Wolfgang Krischke (1999)

Journal of Artificial Societies and Social Simulation vol. 2, no. 3, <https://www.jasss.org/2/3/9.html>

Editors' preface: The following article appeared in the Humanities section of Frankfurter Allgemeine Zeitung (no. 119, May 26, 1999, p. N5), one of the leading German newspapers. It reviews a number of papers published in JASSS, among them (though not explicitly quoted) the JASSS Forum on socionics, in volume 1, issue 3. The article was translated by Klaus G. Troitzsch.

1.1

Dozens of yellow dots move around the screen. Some unite and grow into large clusters, others attack their neighbours or vanish and are not seen any more. They are electronic "actors" (in German: "Akteure") fighting for food, reproducing or dying, trading, forming coalitions, hierarchies and rules for living together --- processes of an archaic society simulated by a computer. "Breeding" such artificial creations is part of socionics, a research approach which melds sociology and computer science. Socionics aims to trace social mechanisms by testing social science theory and hypotheses by means of simulation.

1.2

The germ of an artificial society is a computer program which fixes the initial conditions. Its citizens are intentional agents -- developed from the programs based on search engines on the Internet or office assistants on a PC -- which operate as if they had intentions and goals. In the same way as "classical", cognition-oriented artificial intelligence research (AI) breathed life into old debates about mind and consciousness, "distributed artificial intelligence" (DAI), which defines intelligence socially, seems to revive the interest in the fundamental question of sociology: How does an aggregation of individuals bring about the complex system of hierarchies, conventions and values which makes up society --- that powerful "something" of which Fontane's Baron von Instetten speaks on the eve of the duel. Thus socionics people analyse the role of institutions and their impact on the agents' behaviour by means of computer experiments (Jose Castro Caldas, Helder Coelho: The origins of institutions: socio-economic processes, choice, norms and conventions, JASSS 2/2/1, 1999).

1.3

The trust of many theoreticians in the "invisible hand" by which society, if it is left alone, develops automatically to everyone's advantage, was not justified. To be sure, institutions and conventions emerged from the actors' interplay, i.e. without outside influence, but they did not lead the electronic societies into a "natural equilibrium", but into coercion and extreme inequality. Simulations which test the effects of elementary property norms, such as the principle that everyone is allowed to retain what they find, come to similar results. In simply modelled gatherer and hunter societies this norm indeed prevents aggression and guarantees social harmonisation. But as soon as some actors get small initial advantages, e.g. from inherited property, this norm has just the opposite effect in the long run (Andreas Harrer, Nicole Saam: Simulating Norms, Social Inequality, and Functional Change in Artificial Societies, JASSS 2/1/2, 1999).

1.4

How do norms such as honesty succeed in societies in which everyone first thinks about their own advantage? And what happens if part of a society sticks to such a norm while another part does not, and instead exploits the honesty of their co-citizens? Computer simulation shows how an important societal corrective arises: reputation. Law-abiding actors discuss among themselves. They communicate about the honesty (or lack of it) of their co-actors. Actors with bad reputations are isolated (Cristiano Castelfranchi, Rosaria Conte and Mario Paolucci: Normative Reputation and the Costs of Compliance, JASSS 1/3/3, 1998).

1.5

As in classical AI, socionics uses computers as test instruments for the logical consistency of theories. Sociologists at Hamburg-Harburg Technical University and computer scientists at Munich Technical University will compare Niklas Luhmann's systems theory to Hans Joas's action pragmatism: Is the communication between two individuals that are "black boxes" to each other and hence react towards each other in a trial-and-error manner, the initial condition for the evolution of social structures? Or is it action routines which break down and thus produce what is new in society?

1.6

Computer scientists hope for inspiration for new software in the distributed artificial intelligence area from studies of social processes such as cooperation, negotiation and exchange. A field of application is digital expert systems which analyse technical problems in large plants. Since these are composed of many modules, diagnostic programs often reach their limits when they are to analyse all possible disturbances with their branches sequentially and on one processor. Solutions are seen in the cooperation of several smaller and more specialised expert systems. The standard example is not the lonely super brain controlling everything, but a distributed, decentralised society with its differently distributed capabilities and tasks.

1.7

Socionic simulation is also used to tackle the question of the rise and fall of historical societies. The Santa Fe Institute in the USA hopes to to reveal the secret of the Anasazi, an Indian pueblo tribe which vanished in the fourteenth century, together with its thriving culture. The researchers fed their computer with archeological data and with hypotheses on climate and social change, hoping that they will find traces of the historical processes which led to the Anasazi's extinction. But it is the future which belongs to socionics --- this is what optimistic scholars in the US think. They believe that their forecasts will play a significant role in political and economic planning and decision. Will these dreams come true? Sociological computer games cannot yet represent the complexity of human societies. But socionics has already proven to be a catalyst for research.

Return to Contents of this issue

© Copyright Journal of Artificial Societies and Social Simulation, 1999