REQUISITE VARIETY 11/21 



is now "that which disturbs", and T has no third input; so the 

 formulation agrees with that of S.l 1/4. 



There is, of course, no suggestion here that the noise, as a distur- 

 bance, can be allowed for magically by merely thinking differently 

 about it. The suggestion is that if we start again from the beginning, 

 and re-define D and T then some new transformation of D may be 

 able to restore regulation. The new transformation will, of course, 

 have to be more complex than the old, for D will have more 

 components. 



11/19. Initial states. A related case occurs when T is some machine 

 that shows its behaviour by a trajectory, with the outcome E 

 depending on the properties of T's trajectory. The outcomes will 

 then usually be affected by which of T's states is the initial one. 

 How does r's initial state come into the basic formulation of S.l 1/4? 



If the initial state can be controlled, so that the trajectory can be 

 started always from some standardised state, then no difficulty 

 arises. (In this connexion the method of S.7/25 may be useful.) 

 It may however happen, especially if the system is very large, that 

 r's initial state cannot be standardised. Does the basic formulation 

 include this case? 



It does; for D, as a vector, can be re-defined to include T's initial 

 state. Then the variety brought to E by the variety in T's initial 

 state is allotted its proper place in the formulation. 



11/20. Compound target. It may happen that the acceptable 

 states rjdii E may have more than one condition. Thus of a thermo- 

 stat it might be demanded that 



(i) it shall usually stay between 36° and 37°C; 

 (ii) if displaced by ±10" it shall return to the allowed range within 

 one minute. 



This difficulty can be dealt with by the same method as in S. 11/17, 

 by recognising that E may be a vector, with more than one compo- 

 nent, and that what is acceptable (-17) may be given in the form of 

 separate specifications for each component. 



Thus, by allowing E to become a vector, the basic formulation of 

 S.11/4 can be made to include all cases in which the target is complex, 

 or conditional, or qualified. 



11/21. Internal complexities. As a last example, showing how 

 comprehensive the basic formulation really is, consider the case in 

 which the major problem seems to be not so much a regulation as an 



217 



