A Creative Dance: Symbols, Action and the Bringing Forth of Meaning

In our attempts to understand the evolution of biological, cognitive and cultural systems, critical questions arise concerning the origin of meaning. I argue that the key to success in attempts to create computational systems that exhibit the same capacities as their natural counterparts to evolve new and creative ways of interacting with their environment, beyond that which is simply “programmed into” the system from the start, lies in answering these questions. The nature of the problem is laid bare when we consider the origin and evolution of life. A fundamental question is this: how is it possible for organisms, that follow their own goals and behave according to their own rules, to emerge in a world governed by the laws of physics and chemistry? More generally, how can agents and agency emerge in a system governed by universal laws? And even once our agents have emerged, how can the evolutionary process produce new agents that interact with their environment through previously unexploited modalities? That is, how can new sensors and actuators emerge (such as the origin of a magnetic sense in migratory birds, or the ability of some species of fish to generate electric fields), where the modalities upon which they act were not exploited in earlier species? This second question may perhaps have more significance for attempts to build artificial creative systems; in the biological world the solution is largely related to the fact that the organs and structures that comprise organisms possess multi-functional properties across different modalities (e.g. strength, elasticity, electrical conductivity, sensitivity to light, etc.), and that a structure evolved for one function may turn out (initially by accident) to have beneficial properties in a different modality which may then be exploited by evolution (a phenomenon known variously in the biological literature as exaptation, cooption or preadaptation). However, when we create artificial systems, particularly software systems, the building blocks of the system are not generally multi-functional, and this is one of the reasons why the evolution of new sensors and actuators has become an important issue in this context.1 1See (Dautenhahn et al., 2001) for a special issue of the ArtifiHowever, the solution to the problem of evolving new forms of interaction requires more than just using multifunctional building blocks. To make progress in this area, and on the more fundamental question of the origin of agency, requires a shift of focus away from the properties of individual agents, to consider the dynamic nature of the environment and the relationship between agents and environment. In a previous paper I set out an argument for modelling agents and environment as a single dynamical system (Taylor, 2004) [see also (Taylor, 2001)]. The argument was influenced by the biosemiotics literature, and, in particular, the work of Howard Pattee; biosemiotics offers important insights for addressing these issues.2 Pattee argues that the distinction between the material and symbolic aspects of living organisms, seen as an example of the more general epistemological distinction between laws and initial conditions, is a defining feature of life, and also a necessary condition for open-ended evolution (Pattee, 1995a; Pattee, 1995b). He explains the relationship between the two as follows: