REGULATING THE VERY LARGE SYSTEM 13/11 



verified that this sequence, as a protocol, defines the machine with 

 input: 



Thus when the Grand Disturbance is a vector whose components 

 are all from a basic set of disturbances, the Grand Response can 

 either be a vector of equal variety or the output of a suitable 

 machine with input. 



13/10. Suppose that the regulation discussed throughout Part III 

 is the responsibility of some entity Q, often the possessor of the 

 essential variables E. Through the previous chapters we have 

 studied how the regulator R must behave. We have now seen that 

 in the case when the disturbances are repetitive, Q has the option 

 of either being the regulator (i.e. acting as K) or of building a machine 

 that, once built, will act as R and will carry out a regulation of 

 indefinite length without further action by i3. We have thus arrived 

 at the question: should Q achieve the regulation directly, by his 

 own actions, or should he build a machine to undertake the work ? 

 The question would also have arisen for another reason. From 

 the beginning of Part III we took for granted that the regulator 

 existed, and we then asked what properties it must have. Nothing 

 was said about how the regulator came to be made, about the factors 

 that brought it into existence. Thus, having seen in S.10/5 how 

 advantageous it would be if the organism could have a regulator, 

 we showed no means by which the advantage could be gained. 



For both these reasons we must now start to consider how a 

 regulatory machine is actually to be designed and made. Here we 

 shall be thinking not so much of the engineer at his bench as of the 

 brain that, if it is to achieve regulation in its learned reactions, must 

 somehow cause the development of regulatory machinery within 

 the nervous material available; or of the sociologist who wants a 

 regulatory organisation to bring harmony into society. 



To understand what is involved, we must look more closely at 

 what is impUed, in principle, in the "designing" of a regulatory 

 machine. 



DESIGNING THE REGULATOR 



13/11. Design as convnunication. Let us forget, temporarily, all 

 about "regulation", and turn simply to certain questions related to 

 the design and construction of a machine, any machine. 



251 



