appendix provides an elaboration of SGT, concluding that one of SGT's major contributions is the rule based conceptualization of games as socially embedded with agents in social roles and role relationships and subject to cognitive-normative and agential regulation. SGT rules and rule complexes are based on contemporary developments relating to granular computing and Artificial Intelligence in general. [ABSTRACT FROM AUTHOR]

In this paper, I investigate a connection between a common characterisation of freedom and how uncertainty is managed in a Bayesian hierarchical model. To do this, I consider a distributed factorization of a group's optimization of free energy, in which each agent is attempting to align with the group and with its own model. I show how this can lead to equilibria for groups, defined by the capacity of the model being used, essentially how many different datasets it can handle. In particular, I show that there is a "sweet spot" in the capacity of a normal model in each agent's decentralized optimization, and that this "sweet spot" corresponds to minimal free energy for the group. At the sweet spot, an agent can predict what the group will do and the group is not surprised by the agent. However, there is an asymmetry. A higher capacity model for an agent makes it harder for the individual to learn, as there are more parameters. Simultaneously, a higher capacity model for the group, implemented as a higher capacity model for each member agent, makes it easier for a group to integrate a new member. To optimize for a group of agents then requires one to make a trade-off in capacity, as each individual agent seeks to decrease capacity, but there is pressure from the group to increase capacity of all members. This pressure exists because as individual agent's capacities are reduced, so too are their abilities to model other agents, and thereby to establish pro-social behavioural patterns. I then consider a basic two-level (dual process) Bayesian model of social reasoning and a set of three parameters of capacity that are required to implement such a model. Considering these three capacities as dependent elements in a free energy minimization for a group leads to a "sweet surface" in a three-dimensional space defining the triplet of parameters that each agent must use should they hope to minimize free energy as a group. Finally, I relate these three parameters to three notions of freedom and equality in human social organization, and postulate a correspondence between freedom and model capacity. That is, models with higher capacity, have more freedom as they can interact with more datasets. [ABSTRACT FROM AUTHOR]