THE MEANING OF PERCEPTION 173 



always be elicited. It is true of habitual, learned, and mechani- 

 cally repeated actions which have, so to speak, become customary 

 to the animal that makes them. And it is true (again in general) 

 of the instinctive actions which make up a considerable part of 

 the behaviour of invertebrate animals in particular, and even of 

 the higher vertebrates. Yet habits can be varied, and instinctive 

 actions are not quite invariable. Thus some animals, such as 

 bees, which build structures of certain materials, may alter their 

 methods when different materials are supplied to them. Clearly 

 the responses made even by intact, normal animals are not rigidly 

 determined ones capable in all circumstances of being predicted. 

 And such behaviour as we have indicated in Chapter VIII. in 

 dealing with the functioning of the nervous system shows that 

 there may be spontaneity of behaviour on the part of all animals 

 that have been studied with sufficient care. They react to 

 stimuli sometimes in one way, sometimes in others, and again 

 they may not react at all, or they may act in unpredictable 

 ways even when there are no apparent stimuli. One can see 

 this in men and women: a slight irritation may cause a man to 

 cough again and again, but he may easily repress the response, 

 even when the same physical stimulus recurs. Now in all such 

 cases of variable behaviour to the same kind of stimulus it is 

 . open to us to say that there may have been other unrecognised 

 stimuli which modified the response; that in cases where the 

 behaviour seemed to be absolutely spontaneous there " must have 

 been " some causes which we could not trace, or that there were 

 " physiological conditions " introducing elements of uncertainty, 

 and so on. All this may be so ; still, the fact is that in a great many 

 cases we cannot say that organic actions are preceded by definite 

 specifiable causes, and when we say that a, response " must " 

 always have its appropriate stimulus we simply dogmatise. 



It hardly seems necessary to lay stress on the conclusion that 

 we are about to make ; nevertheless, it is well to do so. Inanimate 

 occurrences are always " more or less " determined, and equivocal 

 results invariably suggest defects in the experimental methods 

 employed, or neglect of some condition or other, or a degree of 

 complexity incapable of complete analysis. That a result can 

 be predicted is a kind of test of a successful physical investigation 

 — in fact, we disregard all other results; they are not useful to 

 us. To some extent this physical determinism is true of the 

 functioning of the lower animals and of the higher ones that 



