564 BRAIN MECHANISMS AND LEARNING 



tasks is suggested by the fact, well known in conditioning experiments, 

 that extraneous stimuli tend to disinhibit weakly inhibited responses. 



It must of course be emphasized that no attempt has been made here to 

 work out these three hypothetical mechanisms in any rigorous fashion. 

 The point is simply that on the basis of available neurophysiological 

 evidence, together with the abnormal symptoms observed in the frontal 

 animal's general behaviour, a reasonably strong argument can be deve- 

 loped in favour of each of them. It is this very multiplicity of possible 

 mechanisms, however, which is the chief source of weakness in the theory. 

 An uncritical application of the general concept of a loss of inhibition, 

 without specifying the type, can probably be used to explain any complex 

 behavioural effect of frontal lobe damage. On the other hand, when care 

 is taken to specify which type of inhibitory defect is being considered it is 

 found that there is evidence from discrimination learning experiments 

 that is inconsistent with each of them. 



Consider first certain predictions which follow from the hypothesis that 

 frontal lesions result in a loss of response inhibition. Suppose that two 

 stimuli A and B arc presented at random on separate trials. A signals food 

 on the left, and B signals food on the right. Since both responses are some- 

 times rewarded, the correct response of going left to A on any particular 

 trail necessitates inhibiting the built-in response tendency to the right. 

 Similarly, going right to B involves inhibiting a strong response-tendency 

 to the left. As noted by Stanley and Jaynes (1949) in the development of 

 their theory, a loss of response inhibition should lead to an impairment on 

 this 'go left-go right' type of descrimination test; yet, in experiments of 

 this kind by Pribram and Mishkin (1955), in vision, and by Battig, 

 Rosvold and Mishkin (unpublished), in audition, no such impairment was 

 found. Now for an opposite example, that is, one in which this particular 

 hypothesis would probably not predict impairment. According to our 

 interpretation, a loss of suppression of the somatic motor systems should 

 produce difficulty in inhibiting well-established locomotor and manipula- 

 tory acts. On the other hand, it should not affect the inhibition of such 

 conditioned autonomic reflexes as salivation. However, in experiments by 

 Brutkowski (1957) in which dogs with frontal lesions were trained to 

 differentiate positive from negative conditioned stimuli, a severe impair- 

 ment in inhibiting salivation to the negative stimulus was observed. 



It will be recognized that the preceding illustration, while weakening 

 the hypothesis of a loss of response inhibition, simultaneously strengthens 

 the hypothesis of a loss in the inhibition of appetitive mechanisms. Other 

 evidence, however, raises dii^iculties for this alternative conception. For 



