Chaos Theory as a New Paradigm in Psychoanalysis a Contribution to the Discussion of Models

This panel addressed chaos theory and, more specifically, non-random chaos theory. Since this is an area not familiar to many psycho-analysts, a significant portion of the panel was devoted to an exploration of a number of the fundamental principles of this idea. Chaos theory is to be distinguished from general systems theory, although they are in many ways consistent. General systems theory is derived from biology while chaos theory is derived from mathematical and computer models. Chaos theory is also differentiated from catastrophe theory although it may have been partially derived from it. Following this exposition the panel shifted towards efforts to demonstrate the clinical utility of chaos theory (Savage-Scharff), especially with application to object-relational approaches, and with an even more specific focus on transitional states (Quinodoz, Pragier). An overriding question of the panel was whether or not chaos theory provides a new paradigm for psychoanalysis.

The language and concepts utilised in chaos theory are substantially different from those in traditional psychoanalysis so a number of the basic principles were explored in some detail. Eight were specifically identified (Scharff).

1. Complex, dynamic systems are characterised by continuous feedback. Such systems can be described mathematically with iterated equations. When an iterated equation is solved, the answer is the starting point for the next run of the equation. Thus iterated systems always begin the next cycle at a place determined by the solution of the previous cycle. All biological life systems work as iterated systems.

2. In the reporter's view, this is a particularly key concept: that of sensitive dependence on initial conditions (SDIS). This is sometimes referred to as the ‘butterfly principle’. This was described by Lorenz, a weather researcher, who found that computer simulation of long-term weather forecasting yielded completely different outcomes when the program was run repetitively. In e