Computational General System Theory

One of the first and essential basics for a general system theory is a scheme for all objects of reality that are of scientific interest and should therefore be able to be analyzed. In (Oppenheim 1958) Paul Oppenheim and Hilary Putnam did recommend their so called “Reductive Levels” which were six levels of scientific universes of discourse together with an according set of conditions. This level-scheme did become quite famous in the area of system thinking.

Reductive levels model of Oppenheim and Putnam (left) and the adapted CGST-Reductive-Levels.

We will take this scheme as basis but adapt it on the higher levels for several reasons: Level one to five are consistent regarding to their composition (of related elements) - always the instances of an lower level are the elements of the next higher level and additionally they are the essential elements that influence these systems most. The step from level five to six is not consistent here, because social groups are only one kind of subsystem that can be formed by living things. Furthermore system thinking does not stop at the level of social groups but can be also applied to global systems as whole and astronomical systems too.

Of course social groups are one of the most interesting areas of research work - not least because of the crisis of the social sciences according to a common theory. But just because of this reason it is important to put them into the right context. Niklas Luhmann’s theory of social systems became quite famous focusing on the difference between social systems and their environment and implicitly also uses a scheme with social groups as level of highest complexity (of course Luhmann does not use levels but declares all levels beyond the level of social groups as “environment”). In his most important book (Luhmann 1969) he argues for orienting on a general system theory when designing a theory of social system but against drawing (direct) analogies between social and biological systems for example. Luhmann’s motivation for his approach is to develop something special for social sciences because here - in contrast to the natural sciences - a basic theory as a common starting point for further researches is still missing.

The special element of social systems for Luhmann roots in human communication including all the emergent effects of communication in social groups. And because also Oppenheim and Putnam did see social systems as the most complex instance of system science they introduced a special level for it.  

But having a closer look on the reductive levels scheme, we can find a similar case also on a lower level: The human brain as an organ consists of a lot of cells and therefore has to be put somewhere above level four. But although it is very complex and there are a lot of emergent effects that can not be explained today, it is not a living system of its own but can only exist as a subsystem within the “system” human being. And because of this reason it is also influenced by other subsystems (other organs for example) on the same reductive level.

When Luhmann speaks about social systems he only includes communication and self-reference of the systems into this term - human psyche but also all kind of ecological influences are only “irritations” and do not belong to any social system. To explain the rules and character of mass media as in (Luhmann 1996) this might be useful, but for all kind of complex issues where also lower reductive levels have huge effects on the behaviour of the system, the method of producing a lot of different environments and the missing concept of connecting them together, makes such analyzing procedures too complex for practical applications.

An example would be the appearance of an epidemic: Mass panic and other complex social effects could be analyzed using Luhmann’s theory of social systems. But we would have to create a lot of different environments to map all included effects to our model: The biological system with the behaviour of bacteria and virus elements, its dissemination criteria and its effects on the human body; the climate parameters of a special region and its influence on the pathogen elements; traffic dates about a special region and information about communication structure; and much more.    

And it becomes even more complex if we look at a very popular scientific topic of today’s global research work: Although its not proven, how these effects look like in detail, it is quite sure that human actions have effects on the climate characteristics and atmosphere of our planet. And vice versa climate changes influence human societies. Using an approach with focus on social systems and declaring all other effects as “environment” we will not have any chance to find a practical model to handle this topic. And at this point it gets clear that social systems are not the most complex ones but are subsystems of the true higher level of reduction: ecological systems in general and (in this case) the global ecological system including the atmosphere and all other elements of the whole planet earth. In (Lovelock 1979) James Lovelock presented his so called Gaia-Hypothesis: Planet earth must be seen as a complex living system that - like all other complex systems - has a mechanism of regulation to keep a status of homoeostasis. The planet itself for Lovelock acts as an instance of a living system including emergent effects that occur on a higher level than the level of human (social) groups, least because human societies can be interpreted as a subsystem.      

Except for this adaptation we will take the set of rules for the scheme from (Oppenheim 1958) with some further changes. The original rules are: (1) There must be several levels. (2) The number of levels must be finite. (3) There must be a unique lowest level; this means that success at transforming all the potential micro-reductions connecting these branches into actual mirco-reduction must, ipso facto, mean reduction to a single branch. (4) Any thing of any level except the lowest most possess a decomposition into things belonging to the next lower level. In this sense each level, will be as it were a “common denominator” for the level immediately above it. (5) Nothing on any level should have a part on any higher level. (6) The levels must be selected in a way which is “natural” and justifiable from the standpoint of present-day empirical science.

(1) and (2) are trivial and we will keep it therefore. We will also keep (3), because defining a lowest level makes sense for further work, although it includes a potential for conflicts in scientific discourse because there are a lot of open questions in the area of subatomic research (maybe more than there have been in 1956). (5) is important for our definition of reductive levels too and so we will keep it.

(4) is incomplete according to potential strong  emergent effects. Such effects are defined in (Stephan 1998) as systemic effects that can not be reduced to the properties and behaviours of the elements of the system and its relations at all.  (6) is no real revisable rule because its evaluation always depends on the point of view. Of course it makes sense in a general way and as a kind of hint it might be important for further developments - but because it cannot be formulated clearly, we will drop it.