Information Technology Reference
In-Depth Information
Complex Neuro-Cognitive Systems
Andreas Schierwagen
Institute for Computer Science, Intelligent Systems Department,
University of Leipzig, Leipzig, Germany
schierwa@informatik.uni-leipzig.de
http://www.informatik.uni-leipzig.de/~schierwa
Abstract.
Cognitive functions such as a perception, thinking and acting
are based on the working of the brain, one of the most complex systems
we know. The traditional scientific methodology, however, has proved to
be not sucient to understand the relation between brain and cognition.
The aim of this paper is to review an alternative methodology - non-
linear dynamical analysis - and to demonstrate its benefit for cognitive
neuroscience in cases when the usual reductionist method fails.
1
Introduction
Cognitive science aims at both understanding natural cognition (as in humans
or animals) and creating artificial systems resembling the natural original. There
are basically two established ways of doing research in cognitive science (includ-
ing cognitive and computational neuroscience), either top-down or bottom-up.
Both approaches have their pro's and con's, and none is practised in the pure
form. For example, the main research strategy in computational neuroscience
- reverse engineering - is basically bottom-up. It is informed, however, by top-
down considerations about the goal of the computation performed by the neural
system under study. It is expected that united efforts of this kind will succeed in
providing a theory of cognition that is ultimately grounded in brain processes.
However, objections have been also raised saying that the established methods
do not meet the complexity of the object of study, and that the research method-
olgy must be complemented accordingly. In this paper, I'll first describe briefly
the two traditional ways of doing cognitive (neuro-)science that are commonly
thought to exhaust the possibilities. Then a third way of analysis - nonlinear
dynamical analysis - is described that relies on results of complexity research,
i.e. chaos theory and nonlinear modeling.
2 Top-Down and Bottom-Up Approaches
The top-down approach consists of (1) specifying a cognitive function by focus-
ing on the characterization of the abstract principles that underlie that function.
Ideally, it proposes (2) possible neural algorithms that might subserve this cogni-
tive function, and finally (3) maps these algorithms onto brain circuits. In many
