this paper I shall make a case for a dynamic semiotics. I list a set of phenomena that are difficult to understand in standard theories, and suggest a model borrowed from theories of complex dynamic systems. Since such theories rely on processes of self-organization that often defy analytical treatment, I use small computational models for assessing the empirical consequences of the theories.

this paper is just another small step leading from the old media to the new one.

In this introductory section, I shall discuss the nature of interactive multimedia, and their present mode of development. 1.1. Multimedia are documents, music or film When new media appear they are normally treated as new versions of old media (Madsen 1994): the film posed as theatre, television as filmed radio, and video as electronic film. This is also true of interactive multimedia which are often conceived of as a book or collection of documents with more flexible ways of turning pages (Apple’s Hypercard, Allegiant’s Supercard, Netscape), as a film or piece of music that can be stopped and started (Macromind Director, Apple’s Quicktime) or as a new kind of theatre (Laurel 1991). However, all applications are computer systems, and reveal themselves as such at closer inspection, so why this playing hide-and-seek with the user? Why pose as a book when one is really an object-oriented program? The reason is the paradox inherent in creating a new medium. On the one