Journal on Policy & Complex Systems Volume 1, Number 2, Fall 2014 | Page 17

������������������������������������
Given this context , a regulator is a good one when he or she achieves the goals established for the systems , considering the relationships among its parts and the disturbances .
This proposed scheme is not subject to the assumptions of linearity , continuity , temporality , or even the existence of a system of measurement . However , when the set of possible events is classifiable , the effectiveness of the regulator is higher when his or her own actions are minimal . The proposed theorem is , thus : “… the theorem says that the best regulator of a system is one which is a model of that system in the sense that the regulator ’ s actions are merely the system ’ s actions as seen through a mapping ” ( Conant & Ashby , 1970 , p . 96 ).
The theorem implies that there can be equally effective regulators , but that some might be unnecessarily complex . Besides , Conant and Ashby conclude that the regulator is , in fact , searching for the best model that maps the relations between the system and the events , conditioned to the objective , which is contained in the universe of the system .
The epistemology of complexity with a focus on organizations and the possibilities of modeling is also discussed by Cilliers ( 1998 , 2001 ).
The authors on this section reinforce the idea that models are compulsory if one aims to understand the mechanisms of the system under analysis . Once the processes are understood , one can move forward and attempt to measure , for example , the efficiency of these processes .
VI - Discussion

An earlier attempt to consolidate the

knowledge around the field of complexity was made by Waldrop ( 1992 ), and a more recent one was done by Mitchell ( 2011 ). Our attempt at this paper is more in the sense of reading and interpreting the original papers , rather than composing ourselves a clear , contemporary epistemological compound of the field .
Israel ( 2005 ) criticizes the argument that the concepts presented in this paper together constitute a science of complexity . In fact , he may have a point . Anyway , not constituting a science does not change the fact that for a number of systems classified as complex , the concepts presented bring analytical power and contribute to the advance of disciplines and the cross-fertilization among disciplines . The point to highlight , thus , is that whether a science or not ; the notions of complex systems presented by the classic authors contribute fundamentally to the scientific approach in at least two manners .
On one hand the concepts bring formalism and theoretical scope for the discussions in areas of the science that are less used them . On the other hand , the concepts effectively handle elements that even in the so-called hard sciences are routinely ignored given their frequent mathematical intractability , those are : ( a ) the dynamics of the events and their strong crossed causality ( interaction ) that includes feedback elements , ( b ) the randomness of events that may be relevant and together lead to rupture points and structural change , and ( c ) self-organization without central control .
Such elements should be considered by the scientist every time he or she analyzes phenomena that have heterogeneous agents interacting with each other and with the environment in a dynamic and nonlinear manner , in systems that have self-organization and differentiated characteristics for distinct scales ; and that have the ability to evolve and adapt . In sum , one should proceed with those concepts in mind when studying systems which are observable and frequent in nature and in society . Such scientist should also model the system at hand meaning that
15