13/19 THE SYSTEM WITH LOCAL STABILITIES 



Localisation in the polystable system 



13/18. How will responses to a stimulus be localised in a poly- 

 stable system? — how will the set of the active variables be dis- 

 tributed over the whole set ? 



In such a system, the reaction to a given stimulus, from a given 

 state, will be regular and reproducible, for the whole is state- 

 determined. To this extent its behaviour is lawful. But when 

 the observer notices which variables have shown the activity it 

 will probably seem lawless, for the details of where the activation 

 spreads to have been determined by the sampling process, and 

 the activated variables will probably be scattered over the system 

 apparently haphazardly (subject to S. 13/4). Thus the question 

 1 Is the reaction localised ? ' is ambiguous, for two different 

 answers can be given. In the sense that the activity is restricted 

 to certain variables of the whole system, the answer is ' yes ' ; but 

 in the sense that these variables occur in no simply describable 

 way, the answer is ' no '. 



An illustration that may be helpful is given by the distribution 

 over a town of the chimneys that ' smoke ' (suffer from a forced 

 down-draught) when the wind blows from a particular direction. 

 The smoking or not of a particular chimney will be locally deter- 

 minate ; for a wind of a particular force and direction, striking the 

 adjacent roofs at a particular angle, will regularly produce the 

 same eddies, which will determine the smoking or not of the 

 chimney. But geographically the smoking chimneys are not 

 distributed with any simple regularity ; for if a plan of the town is 

 marked with a black dot for every chimney that smokes in an 

 east wind, and a red dot for every one that smokes in a west wind, 

 the black and red dots will probably be mixed irregularly. The 

 phenomenon of ' smoking ' is thus determined in detail yet 

 distributed geographically at random. 



13/19. Such is the ' localisation ' shown by a polystable system. 

 In so far as the brain, and especially the cerebral cortex, cor- 

 responds to the polystable, we may expect it to show ' localisation ' 

 of the same type. On this hypothesis we would expect the brain 

 to behave as follows. 



The events in the environment will provide a continuous stream 

 of information which will pour through the sense organs into the 



181 



