THE NEURAL BASIS OF LEARNING 



'487 



ECS may also have differential effects upon two 

 behaviors — it may leave one kind entirely unaffected 

 while attenuating or even abolishing another (20; 71, 

 p. 370; 89). For example, a thirsty rat trained to press 

 a lever for drops of water will do so just as well after 

 ECS as before. If, however, it has also been trained to 

 'expect' a painful shock following an auditory stimu- 

 lus, the animal fails after ECS to display any signs of 

 this conditioned emotional response (20; see Chapter 

 LXIII by Brady in this volume). The conditioned 

 emotional response (CER), consisting of immobility 

 and autonomic events like piloerection and defecation, 

 disappears completely for many days, returning, how- 

 ever, without further training in a matter of weeks. 



These experiments specify two novel aspects of the 

 central correlates of learning. In the first place, they 

 are not instantly established; a period of time follows 

 the termination of the actual conditioning situation 

 during which the central changes "gel," so to speak, 

 and the temporary connections become Stabilized. In 

 the second place, central connections behave differ- 

 ently when subjected to disrupting influences; the 

 CER, which by most criteria is an extremely stable 

 response, vanishes with ECS while the operant ( !R 

 remains apparently untouched. Whether this means 

 that the correlates for CER and CR are different in 

 place or in kind remains to be seen, but a real differ- 

 ence between them has been defined by an elegantly 

 simple experiment. 



Summary 



Psychologists recognize many temporal variables in 

 learning and have extensively studied the effects of 

 reward and punishment upon it. The brain stimula- 

 tion experiments, taken as a whole, provide some 

 glimmerings of what central events underlie these phe- 

 nomena. The process of acquisition, it appears, goes 

 on for a considerable time after the actual training is 

 over. Furthermore, the events related to 'memory' 

 can be greatly disturbed, with certain tvpes of learn- 

 ing suffering far more than others in this regard. As 

 for rewards and punishments, the possibility of spe- 

 cific centers for each of these seems open to experi- 

 mental attack through the self-stimulation technique. 



Besides this, the experiments have made it clear 

 that the locus of the temporary connections in learn- 

 ing lies in the brain. This probablv surprises no one, 

 but there is certainly no harm in having the fact ex- 

 perimentally established. The important point is, 

 however, that the places where and the processes 

 wherein CS and US become connected can now be 



explored, if desired, without interference from neural 

 activity in those parts of the system that merely con- 

 duct information from and to the bodv surface. There 

 is, finally, good reason to believe from self-stimulation 

 experiments and similar evidence that the central re- 

 gions related to CS and US have significantly different 

 properties. 



PSYCHOPHARMACOLOGV 



Until recently the use of drugs as tools for dissecting 

 the learning process has been somewhat unrewarding. 

 Two bibliographies assembling many hundreds of 

 publications attest to the devotion with which investi- 

 gators have attempted to show specific effects of such 

 agents upon acquisition, retention and extinction 

 (8, 18). The failure as yet of any system to emerge 

 from all this effort may be hard to understand at first 

 glance, for it has long been obvious that chemical 

 substances in the blood stream ma) have really pro- 

 found effects upon learning. For example, relatively 

 small amounts of anesthetics unfailingly reduce ani- 

 mals and men to a state where no learning whatever 

 is possible. One might naively expeel thai between 

 the stages of complete anesthesia and none at all a 

 level would be reached where the learning process 

 was, sav, only half impaired. Such a -I. me has, how- 

 ever, never been defined. Similarly, no one questions 

 that a child with thyroid gland deficiency learns 

 poorly and that specific replacement therapv goes far 

 toward restoring him to normal, the explanation for 

 this is, however, still a mystery. Discussions showing 

 the inconclusiveness of studies with hormones (71, 

 p. 386) and other biochemical factors (163, p. 532) 

 in learning arc available, Wikler's summary (244) 

 should prove especially useful. 



Within the last lew years research using drugs in 

 behavioral studies has accelerated remarkably, due in 

 pail at least to the advent of new synthetic and natural 

 substances having powerful effects upon behavior 

 (e.g. "tranquilizers*). Recent symposia (50, 62, 109) 

 summarize the present status of the new discipline of 

 psvchopharmacology that is in the process of evolu- 

 tion. In one of these (50) six papers deal specifically 

 with drugs and learning. Sidman (218), for example, 

 fully discusses the interaction of behavioral and drug 

 variables, and demonstrates with numerous examples 

 how a given drug may lie assayed for its action upon 

 many different learning situations in the same animal. 

 While no general principles of the action of drugs in 

 learning can vet be stated with certainty, the results 



