CHAPTER 15 



Dispersion 



15/1. Systems of part-functions have the fundamental property 

 that each line of behaviour may leave some of the variables 

 inactive. Dispersion occurs when the set of variables made 

 active by one line of behaviour differs from the set made active 

 by another. We will begin to consider the physiological applica- 

 tions of this fact. 



First consider a system of full-functions. Suppose we record 

 a few of its variables' behaviours while it traverses first one line 

 of behaviour and then another. The records would show the 

 variables always fluctuating, and the two records would differ 

 only in their patterns of fluctuation. Now suppose we have a 

 system of part-functions. Again we record some of the variables' 

 behaviours. It may happen that from one initial state the line 

 of behaviour leaves all the recorded variables inactive, while the 

 line from another shows some activity. Since, by S. 6/6, the 

 change of initial state corresponds to ' applying a stimulus ', a 

 by-standing physiologist would describe the affair as a simple 

 case of a mechanism ' responding ' to a stimulus. Since living 

 organisms' responses to stimuli have been sometimes offered as 

 proof that the organism has some power not possessed by 

 mechanisms, we must examine these reactions more closely. 



In S. 6/6 we saw that the most general representation of a 

 1 stimulus ' was a change from one initial state to another. Now 

 in general, even though the system is absolute, the course of 

 the line of behaviour from one initial state puts no restriction 

 on the course from another initial state. From this lack of 

 restriction follow several consequences. 



15/2. The first consequence is that, in a system known only 

 to be complex, however small the difference between the initial 

 states — however slight or simple the stimulus — we can put no 

 limit to the greatness of the difference between the subsequent 



166 



