WARNING (Early)
| Collection | International Encyclopedia of Systems and Cybernetics |
|---|---|
| Year | 2004 |
| Vol. (num.) | 2(2) |
| ID | ◀ 3758 ▶ |
| Object type | General information |
Any symptom of probable change that can only be observed and understood within a systemic framework
Early warnings offer some predictability in nonlinear systems, even if within limits.
Predictability in linear processes is theoretically reliable. However, linear processes are exceptional and generally quite unstable in time.
Nonlinear systems are submitted to different short, median and long term trends, which are imbricated and mutually constraining. Long term trends are difficult to discover, because the needed data must cover a span that easily exceeds the professional life of a scientist, or manager, and even the accumulated historical information. Examples are long term reversals in stock markets; the movement of geological faults; multisecular mega-floods in river basins, etc.
However, long term trends, once established, constitute a more or less reliable frame for median and short term oscillations. This means that any strong or sudden deviation in these lesser trends probably heralds a reversal of the phenomenon, or behavior of the system, possibly in a catastrophic or chaotic way.
The development of a good general theory of early warnings could become a very significant contribution of systemics.