DESIGN FOR A BRAIN 15/3 



round a corner should be preceded both by a slowing down and 

 by a change of gear. The whole system thus shows that temporary 

 and conditional division into subsystems that is typical of the 

 whole that is composed largely of part-functions. 



Thus the terrestrial environment conforms largely to the poly- 

 stable type. 



15/3. To study how ultrastability will act when the environment 

 is not fully joined, we shall have to use the strategy of S. 2/17 and 

 pick out certain cases as type-forms. We will therefore consider 

 environments with four degrees of connectedness. 



First we will consider (in S. 15/4-7) the ' whole ' in which the 

 connexion between the parts is actually zero — the limiting case 

 as the connexions get less and less. 



In S. 15/8-11 we will consider the case in which actual con- 

 nexions exist, but in which the subsystems are connected in a 

 chain, without feedback between subsystems. These two cases 

 will suffice to demonstrate certain basic properties. 



In the next chapter we will consider the more realistic case in 

 which the subsystems are joined unrestrictedly in direction, so 

 that feedback occurs between the subsystems. This case will be 

 considered in two stages: first, in S. 16/2-4 we will dispose of the 

 case in which the connexions are rich; and then, from S. 16/5 

 onwards, we will consider the most interesting case, that in which 

 the connexions are in all directions, so that feedback occurs 

 between the subsystems, but in which the connexions are not rich 

 so that the whole can be regarded as formed from subsystems 

 each of which is richly connected internally, joined by connexions 

 (between the subsystems) that are much poorer — the case, in fact, 

 of the system that is neither richly joined nor unjoined. 



Adaptation in iterated systems 



15/4. The first case to be considered is that in which the whole 

 system, of organism and environment, is actually divided into 

 subsystems that (at least during the time of observation) do not 

 have any effective action on one another. Thus instead of A 

 in Figure 15/4/1 we are considering B. (For simplicity, the 

 diagram shows lines instead of arrows.) If the whole system 

 consists of organism and environment, the actual division between 



196 



