2006-06-02

Non-reducible systems

Some thoughts have come recently... about why we cannot reduce the behaviour of complex systems to the behaviour of their components. There are reducible systems where you can do this, e.g. where properties of a system are given by the average or sum of the properties of the parts. But, where does the emergent "more than the sum" comes from?
The answer is:Interactions.
When two elements interact, new information can be produced. This information is not included in the information of the elements. Thus, it cannot be predicted from that. The system needs to "run" to express its properties, generated only when elements interact. Emergent properties are a posteriori.
Because of this, reductionism is doomed. It simply cannot predict the precise information that will be produced by interactions that cannot be specified before they are observed. They need to be observed first to be described and understood. Feedback between experience and reason.

Post a Comment