DESIGN FOR A BRAIN 9/17 



mammal and its consequent sexual goal-seeking behaviour. A 

 simple alternation between ' present ' and ' absent ', or between 

 two levels with a threshold, would be sufficient to account for 

 any degree of complexity in the two behaviours, for the com- 

 plexity is not to be related to the hormone-parameter but to the 

 nervous system that is affected by it. Since the mammalian 

 nervous system is extremely complex, and since it is, at almost 

 every point, sensitive to both physical and chemical influences, 

 there seems to be no reason to suppose that the directiveness of 

 the sex-hormones on the brain's behaviour is essentially different 

 from that of any parameter on the system it controls. (That 

 the sex-hormones evoke specifically sexual behaviour is, of course, 

 explicable by the fact that evolution, through natural selection, 

 has constructed specific mechanisms that react to the hormone 

 in the specific way.) 



The gene-pattern and ultrastability 



9/17. We can now return to the questions of S. 1/9 and ask 

 what part is played by the gene-pattern in the determination of 

 the process of adaptation. 



Taking the diagram of immediate effects (Figure 7/5/1) as 

 basis, the question is answerable without difficulty if we take the 

 system part by part and channel by channel. 



The environment is, of course, assumed to be given arbitrarily; 

 so is the channel by which the environment affects the essential 

 variables (S. 7/3). The essential variables and their limits are 

 determined by the gene-pattern (S. 3/14); for these are species' 

 characteristics. 



In the living organism, the reacting part R has, in effect, three 

 types of ' input ' : there is the sensory input from the environment, 

 there are the values of its parameters in S, and there are those 

 parameters that were determined genetically during embryonic 

 development. (That all three may be regarded as ' input ' has 

 been shown in /. to C, S. 13/11). These three sets of parameters 

 vary on very different time-scales: the genetic parameters, those 

 that make this a dog-brain and that a bird-brain, are in evidence 

 only at one period in a lifetime ; the parameters in S, if the adapta- 

 tion proceeds by clearly marked trials, change only between trial 

 and trial; and the parameters at the sensory input vary more or 



134 



