Tschacher, W and Dauwalder, JP (eds) (1999): Dynamics, Synergetics, Autonomous Agents. Nonlinear Systems Approaches to Cognitive Psychology and Cognitive Science. Singapore: World Scientific.
Contents of Dynamics, Synergetics, Autonomous Agents
From the Introduction to Dynamics, Synergetics, Autonomous Agents
Cognition is at the heart of both psychology and computer science. Not too long ago, psychologists would have had considerable doubts about this statement—are not behavior and action closer to the scientific core of their discipline? Computer science has not been prone to such ambivalence, provided that cognitive processes were understood as the processing of symbolic tokens ready to be implemented on any material substrate. This situation underwent an ironic change in recent decades. Psychology was deeply impressed by the artificial intelligence engineers’ success in performing information processing on machines, and the mainstream of psychology even adopted a computational foundation for cognitive psychology. But almost at the same time, computer scientists grew increasingly disenchanted with their approach towards intelligence. Thus, while psychology turned more and more computationalist, computer science had, in a way, become more psychological, and had started to consider concepts such as embodiment, semantic grounding of symbols and the different facets of the mind-body problem.
Our intention with this volume is to help synchronize and integrate the progress made in both cognitive psychology and new artificial intelligence. A guiding idea is that complexity theory and dynamics may considerably facilitate this endeavour. The dynamical approach introduces the language and method of dynamical systems theory—with key concepts such as attractor, bifurcation, etc—to cognition research. Complexity theory, in the shapes of self-organization theory and synergetics, adds to this method by considering evolution, i.e. the emergence of patterns in open, nonlinear systems. Order may evolve from the cooperation, the ‘synergy’ of microscopic components of a system.
The concept of autonomous agents put forward by designers of robots has many parallels with this dynamical approach. Autonomous agents can never be completely pre-programmed so that much thought must be invested on their embeddedness in an environment. The question arises of how stable and adaptive behavior can grow out of this interaction between agent and environment. This may be rephrased in a synergetic manner to address the interaction between a complex open system and environmental control parameters, which results in pattern formation. In this way, the models of synergetics can be applied to autonomous agents. In turn, the problem of autonomous agents is obviously analogous to the problem of action control and volition. This is why psychology is directly addressed, too.
This volume—in negotiating the topic of cognition between the perspectives of cognitive psychology, AI, and dynamics—is divided into four sections. The first of these, ‘Theory and Concepts’, is initiated by Hermann Haken’s treatment of how perception, cognition and decision-making may be modeled by synergetics. Tim van Gelder disputes the various objections that may be made to the ‘dynamical hypothesis’ in cognitive science. Herbert Jaeger proposes the concept of ‘dynamical symbols’, i.e. dynamical events that have the qualities of symbols. The contribution of John Shiner, Matt Davison, and Peter Landsberg centers on methodology; they investigate an algorithm to assess order (e.g. in psychological data) which circumvents the problems of complexity measures. Helena Knyazeva and Hermann Haken explore the heuristic value of synergetics by applying it to human creativity and innovative thinking.
The second section, ‘Dynamical Concepts in Cognitive Psychology’, gives an account of cognitive dynamics as viewed from the perspective of psychology. Wolfgang Tschacher and Jean-Pierre Dauwalder introduce this section by pointing out the shortcomings of the standard approaches of information processing and action theory, and by suggesting a dynamical, situated model. The contribution of Richard Eiser puts forward an alternative to core concepts in social psychology, such as attitude and self, by advocating a self-organized systems view. Luc Ciompi argues for an emphasis on the significance of affects in mental functioning, thereby laying the foundation for the approach of ‘fractal affect-logic’. Thomas Bröcker and Jürgen Kriz elaborate Bartlett’s and Piaget’s schema theories in terms of synergetics. Taking a developmental perspective, Fernand Gobet is also concerned with the Piagetian concepts of assimilation and accommodation, which he then models in a symbolic framework.
The third section concerns ‘Autonomous Agents’. Christian Scheier and Rolf Pfeifer outline their research program of ‘embodied cognitive science’ by elaborating principles for the design of artificial autonomous agents. Kazuo Hiraki, Akio Sashima and Steven Phillips investigate how self-locomotion interacts with the development of spatial cognition in a robot experiment. Norbert Glaser and Philippe Morignot lay the conceptual foundation for the study of societies of autonomous agents. Kerstin Dautenhahn emphasizes the importance of a concept of the body—a body image—in social agents. William Sulis presents a conceptualization of ‘collective intelligence’, i.e. adaptive behavior found in swarms or societies of simple autonomous agents.
The concluding fourth section adds (further) ‘Empirical Studies’. Marc Coulson and Stephen Nunn test hypotheses of the nonlinearity of cognition (so-called ‘catastrophe flags’) in experiments on decision-making. Fred Cummins applies the Haken-Kelso-Bunz model of rhythmic coordination to human speech.Ferdinand Keller, Maja Storch and Susanne Bigler operationalize personality changes induced by Jungian psychological intervention using a time series approach. Stephen Guastello studies work behavior in hierarchical organizations in the framework of game theory and chaos theory.
A major incentive for this volume came from a conference we organized in Gstaad, Switzerland, in March 1997. Several of the contributions are based on lectures given in Gstaad, others have been written in the wake of the conference especially for the scope of this book.
The Gstaad conference was seventh in the series of Herbstakademie symposia that are targeted at elucidating the relationship between self-organization theory and psychology. Aside from providing a platform for the ongoing Herbstakademie discussions, which traditionally are attended by German-speaking researchers, the Gstaad symposium was also the ‘First Joint Conference on Complex Systems in Psychology’ because of our collaboration with the Society for Chaos Theory in Psychology and the Life Sciences. (...)
Bern, December 1998
Wolfgang Tschacher, Jean-Pierre Dauwalder