Options
Naming the Unnamable: Socionics or the sociological turn of/to distributed artificial intelligence
Citation Link: https://doi.org/10.15480/882.112
Publikationstyp
Working Paper
Date Issued
2000-01
Sprache
English
Author(s)
Institut
TORE-DOI
Number in series
2
The roots of socionics stretch back to the late seventies and early eighties when computer scientists on the lookout for new methods and techniques for distributed and coordinated problem-solving began to take an interest in social metaphors and human society. In the course of their explorations they made contact with some sociologists, struck up a dialogue and soon found themselves, to their astonishment, involved in unexpected and strange avenues of research (Strübing 1998) into an unknown territory outside the confines of what Th.S.Kuhn used to call the normal sciences. Some years later they brought out a “white paper” on coordinated problem-solving in sociocomputational systems showing the need for further research and bearing the ominous title “The Unnamable” (Bendifallah et al. 1988). We have called this area of research, which was indeed then nameless, “socionics” (Malsch et al. 1996, Malsch 1997, Malsch 1998a, Müller et al. 1998).
Socionics is a new field of research, a kind of tertium quid between sociology and distributed artificial intelligence (DAI). Using an approach similar to that adopted by bionics in which biological phenomena serve as a source of inspiration for new technologies, socionics seeks to address the question how to exploit models from the social world for the development of intelligent computer technologies, specifically multiagent systems (MAS). To discover the borderland between sociology and DAI means to pursue the following questions: In what exactly do the characteristics of modern society consist; what makes social systems so resilient, adaptable and innovative; how may these features of modern society be translated into intelligent computer technologies; and what is the impact of sociology-based technologies on society? This set of questions has very much in common, but is by no means identical, with DAI research or with DAI-based social simulation. Instead, socionics is essentially addressed to the conceptual apparatus by which sociologists seek to observe, describe and explain
modern society and from here - and only from here - it tries to build the bridge to the multiagent systems of DAI. It is an invitation to an unusual Gedankenexperiment where sociologists are requested to read multiagent technology as though it were a sociological text (Woolgar 1991), where computer scientists are asked to read sociological theory as though it were a technological design and and where both groups are required to familiarize themselves with the paradox of agent societies „out of control“.
Thus, three different tenets are writ large on the agenda: (1) The first concerns the use of computer models in sociological theorizing and deals with the constitution of social order and the dynamics of social transformation. Here the claims of socionics must be substantiated in the arena of „sociological reference“. (2) The second – the lynchpin of the socionics research enterprise - is concerned with the development of new techniques and methods in DAI and investigates the role of sociological foundations in the construction of large-scale multiagent systems. Here the rules of the game are the criteria of „computational reference“. (3) The third examines the social impact of hybrid artificial societies composed of both human beings and technical agents – with possibly far-reaching consequences for our own human self-image and our very existence as social beings. And here socionics must assert itself against standards of „praxis reference“. The central issue, however, is whether and how socionics will be capable of transforming sociological theories, and not just social metaphors or naive theories of sociality, into new technological potentials.
Last no least a note of caution must be made to the reader: What follows are programmatic reflections from a sociological perspective rather than research results confirmed by both disciplines; many questions are raised but remain unanswered; and where a more elaborated argumentation should be expected often only a rough outline or a tentative explication is given; and, of course, the persepective on socionics given in this paper is not the only possible one. However, introducing an unusual topic justifies a programmatic presentation of the basic ideas and the general scope of the new research enterprise.
Socionics is a new field of research, a kind of tertium quid between sociology and distributed artificial intelligence (DAI). Using an approach similar to that adopted by bionics in which biological phenomena serve as a source of inspiration for new technologies, socionics seeks to address the question how to exploit models from the social world for the development of intelligent computer technologies, specifically multiagent systems (MAS). To discover the borderland between sociology and DAI means to pursue the following questions: In what exactly do the characteristics of modern society consist; what makes social systems so resilient, adaptable and innovative; how may these features of modern society be translated into intelligent computer technologies; and what is the impact of sociology-based technologies on society? This set of questions has very much in common, but is by no means identical, with DAI research or with DAI-based social simulation. Instead, socionics is essentially addressed to the conceptual apparatus by which sociologists seek to observe, describe and explain
modern society and from here - and only from here - it tries to build the bridge to the multiagent systems of DAI. It is an invitation to an unusual Gedankenexperiment where sociologists are requested to read multiagent technology as though it were a sociological text (Woolgar 1991), where computer scientists are asked to read sociological theory as though it were a technological design and and where both groups are required to familiarize themselves with the paradox of agent societies „out of control“.
Thus, three different tenets are writ large on the agenda: (1) The first concerns the use of computer models in sociological theorizing and deals with the constitution of social order and the dynamics of social transformation. Here the claims of socionics must be substantiated in the arena of „sociological reference“. (2) The second – the lynchpin of the socionics research enterprise - is concerned with the development of new techniques and methods in DAI and investigates the role of sociological foundations in the construction of large-scale multiagent systems. Here the rules of the game are the criteria of „computational reference“. (3) The third examines the social impact of hybrid artificial societies composed of both human beings and technical agents – with possibly far-reaching consequences for our own human self-image and our very existence as social beings. And here socionics must assert itself against standards of „praxis reference“. The central issue, however, is whether and how socionics will be capable of transforming sociological theories, and not just social metaphors or naive theories of sociality, into new technological potentials.
Last no least a note of caution must be made to the reader: What follows are programmatic reflections from a sociological perspective rather than research results confirmed by both disciplines; many questions are raised but remain unanswered; and where a more elaborated argumentation should be expected often only a rough outline or a tentative explication is given; and, of course, the persepective on socionics given in this paper is not the only possible one. However, introducing an unusual topic justifies a programmatic presentation of the basic ideas and the general scope of the new research enterprise.
Subjects
socionic
sociology
distributed artificial intelligence
DDC Class
300: Sozialwissenschaften, Soziologie