The contribution of the computational methodology to the empirical investigation of complex systems

摘要

The framework of this article is the research into the methodology of complex systems' investigation, as defined in the complex epistemology developed by Edgar MORIN. A complex system is a meta-stable dynamical phenomenon implying many entities. It is produced and continuously maintained by circular processes. The circularity of those processes generates emerging properties among which the first is self-regulation and consequently a certain level of autonomy. Complex systems are very common phenomena in all day's life, but they remain unusual science objects. An overview of some influencing analytic and synthetic viewpoints in science and philosophy allows us presenting a few reasons of their inadequacy for dealing with complex systems. Empirical and social sciences have been especially subject to meet practical and theoretical difficulties in their investigations of such systems. We present a methodology where the investigation of a complex system is conducted progressively using theoretical and empirical knowledge about the real system, in conjunction with the construction of the simulation model and its results. That way, computerised simulation allows conducting further the hypothetical-deductive procedure and thus increasing the potential theoretical development of the empirical sciences dealing with complex systems. We call this procedure the computational methodology.