574 BRAIN MECHANISMS AND LEARNING 



learns even dirt'icult visual discriminations just as rapidly as its unoperatcd control. 



While, for simultaneous discrimination, Konorski has offered an explanation 

 which we have argued applies to retention but not to learning, he suggests an 

 explanation for the symmetrically rewarded 'go-no go' discrimination which 

 seems to us to apply to learning but not retention. His argument here is as follows. 

 The stimulus to which the animal's response was extinguished (the 'no go' stimulus) 

 also inhibited the animal's ahmcntary drive, despite the fact that this stimulus was 

 associated with reward whenever the animal did refrain irom responding to it. 

 Since inhibition of drive normally aids in establishing an inhibitory instrumental 

 response, a loss ot drive inhibition should impair this conditioning process and so 

 produce the reported impairment in discrimination learnnig. But this does not 

 account for the impairment found in the animals that had learned the task before 

 operation. Early in pre-operative training the 'no go' stimulus may have indeed 

 inhibited the alimentary drive because the animals responded to it regularly, and 

 so were rewarded only rarely. As training progressed, however, the animals 

 tended to respond to the 'no go' stimulus less and less frequently, and as a result 

 were rewarded more and more often. Finally, when the animals learned to inhibit 

 perfectly, the 'no go' stimulus was associated with reward on every occasion. 

 Surely at this stage of the experiment the animals had learned to 'expect' food (or, 

 in conditioning terms, to salivate) at the sound of the 'no go' stimulus. This stimulus 

 should no longer have been inhibiting the drive, in which case a post-operative 

 disinhibition of drive could not have caused the loss in post-operative retention. 



Our re-examination of the evidence relates so far only to the hypothesis that 

 //'/;//rt'^ prefrontal lesions produce a loss of drive inhibition. Konorski has suggested, 

 however, that with somewhat more extensive lesions, those invading also the 

 pericruciate area in dogs, a second type of inhibitory defect is produced, viz. loss of 

 response inhibition. But would this combination of disinhibition hypotheses help 

 to explain results which each hypothesis taken singly docs not ? It seems to us that 

 it would, in fact, make it easier tor Konorski to account for his own data demon- 

 strating disinhibition of instrumental (in addition to classical) inhibitory reflexes, 

 and also to account for the impairment on the symmetrically rewarded 'go-no go' 

 discrimination, just discussed. At the same time, however, we believe it would not 

 explain away the absence of impairment on simultaneous discriminations, nor on 

 the 'go left-go right' successive discriminations, which we described in our paper. 



Grastyan. Some years ago Dr Shuskin pubhshed some work about this investi- 

 gation concerning the influence of extirpation of frontal lobes on the behavioural 

 aspect of conditioned dogs. If I am not mistaken, Dr Shuskin's data from many 

 points of view arc in accordance with the data of Dr Rosvold. 



RosvoLD. I don't know these studies. I am very happy to have this reference. 



EsTABLE. Do conditioned reflexes increase or decrease the capacity for discrimina- 

 tion? I ask this question because discrimination is used in a very particular sense. It I 

 have ten objects one of which is of iron and a magnet, only the one of iron is 

 attached by the magnet. This is not because there has been any discrimination but 

 simply because it could not be otherwise. In the case of animals with conditioned 

 reflexes, I wonder whether they respond in a specific way to a stimulus which they 

 have discriminated or whether they react because they cannot do otherwise. Does 

 discrimination occur or is it simply a way of reacting = 



Anokhin. During a period of years, from 1935 to 1949, one ot my associates. 



