Comments on the cybernetics of stability and regulation in social systems

Abstract

The methods and principles of cybernetics are applied to a discussion of stability and regulation in social systems taking a global viewpoint. The fundamental but still classical notion of stability as applied to homeostatic and ultrastable systems is discussed, with a particular reference to a specific well-studied example of a closed social group (the Tsembaga studied by Roy Rappaport in New Guinea).

The discussion extends to the problem of evolution in large systems and the question of regulating evolution is addressed without special qualifications. A more comprehensive idea of stability is introduced as the argument turns to the problem of evolution for viability in general.

Concepts pertaining to the problem of evolution are exemplified by a computer simulation model of an abstractly defined ecosystem in which various dynamic processes occur allowing the study of adaptive and evolutionary behaviour. In particular, the role of coalition formation and cooperative behaviour is stressed as a key factor in the evolution of complexity.

The model consists of a population of several species of dimensionless automata inhabiting a geometrically defined environment in which a commodity essential for metabolic requirements (food) appears. Automata can sense properties of their environment, move about it, compete for food, reproduce or combine into coalitions thus forming new and more complex species. Each species is associated with a specific genotype from which the species’ behavioural characteristics (its phenotype) are derived. Complexity and survival efficiency of species increases through coalition formation, an event which occurs when automata are faced with an “undecidable” situation that is resolvable only by forming a new and more complex organization.

Exogenous manipulation of the food distribution pattern and other critical factors produces different environmental conditions resulting in different behaviour patterns of automata and in different evolutionary “pathways.”

Eve-1, the computer program developed to implement this model, accepts a high-level command language which allows for the setting of parameters, definition of initial configurations, and control of output formats. Results of simulation are produced graphically and include various pertinent tables. The program was given a modular hierarchical structure which allows easy generation of new versions incorporating different sets of rules.

The model strives to capture the essence of the evolution of complexity viewed as a general process rather than to describe the evolution of a particular “real” system. In this respect it is not context-specific, and the behaviours which are observable in different runs can receive various interpretation depending on specific identifications. Of these, biological, ecological, and sociological interpretations are the most obvious and the latter, in particular, is stressed.