PARALLELISM 1)2)
← Back
The simultaneous transformations of many relationships among many elements within a system.
Such a situation implies the existence of a significant number of coexisting processes and, as a consequence a certain leeway in the behavior of the system as a whole. It is however important to understand that the very concept of system implies also a global general – but not absolutely constraining – coherence.
Parallelism is altogether the condition for adaptive variation because it reflects a relative autonomy in space and time of different local behavioral lines in the system.
J. HOLLAND defines parallelism as the "concurrent activity of many rules" (1992, p.174) and describes it as a fundamental characteristic of what he calls "classifiers", which are in fact all adaptive and evolutive complex systems, i.e. not the simple monocausal processes. Examples of this type of systems can be ecosystems, the human brain, automata endowed with adaptive behavior.
He comments: "Parallelism makes it possible for the system to combine rules into clusters that model the environment, providing two important advantages:
"1. Combinatorics work for the system instead of against it. The system builds a "picture" of the situation from parts, rather than treating it as a monolithic whole…
"2. Experience can be transferred to novel situations, such as a "red car on the side of the road with a flat tire", the system activates several relevant rules, such as those for "red", "car", "flat tire", etc.
"When "building-block" rules, such as those for "car", have proved useful in past combinations, it is at least plausible that they will prove useful in new similar combinations. To exploit these possibilities, the rules must be organized in a way that permits clusters of rules to be activated in changing combinations. Building-block rules then give the system a capacity for transfering experience to new situations" (Ibid).
The whole of HOLLAND's book is dedicated to explain in detail how such rules work in natural systems and how they could be introduced in complex artificial systems. This last field is now in full development (f. ex. "artificial life" and "genetic algorithms").
HOLLAND distinguishes "intrinsic" parallelism, proper to the system itself, and "implicit" parallelism, as the "implicit" workings of the algorithm (Ibid, p.X).
Categories
- 1) General information
- 2) Methodology or model
- 3) Epistemology, ontology and semantics
- 4) Human sciences
- 5) Discipline oriented
Publisher
Bertalanffy Center for the Study of Systems Science(2020).
To cite this page, please use the following information:
Bertalanffy Center for the Study of Systems Science (2020). Title of the entry. In Charles François (Ed.), International Encyclopedia of Systems and Cybernetics (2). Retrieved from www.systemspedia.org/[full/url]
We thank the following partners for making the open access of this volume possible: