Sympoetic.com
creating with...
CIRCULARITY
People distinguish many forms of circularity, or “circular causality”.
All of them are processes, such that the phenomenon that results happens in a new domain that only arises as the process operates.


Cycles
The simplest form of circularity is cyclicity.
The wheels of a car go round and round, clocks repeat a circular motion, the earthy cycles the sun ... and there are many other cycles.
As long as the cycle does not connect with any other process, it is simply a repetition. If it couples with displacement in any form, an epigenic process can arise and things change. The seasonal cycles result in cumulative changes to the the earth.
Mararie, Flickr Creative Commons
Abstractions From a Matrix
First of all I have to point out that the commonly used notion of linear causality, ie. A causes B (which may be stated in ever more complicated forms, e.g. A and D cause B which causes C) is inherently a selection or abstraction of some relations out of other possible relations. To take action, it is useful to know what things are directly and indirectly connected to what we do. There is no need to abandon this useful simplification.
What is less evident is that circularities are similarly abstracted from a much much more intricate and dimensional matrix of relations than what we can directly perceive.
Yet, it is also very useful to understand circularities.


Feedback
A useful, and well known circularity is “feedback”... i.e. where the consequence “feeds back” and changes the thing that first caused the change.
We speak loosely of feedback in conversation, where what is heard is repeated back. Yes, this can become a loop, but one iteration is usually sufficient for coordination.I find that people who are only familiar with conversational feedback often misunderstand the systems concepts of “negative feedback” and “positive feedback.” They think that negative means bad or unkind.
Reinforcing or Balancing
A better choice of words for positive feedback is “reinforcing feedback” - that is the circularity leads to more and more of some change. The change could be increase, but it could be decrease, and it could be about any attribute (size, brightness, rate of flow, strength, anger, etc.)
Conversely “negative feedback” is better named “balancing feedback”. In this situation further changes in a system act to bring the change back towards the original condition. “Negative feedback” is usually a very good thing, i.e. has positive consequences. The balancing effect enables homeostasis; that is keeping a system within appropriate bounds. Homeostasis is the consequence of all the circularities that maintain the body temperature of mammals.


In some lineages of systems studies influences that cause increase are shown with a +sign and those that cause decrease are shown with a -sign.
The above figure has one plus and one minus arrow, thus depicting balancing feedback. The figure to the left has loops with two of the same sign. The plus sign based reinforcing feedback leads to increase, and if the influences are proportional to current value it leads exponential growth. The loop with two minus signs leads to decline or even extinction. Note that in practice either process may be contained by some other influence coming into play.
It is sometimes claimed that one can simply count the negative and positive
signs and predict whether a complex system is going to be balanced or reinforcing. This is an oversimplification as the magnitude of the effects is also relevant
I would like to reiterate that both influence chains and circular relations are abstractions. Both comprise a focused look at some aspect of a network. Natural systems, and complex designed systems, are networks of relations where many different things may influence one part, and that part in turn may influence many others. A complex web results.
Yet we understand such webs by paying attention to what we discern to be the strongest relationships, and we may do this by postulating chains, circles, or a combination of chains and circularities.
