Self-referential systems theory does not provide for a concept of intelligence. There is even a certain resistance to intelligence that seems to block any explicit exchange of concepts with artificial systems theory. The paper describes the intelligence service in self-referential systems as the self-referential and, hence, paradoxical switching from the self-reference of these systems to other-reference. How this might work is shown by means of G. Spencer Brown’s calculus of indications and Heinz von Foerster’s notion of double closure.

Key words: george spencer-brown, heinz von foerster, intelligence, self-reference, systems

The concept of self-organization is reviewed and its implications are explored in relation to management processes and social systems. A world view is taken, emphasizing a descriptive distinction of levels associated with the physical, biological, social, and mental. Self-organization principles, it is argued, are operative in all levels of such a stratified scheme, but they are manifest in different mechanisms and different embodiments. \\Management, planning, design, and other “intervention” type of activities are among the processes through which self-organization is manifest in the social domain. Ultimately they have to do with maintaining, enriching, and amplifying the potential variety of the systems concerned. The operationally critical question involved, it is suggested, is not whether management activities are “man-made” or “natural,” spontaneous” or “planned,” but rather, whether they enhance or supress the potential variety of a system under consideration.

Self-reference and recursion characterize a vast range of dynamic phenomena, particularly biological automata. In this paper we investigate the dynamics of self-referent phenomena using the Extended Calculus of Indications (ECI) of Kauffman and Varela, who have applied the ECI to mathematics, physics, linguistics, perception, and cognition. Previous studies have focused on the algebraic structure of the ECI, and on form dynamics using only the arithmetic of Spencer-Brown. We here examine the temporal behavior of self-referent or reentrant forms using the full power of the ECI to represent tangled hierarchies and multiple enfolded dimensions of space-time. Further, we explore the temporal convolution of static and recursive states in coherent fluctuation, providing a foundation for going beyond the Turing model of computation in finite automata. Novel results are presented on the structure of reentrant forms and the canonical elements of form eigenbehavior, the characteristic self-determined dynamic inherent in reentrant forms.

A problem of action selection emerges in complex and even not so complex interactive agents: what to do next? The problem of action selection occurs equally for natural and for artificial agents for any embodied agent. The obvious solution to this problem constitutes a form of representation, interactive representation, that is arguably the fundamental form of representation. More carefully, interactive representation satisfies a criterion for representation that no other model of representation in the literature can satisfy or even attempts to address: the possibility of systemdetectable representational error. It also resolves and avoids myriad other problematics of representation and integrates or opens the door to many additional mental processes and phenomena, such as motivation.

The aim of the paper is: a) to gain some knowledge of the so-called ‘natural systems’ as interpreted or defined by modern systems scientists; b) to discuss these descriptions and definitions from the viewpoint of modern philosophy of science. In the course of both a) and b) the interwovenness of the classes of natural systems and the controversial issues connected therewith (a.o. their interwovenness with the artificial systems) will be touched upon.

This paper discusses the phenomenological dimension of social understanding. The author’s general hypothesis is that complex forms of social understanding that biological agents especially humans show are based on two mechanisms: 1 the bodily, experiential dynamics of emphatic resonance and 2 the biographic reconstruction of a communication situation. The latter requires the agent’s bodily experiences as the point of reference for the reconstruction process. This hypothesis is derived from discussions in philosophy, natural sciences, and cognitive science on the social embodiment of cognition and understanding. Evidence comes from studies on social cognition in primates, infants, and autistic people that are interpreted in terms of the “mind-experiencing” hypothesis. The second part of the paper sketches an '‘interactive’’ experiment that investigates the dynamic coupling of a robot with its environment. This example is used to discuss the role of the human observer and designer as an active, embodied agent who is biased toward interpreting the world in terms of intentionality and explanation. The paper describes how this aspect can influence the processes of understanding and interpretation of the behavior of autonomous robotic agents. The author concludes by stressing the need to overcome the distinction between computationalism and phenomenology in order to develop complex artificial systems.

How is the field of systems science different from other scientific fields, and how can we distinguish the various traditions within systems science? We propose that there is a set of underlying assumptions which are generally shared within systems science but are less common in other scientific fields. Furthermore, the various traditions within systems science have adopted different combinations of these assumptions. We examine six traditions within systems science – cybernetics, operations research, general systems theory, system dynamics, total quality management, and organizational learning. We then consider eight underlying assumptions – observation, causality, reflexivity, self-organization, determinism, environment, relationships, and holism. We then assess where each tradition stands with respect to each of the underlying

The pervasiveness of metaphor in our conceptual system suggests a central and basic role in the underlying architecture of thought. Metaphor represents the ability to understand one thing in terms of another as we ascribe an understood pattern to an unknown phenomena and perceive their structural integrity within the environment of our experience. We can then begin to perceive the environment of learning as one in which analogical thinking serves as architecture, analytical thinking serves as engineering and the imagination ensures that the interactions which create life and meaning are always being realized anew. The implications for this approach to applied epistemology provides insight into the design and development of learning systems that support the creative nature of learning.

This paper is primarily concerned with answering two questions: What are necessary elements of embodied architectures? How are we to proceed in a science of embodied systems? Autonomous agents, more specifically cogni- tive agents, are offered as the appropriate objects of study for embodied AI. The necessary elements of the architectures of these agents are then those of embodied AI as well. A concrete proposal is presented as to how to proceed with such a study. This proposal includes a synergistic parallel employment of an engineering approach and a scientific approach. It also supports the exploration of design space and of niche space. A general architecture for a cognitive agent is outlined and discussed.

The paper discusses aspects of a project that strives to base an understanding of what economics call “productivity” on a complexity theoretic foundation. The core thesis of this project is that productivity can best be grasped by referring to two features commonly associated with knowledge – “nonreducibility in consumption” and “time preference”. The paper in hand focuses on theoretical aspects concerning the ontological status of the observer in defining productivity alongside these features. Oriented on the Theory of Social Systems by Niklas Luhmann and methodologically drawing on Multi-Agent-Simulation, it investigates the thesis that the observer itself – circularly – can be conceptualized as a consequence of these two features.