DESIGN FOR A BRAIN 10/12 



The sleeve passing near open jam is another situation, P 2 ; adapta- 

 tion to this, too, can occur. And the alterations necessary in 

 adaptation to P 2 will not, in our present system, cause loss of 

 the adaptation to P v 



Whether the whole situation can be adapted to by such a 

 sequence of sub-adaptations (to P v to P 2 , etc.) depends on the 

 environment: only certain types of environment will allow such 

 fragmentation. If such types of environment are frequent in an 

 organism's life, then there will be advantage in evolution if the 

 species changes so that each organism is provided, genetically, 

 with a mechanism similar to that of Figure 10/9/1. 



10/12. To amplify the point, we may consider the case of an 

 organism that lives in an environment that consists of many 

 sensorily-different situations, and such that each situation is 

 adequately met by one of two reactions, eat or flee say, so that 

 the organism's problem in life is to allot one of the two reactions 

 to each of many situations. In such an environment, the reacting 

 part R of Figures 7/5/1 or 10/9/1 could be quite small, for it 

 requires only mechanism capable of performing two reflexes. The 

 stores of step-mechanisms, however, would have to be large, and 

 the gating mechanism r perhaps elaborate; for here would have 

 to be as many records as there are sensory situations. Each 

 would require its own locus of storage, and the gating mechanism 

 would have to be able to ensure that each situation led to its 

 particular locus. In such a world we would, therefore, expect the 

 organism to have a differently proportioned brain from one that 

 lived in a world such as was presented to the Homeostat. 



It should not be beyond the biologist's powers to identify a 

 species with such an environment. Examination of the organism's 

 nervous system might then enable some fundamental identifica- 

 tions to be made. 



10/13. The objection may be raised that the specification deduced 

 in this chapter is too vague to be of use to the worker who wants 

 to find the corresponding mechanism in, say, the human brain. 

 The reply must be that the specification is right to be vague; for 

 what is given — that adaptation occurs with accumulation — 

 specifies an extremely broad class of mechanisms, so that a very 

 great diversity of actual machines could all show adaptation with 



146 



