H. ENGER ROSVOLD AND MORTIMER MISHKIN 561 



be seen that in auditory discrimination, also, the frontal animal is deficient 

 with one training procedure but not with another. The tn'st graph 

 represents the performance level achieved m the last lOO of over 2000 

 training trials in which responses to the negative stimulus were deterred 

 only by the usual extinction procedure. Up to this point the frontal 

 animal appeared to be completely unable to discriminate the auditory 

 cues. However, as soon as responses to the negative stimulus were 

 severely punished by the introduction of strong electric shocks, this monkey 

 discriminated very well indeed. Within 100 trials it achieved almost 

 perfect discrnnination, equalling the performance of the normal. An even 

 more surprising result is illustrated in the last two graphs of the figure. 

 The frontal animal's deficit could be turned on and off, sometimes even 

 within the same testing session, by the simple expedient of removing or 

 reapplying the shock. Such an impairment can harcily be attributed to 

 sensory effects of the lesion. 



Let us consider what the nature of the non-sensory deficit might be. 

 The general principle which emerges from the evidence as it has been 

 developed to this point can be stated briefly as follows. The discrimination 

 learning of an animal with a frontal lesion is not impaired when its initial 

 response tendency to the negative stimulus is weak or the deterrent to such 

 a response is particularly strong. Conversely, when a strong tendency to 

 respond to the negative stinuilus is coupled with only a weak deterrent, a 

 clear-cut deficit appears. If the principle is correct it provides strong 

 support for the view that damage to the frontal lobes results in a loss of 

 inhibition. This conception of a loss of inhibition following prefrontal 

 damage is a recurrent theme which can be traced at least as far back as 

 Schafer's (1898) Textbook of Pliysiolo(iy. Most recently, a sophisticated 

 variation of this hypothesis has been proposed by Konorski and his 

 associates (Brutkowski, Konorski, Lawicka, St^pieh and St^^pien, 1956). 



A particularly clear example of a c^cficit which might be labelled loss of 

 inhibition is shown in Fig. 6. In this task, a 'go-no go' discrimination, 

 operated monkeys and controls were trained to displace a baited food-cup 

 when they heard one sound, but to refrain from displacing it when they 

 heard another. If an animal made an error by responding to the negative 

 sound, or by failing to respond to the positive one, the same signal was 

 presented again and again until the animal did respond to it correctly. The 

 figure is a plot of the number of extra times the negative stimulus had to 

 be presented minus the number of extra positives, over blocks of 200 

 trials. Now, the preliminary training on this task develops in the animal a 

 strong tendency to respond on every trial. This is indicated by the fact 



