(Computational) Semiotic Systems

With this chapter we reach the point where the final vision of the engineering of intelligent evolutionary semiotic systems has to be realized.

As one could see in the preceding paragraphs the model of the simple but elegant genetic algorithm is quite powerful to describe many different search and optimization tasks sufficiently well. But its expressiveness is too limited for the task of modeling more elaborated 'information structures' like those needed for a more advanced phenotype of a biological system, especially in the case of animals.

Applying in the more complex cases the more advanced model of learning classifier systems can help to a certain degree. But even this more complex model is far to simple to allow the modeling of really interesting cases.

As a next 'stage' in the complexity of systems we can conceive so called semiotic systems. These are systems which can handle 'signs' according to their 'meaning' and according to their 'practical usage' within a 'community of sign users'. While the dimension of the 'community of sign users' is an important aspect of semiotics we will postpone the explicit discussion of communities to the next chapter of distributed cognition. In this chapter we will focus on the central concept of a 'sign' and of individual structures of a system handling signs 'appropriately'.